{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 21520, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 4.646840148698885e-05, "grad_norm": 0.0, "learning_rate": 0.0, "loss": 1.1822, "step": 1 }, { "epoch": 9.29368029739777e-05, "grad_norm": 0.0, "learning_rate": 0.0, "loss": 1.095, "step": 2 }, { "epoch": 0.00013940520446096655, "grad_norm": 0.0, "learning_rate": 0.0, "loss": 1.274, "step": 3 }, { "epoch": 0.0001858736059479554, "grad_norm": 12.325597473195758, "learning_rate": 1.5479876160990715e-08, "loss": 1.6116, "step": 4 }, { "epoch": 0.00023234200743494423, "grad_norm": 12.325597473195758, "learning_rate": 1.5479876160990715e-08, "loss": 1.9589, "step": 5 }, { "epoch": 0.0002788104089219331, "grad_norm": 13.139254183641967, "learning_rate": 3.095975232198143e-08, "loss": 1.5575, "step": 6 }, { "epoch": 0.0003252788104089219, "grad_norm": 9.090864706089535, "learning_rate": 4.6439628482972136e-08, "loss": 1.3278, "step": 7 }, { "epoch": 0.0003717472118959108, "grad_norm": 9.090864706089535, "learning_rate": 4.6439628482972136e-08, "loss": 1.442, "step": 8 }, { "epoch": 0.00041821561338289964, "grad_norm": 11.965991671527695, "learning_rate": 6.191950464396286e-08, "loss": 1.4359, "step": 9 }, { "epoch": 0.00046468401486988845, "grad_norm": 11.260957000947945, "learning_rate": 7.739938080495356e-08, "loss": 1.4151, "step": 10 }, { "epoch": 0.0005111524163568773, "grad_norm": 13.652915880647445, "learning_rate": 9.287925696594427e-08, "loss": 1.5249, "step": 11 }, { "epoch": 0.0005576208178438662, "grad_norm": 11.70158356418506, "learning_rate": 1.08359133126935e-07, "loss": 1.3192, "step": 12 }, { "epoch": 0.0006040892193308551, "grad_norm": 14.47467083168644, "learning_rate": 1.2383900928792572e-07, "loss": 1.8444, "step": 13 }, { "epoch": 0.0006505576208178438, "grad_norm": 11.696617740217153, "learning_rate": 1.393188854489164e-07, "loss": 1.5033, "step": 14 }, { "epoch": 0.0006970260223048327, "grad_norm": 11.696617740217153, "learning_rate": 1.393188854489164e-07, "loss": 2.0322, "step": 15 }, { "epoch": 0.0007434944237918215, "grad_norm": 13.136558849066638, "learning_rate": 1.5479876160990712e-07, "loss": 1.6248, "step": 16 }, { "epoch": 0.0007899628252788104, "grad_norm": 17.101301593745493, "learning_rate": 1.7027863777089783e-07, "loss": 2.0198, "step": 17 }, { "epoch": 0.0008364312267657993, "grad_norm": 7.46912117691421, "learning_rate": 1.8575851393188855e-07, "loss": 1.1708, "step": 18 }, { "epoch": 0.0008828996282527881, "grad_norm": 10.085073971982775, "learning_rate": 2.0123839009287928e-07, "loss": 1.3444, "step": 19 }, { "epoch": 0.0009293680297397769, "grad_norm": 15.280444541665624, "learning_rate": 2.1671826625387e-07, "loss": 1.8961, "step": 20 }, { "epoch": 0.0009758364312267658, "grad_norm": 8.579708479916338, "learning_rate": 2.321981424148607e-07, "loss": 1.2469, "step": 21 }, { "epoch": 0.0010223048327137546, "grad_norm": 13.5743499989944, "learning_rate": 2.4767801857585145e-07, "loss": 1.6971, "step": 22 }, { "epoch": 0.0010687732342007435, "grad_norm": 10.189569752892938, "learning_rate": 2.6315789473684213e-07, "loss": 1.3486, "step": 23 }, { "epoch": 0.0011152416356877324, "grad_norm": 21.005501134707433, "learning_rate": 2.786377708978328e-07, "loss": 2.0465, "step": 24 }, { "epoch": 0.0011617100371747212, "grad_norm": 11.265041151122636, "learning_rate": 2.9411764705882356e-07, "loss": 1.3697, "step": 25 }, { "epoch": 0.0012081784386617101, "grad_norm": 10.340448658879515, "learning_rate": 3.0959752321981424e-07, "loss": 1.3508, "step": 26 }, { "epoch": 0.001254646840148699, "grad_norm": 10.340448658879515, "learning_rate": 3.0959752321981424e-07, "loss": 2.8984, "step": 27 }, { "epoch": 0.0013011152416356876, "grad_norm": 9.119218126939007, "learning_rate": 3.25077399380805e-07, "loss": 1.2323, "step": 28 }, { "epoch": 0.0013475836431226765, "grad_norm": 12.785776400838552, "learning_rate": 3.4055727554179567e-07, "loss": 1.4178, "step": 29 }, { "epoch": 0.0013940520446096654, "grad_norm": 9.064615995419318, "learning_rate": 3.560371517027864e-07, "loss": 1.2611, "step": 30 }, { "epoch": 0.0014405204460966542, "grad_norm": 9.879534105559147, "learning_rate": 3.715170278637771e-07, "loss": 1.3302, "step": 31 }, { "epoch": 0.001486988847583643, "grad_norm": 8.821025160599655, "learning_rate": 3.869969040247679e-07, "loss": 1.0473, "step": 32 }, { "epoch": 0.001533457249070632, "grad_norm": 6.920865692857646, "learning_rate": 4.0247678018575857e-07, "loss": 1.0121, "step": 33 }, { "epoch": 0.0015799256505576208, "grad_norm": 9.444293725913676, "learning_rate": 4.1795665634674925e-07, "loss": 1.108, "step": 34 }, { "epoch": 0.0016263940520446097, "grad_norm": 7.192499319819454, "learning_rate": 4.3343653250774e-07, "loss": 1.1552, "step": 35 }, { "epoch": 0.0016728624535315986, "grad_norm": 7.611596324690644, "learning_rate": 4.489164086687307e-07, "loss": 1.1051, "step": 36 }, { "epoch": 0.0017193308550185874, "grad_norm": 10.576212423314916, "learning_rate": 4.643962848297214e-07, "loss": 1.176, "step": 37 }, { "epoch": 0.0017657992565055763, "grad_norm": 40.486057594085906, "learning_rate": 4.798761609907122e-07, "loss": 2.155, "step": 38 }, { "epoch": 0.001812267657992565, "grad_norm": 6.412687442831512, "learning_rate": 4.953560371517029e-07, "loss": 0.9457, "step": 39 }, { "epoch": 0.0018587360594795538, "grad_norm": 9.000782736019625, "learning_rate": 5.108359133126935e-07, "loss": 1.1815, "step": 40 }, { "epoch": 0.0019052044609665427, "grad_norm": 13.64589440716161, "learning_rate": 5.263157894736843e-07, "loss": 1.4082, "step": 41 }, { "epoch": 0.0019516728624535316, "grad_norm": 9.096533466669271, "learning_rate": 5.41795665634675e-07, "loss": 1.1437, "step": 42 }, { "epoch": 0.0019981412639405206, "grad_norm": 6.921298062110945, "learning_rate": 5.572755417956656e-07, "loss": 0.9043, "step": 43 }, { "epoch": 0.0020446096654275093, "grad_norm": 6.183986383079942, "learning_rate": 5.727554179566564e-07, "loss": 0.8824, "step": 44 }, { "epoch": 0.002091078066914498, "grad_norm": 12.539391339098984, "learning_rate": 5.882352941176471e-07, "loss": 1.345, "step": 45 }, { "epoch": 0.002137546468401487, "grad_norm": 10.517792648255472, "learning_rate": 6.037151702786379e-07, "loss": 1.1673, "step": 46 }, { "epoch": 0.0021840148698884757, "grad_norm": 11.104509403022396, "learning_rate": 6.191950464396285e-07, "loss": 1.1335, "step": 47 }, { "epoch": 0.0022304832713754648, "grad_norm": 10.562598403967508, "learning_rate": 6.346749226006192e-07, "loss": 0.7548, "step": 48 }, { "epoch": 0.0022769516728624534, "grad_norm": 6.216138725576131, "learning_rate": 6.5015479876161e-07, "loss": 0.8764, "step": 49 }, { "epoch": 0.0023234200743494425, "grad_norm": 8.397004093610036, "learning_rate": 6.656346749226007e-07, "loss": 0.9599, "step": 50 }, { "epoch": 0.002369888475836431, "grad_norm": 6.939638470355238, "learning_rate": 6.811145510835913e-07, "loss": 0.6645, "step": 51 }, { "epoch": 0.0024163568773234202, "grad_norm": 11.07867602223368, "learning_rate": 6.965944272445821e-07, "loss": 0.7248, "step": 52 }, { "epoch": 0.002462825278810409, "grad_norm": 8.042166963547556, "learning_rate": 7.120743034055728e-07, "loss": 0.8974, "step": 53 }, { "epoch": 0.002509293680297398, "grad_norm": 19.32660094949193, "learning_rate": 7.275541795665634e-07, "loss": 0.9962, "step": 54 }, { "epoch": 0.0025557620817843866, "grad_norm": 6.675628604151682, "learning_rate": 7.430340557275542e-07, "loss": 0.7657, "step": 55 }, { "epoch": 0.0026022304832713753, "grad_norm": 8.761071689534191, "learning_rate": 7.585139318885449e-07, "loss": 0.6587, "step": 56 }, { "epoch": 0.0026486988847583643, "grad_norm": 5.617220490043353, "learning_rate": 7.739938080495358e-07, "loss": 0.639, "step": 57 }, { "epoch": 0.002695167286245353, "grad_norm": 17.031475228248567, "learning_rate": 7.894736842105263e-07, "loss": 0.8316, "step": 58 }, { "epoch": 0.002741635687732342, "grad_norm": 9.150497679816867, "learning_rate": 8.049535603715171e-07, "loss": 0.8141, "step": 59 }, { "epoch": 0.0027881040892193307, "grad_norm": 4.268356243105394, "learning_rate": 8.204334365325079e-07, "loss": 0.5707, "step": 60 }, { "epoch": 0.00283457249070632, "grad_norm": 7.372949576795533, "learning_rate": 8.359133126934985e-07, "loss": 0.7443, "step": 61 }, { "epoch": 0.0028810408921933085, "grad_norm": 5.229498051068837, "learning_rate": 8.513931888544892e-07, "loss": 0.5481, "step": 62 }, { "epoch": 0.0029275092936802975, "grad_norm": 6.425877908642457, "learning_rate": 8.6687306501548e-07, "loss": 0.5252, "step": 63 }, { "epoch": 0.002973977695167286, "grad_norm": 5.748295238901623, "learning_rate": 8.823529411764707e-07, "loss": 0.4958, "step": 64 }, { "epoch": 0.0030204460966542753, "grad_norm": 6.249941177420096, "learning_rate": 8.978328173374614e-07, "loss": 0.5644, "step": 65 }, { "epoch": 0.003066914498141264, "grad_norm": 5.841414713773794, "learning_rate": 9.133126934984521e-07, "loss": 0.5451, "step": 66 }, { "epoch": 0.0031133828996282526, "grad_norm": 6.403838810945664, "learning_rate": 9.287925696594428e-07, "loss": 0.4816, "step": 67 }, { "epoch": 0.0031598513011152417, "grad_norm": 6.345696881701124, "learning_rate": 9.442724458204335e-07, "loss": 0.4969, "step": 68 }, { "epoch": 0.0032063197026022303, "grad_norm": 3.5949055137522046, "learning_rate": 9.597523219814243e-07, "loss": 0.4296, "step": 69 }, { "epoch": 0.0032527881040892194, "grad_norm": 5.72764634440871, "learning_rate": 9.75232198142415e-07, "loss": 0.4528, "step": 70 }, { "epoch": 0.003299256505576208, "grad_norm": 4.84481405987313, "learning_rate": 9.907120743034058e-07, "loss": 0.4138, "step": 71 }, { "epoch": 0.003345724907063197, "grad_norm": 4.251499718497475, "learning_rate": 1.0061919504643964e-06, "loss": 0.4917, "step": 72 }, { "epoch": 0.0033921933085501858, "grad_norm": 5.174719164876974, "learning_rate": 1.021671826625387e-06, "loss": 0.3646, "step": 73 }, { "epoch": 0.003438661710037175, "grad_norm": 5.006390947788375, "learning_rate": 1.037151702786378e-06, "loss": 0.514, "step": 74 }, { "epoch": 0.0034851301115241635, "grad_norm": 4.980366082937228, "learning_rate": 1.0526315789473685e-06, "loss": 0.5144, "step": 75 }, { "epoch": 0.0035315985130111526, "grad_norm": 5.299354137994818, "learning_rate": 1.0681114551083592e-06, "loss": 0.5013, "step": 76 }, { "epoch": 0.0035780669144981412, "grad_norm": 5.969615473195041, "learning_rate": 1.08359133126935e-06, "loss": 0.4063, "step": 77 }, { "epoch": 0.00362453531598513, "grad_norm": 5.377057463002913, "learning_rate": 1.0990712074303406e-06, "loss": 0.4668, "step": 78 }, { "epoch": 0.003671003717472119, "grad_norm": 5.338364919920044, "learning_rate": 1.1145510835913313e-06, "loss": 0.4363, "step": 79 }, { "epoch": 0.0037174721189591076, "grad_norm": 4.784150909743669, "learning_rate": 1.1300309597523221e-06, "loss": 0.4636, "step": 80 }, { "epoch": 0.0037639405204460967, "grad_norm": 5.435279768223556, "learning_rate": 1.1455108359133127e-06, "loss": 0.4045, "step": 81 }, { "epoch": 0.0038104089219330854, "grad_norm": 5.336301439271362, "learning_rate": 1.1609907120743034e-06, "loss": 0.4324, "step": 82 }, { "epoch": 0.0038568773234200745, "grad_norm": 4.088534986074381, "learning_rate": 1.1764705882352942e-06, "loss": 0.4008, "step": 83 }, { "epoch": 0.003903345724907063, "grad_norm": 5.005445298773726, "learning_rate": 1.1919504643962849e-06, "loss": 0.4605, "step": 84 }, { "epoch": 0.003949814126394052, "grad_norm": 7.048509297934488, "learning_rate": 1.2074303405572757e-06, "loss": 0.3446, "step": 85 }, { "epoch": 0.003996282527881041, "grad_norm": 4.939976425625857, "learning_rate": 1.2229102167182663e-06, "loss": 0.3748, "step": 86 }, { "epoch": 0.00404275092936803, "grad_norm": 5.791074137667183, "learning_rate": 1.238390092879257e-06, "loss": 0.4094, "step": 87 }, { "epoch": 0.004089219330855019, "grad_norm": 6.389807612411811, "learning_rate": 1.2538699690402478e-06, "loss": 0.4561, "step": 88 }, { "epoch": 0.004135687732342007, "grad_norm": 5.578065959262816, "learning_rate": 1.2693498452012384e-06, "loss": 0.4924, "step": 89 }, { "epoch": 0.004182156133828996, "grad_norm": 3.9181817058544994, "learning_rate": 1.2848297213622293e-06, "loss": 0.3384, "step": 90 }, { "epoch": 0.004228624535315985, "grad_norm": 5.151525834767214, "learning_rate": 1.30030959752322e-06, "loss": 0.4153, "step": 91 }, { "epoch": 0.004275092936802974, "grad_norm": 7.474921411734634, "learning_rate": 1.3157894736842106e-06, "loss": 0.3272, "step": 92 }, { "epoch": 0.004321561338289963, "grad_norm": 4.250293825190362, "learning_rate": 1.3312693498452014e-06, "loss": 0.3115, "step": 93 }, { "epoch": 0.004368029739776951, "grad_norm": 3.8150045522772285, "learning_rate": 1.346749226006192e-06, "loss": 0.3488, "step": 94 }, { "epoch": 0.004414498141263941, "grad_norm": 4.337992408383516, "learning_rate": 1.3622291021671827e-06, "loss": 0.346, "step": 95 }, { "epoch": 0.0044609665427509295, "grad_norm": 6.047080438883901, "learning_rate": 1.3777089783281735e-06, "loss": 0.4013, "step": 96 }, { "epoch": 0.004507434944237918, "grad_norm": 5.9090798130256905, "learning_rate": 1.3931888544891641e-06, "loss": 0.3189, "step": 97 }, { "epoch": 0.004553903345724907, "grad_norm": 4.77216640368097, "learning_rate": 1.4086687306501548e-06, "loss": 0.3352, "step": 98 }, { "epoch": 0.004600371747211896, "grad_norm": 6.945563081983961, "learning_rate": 1.4241486068111456e-06, "loss": 0.3968, "step": 99 }, { "epoch": 0.004646840148698885, "grad_norm": 4.406316175028041, "learning_rate": 1.4396284829721363e-06, "loss": 0.3415, "step": 100 }, { "epoch": 0.004693308550185874, "grad_norm": 7.304331653908352, "learning_rate": 1.4551083591331269e-06, "loss": 0.3628, "step": 101 }, { "epoch": 0.004739776951672862, "grad_norm": 4.075262289798682, "learning_rate": 1.4705882352941177e-06, "loss": 0.3358, "step": 102 }, { "epoch": 0.004786245353159851, "grad_norm": 5.04406745839782, "learning_rate": 1.4860681114551084e-06, "loss": 0.3218, "step": 103 }, { "epoch": 0.0048327137546468404, "grad_norm": 5.311276494307027, "learning_rate": 1.5015479876160994e-06, "loss": 0.4167, "step": 104 }, { "epoch": 0.004879182156133829, "grad_norm": 6.588847027618625, "learning_rate": 1.5170278637770898e-06, "loss": 0.2912, "step": 105 }, { "epoch": 0.004925650557620818, "grad_norm": 3.779544976811759, "learning_rate": 1.5325077399380805e-06, "loss": 0.3432, "step": 106 }, { "epoch": 0.004972118959107806, "grad_norm": 6.37494952982739, "learning_rate": 1.5479876160990715e-06, "loss": 0.3849, "step": 107 }, { "epoch": 0.005018587360594796, "grad_norm": 3.5871481197724124, "learning_rate": 1.5634674922600622e-06, "loss": 0.2473, "step": 108 }, { "epoch": 0.0050650557620817846, "grad_norm": 4.014619254585085, "learning_rate": 1.5789473684210526e-06, "loss": 0.2884, "step": 109 }, { "epoch": 0.005111524163568773, "grad_norm": 4.928187056466436, "learning_rate": 1.5944272445820436e-06, "loss": 0.3036, "step": 110 }, { "epoch": 0.005157992565055762, "grad_norm": 4.933908991189177, "learning_rate": 1.6099071207430343e-06, "loss": 0.4057, "step": 111 }, { "epoch": 0.0052044609665427505, "grad_norm": 10.113534599838301, "learning_rate": 1.6253869969040247e-06, "loss": 0.4564, "step": 112 }, { "epoch": 0.00525092936802974, "grad_norm": 6.326296041840831, "learning_rate": 1.6408668730650157e-06, "loss": 0.3735, "step": 113 }, { "epoch": 0.005297397769516729, "grad_norm": 4.134760990612657, "learning_rate": 1.6563467492260064e-06, "loss": 0.3562, "step": 114 }, { "epoch": 0.005343866171003717, "grad_norm": 4.244666093657975, "learning_rate": 1.671826625386997e-06, "loss": 0.2838, "step": 115 }, { "epoch": 0.005390334572490706, "grad_norm": 3.3509996050828246, "learning_rate": 1.6873065015479879e-06, "loss": 0.2721, "step": 116 }, { "epoch": 0.0054368029739776955, "grad_norm": 7.06285194443431, "learning_rate": 1.7027863777089785e-06, "loss": 0.346, "step": 117 }, { "epoch": 0.005483271375464684, "grad_norm": 3.9543266412569618, "learning_rate": 1.7182662538699693e-06, "loss": 0.3393, "step": 118 }, { "epoch": 0.005529739776951673, "grad_norm": 3.21244849392471, "learning_rate": 1.73374613003096e-06, "loss": 0.297, "step": 119 }, { "epoch": 0.0055762081784386614, "grad_norm": 4.95393080710742, "learning_rate": 1.7492260061919506e-06, "loss": 0.3938, "step": 120 }, { "epoch": 0.005622676579925651, "grad_norm": 5.917105807224683, "learning_rate": 1.7647058823529414e-06, "loss": 0.3109, "step": 121 }, { "epoch": 0.00566914498141264, "grad_norm": 3.7434047656062375, "learning_rate": 1.780185758513932e-06, "loss": 0.291, "step": 122 }, { "epoch": 0.005715613382899628, "grad_norm": 6.087472300731493, "learning_rate": 1.7956656346749227e-06, "loss": 0.4821, "step": 123 }, { "epoch": 0.005762081784386617, "grad_norm": 4.58468629766215, "learning_rate": 1.8111455108359136e-06, "loss": 0.3771, "step": 124 }, { "epoch": 0.005808550185873606, "grad_norm": 6.358131644768691, "learning_rate": 1.8266253869969042e-06, "loss": 0.5175, "step": 125 }, { "epoch": 0.005855018587360595, "grad_norm": 4.498009834980018, "learning_rate": 1.8421052631578948e-06, "loss": 0.3298, "step": 126 }, { "epoch": 0.005901486988847584, "grad_norm": 3.385641289200555, "learning_rate": 1.8575851393188857e-06, "loss": 0.3139, "step": 127 }, { "epoch": 0.005947955390334572, "grad_norm": 3.7186469595051164, "learning_rate": 1.8730650154798763e-06, "loss": 0.363, "step": 128 }, { "epoch": 0.005994423791821561, "grad_norm": 4.910705600866876, "learning_rate": 1.888544891640867e-06, "loss": 0.3306, "step": 129 }, { "epoch": 0.0060408921933085506, "grad_norm": 6.081628700489746, "learning_rate": 1.9040247678018578e-06, "loss": 0.4115, "step": 130 }, { "epoch": 0.006087360594795539, "grad_norm": 4.326970746758124, "learning_rate": 1.9195046439628486e-06, "loss": 0.3133, "step": 131 }, { "epoch": 0.006133828996282528, "grad_norm": 6.108280907215603, "learning_rate": 1.9349845201238395e-06, "loss": 0.3201, "step": 132 }, { "epoch": 0.0061802973977695165, "grad_norm": 3.640985810831038, "learning_rate": 1.95046439628483e-06, "loss": 0.3299, "step": 133 }, { "epoch": 0.006226765799256505, "grad_norm": 4.976543708238896, "learning_rate": 1.9659442724458207e-06, "loss": 0.2771, "step": 134 }, { "epoch": 0.006273234200743495, "grad_norm": 9.80680818073562, "learning_rate": 1.9814241486068116e-06, "loss": 0.3637, "step": 135 }, { "epoch": 0.006319702602230483, "grad_norm": 4.330301431097096, "learning_rate": 1.996904024767802e-06, "loss": 0.3286, "step": 136 }, { "epoch": 0.006366171003717472, "grad_norm": 3.2347001622371416, "learning_rate": 2.012383900928793e-06, "loss": 0.3234, "step": 137 }, { "epoch": 0.006412639405204461, "grad_norm": 4.020385814217698, "learning_rate": 2.0278637770897837e-06, "loss": 0.2916, "step": 138 }, { "epoch": 0.00645910780669145, "grad_norm": 4.9508330047142595, "learning_rate": 2.043343653250774e-06, "loss": 0.3227, "step": 139 }, { "epoch": 0.006505576208178439, "grad_norm": 6.032033982014659, "learning_rate": 2.058823529411765e-06, "loss": 0.3205, "step": 140 }, { "epoch": 0.0065520446096654274, "grad_norm": 4.668423013484358, "learning_rate": 2.074303405572756e-06, "loss": 0.3745, "step": 141 }, { "epoch": 0.006598513011152416, "grad_norm": 5.34773343863283, "learning_rate": 2.0897832817337462e-06, "loss": 0.3118, "step": 142 }, { "epoch": 0.006644981412639406, "grad_norm": 5.653113958750201, "learning_rate": 2.105263157894737e-06, "loss": 0.3013, "step": 143 }, { "epoch": 0.006691449814126394, "grad_norm": 5.0772228224447105, "learning_rate": 2.120743034055728e-06, "loss": 0.3052, "step": 144 }, { "epoch": 0.006737918215613383, "grad_norm": 4.588700544859751, "learning_rate": 2.1362229102167183e-06, "loss": 0.2651, "step": 145 }, { "epoch": 0.0067843866171003716, "grad_norm": 3.6072106132130486, "learning_rate": 2.151702786377709e-06, "loss": 0.3151, "step": 146 }, { "epoch": 0.00683085501858736, "grad_norm": 4.014865761327939, "learning_rate": 2.1671826625387e-06, "loss": 0.239, "step": 147 }, { "epoch": 0.00687732342007435, "grad_norm": 3.5037255346130762, "learning_rate": 2.1826625386996904e-06, "loss": 0.3127, "step": 148 }, { "epoch": 0.006923791821561338, "grad_norm": 5.9701740013794975, "learning_rate": 2.1981424148606813e-06, "loss": 0.2636, "step": 149 }, { "epoch": 0.006970260223048327, "grad_norm": 4.517701414530341, "learning_rate": 2.213622291021672e-06, "loss": 0.2583, "step": 150 }, { "epoch": 0.007016728624535316, "grad_norm": 5.021329721981631, "learning_rate": 2.2291021671826625e-06, "loss": 0.3798, "step": 151 }, { "epoch": 0.007063197026022305, "grad_norm": 4.829758958100906, "learning_rate": 2.2445820433436534e-06, "loss": 0.2416, "step": 152 }, { "epoch": 0.007109665427509294, "grad_norm": 8.049408570514936, "learning_rate": 2.2600619195046442e-06, "loss": 0.2725, "step": 153 }, { "epoch": 0.0071561338289962825, "grad_norm": 4.791752146685399, "learning_rate": 2.2755417956656347e-06, "loss": 0.2735, "step": 154 }, { "epoch": 0.007202602230483271, "grad_norm": 4.380540468085267, "learning_rate": 2.2910216718266255e-06, "loss": 0.2844, "step": 155 }, { "epoch": 0.00724907063197026, "grad_norm": 3.9269110949110897, "learning_rate": 2.3065015479876163e-06, "loss": 0.3136, "step": 156 }, { "epoch": 0.007295539033457249, "grad_norm": 4.180958889103933, "learning_rate": 2.3219814241486068e-06, "loss": 0.2769, "step": 157 }, { "epoch": 0.007342007434944238, "grad_norm": 4.298809561806176, "learning_rate": 2.3374613003095976e-06, "loss": 0.3269, "step": 158 }, { "epoch": 0.007388475836431227, "grad_norm": 6.410040954569872, "learning_rate": 2.3529411764705885e-06, "loss": 0.3886, "step": 159 }, { "epoch": 0.007434944237918215, "grad_norm": 5.132474247182361, "learning_rate": 2.368421052631579e-06, "loss": 0.4471, "step": 160 }, { "epoch": 0.007481412639405205, "grad_norm": 3.501909536148889, "learning_rate": 2.3839009287925697e-06, "loss": 0.2635, "step": 161 }, { "epoch": 0.007527881040892193, "grad_norm": 3.1348039390388527, "learning_rate": 2.3993808049535606e-06, "loss": 0.4067, "step": 162 }, { "epoch": 0.007574349442379182, "grad_norm": 4.503126826937322, "learning_rate": 2.4148606811145514e-06, "loss": 0.296, "step": 163 }, { "epoch": 0.007620817843866171, "grad_norm": 4.754204586290884, "learning_rate": 2.430340557275542e-06, "loss": 0.2498, "step": 164 }, { "epoch": 0.00766728624535316, "grad_norm": 2.8384372412810865, "learning_rate": 2.4458204334365327e-06, "loss": 0.2806, "step": 165 }, { "epoch": 0.007713754646840149, "grad_norm": 6.8146168814734756, "learning_rate": 2.4613003095975235e-06, "loss": 0.2776, "step": 166 }, { "epoch": 0.0077602230483271376, "grad_norm": 6.147484546215495, "learning_rate": 2.476780185758514e-06, "loss": 0.3765, "step": 167 }, { "epoch": 0.007806691449814126, "grad_norm": 4.875071426163506, "learning_rate": 2.4922600619195048e-06, "loss": 0.2964, "step": 168 }, { "epoch": 0.007853159851301115, "grad_norm": 3.5460465466118536, "learning_rate": 2.5077399380804956e-06, "loss": 0.2366, "step": 169 }, { "epoch": 0.007899628252788104, "grad_norm": 3.6907687746552487, "learning_rate": 2.5232198142414865e-06, "loss": 0.1395, "step": 170 }, { "epoch": 0.007946096654275092, "grad_norm": 4.107726929469841, "learning_rate": 2.538699690402477e-06, "loss": 0.2553, "step": 171 }, { "epoch": 0.007992565055762083, "grad_norm": 3.805437578987958, "learning_rate": 2.5541795665634677e-06, "loss": 0.3197, "step": 172 }, { "epoch": 0.008039033457249071, "grad_norm": 4.313026180793064, "learning_rate": 2.5696594427244586e-06, "loss": 0.3098, "step": 173 }, { "epoch": 0.00808550185873606, "grad_norm": 5.7339518054876315, "learning_rate": 2.585139318885449e-06, "loss": 0.2481, "step": 174 }, { "epoch": 0.008131970260223048, "grad_norm": 4.902922893766478, "learning_rate": 2.60061919504644e-06, "loss": 0.286, "step": 175 }, { "epoch": 0.008178438661710037, "grad_norm": 4.762448751615562, "learning_rate": 2.6160990712074307e-06, "loss": 0.2929, "step": 176 }, { "epoch": 0.008224907063197026, "grad_norm": 3.99766245906578, "learning_rate": 2.631578947368421e-06, "loss": 0.297, "step": 177 }, { "epoch": 0.008271375464684014, "grad_norm": 3.461029963252058, "learning_rate": 2.647058823529412e-06, "loss": 0.2802, "step": 178 }, { "epoch": 0.008317843866171003, "grad_norm": 4.640225118981779, "learning_rate": 2.662538699690403e-06, "loss": 0.3061, "step": 179 }, { "epoch": 0.008364312267657992, "grad_norm": 3.5110710218700953, "learning_rate": 2.6780185758513932e-06, "loss": 0.2525, "step": 180 }, { "epoch": 0.008410780669144982, "grad_norm": 6.231431395036645, "learning_rate": 2.693498452012384e-06, "loss": 0.2753, "step": 181 }, { "epoch": 0.00845724907063197, "grad_norm": 4.7487376270976265, "learning_rate": 2.708978328173375e-06, "loss": 0.2013, "step": 182 }, { "epoch": 0.00850371747211896, "grad_norm": 5.9888934320070195, "learning_rate": 2.7244582043343653e-06, "loss": 0.2828, "step": 183 }, { "epoch": 0.008550185873605948, "grad_norm": 3.349658067912409, "learning_rate": 2.739938080495356e-06, "loss": 0.2361, "step": 184 }, { "epoch": 0.008596654275092937, "grad_norm": 6.016994062820489, "learning_rate": 2.755417956656347e-06, "loss": 0.2909, "step": 185 }, { "epoch": 0.008643122676579925, "grad_norm": 3.8610488166020414, "learning_rate": 2.7708978328173374e-06, "loss": 0.2831, "step": 186 }, { "epoch": 0.008689591078066914, "grad_norm": 5.351500844743597, "learning_rate": 2.7863777089783283e-06, "loss": 0.3268, "step": 187 }, { "epoch": 0.008736059479553903, "grad_norm": 4.962011046727305, "learning_rate": 2.801857585139319e-06, "loss": 0.2027, "step": 188 }, { "epoch": 0.008782527881040891, "grad_norm": 8.87357678620111, "learning_rate": 2.8173374613003096e-06, "loss": 0.3323, "step": 189 }, { "epoch": 0.008828996282527882, "grad_norm": 9.867260834143663, "learning_rate": 2.8328173374613004e-06, "loss": 0.4093, "step": 190 }, { "epoch": 0.00887546468401487, "grad_norm": 4.848598269466753, "learning_rate": 2.8482972136222912e-06, "loss": 0.3236, "step": 191 }, { "epoch": 0.008921933085501859, "grad_norm": 3.489125858456522, "learning_rate": 2.8637770897832817e-06, "loss": 0.2232, "step": 192 }, { "epoch": 0.008968401486988848, "grad_norm": 4.663827188882804, "learning_rate": 2.8792569659442725e-06, "loss": 0.2701, "step": 193 }, { "epoch": 0.009014869888475836, "grad_norm": 4.723479407362323, "learning_rate": 2.8947368421052634e-06, "loss": 0.2683, "step": 194 }, { "epoch": 0.009061338289962825, "grad_norm": 4.838183660849908, "learning_rate": 2.9102167182662538e-06, "loss": 0.3504, "step": 195 }, { "epoch": 0.009107806691449814, "grad_norm": 3.478495605443216, "learning_rate": 2.9256965944272446e-06, "loss": 0.3048, "step": 196 }, { "epoch": 0.009154275092936802, "grad_norm": 7.270941251909675, "learning_rate": 2.9411764705882355e-06, "loss": 0.3179, "step": 197 }, { "epoch": 0.009200743494423793, "grad_norm": 3.1936205703098977, "learning_rate": 2.9566563467492267e-06, "loss": 0.2152, "step": 198 }, { "epoch": 0.009247211895910781, "grad_norm": 4.414777600753178, "learning_rate": 2.9721362229102167e-06, "loss": 0.3147, "step": 199 }, { "epoch": 0.00929368029739777, "grad_norm": 6.7512128430874085, "learning_rate": 2.9876160990712076e-06, "loss": 0.3568, "step": 200 }, { "epoch": 0.009340148698884759, "grad_norm": 29.332219814350953, "learning_rate": 3.003095975232199e-06, "loss": 0.7697, "step": 201 }, { "epoch": 0.009386617100371747, "grad_norm": 4.454593500111261, "learning_rate": 3.018575851393189e-06, "loss": 0.477, "step": 202 }, { "epoch": 0.009433085501858736, "grad_norm": 3.91162092195588, "learning_rate": 3.0340557275541797e-06, "loss": 0.2615, "step": 203 }, { "epoch": 0.009479553903345725, "grad_norm": 4.867592817095189, "learning_rate": 3.049535603715171e-06, "loss": 0.3509, "step": 204 }, { "epoch": 0.009526022304832713, "grad_norm": 4.576822591865502, "learning_rate": 3.065015479876161e-06, "loss": 0.3277, "step": 205 }, { "epoch": 0.009572490706319702, "grad_norm": 4.770576393741442, "learning_rate": 3.080495356037152e-06, "loss": 0.197, "step": 206 }, { "epoch": 0.009618959107806692, "grad_norm": 5.204052887607754, "learning_rate": 3.095975232198143e-06, "loss": 0.376, "step": 207 }, { "epoch": 0.009665427509293681, "grad_norm": 4.213799299143785, "learning_rate": 3.111455108359133e-06, "loss": 0.2513, "step": 208 }, { "epoch": 0.00971189591078067, "grad_norm": 4.546150423326507, "learning_rate": 3.1269349845201243e-06, "loss": 0.2552, "step": 209 }, { "epoch": 0.009758364312267658, "grad_norm": 19.555734929985842, "learning_rate": 3.142414860681115e-06, "loss": 0.2828, "step": 210 }, { "epoch": 0.009804832713754647, "grad_norm": 5.408111143391318, "learning_rate": 3.157894736842105e-06, "loss": 0.3195, "step": 211 }, { "epoch": 0.009851301115241635, "grad_norm": 3.414806440700691, "learning_rate": 3.1733746130030964e-06, "loss": 0.2916, "step": 212 }, { "epoch": 0.009897769516728624, "grad_norm": 3.8891854290479806, "learning_rate": 3.1888544891640873e-06, "loss": 0.3008, "step": 213 }, { "epoch": 0.009944237918215613, "grad_norm": 3.943369362750631, "learning_rate": 3.2043343653250773e-06, "loss": 0.2343, "step": 214 }, { "epoch": 0.009990706319702601, "grad_norm": 4.350473403517226, "learning_rate": 3.2198142414860685e-06, "loss": 0.3489, "step": 215 }, { "epoch": 0.010037174721189592, "grad_norm": 4.108501807385164, "learning_rate": 3.2352941176470594e-06, "loss": 0.2424, "step": 216 }, { "epoch": 0.01008364312267658, "grad_norm": 3.1283408388367713, "learning_rate": 3.2507739938080494e-06, "loss": 0.2454, "step": 217 }, { "epoch": 0.010130111524163569, "grad_norm": 4.262212590958057, "learning_rate": 3.2662538699690407e-06, "loss": 0.2483, "step": 218 }, { "epoch": 0.010176579925650558, "grad_norm": 4.606365657309092, "learning_rate": 3.2817337461300315e-06, "loss": 0.2719, "step": 219 }, { "epoch": 0.010223048327137546, "grad_norm": 5.644472665705274, "learning_rate": 3.297213622291022e-06, "loss": 0.2639, "step": 220 }, { "epoch": 0.010269516728624535, "grad_norm": 5.392063183817529, "learning_rate": 3.3126934984520128e-06, "loss": 0.4247, "step": 221 }, { "epoch": 0.010315985130111524, "grad_norm": 6.403506470950756, "learning_rate": 3.3281733746130036e-06, "loss": 0.3307, "step": 222 }, { "epoch": 0.010362453531598512, "grad_norm": 6.060498251345855, "learning_rate": 3.343653250773994e-06, "loss": 0.1866, "step": 223 }, { "epoch": 0.010408921933085501, "grad_norm": 4.44675077391161, "learning_rate": 3.359133126934985e-06, "loss": 0.277, "step": 224 }, { "epoch": 0.010455390334572491, "grad_norm": 3.016944332856561, "learning_rate": 3.3746130030959757e-06, "loss": 0.3074, "step": 225 }, { "epoch": 0.01050185873605948, "grad_norm": 5.291578506325065, "learning_rate": 3.390092879256966e-06, "loss": 0.3047, "step": 226 }, { "epoch": 0.010548327137546469, "grad_norm": 4.07423710718693, "learning_rate": 3.405572755417957e-06, "loss": 0.2231, "step": 227 }, { "epoch": 0.010594795539033457, "grad_norm": 5.977876698667395, "learning_rate": 3.421052631578948e-06, "loss": 0.3855, "step": 228 }, { "epoch": 0.010641263940520446, "grad_norm": 4.748883930292941, "learning_rate": 3.4365325077399387e-06, "loss": 0.2634, "step": 229 }, { "epoch": 0.010687732342007435, "grad_norm": 3.9513173516935045, "learning_rate": 3.452012383900929e-06, "loss": 0.1888, "step": 230 }, { "epoch": 0.010734200743494423, "grad_norm": 2.874650931450424, "learning_rate": 3.46749226006192e-06, "loss": 0.1849, "step": 231 }, { "epoch": 0.010780669144981412, "grad_norm": 5.041962657863761, "learning_rate": 3.4829721362229108e-06, "loss": 0.3191, "step": 232 }, { "epoch": 0.010827137546468402, "grad_norm": 3.778700180262227, "learning_rate": 3.498452012383901e-06, "loss": 0.1799, "step": 233 }, { "epoch": 0.010873605947955391, "grad_norm": 3.321712356837421, "learning_rate": 3.513931888544892e-06, "loss": 0.2166, "step": 234 }, { "epoch": 0.01092007434944238, "grad_norm": 3.802167856862224, "learning_rate": 3.529411764705883e-06, "loss": 0.2627, "step": 235 }, { "epoch": 0.010966542750929368, "grad_norm": 4.808705424202439, "learning_rate": 3.5448916408668733e-06, "loss": 0.2381, "step": 236 }, { "epoch": 0.011013011152416357, "grad_norm": 3.568460134673285, "learning_rate": 3.560371517027864e-06, "loss": 0.2421, "step": 237 }, { "epoch": 0.011059479553903346, "grad_norm": 3.897928938990108, "learning_rate": 3.575851393188855e-06, "loss": 0.2001, "step": 238 }, { "epoch": 0.011105947955390334, "grad_norm": 2.9943794549469565, "learning_rate": 3.5913312693498454e-06, "loss": 0.2529, "step": 239 }, { "epoch": 0.011152416356877323, "grad_norm": 4.618331870327204, "learning_rate": 3.6068111455108363e-06, "loss": 0.2372, "step": 240 }, { "epoch": 0.011198884758364312, "grad_norm": 19.59378502002757, "learning_rate": 3.622291021671827e-06, "loss": 0.4572, "step": 241 }, { "epoch": 0.011245353159851302, "grad_norm": 5.720903900352172, "learning_rate": 3.6377708978328175e-06, "loss": 0.2052, "step": 242 }, { "epoch": 0.01129182156133829, "grad_norm": 5.2449937284251735, "learning_rate": 3.6532507739938084e-06, "loss": 0.2272, "step": 243 }, { "epoch": 0.01133828996282528, "grad_norm": 7.502722025497967, "learning_rate": 3.6687306501547992e-06, "loss": 0.2553, "step": 244 }, { "epoch": 0.011384758364312268, "grad_norm": 10.513455608546703, "learning_rate": 3.6842105263157896e-06, "loss": 0.2623, "step": 245 }, { "epoch": 0.011431226765799257, "grad_norm": 9.538913151978225, "learning_rate": 3.6996904024767805e-06, "loss": 0.5216, "step": 246 }, { "epoch": 0.011477695167286245, "grad_norm": 7.226671843064341, "learning_rate": 3.7151702786377713e-06, "loss": 0.3066, "step": 247 }, { "epoch": 0.011524163568773234, "grad_norm": 5.383889900516751, "learning_rate": 3.7306501547987618e-06, "loss": 0.4135, "step": 248 }, { "epoch": 0.011570631970260222, "grad_norm": 7.591144007631924, "learning_rate": 3.7461300309597526e-06, "loss": 0.3601, "step": 249 }, { "epoch": 0.011617100371747211, "grad_norm": 5.230520562276688, "learning_rate": 3.7616099071207434e-06, "loss": 0.2369, "step": 250 }, { "epoch": 0.011663568773234202, "grad_norm": 4.064158273472371, "learning_rate": 3.777089783281734e-06, "loss": 0.2317, "step": 251 }, { "epoch": 0.01171003717472119, "grad_norm": 7.4524153816587635, "learning_rate": 3.7925696594427247e-06, "loss": 0.2558, "step": 252 }, { "epoch": 0.011756505576208179, "grad_norm": 3.9675892496349956, "learning_rate": 3.8080495356037155e-06, "loss": 0.2686, "step": 253 }, { "epoch": 0.011802973977695167, "grad_norm": 4.411427572053885, "learning_rate": 3.8235294117647055e-06, "loss": 0.2444, "step": 254 }, { "epoch": 0.011849442379182156, "grad_norm": 3.4060376461528534, "learning_rate": 3.839009287925697e-06, "loss": 0.1875, "step": 255 }, { "epoch": 0.011895910780669145, "grad_norm": 5.222829524190345, "learning_rate": 3.854489164086688e-06, "loss": 0.2719, "step": 256 }, { "epoch": 0.011942379182156133, "grad_norm": 4.244693183626929, "learning_rate": 3.869969040247679e-06, "loss": 0.2701, "step": 257 }, { "epoch": 0.011988847583643122, "grad_norm": 6.648368252517687, "learning_rate": 3.885448916408669e-06, "loss": 0.3784, "step": 258 }, { "epoch": 0.01203531598513011, "grad_norm": 5.383648148956278, "learning_rate": 3.90092879256966e-06, "loss": 0.2801, "step": 259 }, { "epoch": 0.012081784386617101, "grad_norm": 3.8414587795090145, "learning_rate": 3.916408668730651e-06, "loss": 0.2133, "step": 260 }, { "epoch": 0.01212825278810409, "grad_norm": 4.710074547249099, "learning_rate": 3.9318885448916415e-06, "loss": 0.2559, "step": 261 }, { "epoch": 0.012174721189591078, "grad_norm": 4.971207346259624, "learning_rate": 3.947368421052632e-06, "loss": 0.3131, "step": 262 }, { "epoch": 0.012221189591078067, "grad_norm": 5.646374310239427, "learning_rate": 3.962848297213623e-06, "loss": 0.3259, "step": 263 }, { "epoch": 0.012267657992565056, "grad_norm": 4.8289505163183115, "learning_rate": 3.978328173374613e-06, "loss": 0.2375, "step": 264 }, { "epoch": 0.012314126394052044, "grad_norm": 5.312627751492137, "learning_rate": 3.993808049535604e-06, "loss": 0.2348, "step": 265 }, { "epoch": 0.012360594795539033, "grad_norm": 3.5401341465241285, "learning_rate": 4.009287925696595e-06, "loss": 0.2086, "step": 266 }, { "epoch": 0.012407063197026022, "grad_norm": 4.492369173303594, "learning_rate": 4.024767801857586e-06, "loss": 0.2613, "step": 267 }, { "epoch": 0.01245353159851301, "grad_norm": 4.677716399165601, "learning_rate": 4.0402476780185765e-06, "loss": 0.2997, "step": 268 }, { "epoch": 0.0125, "grad_norm": 4.1518392667758945, "learning_rate": 4.055727554179567e-06, "loss": 0.2737, "step": 269 }, { "epoch": 0.01254646840148699, "grad_norm": 5.3712605794139945, "learning_rate": 4.071207430340557e-06, "loss": 0.2899, "step": 270 }, { "epoch": 0.012592936802973978, "grad_norm": 3.071551774215343, "learning_rate": 4.086687306501548e-06, "loss": 0.3149, "step": 271 }, { "epoch": 0.012639405204460967, "grad_norm": 3.972909773696716, "learning_rate": 4.102167182662539e-06, "loss": 0.2298, "step": 272 }, { "epoch": 0.012685873605947955, "grad_norm": 6.225414564278348, "learning_rate": 4.11764705882353e-06, "loss": 0.2837, "step": 273 }, { "epoch": 0.012732342007434944, "grad_norm": 3.8804973658364093, "learning_rate": 4.133126934984521e-06, "loss": 0.1977, "step": 274 }, { "epoch": 0.012778810408921933, "grad_norm": 7.758694448354581, "learning_rate": 4.148606811145512e-06, "loss": 0.2514, "step": 275 }, { "epoch": 0.012825278810408921, "grad_norm": 4.06702217056824, "learning_rate": 4.164086687306502e-06, "loss": 0.3447, "step": 276 }, { "epoch": 0.012871747211895912, "grad_norm": 4.169509582214421, "learning_rate": 4.1795665634674924e-06, "loss": 0.2247, "step": 277 }, { "epoch": 0.0129182156133829, "grad_norm": 4.127499365264189, "learning_rate": 4.195046439628483e-06, "loss": 0.2381, "step": 278 }, { "epoch": 0.012964684014869889, "grad_norm": 3.939203911187921, "learning_rate": 4.210526315789474e-06, "loss": 0.1925, "step": 279 }, { "epoch": 0.013011152416356878, "grad_norm": 5.936644819658083, "learning_rate": 4.226006191950465e-06, "loss": 0.3653, "step": 280 }, { "epoch": 0.013057620817843866, "grad_norm": 3.308808803570643, "learning_rate": 4.241486068111456e-06, "loss": 0.1774, "step": 281 }, { "epoch": 0.013104089219330855, "grad_norm": 4.329301059546778, "learning_rate": 4.256965944272446e-06, "loss": 0.2679, "step": 282 }, { "epoch": 0.013150557620817844, "grad_norm": 4.235540976616256, "learning_rate": 4.272445820433437e-06, "loss": 0.1782, "step": 283 }, { "epoch": 0.013197026022304832, "grad_norm": 3.11264645215545, "learning_rate": 4.2879256965944275e-06, "loss": 0.1834, "step": 284 }, { "epoch": 0.01324349442379182, "grad_norm": 4.562133158481097, "learning_rate": 4.303405572755418e-06, "loss": 0.2823, "step": 285 }, { "epoch": 0.013289962825278811, "grad_norm": 3.703879860284813, "learning_rate": 4.318885448916409e-06, "loss": 0.2928, "step": 286 }, { "epoch": 0.0133364312267658, "grad_norm": 4.798044267687732, "learning_rate": 4.3343653250774e-06, "loss": 0.2186, "step": 287 }, { "epoch": 0.013382899628252789, "grad_norm": 3.532492038468166, "learning_rate": 4.349845201238391e-06, "loss": 0.1825, "step": 288 }, { "epoch": 0.013429368029739777, "grad_norm": 4.279384296527036, "learning_rate": 4.365325077399381e-06, "loss": 0.163, "step": 289 }, { "epoch": 0.013475836431226766, "grad_norm": 6.744290955423294, "learning_rate": 4.380804953560372e-06, "loss": 0.2568, "step": 290 }, { "epoch": 0.013522304832713754, "grad_norm": 4.97798816315067, "learning_rate": 4.3962848297213626e-06, "loss": 0.2163, "step": 291 }, { "epoch": 0.013568773234200743, "grad_norm": 9.723047159757622, "learning_rate": 4.411764705882353e-06, "loss": 0.1901, "step": 292 }, { "epoch": 0.013615241635687732, "grad_norm": 3.779274277603641, "learning_rate": 4.427244582043344e-06, "loss": 0.2136, "step": 293 }, { "epoch": 0.01366171003717472, "grad_norm": 3.5826505781736273, "learning_rate": 4.442724458204335e-06, "loss": 0.1838, "step": 294 }, { "epoch": 0.01370817843866171, "grad_norm": 2.69329898107229, "learning_rate": 4.458204334365325e-06, "loss": 0.2251, "step": 295 }, { "epoch": 0.0137546468401487, "grad_norm": 5.002252060451676, "learning_rate": 4.473684210526316e-06, "loss": 0.2933, "step": 296 }, { "epoch": 0.013801115241635688, "grad_norm": 3.766992127312884, "learning_rate": 4.489164086687307e-06, "loss": 0.2076, "step": 297 }, { "epoch": 0.013847583643122677, "grad_norm": 6.534862741438805, "learning_rate": 4.504643962848298e-06, "loss": 0.2741, "step": 298 }, { "epoch": 0.013894052044609665, "grad_norm": 5.8784996958166, "learning_rate": 4.5201238390092885e-06, "loss": 0.2979, "step": 299 }, { "epoch": 0.013940520446096654, "grad_norm": 3.6502038490093667, "learning_rate": 4.535603715170279e-06, "loss": 0.2609, "step": 300 }, { "epoch": 0.013986988847583643, "grad_norm": 2.6603488841785308, "learning_rate": 4.551083591331269e-06, "loss": 0.1889, "step": 301 }, { "epoch": 0.014033457249070631, "grad_norm": 8.516402272713751, "learning_rate": 4.56656346749226e-06, "loss": 0.2595, "step": 302 }, { "epoch": 0.01407992565055762, "grad_norm": 3.4453211377297075, "learning_rate": 4.582043343653251e-06, "loss": 0.2777, "step": 303 }, { "epoch": 0.01412639405204461, "grad_norm": 16.578860183047315, "learning_rate": 4.597523219814242e-06, "loss": 0.4174, "step": 304 }, { "epoch": 0.014172862453531599, "grad_norm": 4.916703429011074, "learning_rate": 4.613003095975233e-06, "loss": 0.2608, "step": 305 }, { "epoch": 0.014219330855018588, "grad_norm": 3.3266489182011565, "learning_rate": 4.6284829721362235e-06, "loss": 0.1685, "step": 306 }, { "epoch": 0.014265799256505576, "grad_norm": 2.2688822638578405, "learning_rate": 4.6439628482972135e-06, "loss": 0.1461, "step": 307 }, { "epoch": 0.014312267657992565, "grad_norm": 4.621654619352267, "learning_rate": 4.659442724458204e-06, "loss": 0.2144, "step": 308 }, { "epoch": 0.014358736059479554, "grad_norm": 4.3330262275988645, "learning_rate": 4.674922600619195e-06, "loss": 0.2288, "step": 309 }, { "epoch": 0.014405204460966542, "grad_norm": 2.8437364928949176, "learning_rate": 4.690402476780186e-06, "loss": 0.179, "step": 310 }, { "epoch": 0.014451672862453531, "grad_norm": 3.3964776177330145, "learning_rate": 4.705882352941177e-06, "loss": 0.1797, "step": 311 }, { "epoch": 0.01449814126394052, "grad_norm": 3.6995653764428114, "learning_rate": 4.721362229102168e-06, "loss": 0.1463, "step": 312 }, { "epoch": 0.01454460966542751, "grad_norm": 3.372665413777781, "learning_rate": 4.736842105263158e-06, "loss": 0.1618, "step": 313 }, { "epoch": 0.014591078066914499, "grad_norm": 5.260950999292075, "learning_rate": 4.752321981424149e-06, "loss": 0.342, "step": 314 }, { "epoch": 0.014637546468401487, "grad_norm": 5.237702010626586, "learning_rate": 4.7678018575851394e-06, "loss": 0.2511, "step": 315 }, { "epoch": 0.014684014869888476, "grad_norm": 4.991343349031539, "learning_rate": 4.78328173374613e-06, "loss": 0.2645, "step": 316 }, { "epoch": 0.014730483271375465, "grad_norm": 4.8979418405893345, "learning_rate": 4.798761609907121e-06, "loss": 0.2374, "step": 317 }, { "epoch": 0.014776951672862453, "grad_norm": 4.616369009067748, "learning_rate": 4.814241486068112e-06, "loss": 0.3525, "step": 318 }, { "epoch": 0.014823420074349442, "grad_norm": 2.857510792029597, "learning_rate": 4.829721362229103e-06, "loss": 0.1656, "step": 319 }, { "epoch": 0.01486988847583643, "grad_norm": 4.461114028444848, "learning_rate": 4.845201238390093e-06, "loss": 0.2626, "step": 320 }, { "epoch": 0.014916356877323421, "grad_norm": 5.060289199950021, "learning_rate": 4.860681114551084e-06, "loss": 0.2303, "step": 321 }, { "epoch": 0.01496282527881041, "grad_norm": 2.9926106793232177, "learning_rate": 4.876160990712075e-06, "loss": 0.2488, "step": 322 }, { "epoch": 0.015009293680297398, "grad_norm": 6.7466119425174025, "learning_rate": 4.891640866873065e-06, "loss": 0.2733, "step": 323 }, { "epoch": 0.015055762081784387, "grad_norm": 6.032569699493527, "learning_rate": 4.907120743034056e-06, "loss": 0.1852, "step": 324 }, { "epoch": 0.015102230483271376, "grad_norm": 3.765329056131137, "learning_rate": 4.922600619195047e-06, "loss": 0.2751, "step": 325 }, { "epoch": 0.015148698884758364, "grad_norm": 4.592509120831924, "learning_rate": 4.938080495356037e-06, "loss": 0.1789, "step": 326 }, { "epoch": 0.015195167286245353, "grad_norm": 3.1915726626224314, "learning_rate": 4.953560371517028e-06, "loss": 0.1521, "step": 327 }, { "epoch": 0.015241635687732341, "grad_norm": 3.3120482963286118, "learning_rate": 4.9690402476780196e-06, "loss": 0.2876, "step": 328 }, { "epoch": 0.01528810408921933, "grad_norm": 5.132408859662103, "learning_rate": 4.9845201238390096e-06, "loss": 0.2793, "step": 329 }, { "epoch": 0.01533457249070632, "grad_norm": 3.512681252326542, "learning_rate": 5e-06, "loss": 0.1951, "step": 330 }, { "epoch": 0.01538104089219331, "grad_norm": 8.590775459788532, "learning_rate": 5.015479876160991e-06, "loss": 0.2918, "step": 331 }, { "epoch": 0.015427509293680298, "grad_norm": 5.470521446011377, "learning_rate": 5.030959752321982e-06, "loss": 0.2506, "step": 332 }, { "epoch": 0.015473977695167286, "grad_norm": 4.73929604152104, "learning_rate": 5.046439628482973e-06, "loss": 0.2261, "step": 333 }, { "epoch": 0.015520446096654275, "grad_norm": 3.2455876536573807, "learning_rate": 5.061919504643963e-06, "loss": 0.2258, "step": 334 }, { "epoch": 0.015566914498141264, "grad_norm": 3.2159158203899274, "learning_rate": 5.077399380804954e-06, "loss": 0.3134, "step": 335 }, { "epoch": 0.015613382899628252, "grad_norm": 3.3346035951573345, "learning_rate": 5.092879256965945e-06, "loss": 0.1706, "step": 336 }, { "epoch": 0.015659851301115243, "grad_norm": 4.277676400314956, "learning_rate": 5.1083591331269355e-06, "loss": 0.353, "step": 337 }, { "epoch": 0.01570631970260223, "grad_norm": 3.390495578285796, "learning_rate": 5.123839009287926e-06, "loss": 0.2677, "step": 338 }, { "epoch": 0.01575278810408922, "grad_norm": 3.3372945237635463, "learning_rate": 5.139318885448917e-06, "loss": 0.2218, "step": 339 }, { "epoch": 0.015799256505576207, "grad_norm": 3.3372945237635463, "learning_rate": 5.139318885448917e-06, "loss": 0.279, "step": 340 }, { "epoch": 0.015845724907063197, "grad_norm": 2.792332552876037, "learning_rate": 5.154798761609907e-06, "loss": 0.1783, "step": 341 }, { "epoch": 0.015892193308550184, "grad_norm": 3.4781832061462574, "learning_rate": 5.170278637770898e-06, "loss": 0.2169, "step": 342 }, { "epoch": 0.015938661710037175, "grad_norm": 4.628937433223327, "learning_rate": 5.185758513931889e-06, "loss": 0.2108, "step": 343 }, { "epoch": 0.015985130111524165, "grad_norm": 3.600359592519204, "learning_rate": 5.20123839009288e-06, "loss": 0.2268, "step": 344 }, { "epoch": 0.016031598513011152, "grad_norm": 3.06129392679933, "learning_rate": 5.2167182662538705e-06, "loss": 0.188, "step": 345 }, { "epoch": 0.016078066914498142, "grad_norm": 3.567638182410701, "learning_rate": 5.232198142414861e-06, "loss": 0.2468, "step": 346 }, { "epoch": 0.01612453531598513, "grad_norm": 3.1126477526915712, "learning_rate": 5.247678018575851e-06, "loss": 0.1723, "step": 347 }, { "epoch": 0.01617100371747212, "grad_norm": 3.9395216538529674, "learning_rate": 5.263157894736842e-06, "loss": 0.224, "step": 348 }, { "epoch": 0.016217472118959107, "grad_norm": 2.6343756299583747, "learning_rate": 5.278637770897833e-06, "loss": 0.1853, "step": 349 }, { "epoch": 0.016263940520446097, "grad_norm": 3.9785287270035976, "learning_rate": 5.294117647058824e-06, "loss": 0.2462, "step": 350 }, { "epoch": 0.016310408921933084, "grad_norm": 3.8369572633198326, "learning_rate": 5.309597523219815e-06, "loss": 0.1786, "step": 351 }, { "epoch": 0.016356877323420074, "grad_norm": 2.698362044950442, "learning_rate": 5.325077399380806e-06, "loss": 0.1923, "step": 352 }, { "epoch": 0.016403345724907065, "grad_norm": 32.067302990748054, "learning_rate": 5.340557275541796e-06, "loss": 1.0023, "step": 353 }, { "epoch": 0.01644981412639405, "grad_norm": 4.636377789648639, "learning_rate": 5.3560371517027864e-06, "loss": 0.2347, "step": 354 }, { "epoch": 0.016496282527881042, "grad_norm": 3.5930041074994343, "learning_rate": 5.371517027863777e-06, "loss": 0.3012, "step": 355 }, { "epoch": 0.01654275092936803, "grad_norm": 3.7387842219145835, "learning_rate": 5.386996904024768e-06, "loss": 0.2625, "step": 356 }, { "epoch": 0.01658921933085502, "grad_norm": 6.417714103044981, "learning_rate": 5.402476780185759e-06, "loss": 0.261, "step": 357 }, { "epoch": 0.016635687732342006, "grad_norm": 3.7285489035785386, "learning_rate": 5.41795665634675e-06, "loss": 0.1974, "step": 358 }, { "epoch": 0.016682156133828997, "grad_norm": 3.811020148389298, "learning_rate": 5.43343653250774e-06, "loss": 0.2139, "step": 359 }, { "epoch": 0.016728624535315983, "grad_norm": 3.266170297677356, "learning_rate": 5.448916408668731e-06, "loss": 0.224, "step": 360 }, { "epoch": 0.016775092936802974, "grad_norm": 3.1715777877369633, "learning_rate": 5.4643962848297215e-06, "loss": 0.1942, "step": 361 }, { "epoch": 0.016821561338289964, "grad_norm": 5.0923321249876965, "learning_rate": 5.479876160990712e-06, "loss": 0.3608, "step": 362 }, { "epoch": 0.01686802973977695, "grad_norm": 6.640676542401407, "learning_rate": 5.495356037151703e-06, "loss": 0.26, "step": 363 }, { "epoch": 0.01691449814126394, "grad_norm": 5.627773349012163, "learning_rate": 5.510835913312694e-06, "loss": 0.3196, "step": 364 }, { "epoch": 0.01696096654275093, "grad_norm": 4.281902384770439, "learning_rate": 5.526315789473685e-06, "loss": 0.3176, "step": 365 }, { "epoch": 0.01700743494423792, "grad_norm": 4.140516631779172, "learning_rate": 5.541795665634675e-06, "loss": 0.2548, "step": 366 }, { "epoch": 0.017053903345724906, "grad_norm": 5.080270444340366, "learning_rate": 5.557275541795666e-06, "loss": 0.2699, "step": 367 }, { "epoch": 0.017100371747211896, "grad_norm": 2.871381552458416, "learning_rate": 5.5727554179566566e-06, "loss": 0.29, "step": 368 }, { "epoch": 0.017146840148698883, "grad_norm": 4.010429218581986, "learning_rate": 5.588235294117647e-06, "loss": 0.2935, "step": 369 }, { "epoch": 0.017193308550185873, "grad_norm": 6.954357203554264, "learning_rate": 5.603715170278638e-06, "loss": 0.2936, "step": 370 }, { "epoch": 0.017239776951672864, "grad_norm": 4.670614291343831, "learning_rate": 5.619195046439629e-06, "loss": 0.3538, "step": 371 }, { "epoch": 0.01728624535315985, "grad_norm": 3.850733701113996, "learning_rate": 5.634674922600619e-06, "loss": 0.1521, "step": 372 }, { "epoch": 0.01733271375464684, "grad_norm": 2.6763184945029423, "learning_rate": 5.65015479876161e-06, "loss": 0.23, "step": 373 }, { "epoch": 0.017379182156133828, "grad_norm": 4.228854464913015, "learning_rate": 5.665634674922601e-06, "loss": 0.2907, "step": 374 }, { "epoch": 0.01742565055762082, "grad_norm": 4.328184133439037, "learning_rate": 5.681114551083592e-06, "loss": 0.1578, "step": 375 }, { "epoch": 0.017472118959107805, "grad_norm": 3.6586690829603303, "learning_rate": 5.6965944272445825e-06, "loss": 0.2582, "step": 376 }, { "epoch": 0.017518587360594796, "grad_norm": 3.311198698862265, "learning_rate": 5.712074303405573e-06, "loss": 0.1711, "step": 377 }, { "epoch": 0.017565055762081783, "grad_norm": 2.5120596000081927, "learning_rate": 5.727554179566563e-06, "loss": 0.1749, "step": 378 }, { "epoch": 0.017611524163568773, "grad_norm": 2.8198684660300692, "learning_rate": 5.743034055727554e-06, "loss": 0.2497, "step": 379 }, { "epoch": 0.017657992565055763, "grad_norm": 5.13445061331938, "learning_rate": 5.758513931888545e-06, "loss": 0.3148, "step": 380 }, { "epoch": 0.01770446096654275, "grad_norm": 5.592590628263259, "learning_rate": 5.773993808049536e-06, "loss": 0.2234, "step": 381 }, { "epoch": 0.01775092936802974, "grad_norm": 2.272738104327487, "learning_rate": 5.789473684210527e-06, "loss": 0.1799, "step": 382 }, { "epoch": 0.017797397769516728, "grad_norm": 1.7151643941469592, "learning_rate": 5.8049535603715175e-06, "loss": 0.094, "step": 383 }, { "epoch": 0.017843866171003718, "grad_norm": 3.4924289104007213, "learning_rate": 5.8204334365325075e-06, "loss": 0.2441, "step": 384 }, { "epoch": 0.017890334572490705, "grad_norm": 14.113153762971395, "learning_rate": 5.835913312693498e-06, "loss": 0.2963, "step": 385 }, { "epoch": 0.017936802973977695, "grad_norm": 3.4349705445165095, "learning_rate": 5.851393188854489e-06, "loss": 0.1969, "step": 386 }, { "epoch": 0.017983271375464686, "grad_norm": 4.895621357316341, "learning_rate": 5.86687306501548e-06, "loss": 0.2645, "step": 387 }, { "epoch": 0.018029739776951673, "grad_norm": 3.706789512654578, "learning_rate": 5.882352941176471e-06, "loss": 0.2685, "step": 388 }, { "epoch": 0.018076208178438663, "grad_norm": 3.3513324927381745, "learning_rate": 5.897832817337462e-06, "loss": 0.1923, "step": 389 }, { "epoch": 0.01812267657992565, "grad_norm": 3.1974086388993053, "learning_rate": 5.9133126934984535e-06, "loss": 0.1735, "step": 390 }, { "epoch": 0.01816914498141264, "grad_norm": 2.8710100681189883, "learning_rate": 5.928792569659443e-06, "loss": 0.1725, "step": 391 }, { "epoch": 0.018215613382899627, "grad_norm": 4.383604222959407, "learning_rate": 5.9442724458204335e-06, "loss": 0.3054, "step": 392 }, { "epoch": 0.018262081784386618, "grad_norm": 3.9501025433990464, "learning_rate": 5.959752321981424e-06, "loss": 0.209, "step": 393 }, { "epoch": 0.018308550185873605, "grad_norm": 2.835677145590304, "learning_rate": 5.975232198142415e-06, "loss": 0.1638, "step": 394 }, { "epoch": 0.018355018587360595, "grad_norm": 3.228026775303914, "learning_rate": 5.990712074303406e-06, "loss": 0.1807, "step": 395 }, { "epoch": 0.018401486988847585, "grad_norm": 2.451755502307783, "learning_rate": 6.006191950464398e-06, "loss": 0.1243, "step": 396 }, { "epoch": 0.018447955390334572, "grad_norm": 5.024963132010327, "learning_rate": 6.021671826625387e-06, "loss": 0.1779, "step": 397 }, { "epoch": 0.018494423791821563, "grad_norm": 4.506459504895351, "learning_rate": 6.037151702786378e-06, "loss": 0.2045, "step": 398 }, { "epoch": 0.01854089219330855, "grad_norm": 2.69461285743671, "learning_rate": 6.0526315789473685e-06, "loss": 0.1604, "step": 399 }, { "epoch": 0.01858736059479554, "grad_norm": 6.006834197780471, "learning_rate": 6.068111455108359e-06, "loss": 0.3504, "step": 400 }, { "epoch": 0.018633828996282527, "grad_norm": 8.855619913088049, "learning_rate": 6.083591331269351e-06, "loss": 0.2426, "step": 401 }, { "epoch": 0.018680297397769517, "grad_norm": 4.516207146021804, "learning_rate": 6.099071207430342e-06, "loss": 0.3141, "step": 402 }, { "epoch": 0.018726765799256504, "grad_norm": 3.7302178231573593, "learning_rate": 6.114551083591331e-06, "loss": 0.2363, "step": 403 }, { "epoch": 0.018773234200743494, "grad_norm": 2.5490221210724697, "learning_rate": 6.130030959752322e-06, "loss": 0.0998, "step": 404 }, { "epoch": 0.018819702602230485, "grad_norm": 3.5480220851584248, "learning_rate": 6.145510835913313e-06, "loss": 0.2377, "step": 405 }, { "epoch": 0.018866171003717472, "grad_norm": 4.239008015219218, "learning_rate": 6.160990712074304e-06, "loss": 0.2915, "step": 406 }, { "epoch": 0.018912639405204462, "grad_norm": 6.26657517004981, "learning_rate": 6.176470588235295e-06, "loss": 0.1835, "step": 407 }, { "epoch": 0.01895910780669145, "grad_norm": 2.249996093018382, "learning_rate": 6.191950464396286e-06, "loss": 0.1355, "step": 408 }, { "epoch": 0.01900557620817844, "grad_norm": 2.4319053227772502, "learning_rate": 6.207430340557275e-06, "loss": 0.1585, "step": 409 }, { "epoch": 0.019052044609665426, "grad_norm": 4.125219431316329, "learning_rate": 6.222910216718266e-06, "loss": 0.2893, "step": 410 }, { "epoch": 0.019098513011152417, "grad_norm": 2.730564712194672, "learning_rate": 6.238390092879257e-06, "loss": 0.1433, "step": 411 }, { "epoch": 0.019144981412639404, "grad_norm": 2.4330179781455206, "learning_rate": 6.253869969040249e-06, "loss": 0.1897, "step": 412 }, { "epoch": 0.019191449814126394, "grad_norm": 4.526795496216333, "learning_rate": 6.2693498452012395e-06, "loss": 0.3139, "step": 413 }, { "epoch": 0.019237918215613384, "grad_norm": 3.166349681668493, "learning_rate": 6.28482972136223e-06, "loss": 0.2197, "step": 414 }, { "epoch": 0.01928438661710037, "grad_norm": 3.5850788830061333, "learning_rate": 6.3003095975232195e-06, "loss": 0.1705, "step": 415 }, { "epoch": 0.019330855018587362, "grad_norm": 4.5061823347027, "learning_rate": 6.31578947368421e-06, "loss": 0.2923, "step": 416 }, { "epoch": 0.01937732342007435, "grad_norm": 9.312107422838821, "learning_rate": 6.331269349845201e-06, "loss": 0.2536, "step": 417 }, { "epoch": 0.01942379182156134, "grad_norm": 3.7071561837048437, "learning_rate": 6.346749226006193e-06, "loss": 0.2111, "step": 418 }, { "epoch": 0.019470260223048326, "grad_norm": 3.359789847009978, "learning_rate": 6.362229102167184e-06, "loss": 0.2207, "step": 419 }, { "epoch": 0.019516728624535316, "grad_norm": 7.861842788988482, "learning_rate": 6.3777089783281746e-06, "loss": 0.2261, "step": 420 }, { "epoch": 0.019563197026022303, "grad_norm": 4.991386098906213, "learning_rate": 6.393188854489165e-06, "loss": 0.2549, "step": 421 }, { "epoch": 0.019609665427509294, "grad_norm": 3.643993653812466, "learning_rate": 6.4086687306501545e-06, "loss": 0.1356, "step": 422 }, { "epoch": 0.019656133828996284, "grad_norm": 3.778209098486435, "learning_rate": 6.424148606811146e-06, "loss": 0.1924, "step": 423 }, { "epoch": 0.01970260223048327, "grad_norm": 4.629358277182535, "learning_rate": 6.439628482972137e-06, "loss": 0.2244, "step": 424 }, { "epoch": 0.01974907063197026, "grad_norm": 4.228253572346518, "learning_rate": 6.455108359133128e-06, "loss": 0.1916, "step": 425 }, { "epoch": 0.01979553903345725, "grad_norm": 2.683937484868731, "learning_rate": 6.470588235294119e-06, "loss": 0.167, "step": 426 }, { "epoch": 0.01984200743494424, "grad_norm": 6.548799422487378, "learning_rate": 6.48606811145511e-06, "loss": 0.2471, "step": 427 }, { "epoch": 0.019888475836431226, "grad_norm": 3.919983772154325, "learning_rate": 6.501547987616099e-06, "loss": 0.1263, "step": 428 }, { "epoch": 0.019934944237918216, "grad_norm": 2.576361968330469, "learning_rate": 6.5170278637770905e-06, "loss": 0.1878, "step": 429 }, { "epoch": 0.019981412639405203, "grad_norm": 4.928229583355917, "learning_rate": 6.532507739938081e-06, "loss": 0.248, "step": 430 }, { "epoch": 0.020027881040892193, "grad_norm": 3.7559396419655644, "learning_rate": 6.547987616099072e-06, "loss": 0.1955, "step": 431 }, { "epoch": 0.020074349442379184, "grad_norm": 2.530711212452965, "learning_rate": 6.563467492260063e-06, "loss": 0.1328, "step": 432 }, { "epoch": 0.02012081784386617, "grad_norm": 3.2043389527057533, "learning_rate": 6.578947368421054e-06, "loss": 0.1474, "step": 433 }, { "epoch": 0.02016728624535316, "grad_norm": 3.855924795239373, "learning_rate": 6.594427244582044e-06, "loss": 0.1742, "step": 434 }, { "epoch": 0.020213754646840148, "grad_norm": 5.264808419670943, "learning_rate": 6.609907120743035e-06, "loss": 0.3034, "step": 435 }, { "epoch": 0.020260223048327138, "grad_norm": 6.622275643540305, "learning_rate": 6.6253869969040255e-06, "loss": 0.2542, "step": 436 }, { "epoch": 0.020306691449814125, "grad_norm": 4.897651151383281, "learning_rate": 6.640866873065016e-06, "loss": 0.2358, "step": 437 }, { "epoch": 0.020353159851301116, "grad_norm": 3.7353771629323274, "learning_rate": 6.656346749226007e-06, "loss": 0.1939, "step": 438 }, { "epoch": 0.020399628252788102, "grad_norm": 3.5422717648015905, "learning_rate": 6.671826625386998e-06, "loss": 0.2185, "step": 439 }, { "epoch": 0.020446096654275093, "grad_norm": 3.463412992464497, "learning_rate": 6.687306501547988e-06, "loss": 0.2677, "step": 440 }, { "epoch": 0.020492565055762083, "grad_norm": 4.143340929094671, "learning_rate": 6.702786377708979e-06, "loss": 0.1511, "step": 441 }, { "epoch": 0.02053903345724907, "grad_norm": 3.286170762490976, "learning_rate": 6.71826625386997e-06, "loss": 0.2457, "step": 442 }, { "epoch": 0.02058550185873606, "grad_norm": 2.6914915052255552, "learning_rate": 6.733746130030961e-06, "loss": 0.1971, "step": 443 }, { "epoch": 0.020631970260223047, "grad_norm": 4.193226139729231, "learning_rate": 6.7492260061919514e-06, "loss": 0.1446, "step": 444 }, { "epoch": 0.020678438661710038, "grad_norm": 4.473391581748665, "learning_rate": 6.764705882352942e-06, "loss": 0.1928, "step": 445 }, { "epoch": 0.020724907063197025, "grad_norm": 2.1736485399717282, "learning_rate": 6.780185758513932e-06, "loss": 0.1434, "step": 446 }, { "epoch": 0.020771375464684015, "grad_norm": 20.71006637108974, "learning_rate": 6.795665634674923e-06, "loss": 0.4507, "step": 447 }, { "epoch": 0.020817843866171002, "grad_norm": 2.505978216160386, "learning_rate": 6.811145510835914e-06, "loss": 0.2368, "step": 448 }, { "epoch": 0.020864312267657992, "grad_norm": 3.0901663399596093, "learning_rate": 6.826625386996905e-06, "loss": 0.1678, "step": 449 }, { "epoch": 0.020910780669144983, "grad_norm": 6.419283183085034, "learning_rate": 6.842105263157896e-06, "loss": 0.1919, "step": 450 }, { "epoch": 0.02095724907063197, "grad_norm": 3.34810441841821, "learning_rate": 6.8575851393188865e-06, "loss": 0.1944, "step": 451 }, { "epoch": 0.02100371747211896, "grad_norm": 2.9794377261058127, "learning_rate": 6.873065015479877e-06, "loss": 0.2322, "step": 452 }, { "epoch": 0.021050185873605947, "grad_norm": 6.364344344807768, "learning_rate": 6.888544891640867e-06, "loss": 0.2773, "step": 453 }, { "epoch": 0.021096654275092937, "grad_norm": 3.4892323951523774, "learning_rate": 6.904024767801858e-06, "loss": 0.2815, "step": 454 }, { "epoch": 0.021143122676579924, "grad_norm": 3.1742191820337875, "learning_rate": 6.919504643962849e-06, "loss": 0.1675, "step": 455 }, { "epoch": 0.021189591078066915, "grad_norm": 2.727877349871213, "learning_rate": 6.93498452012384e-06, "loss": 0.1729, "step": 456 }, { "epoch": 0.0212360594795539, "grad_norm": 2.4732949524640757, "learning_rate": 6.950464396284831e-06, "loss": 0.1515, "step": 457 }, { "epoch": 0.021282527881040892, "grad_norm": 3.3027170403591457, "learning_rate": 6.9659442724458216e-06, "loss": 0.1876, "step": 458 }, { "epoch": 0.021328996282527882, "grad_norm": 5.084319783824877, "learning_rate": 6.9814241486068116e-06, "loss": 0.2595, "step": 459 }, { "epoch": 0.02137546468401487, "grad_norm": 2.9548704833719386, "learning_rate": 6.996904024767802e-06, "loss": 0.2128, "step": 460 }, { "epoch": 0.02142193308550186, "grad_norm": 11.727658047096886, "learning_rate": 7.012383900928793e-06, "loss": 0.3204, "step": 461 }, { "epoch": 0.021468401486988847, "grad_norm": 3.1853856510918472, "learning_rate": 7.027863777089784e-06, "loss": 0.2198, "step": 462 }, { "epoch": 0.021514869888475837, "grad_norm": 2.8320510999305837, "learning_rate": 7.043343653250775e-06, "loss": 0.1658, "step": 463 }, { "epoch": 0.021561338289962824, "grad_norm": 5.178636905906762, "learning_rate": 7.058823529411766e-06, "loss": 0.1778, "step": 464 }, { "epoch": 0.021607806691449814, "grad_norm": 3.358125381016303, "learning_rate": 7.074303405572756e-06, "loss": 0.1613, "step": 465 }, { "epoch": 0.021654275092936805, "grad_norm": 3.9933484950279787, "learning_rate": 7.089783281733747e-06, "loss": 0.2296, "step": 466 }, { "epoch": 0.02170074349442379, "grad_norm": 3.008193450838977, "learning_rate": 7.1052631578947375e-06, "loss": 0.1729, "step": 467 }, { "epoch": 0.021747211895910782, "grad_norm": 4.546347168826001, "learning_rate": 7.120743034055728e-06, "loss": 0.1609, "step": 468 }, { "epoch": 0.02179368029739777, "grad_norm": 3.0056767089039944, "learning_rate": 7.136222910216719e-06, "loss": 0.1547, "step": 469 }, { "epoch": 0.02184014869888476, "grad_norm": 4.345482847448647, "learning_rate": 7.15170278637771e-06, "loss": 0.1896, "step": 470 }, { "epoch": 0.021886617100371746, "grad_norm": 3.0225749163755773, "learning_rate": 7.1671826625387e-06, "loss": 0.2709, "step": 471 }, { "epoch": 0.021933085501858737, "grad_norm": 3.791365946919784, "learning_rate": 7.182662538699691e-06, "loss": 0.1584, "step": 472 }, { "epoch": 0.021979553903345724, "grad_norm": 5.087414987415018, "learning_rate": 7.198142414860682e-06, "loss": 0.342, "step": 473 }, { "epoch": 0.022026022304832714, "grad_norm": 5.434417952259128, "learning_rate": 7.2136222910216725e-06, "loss": 0.2018, "step": 474 }, { "epoch": 0.022072490706319704, "grad_norm": 3.645921896532503, "learning_rate": 7.229102167182663e-06, "loss": 0.2617, "step": 475 }, { "epoch": 0.02211895910780669, "grad_norm": 4.425599611585742, "learning_rate": 7.244582043343654e-06, "loss": 0.2066, "step": 476 }, { "epoch": 0.02216542750929368, "grad_norm": 4.228285870261609, "learning_rate": 7.260061919504644e-06, "loss": 0.1594, "step": 477 }, { "epoch": 0.02221189591078067, "grad_norm": 4.800745572022538, "learning_rate": 7.275541795665635e-06, "loss": 0.3049, "step": 478 }, { "epoch": 0.02225836431226766, "grad_norm": 3.0198781894965103, "learning_rate": 7.291021671826626e-06, "loss": 0.1611, "step": 479 }, { "epoch": 0.022304832713754646, "grad_norm": 3.2232689105842676, "learning_rate": 7.306501547987617e-06, "loss": 0.1263, "step": 480 }, { "epoch": 0.022351301115241636, "grad_norm": 3.7024130791593226, "learning_rate": 7.321981424148608e-06, "loss": 0.24, "step": 481 }, { "epoch": 0.022397769516728623, "grad_norm": 5.261725661638616, "learning_rate": 7.3374613003095984e-06, "loss": 0.1739, "step": 482 }, { "epoch": 0.022444237918215613, "grad_norm": 4.336939408085075, "learning_rate": 7.352941176470589e-06, "loss": 0.2353, "step": 483 }, { "epoch": 0.022490706319702604, "grad_norm": 3.185097554550057, "learning_rate": 7.368421052631579e-06, "loss": 0.1928, "step": 484 }, { "epoch": 0.02253717472118959, "grad_norm": 4.345805301273989, "learning_rate": 7.38390092879257e-06, "loss": 0.2013, "step": 485 }, { "epoch": 0.02258364312267658, "grad_norm": 4.759090823787239, "learning_rate": 7.399380804953561e-06, "loss": 0.1623, "step": 486 }, { "epoch": 0.022630111524163568, "grad_norm": 2.170719364216724, "learning_rate": 7.414860681114552e-06, "loss": 0.1359, "step": 487 }, { "epoch": 0.02267657992565056, "grad_norm": 2.585008119063628, "learning_rate": 7.430340557275543e-06, "loss": 0.2526, "step": 488 }, { "epoch": 0.022723048327137545, "grad_norm": 8.687948505536552, "learning_rate": 7.4458204334365335e-06, "loss": 0.181, "step": 489 }, { "epoch": 0.022769516728624536, "grad_norm": 3.1822207796417743, "learning_rate": 7.4613003095975235e-06, "loss": 0.1608, "step": 490 }, { "epoch": 0.022815985130111523, "grad_norm": 3.3794348094496, "learning_rate": 7.476780185758514e-06, "loss": 0.1929, "step": 491 }, { "epoch": 0.022862453531598513, "grad_norm": 4.140668751077788, "learning_rate": 7.492260061919505e-06, "loss": 0.1366, "step": 492 }, { "epoch": 0.022908921933085503, "grad_norm": 2.7867972663185396, "learning_rate": 7.507739938080496e-06, "loss": 0.1458, "step": 493 }, { "epoch": 0.02295539033457249, "grad_norm": 3.767579988934783, "learning_rate": 7.523219814241487e-06, "loss": 0.2124, "step": 494 }, { "epoch": 0.02300185873605948, "grad_norm": 3.9192637399817936, "learning_rate": 7.538699690402478e-06, "loss": 0.1834, "step": 495 }, { "epoch": 0.023048327137546468, "grad_norm": 4.857597663474624, "learning_rate": 7.554179566563468e-06, "loss": 0.1958, "step": 496 }, { "epoch": 0.023094795539033458, "grad_norm": 4.650543083884602, "learning_rate": 7.5696594427244586e-06, "loss": 0.1702, "step": 497 }, { "epoch": 0.023141263940520445, "grad_norm": 4.004255720703912, "learning_rate": 7.585139318885449e-06, "loss": 0.2263, "step": 498 }, { "epoch": 0.023187732342007435, "grad_norm": 4.799114421499772, "learning_rate": 7.60061919504644e-06, "loss": 0.2377, "step": 499 }, { "epoch": 0.023234200743494422, "grad_norm": 5.079932287954157, "learning_rate": 7.616099071207431e-06, "loss": 0.2097, "step": 500 }, { "epoch": 0.023280669144981413, "grad_norm": 7.713565538796907, "learning_rate": 7.631578947368423e-06, "loss": 0.2038, "step": 501 }, { "epoch": 0.023327137546468403, "grad_norm": 4.086242160028787, "learning_rate": 7.647058823529411e-06, "loss": 0.1315, "step": 502 }, { "epoch": 0.02337360594795539, "grad_norm": 5.4917028559986925, "learning_rate": 7.662538699690403e-06, "loss": 0.2269, "step": 503 }, { "epoch": 0.02342007434944238, "grad_norm": 3.2478425477776662, "learning_rate": 7.678018575851394e-06, "loss": 0.2705, "step": 504 }, { "epoch": 0.023466542750929367, "grad_norm": 3.241239606376217, "learning_rate": 7.693498452012384e-06, "loss": 0.1339, "step": 505 }, { "epoch": 0.023513011152416358, "grad_norm": 3.523087315480207, "learning_rate": 7.708978328173376e-06, "loss": 0.2411, "step": 506 }, { "epoch": 0.023559479553903345, "grad_norm": 3.7211021128956685, "learning_rate": 7.724458204334366e-06, "loss": 0.2509, "step": 507 }, { "epoch": 0.023605947955390335, "grad_norm": 4.603881619124662, "learning_rate": 7.739938080495358e-06, "loss": 0.2187, "step": 508 }, { "epoch": 0.023652416356877322, "grad_norm": 3.1120266297928114, "learning_rate": 7.755417956656346e-06, "loss": 0.1358, "step": 509 }, { "epoch": 0.023698884758364312, "grad_norm": 3.130709728159283, "learning_rate": 7.770897832817338e-06, "loss": 0.1687, "step": 510 }, { "epoch": 0.023745353159851303, "grad_norm": 3.07835429710872, "learning_rate": 7.78637770897833e-06, "loss": 0.1359, "step": 511 }, { "epoch": 0.02379182156133829, "grad_norm": 4.053994161665693, "learning_rate": 7.80185758513932e-06, "loss": 0.2044, "step": 512 }, { "epoch": 0.02383828996282528, "grad_norm": 3.468698960080912, "learning_rate": 7.817337461300311e-06, "loss": 0.1799, "step": 513 }, { "epoch": 0.023884758364312267, "grad_norm": 2.2924778491151514, "learning_rate": 7.832817337461301e-06, "loss": 0.1392, "step": 514 }, { "epoch": 0.023931226765799257, "grad_norm": 3.150561577694531, "learning_rate": 7.848297213622291e-06, "loss": 0.246, "step": 515 }, { "epoch": 0.023977695167286244, "grad_norm": 3.7630952540882423, "learning_rate": 7.863777089783283e-06, "loss": 0.1527, "step": 516 }, { "epoch": 0.024024163568773235, "grad_norm": 4.03795129348454, "learning_rate": 7.879256965944273e-06, "loss": 0.1556, "step": 517 }, { "epoch": 0.02407063197026022, "grad_norm": 3.9609504007870187, "learning_rate": 7.894736842105265e-06, "loss": 0.1687, "step": 518 }, { "epoch": 0.024117100371747212, "grad_norm": 6.394906967019487, "learning_rate": 7.910216718266255e-06, "loss": 0.3477, "step": 519 }, { "epoch": 0.024163568773234202, "grad_norm": 3.624722998816764, "learning_rate": 7.925696594427246e-06, "loss": 0.18, "step": 520 }, { "epoch": 0.02421003717472119, "grad_norm": 2.7973063592502934, "learning_rate": 7.941176470588236e-06, "loss": 0.2043, "step": 521 }, { "epoch": 0.02425650557620818, "grad_norm": 6.341969151543639, "learning_rate": 7.956656346749226e-06, "loss": 0.1984, "step": 522 }, { "epoch": 0.024302973977695166, "grad_norm": 4.578073015750283, "learning_rate": 7.972136222910218e-06, "loss": 0.2337, "step": 523 }, { "epoch": 0.024349442379182157, "grad_norm": 2.2696772217800074, "learning_rate": 7.987616099071208e-06, "loss": 0.1681, "step": 524 }, { "epoch": 0.024395910780669144, "grad_norm": 3.2839074051123838, "learning_rate": 8.0030959752322e-06, "loss": 0.2677, "step": 525 }, { "epoch": 0.024442379182156134, "grad_norm": 2.510527321102306, "learning_rate": 8.01857585139319e-06, "loss": 0.1367, "step": 526 }, { "epoch": 0.02448884758364312, "grad_norm": 3.5386657545167473, "learning_rate": 8.03405572755418e-06, "loss": 0.2656, "step": 527 }, { "epoch": 0.02453531598513011, "grad_norm": 2.696119404697634, "learning_rate": 8.049535603715171e-06, "loss": 0.1521, "step": 528 }, { "epoch": 0.024581784386617102, "grad_norm": 2.031805398919461, "learning_rate": 8.065015479876161e-06, "loss": 0.0948, "step": 529 }, { "epoch": 0.02462825278810409, "grad_norm": 3.463195745535859, "learning_rate": 8.080495356037153e-06, "loss": 0.1984, "step": 530 }, { "epoch": 0.02467472118959108, "grad_norm": 2.5915914510162144, "learning_rate": 8.095975232198143e-06, "loss": 0.1347, "step": 531 }, { "epoch": 0.024721189591078066, "grad_norm": 3.785564159477594, "learning_rate": 8.111455108359135e-06, "loss": 0.1393, "step": 532 }, { "epoch": 0.024767657992565056, "grad_norm": 2.163051351384502, "learning_rate": 8.126934984520125e-06, "loss": 0.2274, "step": 533 }, { "epoch": 0.024814126394052043, "grad_norm": 2.7879425820663126, "learning_rate": 8.142414860681115e-06, "loss": 0.1751, "step": 534 }, { "epoch": 0.024860594795539034, "grad_norm": 2.9188174856145537, "learning_rate": 8.157894736842106e-06, "loss": 0.1334, "step": 535 }, { "epoch": 0.02490706319702602, "grad_norm": 3.5924873276327833, "learning_rate": 8.173374613003096e-06, "loss": 0.2616, "step": 536 }, { "epoch": 0.02495353159851301, "grad_norm": 4.151783097101983, "learning_rate": 8.188854489164088e-06, "loss": 0.2415, "step": 537 }, { "epoch": 0.025, "grad_norm": 2.790328313796625, "learning_rate": 8.204334365325078e-06, "loss": 0.1902, "step": 538 }, { "epoch": 0.02504646840148699, "grad_norm": 6.108656410354589, "learning_rate": 8.21981424148607e-06, "loss": 0.2368, "step": 539 }, { "epoch": 0.02509293680297398, "grad_norm": 2.877213881919399, "learning_rate": 8.23529411764706e-06, "loss": 0.2462, "step": 540 }, { "epoch": 0.025139405204460966, "grad_norm": 3.7374233175883425, "learning_rate": 8.25077399380805e-06, "loss": 0.1872, "step": 541 }, { "epoch": 0.025185873605947956, "grad_norm": 3.376352730492639, "learning_rate": 8.266253869969041e-06, "loss": 0.2191, "step": 542 }, { "epoch": 0.025232342007434943, "grad_norm": 2.9446212694157676, "learning_rate": 8.281733746130031e-06, "loss": 0.2378, "step": 543 }, { "epoch": 0.025278810408921933, "grad_norm": 7.710128467873351, "learning_rate": 8.297213622291023e-06, "loss": 0.2723, "step": 544 }, { "epoch": 0.02532527881040892, "grad_norm": 2.66320793562459, "learning_rate": 8.312693498452013e-06, "loss": 0.17, "step": 545 }, { "epoch": 0.02537174721189591, "grad_norm": 2.075044176354, "learning_rate": 8.328173374613003e-06, "loss": 0.1321, "step": 546 }, { "epoch": 0.0254182156133829, "grad_norm": 2.2508875790796252, "learning_rate": 8.343653250773995e-06, "loss": 0.1069, "step": 547 }, { "epoch": 0.025464684014869888, "grad_norm": 3.9056797197167508, "learning_rate": 8.359133126934985e-06, "loss": 0.2868, "step": 548 }, { "epoch": 0.02551115241635688, "grad_norm": 10.007181159722194, "learning_rate": 8.374613003095977e-06, "loss": 0.2145, "step": 549 }, { "epoch": 0.025557620817843865, "grad_norm": 2.696998529760011, "learning_rate": 8.390092879256967e-06, "loss": 0.1928, "step": 550 }, { "epoch": 0.025604089219330856, "grad_norm": 3.6084895973416278, "learning_rate": 8.405572755417958e-06, "loss": 0.1728, "step": 551 }, { "epoch": 0.025650557620817842, "grad_norm": 11.199828669633716, "learning_rate": 8.421052631578948e-06, "loss": 0.2116, "step": 552 }, { "epoch": 0.025697026022304833, "grad_norm": 3.1746308513581782, "learning_rate": 8.436532507739938e-06, "loss": 0.0899, "step": 553 }, { "epoch": 0.025743494423791823, "grad_norm": 4.648214465801042, "learning_rate": 8.45201238390093e-06, "loss": 0.2344, "step": 554 }, { "epoch": 0.02578996282527881, "grad_norm": 103.6658694120169, "learning_rate": 8.46749226006192e-06, "loss": 0.375, "step": 555 }, { "epoch": 0.0258364312267658, "grad_norm": 2.775031912407035, "learning_rate": 8.482972136222912e-06, "loss": 0.1361, "step": 556 }, { "epoch": 0.025882899628252787, "grad_norm": 109.5357576969477, "learning_rate": 8.498452012383902e-06, "loss": 0.2195, "step": 557 }, { "epoch": 0.025929368029739778, "grad_norm": 5.8452419391333805, "learning_rate": 8.513931888544892e-06, "loss": 0.2985, "step": 558 }, { "epoch": 0.025975836431226765, "grad_norm": 5.176085304544318, "learning_rate": 8.529411764705883e-06, "loss": 0.1811, "step": 559 }, { "epoch": 0.026022304832713755, "grad_norm": 4.231441128650319, "learning_rate": 8.544891640866873e-06, "loss": 0.1889, "step": 560 }, { "epoch": 0.026068773234200742, "grad_norm": 3.5586267898688053, "learning_rate": 8.560371517027865e-06, "loss": 0.2022, "step": 561 }, { "epoch": 0.026115241635687732, "grad_norm": 3.431694927756613, "learning_rate": 8.575851393188855e-06, "loss": 0.2327, "step": 562 }, { "epoch": 0.026161710037174723, "grad_norm": 3.569273503376163, "learning_rate": 8.591331269349847e-06, "loss": 0.3099, "step": 563 }, { "epoch": 0.02620817843866171, "grad_norm": 5.481761083774663, "learning_rate": 8.606811145510837e-06, "loss": 0.2313, "step": 564 }, { "epoch": 0.0262546468401487, "grad_norm": 5.9483262696522115, "learning_rate": 8.622291021671827e-06, "loss": 0.3915, "step": 565 }, { "epoch": 0.026301115241635687, "grad_norm": 2.9941741470005923, "learning_rate": 8.637770897832818e-06, "loss": 0.1429, "step": 566 }, { "epoch": 0.026347583643122677, "grad_norm": 4.6268692114831556, "learning_rate": 8.653250773993808e-06, "loss": 0.235, "step": 567 }, { "epoch": 0.026394052044609664, "grad_norm": 6.829964478850745, "learning_rate": 8.6687306501548e-06, "loss": 0.1846, "step": 568 }, { "epoch": 0.026440520446096655, "grad_norm": 3.963885820005153, "learning_rate": 8.68421052631579e-06, "loss": 0.3413, "step": 569 }, { "epoch": 0.02648698884758364, "grad_norm": 2.918584627905491, "learning_rate": 8.699690402476782e-06, "loss": 0.2432, "step": 570 }, { "epoch": 0.026533457249070632, "grad_norm": 6.153971835323851, "learning_rate": 8.715170278637772e-06, "loss": 0.3177, "step": 571 }, { "epoch": 0.026579925650557622, "grad_norm": 3.4908181941796794, "learning_rate": 8.730650154798762e-06, "loss": 0.2063, "step": 572 }, { "epoch": 0.02662639405204461, "grad_norm": 3.1100904462128205, "learning_rate": 8.746130030959753e-06, "loss": 0.1885, "step": 573 }, { "epoch": 0.0266728624535316, "grad_norm": 3.517232449107236, "learning_rate": 8.761609907120743e-06, "loss": 0.1999, "step": 574 }, { "epoch": 0.026719330855018587, "grad_norm": 3.6994762537859835, "learning_rate": 8.777089783281735e-06, "loss": 0.1881, "step": 575 }, { "epoch": 0.026765799256505577, "grad_norm": 4.3222483750689324, "learning_rate": 8.792569659442725e-06, "loss": 0.2195, "step": 576 }, { "epoch": 0.026812267657992564, "grad_norm": 5.505583830949338, "learning_rate": 8.808049535603715e-06, "loss": 0.1903, "step": 577 }, { "epoch": 0.026858736059479554, "grad_norm": 6.578926255775128, "learning_rate": 8.823529411764707e-06, "loss": 0.1708, "step": 578 }, { "epoch": 0.02690520446096654, "grad_norm": 4.835930505131338, "learning_rate": 8.839009287925697e-06, "loss": 0.1855, "step": 579 }, { "epoch": 0.02695167286245353, "grad_norm": 3.6710990496930167, "learning_rate": 8.854489164086688e-06, "loss": 0.1259, "step": 580 }, { "epoch": 0.026998141263940522, "grad_norm": 3.0314614789355834, "learning_rate": 8.869969040247678e-06, "loss": 0.2307, "step": 581 }, { "epoch": 0.02704460966542751, "grad_norm": 4.790863002820045, "learning_rate": 8.88544891640867e-06, "loss": 0.2658, "step": 582 }, { "epoch": 0.0270910780669145, "grad_norm": 9.305462312362186, "learning_rate": 8.90092879256966e-06, "loss": 0.2594, "step": 583 }, { "epoch": 0.027137546468401486, "grad_norm": 7.026189244368385, "learning_rate": 8.91640866873065e-06, "loss": 0.25, "step": 584 }, { "epoch": 0.027184014869888477, "grad_norm": 2.265440425155805, "learning_rate": 8.931888544891642e-06, "loss": 0.1051, "step": 585 }, { "epoch": 0.027230483271375464, "grad_norm": 3.7832297910086403, "learning_rate": 8.947368421052632e-06, "loss": 0.2558, "step": 586 }, { "epoch": 0.027276951672862454, "grad_norm": 3.08714239654486, "learning_rate": 8.962848297213624e-06, "loss": 0.1965, "step": 587 }, { "epoch": 0.02732342007434944, "grad_norm": 5.588689416636826, "learning_rate": 8.978328173374614e-06, "loss": 0.2195, "step": 588 }, { "epoch": 0.02736988847583643, "grad_norm": 6.704928007691182, "learning_rate": 8.993808049535604e-06, "loss": 0.2739, "step": 589 }, { "epoch": 0.02741635687732342, "grad_norm": 6.0701152049998255, "learning_rate": 9.009287925696595e-06, "loss": 0.2615, "step": 590 }, { "epoch": 0.02746282527881041, "grad_norm": 3.193745982213489, "learning_rate": 9.024767801857585e-06, "loss": 0.1581, "step": 591 }, { "epoch": 0.0275092936802974, "grad_norm": 2.5686120979231943, "learning_rate": 9.040247678018577e-06, "loss": 0.148, "step": 592 }, { "epoch": 0.027555762081784386, "grad_norm": 4.938305885509285, "learning_rate": 9.055727554179567e-06, "loss": 0.2718, "step": 593 }, { "epoch": 0.027602230483271376, "grad_norm": 5.913579215247392, "learning_rate": 9.071207430340559e-06, "loss": 0.1557, "step": 594 }, { "epoch": 0.027648698884758363, "grad_norm": 5.449530683028655, "learning_rate": 9.086687306501549e-06, "loss": 0.2186, "step": 595 }, { "epoch": 0.027695167286245354, "grad_norm": 2.6633407135329987, "learning_rate": 9.102167182662539e-06, "loss": 0.1713, "step": 596 }, { "epoch": 0.02774163568773234, "grad_norm": 5.166337251927116, "learning_rate": 9.11764705882353e-06, "loss": 0.1324, "step": 597 }, { "epoch": 0.02778810408921933, "grad_norm": 3.283077252560779, "learning_rate": 9.13312693498452e-06, "loss": 0.2155, "step": 598 }, { "epoch": 0.02783457249070632, "grad_norm": 3.9348819739554375, "learning_rate": 9.148606811145512e-06, "loss": 0.2268, "step": 599 }, { "epoch": 0.027881040892193308, "grad_norm": 3.5085652894887636, "learning_rate": 9.164086687306502e-06, "loss": 0.1383, "step": 600 }, { "epoch": 0.0279275092936803, "grad_norm": 7.808519369285745, "learning_rate": 9.179566563467494e-06, "loss": 0.2618, "step": 601 }, { "epoch": 0.027973977695167285, "grad_norm": 6.5371204655463515, "learning_rate": 9.195046439628484e-06, "loss": 0.1425, "step": 602 }, { "epoch": 0.028020446096654276, "grad_norm": 7.34110554212186, "learning_rate": 9.210526315789474e-06, "loss": 0.2668, "step": 603 }, { "epoch": 0.028066914498141263, "grad_norm": 6.705707430309093, "learning_rate": 9.226006191950465e-06, "loss": 0.1945, "step": 604 }, { "epoch": 0.028113382899628253, "grad_norm": 4.858727218339721, "learning_rate": 9.241486068111455e-06, "loss": 0.1797, "step": 605 }, { "epoch": 0.02815985130111524, "grad_norm": 3.7427721361819395, "learning_rate": 9.256965944272447e-06, "loss": 0.1304, "step": 606 }, { "epoch": 0.02820631970260223, "grad_norm": 3.0180249843771496, "learning_rate": 9.272445820433437e-06, "loss": 0.2258, "step": 607 }, { "epoch": 0.02825278810408922, "grad_norm": 15.26995358200986, "learning_rate": 9.287925696594427e-06, "loss": 0.1984, "step": 608 }, { "epoch": 0.028299256505576208, "grad_norm": 4.142765423563331, "learning_rate": 9.303405572755419e-06, "loss": 0.1837, "step": 609 }, { "epoch": 0.028345724907063198, "grad_norm": 3.8302210550531828, "learning_rate": 9.318885448916409e-06, "loss": 0.2252, "step": 610 }, { "epoch": 0.028392193308550185, "grad_norm": 6.592382467678612, "learning_rate": 9.3343653250774e-06, "loss": 0.2093, "step": 611 }, { "epoch": 0.028438661710037175, "grad_norm": 4.5109820996363075, "learning_rate": 9.34984520123839e-06, "loss": 0.1665, "step": 612 }, { "epoch": 0.028485130111524162, "grad_norm": 3.272618732671939, "learning_rate": 9.365325077399382e-06, "loss": 0.1263, "step": 613 }, { "epoch": 0.028531598513011153, "grad_norm": 8.066161359784118, "learning_rate": 9.380804953560372e-06, "loss": 0.1934, "step": 614 }, { "epoch": 0.02857806691449814, "grad_norm": 3.902606481879283, "learning_rate": 9.396284829721362e-06, "loss": 0.2322, "step": 615 }, { "epoch": 0.02862453531598513, "grad_norm": 3.990636941291722, "learning_rate": 9.411764705882354e-06, "loss": 0.1257, "step": 616 }, { "epoch": 0.02867100371747212, "grad_norm": 4.321218741344666, "learning_rate": 9.427244582043344e-06, "loss": 0.2636, "step": 617 }, { "epoch": 0.028717472118959107, "grad_norm": 3.4687555594833244, "learning_rate": 9.442724458204335e-06, "loss": 0.2143, "step": 618 }, { "epoch": 0.028763940520446098, "grad_norm": 6.7218524892771265, "learning_rate": 9.458204334365325e-06, "loss": 0.1763, "step": 619 }, { "epoch": 0.028810408921933085, "grad_norm": 5.114744052224355, "learning_rate": 9.473684210526315e-06, "loss": 0.2442, "step": 620 }, { "epoch": 0.028856877323420075, "grad_norm": 6.931673281254409, "learning_rate": 9.489164086687307e-06, "loss": 0.1995, "step": 621 }, { "epoch": 0.028903345724907062, "grad_norm": 5.291586347141888, "learning_rate": 9.504643962848297e-06, "loss": 0.202, "step": 622 }, { "epoch": 0.028949814126394052, "grad_norm": 2.147872048527835, "learning_rate": 9.520123839009289e-06, "loss": 0.1645, "step": 623 }, { "epoch": 0.02899628252788104, "grad_norm": 4.338588557217105, "learning_rate": 9.535603715170279e-06, "loss": 0.1756, "step": 624 }, { "epoch": 0.02904275092936803, "grad_norm": 3.270070744396621, "learning_rate": 9.55108359133127e-06, "loss": 0.1834, "step": 625 }, { "epoch": 0.02908921933085502, "grad_norm": 6.019257575203314, "learning_rate": 9.56656346749226e-06, "loss": 0.2913, "step": 626 }, { "epoch": 0.029135687732342007, "grad_norm": 2.485405580689049, "learning_rate": 9.58204334365325e-06, "loss": 0.1703, "step": 627 }, { "epoch": 0.029182156133828997, "grad_norm": 12.618013760699007, "learning_rate": 9.597523219814242e-06, "loss": 0.1832, "step": 628 }, { "epoch": 0.029228624535315984, "grad_norm": 3.446288934808853, "learning_rate": 9.613003095975232e-06, "loss": 0.2169, "step": 629 }, { "epoch": 0.029275092936802975, "grad_norm": 3.706221943606096, "learning_rate": 9.628482972136224e-06, "loss": 0.2138, "step": 630 }, { "epoch": 0.02932156133828996, "grad_norm": 9.9733624940107, "learning_rate": 9.643962848297214e-06, "loss": 0.199, "step": 631 }, { "epoch": 0.029368029739776952, "grad_norm": 3.676730157728827, "learning_rate": 9.659442724458206e-06, "loss": 0.2457, "step": 632 }, { "epoch": 0.029414498141263942, "grad_norm": 3.7257676103392283, "learning_rate": 9.674922600619196e-06, "loss": 0.1319, "step": 633 }, { "epoch": 0.02946096654275093, "grad_norm": 4.509209422972252, "learning_rate": 9.690402476780186e-06, "loss": 0.1802, "step": 634 }, { "epoch": 0.02950743494423792, "grad_norm": 2.6207574473034088, "learning_rate": 9.705882352941177e-06, "loss": 0.1274, "step": 635 }, { "epoch": 0.029553903345724906, "grad_norm": 1.5389025419978437, "learning_rate": 9.721362229102167e-06, "loss": 0.1479, "step": 636 }, { "epoch": 0.029600371747211897, "grad_norm": 3.104053467086817, "learning_rate": 9.736842105263159e-06, "loss": 0.2214, "step": 637 }, { "epoch": 0.029646840148698884, "grad_norm": 3.84340558904826, "learning_rate": 9.75232198142415e-06, "loss": 0.1941, "step": 638 }, { "epoch": 0.029693308550185874, "grad_norm": 6.206790222025582, "learning_rate": 9.767801857585139e-06, "loss": 0.2764, "step": 639 }, { "epoch": 0.02973977695167286, "grad_norm": 5.419535606234057, "learning_rate": 9.78328173374613e-06, "loss": 0.2038, "step": 640 }, { "epoch": 0.02978624535315985, "grad_norm": 7.535594210078206, "learning_rate": 9.79876160990712e-06, "loss": 0.261, "step": 641 }, { "epoch": 0.029832713754646842, "grad_norm": 3.5087902536049627, "learning_rate": 9.814241486068112e-06, "loss": 0.2056, "step": 642 }, { "epoch": 0.02987918215613383, "grad_norm": 4.465888674417129, "learning_rate": 9.829721362229104e-06, "loss": 0.2206, "step": 643 }, { "epoch": 0.02992565055762082, "grad_norm": 4.3225339541323935, "learning_rate": 9.845201238390094e-06, "loss": 0.1799, "step": 644 }, { "epoch": 0.029972118959107806, "grad_norm": 3.1293177802711547, "learning_rate": 9.860681114551084e-06, "loss": 0.1852, "step": 645 }, { "epoch": 0.030018587360594796, "grad_norm": 6.19556763512656, "learning_rate": 9.876160990712074e-06, "loss": 0.2954, "step": 646 }, { "epoch": 0.030065055762081783, "grad_norm": 6.29334477724563, "learning_rate": 9.891640866873066e-06, "loss": 0.1808, "step": 647 }, { "epoch": 0.030111524163568774, "grad_norm": 3.5972298160968648, "learning_rate": 9.907120743034056e-06, "loss": 0.1427, "step": 648 }, { "epoch": 0.03015799256505576, "grad_norm": 4.563591097141877, "learning_rate": 9.922600619195047e-06, "loss": 0.298, "step": 649 }, { "epoch": 0.03020446096654275, "grad_norm": 1.9691560809144382, "learning_rate": 9.938080495356039e-06, "loss": 0.1193, "step": 650 }, { "epoch": 0.03025092936802974, "grad_norm": 3.496238809783223, "learning_rate": 9.953560371517027e-06, "loss": 0.1549, "step": 651 }, { "epoch": 0.03029739776951673, "grad_norm": 6.536118510944501, "learning_rate": 9.969040247678019e-06, "loss": 0.3454, "step": 652 }, { "epoch": 0.03034386617100372, "grad_norm": 5.211165533443358, "learning_rate": 9.984520123839009e-06, "loss": 0.3398, "step": 653 }, { "epoch": 0.030390334572490706, "grad_norm": 27.282285373180695, "learning_rate": 1e-05, "loss": 0.1593, "step": 654 }, { "epoch": 0.030436802973977696, "grad_norm": 1.954275955166861, "learning_rate": 9.999999943372369e-06, "loss": 0.1305, "step": 655 }, { "epoch": 0.030483271375464683, "grad_norm": 8.10995379221968, "learning_rate": 9.999999773489477e-06, "loss": 0.1766, "step": 656 }, { "epoch": 0.030529739776951673, "grad_norm": 2.5936835361888635, "learning_rate": 9.999999490351327e-06, "loss": 0.1089, "step": 657 }, { "epoch": 0.03057620817843866, "grad_norm": 4.9697974702038525, "learning_rate": 9.999999093957927e-06, "loss": 0.208, "step": 658 }, { "epoch": 0.03062267657992565, "grad_norm": 2.3058791598937907, "learning_rate": 9.999998584309286e-06, "loss": 0.1172, "step": 659 }, { "epoch": 0.03066914498141264, "grad_norm": 4.1954297816206685, "learning_rate": 9.999997961405412e-06, "loss": 0.1665, "step": 660 }, { "epoch": 0.030715613382899628, "grad_norm": 6.549009677183697, "learning_rate": 9.999997225246323e-06, "loss": 0.2422, "step": 661 }, { "epoch": 0.03076208178438662, "grad_norm": 3.6010714342600836, "learning_rate": 9.999996375832035e-06, "loss": 0.2297, "step": 662 }, { "epoch": 0.030808550185873605, "grad_norm": 3.2765045720115715, "learning_rate": 9.999995413162567e-06, "loss": 0.1668, "step": 663 }, { "epoch": 0.030855018587360596, "grad_norm": 3.7390291315943607, "learning_rate": 9.99999433723794e-06, "loss": 0.2355, "step": 664 }, { "epoch": 0.030901486988847583, "grad_norm": 4.912722877836016, "learning_rate": 9.999993148058178e-06, "loss": 0.1665, "step": 665 }, { "epoch": 0.030947955390334573, "grad_norm": 13.770697723720847, "learning_rate": 9.99999184562331e-06, "loss": 0.2364, "step": 666 }, { "epoch": 0.03099442379182156, "grad_norm": 4.035121177841205, "learning_rate": 9.999990429933362e-06, "loss": 0.256, "step": 667 }, { "epoch": 0.03104089219330855, "grad_norm": 3.102746354159052, "learning_rate": 9.999988900988371e-06, "loss": 0.1944, "step": 668 }, { "epoch": 0.03108736059479554, "grad_norm": 2.978769859472241, "learning_rate": 9.999987258788368e-06, "loss": 0.1757, "step": 669 }, { "epoch": 0.031133828996282528, "grad_norm": 6.67076681327225, "learning_rate": 9.999985503333391e-06, "loss": 0.308, "step": 670 }, { "epoch": 0.031180297397769518, "grad_norm": 20.34555971072534, "learning_rate": 9.99998363462348e-06, "loss": 0.3268, "step": 671 }, { "epoch": 0.031226765799256505, "grad_norm": 19.15263941383422, "learning_rate": 9.999981652658676e-06, "loss": 0.2315, "step": 672 }, { "epoch": 0.031273234200743495, "grad_norm": 27.948734608444944, "learning_rate": 9.999979557439028e-06, "loss": 0.2657, "step": 673 }, { "epoch": 0.031319702602230486, "grad_norm": 5.410037310778006, "learning_rate": 9.99997734896458e-06, "loss": 0.2626, "step": 674 }, { "epoch": 0.03136617100371747, "grad_norm": 2.5570190918284608, "learning_rate": 9.999975027235381e-06, "loss": 0.1299, "step": 675 }, { "epoch": 0.03141263940520446, "grad_norm": 7.394367765292386, "learning_rate": 9.999972592251487e-06, "loss": 0.202, "step": 676 }, { "epoch": 0.03145910780669145, "grad_norm": 4.967261014668897, "learning_rate": 9.99997004401295e-06, "loss": 0.2007, "step": 677 }, { "epoch": 0.03150557620817844, "grad_norm": 5.6168900753910105, "learning_rate": 9.99996738251983e-06, "loss": 0.257, "step": 678 }, { "epoch": 0.03155204460966543, "grad_norm": 3.1293037623674014, "learning_rate": 9.999964607772185e-06, "loss": 0.168, "step": 679 }, { "epoch": 0.031598513011152414, "grad_norm": 6.884129096752474, "learning_rate": 9.999961719770081e-06, "loss": 0.2394, "step": 680 }, { "epoch": 0.031644981412639404, "grad_norm": 3.644645855392024, "learning_rate": 9.999958718513583e-06, "loss": 0.2649, "step": 681 }, { "epoch": 0.031691449814126395, "grad_norm": 3.5683638737419967, "learning_rate": 9.999955604002755e-06, "loss": 0.2194, "step": 682 }, { "epoch": 0.031737918215613385, "grad_norm": 8.94250906416544, "learning_rate": 9.999952376237672e-06, "loss": 0.1437, "step": 683 }, { "epoch": 0.03178438661710037, "grad_norm": 5.1029381431976395, "learning_rate": 9.999949035218403e-06, "loss": 0.3311, "step": 684 }, { "epoch": 0.03183085501858736, "grad_norm": 3.7542588090484537, "learning_rate": 9.999945580945028e-06, "loss": 0.1897, "step": 685 }, { "epoch": 0.03187732342007435, "grad_norm": 2.252902764869809, "learning_rate": 9.999942013417622e-06, "loss": 0.1441, "step": 686 }, { "epoch": 0.03192379182156134, "grad_norm": 4.615047819743459, "learning_rate": 9.999938332636267e-06, "loss": 0.3368, "step": 687 }, { "epoch": 0.03197026022304833, "grad_norm": 3.027701335723442, "learning_rate": 9.999934538601048e-06, "loss": 0.233, "step": 688 }, { "epoch": 0.032016728624535314, "grad_norm": 23.332500077389938, "learning_rate": 9.999930631312047e-06, "loss": 0.123, "step": 689 }, { "epoch": 0.032063197026022304, "grad_norm": 6.810973469220343, "learning_rate": 9.999926610769358e-06, "loss": 0.165, "step": 690 }, { "epoch": 0.032109665427509294, "grad_norm": 2.906695785935884, "learning_rate": 9.999922476973068e-06, "loss": 0.1655, "step": 691 }, { "epoch": 0.032156133828996285, "grad_norm": 4.544797093144928, "learning_rate": 9.999918229923268e-06, "loss": 0.2208, "step": 692 }, { "epoch": 0.03220260223048327, "grad_norm": 3.6427099299002483, "learning_rate": 9.999913869620061e-06, "loss": 0.1607, "step": 693 }, { "epoch": 0.03224907063197026, "grad_norm": 15.32152240663102, "learning_rate": 9.999909396063543e-06, "loss": 0.1679, "step": 694 }, { "epoch": 0.03229553903345725, "grad_norm": 3.246357898340742, "learning_rate": 9.999904809253814e-06, "loss": 0.1849, "step": 695 }, { "epoch": 0.03234200743494424, "grad_norm": 4.092000488385489, "learning_rate": 9.999900109190979e-06, "loss": 0.3057, "step": 696 }, { "epoch": 0.03238847583643123, "grad_norm": 2.4774469431439754, "learning_rate": 9.999895295875144e-06, "loss": 0.1588, "step": 697 }, { "epoch": 0.03243494423791821, "grad_norm": 5.177346253463151, "learning_rate": 9.999890369306417e-06, "loss": 0.1995, "step": 698 }, { "epoch": 0.032481412639405204, "grad_norm": 4.9342775485177635, "learning_rate": 9.999885329484913e-06, "loss": 0.2075, "step": 699 }, { "epoch": 0.032527881040892194, "grad_norm": 2.1364371587290014, "learning_rate": 9.999880176410743e-06, "loss": 0.1399, "step": 700 }, { "epoch": 0.032574349442379184, "grad_norm": 3.984327298906629, "learning_rate": 9.999874910084026e-06, "loss": 0.2387, "step": 701 }, { "epoch": 0.03262081784386617, "grad_norm": 2.9827932743682504, "learning_rate": 9.999869530504877e-06, "loss": 0.1021, "step": 702 }, { "epoch": 0.03266728624535316, "grad_norm": 8.145248013870372, "learning_rate": 9.999864037673424e-06, "loss": 0.2706, "step": 703 }, { "epoch": 0.03271375464684015, "grad_norm": 2.6539104151571, "learning_rate": 9.999858431589786e-06, "loss": 0.3055, "step": 704 }, { "epoch": 0.03276022304832714, "grad_norm": 2.680129143178953, "learning_rate": 9.999852712254094e-06, "loss": 0.17, "step": 705 }, { "epoch": 0.03280669144981413, "grad_norm": 3.09976427993092, "learning_rate": 9.999846879666475e-06, "loss": 0.2073, "step": 706 }, { "epoch": 0.03285315985130111, "grad_norm": 4.96665995681907, "learning_rate": 9.999840933827061e-06, "loss": 0.2534, "step": 707 }, { "epoch": 0.0328996282527881, "grad_norm": 3.6601686606575363, "learning_rate": 9.99983487473599e-06, "loss": 0.143, "step": 708 }, { "epoch": 0.032946096654275094, "grad_norm": 4.192715340533569, "learning_rate": 9.999828702393395e-06, "loss": 0.1663, "step": 709 }, { "epoch": 0.032992565055762084, "grad_norm": 3.193879808022246, "learning_rate": 9.999822416799418e-06, "loss": 0.1861, "step": 710 }, { "epoch": 0.03303903345724907, "grad_norm": 2.805681238285264, "learning_rate": 9.9998160179542e-06, "loss": 0.2214, "step": 711 }, { "epoch": 0.03308550185873606, "grad_norm": 2.7748608237931114, "learning_rate": 9.99980950585789e-06, "loss": 0.1244, "step": 712 }, { "epoch": 0.03313197026022305, "grad_norm": 1.5978263562226203, "learning_rate": 9.999802880510631e-06, "loss": 0.1394, "step": 713 }, { "epoch": 0.03317843866171004, "grad_norm": 6.264033332642128, "learning_rate": 9.999796141912574e-06, "loss": 0.1692, "step": 714 }, { "epoch": 0.03322490706319703, "grad_norm": 2.463132237125856, "learning_rate": 9.999789290063873e-06, "loss": 0.1951, "step": 715 }, { "epoch": 0.03327137546468401, "grad_norm": 2.0670352445104943, "learning_rate": 9.999782324964681e-06, "loss": 0.1832, "step": 716 }, { "epoch": 0.033317843866171, "grad_norm": 2.3507595001141928, "learning_rate": 9.999775246615157e-06, "loss": 0.1302, "step": 717 }, { "epoch": 0.03336431226765799, "grad_norm": 4.287596576080758, "learning_rate": 9.999768055015464e-06, "loss": 0.1843, "step": 718 }, { "epoch": 0.033410780669144984, "grad_norm": 1.8727053069816575, "learning_rate": 9.99976075016576e-06, "loss": 0.123, "step": 719 }, { "epoch": 0.03345724907063197, "grad_norm": 3.548995306418956, "learning_rate": 9.999753332066213e-06, "loss": 0.2366, "step": 720 }, { "epoch": 0.03350371747211896, "grad_norm": 1.7022631603262441, "learning_rate": 9.999745800716992e-06, "loss": 0.1027, "step": 721 }, { "epoch": 0.03355018587360595, "grad_norm": 3.8854789554027063, "learning_rate": 9.999738156118266e-06, "loss": 0.2349, "step": 722 }, { "epoch": 0.03359665427509294, "grad_norm": 3.905581593913924, "learning_rate": 9.999730398270208e-06, "loss": 0.2622, "step": 723 }, { "epoch": 0.03364312267657993, "grad_norm": 2.9334251912692895, "learning_rate": 9.999722527172997e-06, "loss": 0.2193, "step": 724 }, { "epoch": 0.03368959107806691, "grad_norm": 2.2983856559920115, "learning_rate": 9.999714542826806e-06, "loss": 0.1741, "step": 725 }, { "epoch": 0.0337360594795539, "grad_norm": 1.9021208765639235, "learning_rate": 9.99970644523182e-06, "loss": 0.1772, "step": 726 }, { "epoch": 0.03378252788104089, "grad_norm": 5.544105923811797, "learning_rate": 9.99969823438822e-06, "loss": 0.1673, "step": 727 }, { "epoch": 0.03382899628252788, "grad_norm": 4.0940167035197605, "learning_rate": 9.999689910296193e-06, "loss": 0.2036, "step": 728 }, { "epoch": 0.03387546468401487, "grad_norm": 2.4598218772850844, "learning_rate": 9.999681472955927e-06, "loss": 0.2252, "step": 729 }, { "epoch": 0.03392193308550186, "grad_norm": 2.4746550277545754, "learning_rate": 9.999672922367614e-06, "loss": 0.2266, "step": 730 }, { "epoch": 0.03396840148698885, "grad_norm": 3.193520989267536, "learning_rate": 9.999664258531447e-06, "loss": 0.2431, "step": 731 }, { "epoch": 0.03401486988847584, "grad_norm": 4.5397505308208075, "learning_rate": 9.999655481447624e-06, "loss": 0.3903, "step": 732 }, { "epoch": 0.03406133828996283, "grad_norm": 2.7805921938370965, "learning_rate": 9.999646591116341e-06, "loss": 0.1568, "step": 733 }, { "epoch": 0.03410780669144981, "grad_norm": 2.1881111714651182, "learning_rate": 9.9996375875378e-06, "loss": 0.1382, "step": 734 }, { "epoch": 0.0341542750929368, "grad_norm": 3.5650276232648435, "learning_rate": 9.999628470712206e-06, "loss": 0.1475, "step": 735 }, { "epoch": 0.03420074349442379, "grad_norm": 3.3850950261057076, "learning_rate": 9.999619240639767e-06, "loss": 0.1465, "step": 736 }, { "epoch": 0.03424721189591078, "grad_norm": 2.7435007808552125, "learning_rate": 9.999609897320688e-06, "loss": 0.1488, "step": 737 }, { "epoch": 0.034293680297397766, "grad_norm": 3.0124755583767633, "learning_rate": 9.999600440755183e-06, "loss": 0.1515, "step": 738 }, { "epoch": 0.034340148698884757, "grad_norm": 2.45176869460976, "learning_rate": 9.999590870943467e-06, "loss": 0.2528, "step": 739 }, { "epoch": 0.03438661710037175, "grad_norm": 2.9324799620293995, "learning_rate": 9.999581187885755e-06, "loss": 0.1432, "step": 740 }, { "epoch": 0.03443308550185874, "grad_norm": 1.9895610840704663, "learning_rate": 9.999571391582268e-06, "loss": 0.0892, "step": 741 }, { "epoch": 0.03447955390334573, "grad_norm": 2.581874763545619, "learning_rate": 9.999561482033226e-06, "loss": 0.159, "step": 742 }, { "epoch": 0.03452602230483271, "grad_norm": 2.347365790839522, "learning_rate": 9.999551459238856e-06, "loss": 0.2137, "step": 743 }, { "epoch": 0.0345724907063197, "grad_norm": 5.336520075084788, "learning_rate": 9.999541323199382e-06, "loss": 0.2151, "step": 744 }, { "epoch": 0.03461895910780669, "grad_norm": 3.188967306107003, "learning_rate": 9.999531073915035e-06, "loss": 0.1417, "step": 745 }, { "epoch": 0.03466542750929368, "grad_norm": 2.976106584730563, "learning_rate": 9.999520711386048e-06, "loss": 0.2285, "step": 746 }, { "epoch": 0.034711895910780666, "grad_norm": 1.94039887134798, "learning_rate": 9.999510235612655e-06, "loss": 0.1281, "step": 747 }, { "epoch": 0.034758364312267656, "grad_norm": 2.41299700230099, "learning_rate": 9.999499646595093e-06, "loss": 0.1028, "step": 748 }, { "epoch": 0.034804832713754646, "grad_norm": 2.610385661327798, "learning_rate": 9.999488944333603e-06, "loss": 0.1145, "step": 749 }, { "epoch": 0.03485130111524164, "grad_norm": 3.033642359670474, "learning_rate": 9.999478128828425e-06, "loss": 0.1435, "step": 750 }, { "epoch": 0.03489776951672863, "grad_norm": 3.425736367216075, "learning_rate": 9.999467200079807e-06, "loss": 0.1936, "step": 751 }, { "epoch": 0.03494423791821561, "grad_norm": 2.6070509076686843, "learning_rate": 9.999456158087994e-06, "loss": 0.1126, "step": 752 }, { "epoch": 0.0349907063197026, "grad_norm": 3.966800205330091, "learning_rate": 9.999445002853238e-06, "loss": 0.2236, "step": 753 }, { "epoch": 0.03503717472118959, "grad_norm": 2.659427127831662, "learning_rate": 9.999433734375791e-06, "loss": 0.1032, "step": 754 }, { "epoch": 0.03508364312267658, "grad_norm": 2.7757039472420346, "learning_rate": 9.999422352655908e-06, "loss": 0.1928, "step": 755 }, { "epoch": 0.035130111524163565, "grad_norm": 2.657628952602905, "learning_rate": 9.999410857693845e-06, "loss": 0.1368, "step": 756 }, { "epoch": 0.035176579925650556, "grad_norm": 3.0495325852660233, "learning_rate": 9.999399249489867e-06, "loss": 0.2511, "step": 757 }, { "epoch": 0.035223048327137546, "grad_norm": 2.8001835426056294, "learning_rate": 9.999387528044233e-06, "loss": 0.1942, "step": 758 }, { "epoch": 0.035269516728624536, "grad_norm": 2.067845379687603, "learning_rate": 9.99937569335721e-06, "loss": 0.0808, "step": 759 }, { "epoch": 0.03531598513011153, "grad_norm": 4.1142944218741855, "learning_rate": 9.999363745429065e-06, "loss": 0.1976, "step": 760 }, { "epoch": 0.03536245353159851, "grad_norm": 4.307109728779299, "learning_rate": 9.99935168426007e-06, "loss": 0.176, "step": 761 }, { "epoch": 0.0354089219330855, "grad_norm": 3.9585605559350947, "learning_rate": 9.999339509850498e-06, "loss": 0.2532, "step": 762 }, { "epoch": 0.03545539033457249, "grad_norm": 3.1782657516585857, "learning_rate": 9.999327222200623e-06, "loss": 0.1047, "step": 763 }, { "epoch": 0.03550185873605948, "grad_norm": 6.374006957964657, "learning_rate": 9.999314821310726e-06, "loss": 0.1485, "step": 764 }, { "epoch": 0.03554832713754647, "grad_norm": 2.1655220305846483, "learning_rate": 9.999302307181087e-06, "loss": 0.0976, "step": 765 }, { "epoch": 0.035594795539033455, "grad_norm": 5.220737579792372, "learning_rate": 9.999289679811988e-06, "loss": 0.2788, "step": 766 }, { "epoch": 0.035641263940520446, "grad_norm": 1.969838515102553, "learning_rate": 9.999276939203717e-06, "loss": 0.1178, "step": 767 }, { "epoch": 0.035687732342007436, "grad_norm": 2.082387771873488, "learning_rate": 9.999264085356561e-06, "loss": 0.1454, "step": 768 }, { "epoch": 0.035734200743494426, "grad_norm": 1.9095743449660807, "learning_rate": 9.999251118270813e-06, "loss": 0.1661, "step": 769 }, { "epoch": 0.03578066914498141, "grad_norm": 5.006265375290064, "learning_rate": 9.999238037946764e-06, "loss": 0.2286, "step": 770 }, { "epoch": 0.0358271375464684, "grad_norm": 2.8769641595244995, "learning_rate": 9.999224844384712e-06, "loss": 0.2016, "step": 771 }, { "epoch": 0.03587360594795539, "grad_norm": 1.9793074858688213, "learning_rate": 9.999211537584956e-06, "loss": 0.1601, "step": 772 }, { "epoch": 0.03592007434944238, "grad_norm": 2.5849779486589406, "learning_rate": 9.999198117547797e-06, "loss": 0.1139, "step": 773 }, { "epoch": 0.03596654275092937, "grad_norm": 4.050390861924843, "learning_rate": 9.99918458427354e-06, "loss": 0.2205, "step": 774 }, { "epoch": 0.036013011152416355, "grad_norm": 3.046168672574063, "learning_rate": 9.99917093776249e-06, "loss": 0.2399, "step": 775 }, { "epoch": 0.036059479553903345, "grad_norm": 5.724179683123859, "learning_rate": 9.999157178014956e-06, "loss": 0.2509, "step": 776 }, { "epoch": 0.036105947955390336, "grad_norm": 2.7435640625209423, "learning_rate": 9.999143305031252e-06, "loss": 0.1279, "step": 777 }, { "epoch": 0.036152416356877326, "grad_norm": 5.955768143669971, "learning_rate": 9.999129318811688e-06, "loss": 0.3067, "step": 778 }, { "epoch": 0.03619888475836431, "grad_norm": 3.5824629854305754, "learning_rate": 9.999115219356585e-06, "loss": 0.1804, "step": 779 }, { "epoch": 0.0362453531598513, "grad_norm": 1.8882519234415072, "learning_rate": 9.999101006666263e-06, "loss": 0.1205, "step": 780 }, { "epoch": 0.03629182156133829, "grad_norm": 2.2542384981774455, "learning_rate": 9.999086680741039e-06, "loss": 0.2198, "step": 781 }, { "epoch": 0.03633828996282528, "grad_norm": 2.7275900296889852, "learning_rate": 9.999072241581241e-06, "loss": 0.1453, "step": 782 }, { "epoch": 0.03638475836431227, "grad_norm": 2.1422871833559283, "learning_rate": 9.999057689187195e-06, "loss": 0.174, "step": 783 }, { "epoch": 0.036431226765799254, "grad_norm": 2.5603513877928052, "learning_rate": 9.999043023559232e-06, "loss": 0.1593, "step": 784 }, { "epoch": 0.036477695167286245, "grad_norm": 3.4157698279102493, "learning_rate": 9.999028244697681e-06, "loss": 0.1949, "step": 785 }, { "epoch": 0.036524163568773235, "grad_norm": 2.7646166899225966, "learning_rate": 9.99901335260288e-06, "loss": 0.2138, "step": 786 }, { "epoch": 0.036570631970260226, "grad_norm": 3.408772560936423, "learning_rate": 9.998998347275167e-06, "loss": 0.2401, "step": 787 }, { "epoch": 0.03661710037174721, "grad_norm": 2.2813849949153306, "learning_rate": 9.998983228714878e-06, "loss": 0.0861, "step": 788 }, { "epoch": 0.0366635687732342, "grad_norm": 4.417592849348099, "learning_rate": 9.998967996922358e-06, "loss": 0.1714, "step": 789 }, { "epoch": 0.03671003717472119, "grad_norm": 2.385274764266831, "learning_rate": 9.998952651897951e-06, "loss": 0.1578, "step": 790 }, { "epoch": 0.03675650557620818, "grad_norm": 1.670969352696277, "learning_rate": 9.998937193642006e-06, "loss": 0.1061, "step": 791 }, { "epoch": 0.03680297397769517, "grad_norm": 1.6488363862738022, "learning_rate": 9.99892162215487e-06, "loss": 0.1045, "step": 792 }, { "epoch": 0.036849442379182154, "grad_norm": 1.8629716374300491, "learning_rate": 9.998905937436901e-06, "loss": 0.1414, "step": 793 }, { "epoch": 0.036895910780669144, "grad_norm": 2.4504447400941314, "learning_rate": 9.99889013948845e-06, "loss": 0.1232, "step": 794 }, { "epoch": 0.036942379182156135, "grad_norm": 2.705915526873342, "learning_rate": 9.998874228309876e-06, "loss": 0.2018, "step": 795 }, { "epoch": 0.036988847583643125, "grad_norm": 6.424480951913342, "learning_rate": 9.99885820390154e-06, "loss": 0.2546, "step": 796 }, { "epoch": 0.03703531598513011, "grad_norm": 2.822696756830644, "learning_rate": 9.998842066263806e-06, "loss": 0.1597, "step": 797 }, { "epoch": 0.0370817843866171, "grad_norm": 2.8745151416035943, "learning_rate": 9.998825815397036e-06, "loss": 0.1196, "step": 798 }, { "epoch": 0.03712825278810409, "grad_norm": 4.749300942997987, "learning_rate": 9.998809451301598e-06, "loss": 0.2511, "step": 799 }, { "epoch": 0.03717472118959108, "grad_norm": 2.7268768084662294, "learning_rate": 9.99879297397787e-06, "loss": 0.1082, "step": 800 }, { "epoch": 0.03722118959107807, "grad_norm": 4.90386962393825, "learning_rate": 9.998776383426217e-06, "loss": 0.1593, "step": 801 }, { "epoch": 0.037267657992565054, "grad_norm": 3.177039117298357, "learning_rate": 9.998759679647017e-06, "loss": 0.1315, "step": 802 }, { "epoch": 0.037314126394052044, "grad_norm": 5.099822270088214, "learning_rate": 9.99874286264065e-06, "loss": 0.2699, "step": 803 }, { "epoch": 0.037360594795539034, "grad_norm": 2.207850869195552, "learning_rate": 9.998725932407497e-06, "loss": 0.1059, "step": 804 }, { "epoch": 0.037407063197026025, "grad_norm": 3.047676728130234, "learning_rate": 9.998708888947939e-06, "loss": 0.1261, "step": 805 }, { "epoch": 0.03745353159851301, "grad_norm": 2.216803806302805, "learning_rate": 9.998691732262365e-06, "loss": 0.1211, "step": 806 }, { "epoch": 0.0375, "grad_norm": 2.7634139277881142, "learning_rate": 9.99867446235116e-06, "loss": 0.2082, "step": 807 }, { "epoch": 0.03754646840148699, "grad_norm": 4.759868180981096, "learning_rate": 9.99865707921472e-06, "loss": 0.2554, "step": 808 }, { "epoch": 0.03759293680297398, "grad_norm": 4.480958710917975, "learning_rate": 9.998639582853435e-06, "loss": 0.1559, "step": 809 }, { "epoch": 0.03763940520446097, "grad_norm": 2.5551293714974213, "learning_rate": 9.998621973267704e-06, "loss": 0.1391, "step": 810 }, { "epoch": 0.03768587360594795, "grad_norm": 3.12403930049813, "learning_rate": 9.998604250457922e-06, "loss": 0.1854, "step": 811 }, { "epoch": 0.037732342007434944, "grad_norm": 5.38473115249965, "learning_rate": 9.998586414424494e-06, "loss": 0.241, "step": 812 }, { "epoch": 0.037778810408921934, "grad_norm": 1.5144704769267234, "learning_rate": 9.998568465167822e-06, "loss": 0.0802, "step": 813 }, { "epoch": 0.037825278810408924, "grad_norm": 3.2462138520923194, "learning_rate": 9.998550402688314e-06, "loss": 0.1887, "step": 814 }, { "epoch": 0.03787174721189591, "grad_norm": 3.4255871786872074, "learning_rate": 9.99853222698638e-06, "loss": 0.1809, "step": 815 }, { "epoch": 0.0379182156133829, "grad_norm": 2.1871850786114346, "learning_rate": 9.998513938062428e-06, "loss": 0.1124, "step": 816 }, { "epoch": 0.03796468401486989, "grad_norm": 7.810269669444946, "learning_rate": 9.998495535916874e-06, "loss": 0.2277, "step": 817 }, { "epoch": 0.03801115241635688, "grad_norm": 2.7139873181256085, "learning_rate": 9.998477020550137e-06, "loss": 0.1589, "step": 818 }, { "epoch": 0.03805762081784387, "grad_norm": 4.033008217404078, "learning_rate": 9.998458391962634e-06, "loss": 0.2418, "step": 819 }, { "epoch": 0.03810408921933085, "grad_norm": 3.057233797649618, "learning_rate": 9.998439650154788e-06, "loss": 0.1807, "step": 820 }, { "epoch": 0.03815055762081784, "grad_norm": 2.480131994658339, "learning_rate": 9.998420795127023e-06, "loss": 0.108, "step": 821 }, { "epoch": 0.038197026022304834, "grad_norm": 2.1112145258691606, "learning_rate": 9.998401826879767e-06, "loss": 0.151, "step": 822 }, { "epoch": 0.038243494423791824, "grad_norm": 3.5312816058594727, "learning_rate": 9.998382745413447e-06, "loss": 0.1831, "step": 823 }, { "epoch": 0.03828996282527881, "grad_norm": 2.8010842884044327, "learning_rate": 9.998363550728497e-06, "loss": 0.1415, "step": 824 }, { "epoch": 0.0383364312267658, "grad_norm": 3.0780818573873776, "learning_rate": 9.998344242825355e-06, "loss": 0.1886, "step": 825 }, { "epoch": 0.03838289962825279, "grad_norm": 2.9937753217221577, "learning_rate": 9.998324821704453e-06, "loss": 0.1394, "step": 826 }, { "epoch": 0.03842936802973978, "grad_norm": 1.8229587393536613, "learning_rate": 9.998305287366233e-06, "loss": 0.1244, "step": 827 }, { "epoch": 0.03847583643122677, "grad_norm": 2.063627021077833, "learning_rate": 9.998285639811139e-06, "loss": 0.1262, "step": 828 }, { "epoch": 0.03852230483271375, "grad_norm": 2.0287028385503203, "learning_rate": 9.998265879039613e-06, "loss": 0.1615, "step": 829 }, { "epoch": 0.03856877323420074, "grad_norm": 2.684945452224348, "learning_rate": 9.998246005052105e-06, "loss": 0.1752, "step": 830 }, { "epoch": 0.03861524163568773, "grad_norm": 1.9364030878822889, "learning_rate": 9.998226017849064e-06, "loss": 0.0977, "step": 831 }, { "epoch": 0.038661710037174724, "grad_norm": 2.0590696932534485, "learning_rate": 9.998205917430944e-06, "loss": 0.1361, "step": 832 }, { "epoch": 0.03870817843866171, "grad_norm": 1.867231203738669, "learning_rate": 9.9981857037982e-06, "loss": 0.1323, "step": 833 }, { "epoch": 0.0387546468401487, "grad_norm": 1.7198667142989301, "learning_rate": 9.998165376951287e-06, "loss": 0.1648, "step": 834 }, { "epoch": 0.03880111524163569, "grad_norm": 2.795370535816204, "learning_rate": 9.99814493689067e-06, "loss": 0.187, "step": 835 }, { "epoch": 0.03884758364312268, "grad_norm": 2.53252777241959, "learning_rate": 9.998124383616808e-06, "loss": 0.1814, "step": 836 }, { "epoch": 0.03889405204460967, "grad_norm": 1.9220521846606429, "learning_rate": 9.99810371713017e-06, "loss": 0.1429, "step": 837 }, { "epoch": 0.03894052044609665, "grad_norm": 3.4753989636584315, "learning_rate": 9.998082937431221e-06, "loss": 0.2826, "step": 838 }, { "epoch": 0.03898698884758364, "grad_norm": 3.738147380417305, "learning_rate": 9.998062044520434e-06, "loss": 0.2025, "step": 839 }, { "epoch": 0.03903345724907063, "grad_norm": 2.4730717435573024, "learning_rate": 9.99804103839828e-06, "loss": 0.1425, "step": 840 }, { "epoch": 0.03907992565055762, "grad_norm": 3.197534296542115, "learning_rate": 9.998019919065239e-06, "loss": 0.2611, "step": 841 }, { "epoch": 0.03912639405204461, "grad_norm": 2.4202818452740957, "learning_rate": 9.997998686521784e-06, "loss": 0.0874, "step": 842 }, { "epoch": 0.0391728624535316, "grad_norm": 1.6463143563342053, "learning_rate": 9.997977340768399e-06, "loss": 0.1905, "step": 843 }, { "epoch": 0.03921933085501859, "grad_norm": 1.8933780259750497, "learning_rate": 9.997955881805567e-06, "loss": 0.244, "step": 844 }, { "epoch": 0.03926579925650558, "grad_norm": 1.5832810163466844, "learning_rate": 9.997934309633774e-06, "loss": 0.1068, "step": 845 }, { "epoch": 0.03931226765799257, "grad_norm": 4.314968026014301, "learning_rate": 9.997912624253508e-06, "loss": 0.1459, "step": 846 }, { "epoch": 0.03935873605947955, "grad_norm": 1.7839497222478473, "learning_rate": 9.997890825665261e-06, "loss": 0.1282, "step": 847 }, { "epoch": 0.03940520446096654, "grad_norm": 3.879910954792317, "learning_rate": 9.997868913869529e-06, "loss": 0.1625, "step": 848 }, { "epoch": 0.03945167286245353, "grad_norm": 3.2015120528721637, "learning_rate": 9.997846888866802e-06, "loss": 0.1713, "step": 849 }, { "epoch": 0.03949814126394052, "grad_norm": 8.513844562396269, "learning_rate": 9.997824750657586e-06, "loss": 0.1801, "step": 850 }, { "epoch": 0.039544609665427506, "grad_norm": 2.046991638116981, "learning_rate": 9.997802499242378e-06, "loss": 0.1708, "step": 851 }, { "epoch": 0.0395910780669145, "grad_norm": 1.3735540648135585, "learning_rate": 9.997780134621684e-06, "loss": 0.0933, "step": 852 }, { "epoch": 0.03963754646840149, "grad_norm": 2.557404689691375, "learning_rate": 9.997757656796008e-06, "loss": 0.1806, "step": 853 }, { "epoch": 0.03968401486988848, "grad_norm": 1.4535241331013877, "learning_rate": 9.997735065765864e-06, "loss": 0.0928, "step": 854 }, { "epoch": 0.03973048327137547, "grad_norm": 3.1093640785004566, "learning_rate": 9.997712361531758e-06, "loss": 0.2031, "step": 855 }, { "epoch": 0.03977695167286245, "grad_norm": 2.0623349105329467, "learning_rate": 9.99768954409421e-06, "loss": 0.176, "step": 856 }, { "epoch": 0.03982342007434944, "grad_norm": 2.2281396085649057, "learning_rate": 9.99766661345373e-06, "loss": 0.1599, "step": 857 }, { "epoch": 0.03986988847583643, "grad_norm": 2.1356249749434886, "learning_rate": 9.997643569610844e-06, "loss": 0.1684, "step": 858 }, { "epoch": 0.03991635687732342, "grad_norm": 3.2277416970379806, "learning_rate": 9.997620412566071e-06, "loss": 0.2244, "step": 859 }, { "epoch": 0.039962825278810406, "grad_norm": 2.2169370271485547, "learning_rate": 9.997597142319935e-06, "loss": 0.2854, "step": 860 }, { "epoch": 0.040009293680297396, "grad_norm": 1.616668139934371, "learning_rate": 9.997573758872962e-06, "loss": 0.1005, "step": 861 }, { "epoch": 0.040055762081784387, "grad_norm": 1.6787320711844278, "learning_rate": 9.997550262225687e-06, "loss": 0.2004, "step": 862 }, { "epoch": 0.04010223048327138, "grad_norm": 3.5775807799822847, "learning_rate": 9.997526652378638e-06, "loss": 0.1583, "step": 863 }, { "epoch": 0.04014869888475837, "grad_norm": 4.023029222983887, "learning_rate": 9.997502929332348e-06, "loss": 0.1897, "step": 864 }, { "epoch": 0.04019516728624535, "grad_norm": 1.9487664397854523, "learning_rate": 9.99747909308736e-06, "loss": 0.1509, "step": 865 }, { "epoch": 0.04024163568773234, "grad_norm": 3.6260533881806425, "learning_rate": 9.997455143644208e-06, "loss": 0.252, "step": 866 }, { "epoch": 0.04028810408921933, "grad_norm": 1.8881012721341404, "learning_rate": 9.99743108100344e-06, "loss": 0.1657, "step": 867 }, { "epoch": 0.04033457249070632, "grad_norm": 9.874141810318001, "learning_rate": 9.997406905165596e-06, "loss": 0.1135, "step": 868 }, { "epoch": 0.040381040892193305, "grad_norm": 1.9429710021984994, "learning_rate": 9.997382616131228e-06, "loss": 0.1639, "step": 869 }, { "epoch": 0.040427509293680296, "grad_norm": 1.3561147045608346, "learning_rate": 9.997358213900883e-06, "loss": 0.0848, "step": 870 }, { "epoch": 0.040473977695167286, "grad_norm": 2.650274189636799, "learning_rate": 9.997333698475115e-06, "loss": 0.168, "step": 871 }, { "epoch": 0.040520446096654276, "grad_norm": 3.597284603227582, "learning_rate": 9.99730906985448e-06, "loss": 0.1892, "step": 872 }, { "epoch": 0.04056691449814127, "grad_norm": 2.746827449851567, "learning_rate": 9.997284328039534e-06, "loss": 0.1454, "step": 873 }, { "epoch": 0.04061338289962825, "grad_norm": 3.475492777038183, "learning_rate": 9.997259473030838e-06, "loss": 0.1915, "step": 874 }, { "epoch": 0.04065985130111524, "grad_norm": 2.0702434281247744, "learning_rate": 9.997234504828956e-06, "loss": 0.1498, "step": 875 }, { "epoch": 0.04070631970260223, "grad_norm": 2.964853102593617, "learning_rate": 9.997209423434453e-06, "loss": 0.2031, "step": 876 }, { "epoch": 0.04075278810408922, "grad_norm": 2.089057103480083, "learning_rate": 9.997184228847897e-06, "loss": 0.1643, "step": 877 }, { "epoch": 0.040799256505576205, "grad_norm": 5.363916929651491, "learning_rate": 9.997158921069857e-06, "loss": 0.2346, "step": 878 }, { "epoch": 0.040845724907063195, "grad_norm": 3.213699058761861, "learning_rate": 9.99713350010091e-06, "loss": 0.2293, "step": 879 }, { "epoch": 0.040892193308550186, "grad_norm": 3.4244359679568923, "learning_rate": 9.99710796594163e-06, "loss": 0.1243, "step": 880 }, { "epoch": 0.040938661710037176, "grad_norm": 4.198510651069142, "learning_rate": 9.997082318592595e-06, "loss": 0.2085, "step": 881 }, { "epoch": 0.040985130111524166, "grad_norm": 10.763260036648933, "learning_rate": 9.997056558054387e-06, "loss": 0.2346, "step": 882 }, { "epoch": 0.04103159851301115, "grad_norm": 2.3369134923293964, "learning_rate": 9.997030684327586e-06, "loss": 0.156, "step": 883 }, { "epoch": 0.04107806691449814, "grad_norm": 2.9571845523030453, "learning_rate": 9.997004697412785e-06, "loss": 0.2492, "step": 884 }, { "epoch": 0.04112453531598513, "grad_norm": 7.0860507068439205, "learning_rate": 9.996978597310564e-06, "loss": 0.2526, "step": 885 }, { "epoch": 0.04117100371747212, "grad_norm": 2.2308808482148565, "learning_rate": 9.99695238402152e-06, "loss": 0.1534, "step": 886 }, { "epoch": 0.041217472118959105, "grad_norm": 2.006784163352565, "learning_rate": 9.996926057546246e-06, "loss": 0.0748, "step": 887 }, { "epoch": 0.041263940520446095, "grad_norm": 2.4071649978071767, "learning_rate": 9.996899617885337e-06, "loss": 0.2147, "step": 888 }, { "epoch": 0.041310408921933085, "grad_norm": 2.4875402746648176, "learning_rate": 9.996873065039393e-06, "loss": 0.1304, "step": 889 }, { "epoch": 0.041356877323420076, "grad_norm": 5.284775649719892, "learning_rate": 9.996846399009015e-06, "loss": 0.18, "step": 890 }, { "epoch": 0.041403345724907066, "grad_norm": 2.1909819411210814, "learning_rate": 9.996819619794806e-06, "loss": 0.1157, "step": 891 }, { "epoch": 0.04144981412639405, "grad_norm": 2.1054233891037217, "learning_rate": 9.996792727397374e-06, "loss": 0.0886, "step": 892 }, { "epoch": 0.04149628252788104, "grad_norm": 2.4604310753719245, "learning_rate": 9.996765721817327e-06, "loss": 0.1872, "step": 893 }, { "epoch": 0.04154275092936803, "grad_norm": 2.5857972948599546, "learning_rate": 9.996738603055279e-06, "loss": 0.1578, "step": 894 }, { "epoch": 0.04158921933085502, "grad_norm": 1.9275552225250403, "learning_rate": 9.996711371111842e-06, "loss": 0.1622, "step": 895 }, { "epoch": 0.041635687732342004, "grad_norm": 1.7864114245349378, "learning_rate": 9.996684025987635e-06, "loss": 0.1077, "step": 896 }, { "epoch": 0.041682156133828994, "grad_norm": 2.7270281578911395, "learning_rate": 9.996656567683274e-06, "loss": 0.168, "step": 897 }, { "epoch": 0.041728624535315985, "grad_norm": 1.2718256467627285, "learning_rate": 9.996628996199383e-06, "loss": 0.0828, "step": 898 }, { "epoch": 0.041775092936802975, "grad_norm": 2.459996482987762, "learning_rate": 9.996601311536586e-06, "loss": 0.1747, "step": 899 }, { "epoch": 0.041821561338289966, "grad_norm": 26.13036925950346, "learning_rate": 9.99657351369551e-06, "loss": 0.1345, "step": 900 }, { "epoch": 0.04186802973977695, "grad_norm": 2.380496070420489, "learning_rate": 9.996545602676788e-06, "loss": 0.0908, "step": 901 }, { "epoch": 0.04191449814126394, "grad_norm": 2.682722895990996, "learning_rate": 9.996517578481046e-06, "loss": 0.1666, "step": 902 }, { "epoch": 0.04196096654275093, "grad_norm": 5.213323988141187, "learning_rate": 9.996489441108924e-06, "loss": 0.1406, "step": 903 }, { "epoch": 0.04200743494423792, "grad_norm": 2.7940983229746315, "learning_rate": 9.996461190561058e-06, "loss": 0.1063, "step": 904 }, { "epoch": 0.042053903345724904, "grad_norm": 3.316210722515401, "learning_rate": 9.996432826838087e-06, "loss": 0.1766, "step": 905 }, { "epoch": 0.042100371747211894, "grad_norm": 2.273881700396464, "learning_rate": 9.996404349940651e-06, "loss": 0.1818, "step": 906 }, { "epoch": 0.042146840148698884, "grad_norm": 2.6520770636662134, "learning_rate": 9.996375759869401e-06, "loss": 0.1106, "step": 907 }, { "epoch": 0.042193308550185875, "grad_norm": 6.373358094576792, "learning_rate": 9.99634705662498e-06, "loss": 0.2586, "step": 908 }, { "epoch": 0.042239776951672865, "grad_norm": 2.814515540252412, "learning_rate": 9.996318240208041e-06, "loss": 0.1562, "step": 909 }, { "epoch": 0.04228624535315985, "grad_norm": 1.9559613733563679, "learning_rate": 9.996289310619234e-06, "loss": 0.1475, "step": 910 }, { "epoch": 0.04233271375464684, "grad_norm": 5.528302323840256, "learning_rate": 9.996260267859216e-06, "loss": 0.263, "step": 911 }, { "epoch": 0.04237918215613383, "grad_norm": 2.6544186296354297, "learning_rate": 9.996231111928645e-06, "loss": 0.157, "step": 912 }, { "epoch": 0.04242565055762082, "grad_norm": 2.8405523187411372, "learning_rate": 9.996201842828181e-06, "loss": 0.132, "step": 913 }, { "epoch": 0.0424721189591078, "grad_norm": 2.022019580917888, "learning_rate": 9.996172460558486e-06, "loss": 0.1965, "step": 914 }, { "epoch": 0.042518587360594794, "grad_norm": 2.7642844635331376, "learning_rate": 9.996142965120225e-06, "loss": 0.2232, "step": 915 }, { "epoch": 0.042565055762081784, "grad_norm": 2.3246421326552476, "learning_rate": 9.99611335651407e-06, "loss": 0.2584, "step": 916 }, { "epoch": 0.042611524163568774, "grad_norm": 2.7195066542752384, "learning_rate": 9.99608363474069e-06, "loss": 0.1265, "step": 917 }, { "epoch": 0.042657992565055765, "grad_norm": 2.6688354673198145, "learning_rate": 9.996053799800754e-06, "loss": 0.2004, "step": 918 }, { "epoch": 0.04270446096654275, "grad_norm": 3.6927834893234954, "learning_rate": 9.996023851694942e-06, "loss": 0.247, "step": 919 }, { "epoch": 0.04275092936802974, "grad_norm": 2.206320608116945, "learning_rate": 9.995993790423934e-06, "loss": 0.1198, "step": 920 }, { "epoch": 0.04279739776951673, "grad_norm": 7.192578866078779, "learning_rate": 9.995963615988406e-06, "loss": 0.1185, "step": 921 }, { "epoch": 0.04284386617100372, "grad_norm": 3.1172875871847876, "learning_rate": 9.995933328389047e-06, "loss": 0.2606, "step": 922 }, { "epoch": 0.0428903345724907, "grad_norm": 2.4463948709733843, "learning_rate": 9.995902927626535e-06, "loss": 0.1772, "step": 923 }, { "epoch": 0.04293680297397769, "grad_norm": 2.5887746122030766, "learning_rate": 9.995872413701568e-06, "loss": 0.156, "step": 924 }, { "epoch": 0.042983271375464684, "grad_norm": 2.070231650980679, "learning_rate": 9.99584178661483e-06, "loss": 0.1568, "step": 925 }, { "epoch": 0.043029739776951674, "grad_norm": 1.6243818206189087, "learning_rate": 9.99581104636702e-06, "loss": 0.0851, "step": 926 }, { "epoch": 0.043076208178438664, "grad_norm": 2.397126544195999, "learning_rate": 9.99578019295883e-06, "loss": 0.1465, "step": 927 }, { "epoch": 0.04312267657992565, "grad_norm": 3.9930740795696447, "learning_rate": 9.99574922639096e-06, "loss": 0.1455, "step": 928 }, { "epoch": 0.04316914498141264, "grad_norm": 2.5552281758040456, "learning_rate": 9.995718146664115e-06, "loss": 0.1436, "step": 929 }, { "epoch": 0.04321561338289963, "grad_norm": 2.7223579489372636, "learning_rate": 9.995686953778994e-06, "loss": 0.1544, "step": 930 }, { "epoch": 0.04326208178438662, "grad_norm": 4.913478512466867, "learning_rate": 9.995655647736307e-06, "loss": 0.3182, "step": 931 }, { "epoch": 0.04330855018587361, "grad_norm": 2.5719180573245195, "learning_rate": 9.99562422853676e-06, "loss": 0.1027, "step": 932 }, { "epoch": 0.04335501858736059, "grad_norm": 2.5071799741684693, "learning_rate": 9.995592696181066e-06, "loss": 0.1509, "step": 933 }, { "epoch": 0.04340148698884758, "grad_norm": 3.9840146551896245, "learning_rate": 9.995561050669943e-06, "loss": 0.1017, "step": 934 }, { "epoch": 0.043447955390334574, "grad_norm": 1.604749641531468, "learning_rate": 9.9955292920041e-06, "loss": 0.1497, "step": 935 }, { "epoch": 0.043494423791821564, "grad_norm": 2.3959905066329363, "learning_rate": 9.995497420184264e-06, "loss": 0.1082, "step": 936 }, { "epoch": 0.04354089219330855, "grad_norm": 2.5100736076659005, "learning_rate": 9.995465435211153e-06, "loss": 0.2249, "step": 937 }, { "epoch": 0.04358736059479554, "grad_norm": 2.7775625201025096, "learning_rate": 9.995433337085492e-06, "loss": 0.1379, "step": 938 }, { "epoch": 0.04363382899628253, "grad_norm": 2.4132091177684454, "learning_rate": 9.995401125808008e-06, "loss": 0.1105, "step": 939 }, { "epoch": 0.04368029739776952, "grad_norm": 5.288125496307036, "learning_rate": 9.995368801379433e-06, "loss": 0.1713, "step": 940 }, { "epoch": 0.04372676579925651, "grad_norm": 3.558527687573744, "learning_rate": 9.995336363800494e-06, "loss": 0.2563, "step": 941 }, { "epoch": 0.04377323420074349, "grad_norm": 2.6208644729398847, "learning_rate": 9.995303813071931e-06, "loss": 0.2111, "step": 942 }, { "epoch": 0.04381970260223048, "grad_norm": 3.3065634279716583, "learning_rate": 9.995271149194479e-06, "loss": 0.2799, "step": 943 }, { "epoch": 0.04386617100371747, "grad_norm": 5.373793898190635, "learning_rate": 9.995238372168877e-06, "loss": 0.1319, "step": 944 }, { "epoch": 0.043912639405204464, "grad_norm": 4.694044755281646, "learning_rate": 9.995205481995868e-06, "loss": 0.1975, "step": 945 }, { "epoch": 0.04395910780669145, "grad_norm": 2.2953863130203684, "learning_rate": 9.9951724786762e-06, "loss": 0.2153, "step": 946 }, { "epoch": 0.04400557620817844, "grad_norm": 4.918614233469083, "learning_rate": 9.995139362210616e-06, "loss": 0.2089, "step": 947 }, { "epoch": 0.04405204460966543, "grad_norm": 2.865941545527704, "learning_rate": 9.995106132599869e-06, "loss": 0.159, "step": 948 }, { "epoch": 0.04409851301115242, "grad_norm": 1.7531303875778386, "learning_rate": 9.99507278984471e-06, "loss": 0.1365, "step": 949 }, { "epoch": 0.04414498141263941, "grad_norm": 2.8392709756053502, "learning_rate": 9.995039333945896e-06, "loss": 0.1091, "step": 950 }, { "epoch": 0.04419144981412639, "grad_norm": 2.286319166252824, "learning_rate": 9.995005764904184e-06, "loss": 0.1257, "step": 951 }, { "epoch": 0.04423791821561338, "grad_norm": 1.7183644179097115, "learning_rate": 9.994972082720333e-06, "loss": 0.1125, "step": 952 }, { "epoch": 0.04428438661710037, "grad_norm": 1.6829586254593334, "learning_rate": 9.994938287395108e-06, "loss": 0.1539, "step": 953 }, { "epoch": 0.04433085501858736, "grad_norm": 1.9695227310800174, "learning_rate": 9.994904378929275e-06, "loss": 0.1294, "step": 954 }, { "epoch": 0.04437732342007435, "grad_norm": 1.8140568830578099, "learning_rate": 9.994870357323598e-06, "loss": 0.2261, "step": 955 }, { "epoch": 0.04442379182156134, "grad_norm": 2.8558500832427343, "learning_rate": 9.994836222578854e-06, "loss": 0.2178, "step": 956 }, { "epoch": 0.04447026022304833, "grad_norm": 2.018902314882039, "learning_rate": 9.994801974695809e-06, "loss": 0.1359, "step": 957 }, { "epoch": 0.04451672862453532, "grad_norm": 5.072319368366155, "learning_rate": 9.994767613675244e-06, "loss": 0.1712, "step": 958 }, { "epoch": 0.04456319702602231, "grad_norm": 4.102354666764538, "learning_rate": 9.994733139517935e-06, "loss": 0.1192, "step": 959 }, { "epoch": 0.04460966542750929, "grad_norm": 4.741056171715995, "learning_rate": 9.994698552224664e-06, "loss": 0.1731, "step": 960 }, { "epoch": 0.04465613382899628, "grad_norm": 2.8805383646796887, "learning_rate": 9.994663851796215e-06, "loss": 0.1082, "step": 961 }, { "epoch": 0.04470260223048327, "grad_norm": 2.861311700205426, "learning_rate": 9.994629038233372e-06, "loss": 0.0943, "step": 962 }, { "epoch": 0.04474907063197026, "grad_norm": 2.660444933526591, "learning_rate": 9.994594111536924e-06, "loss": 0.1563, "step": 963 }, { "epoch": 0.044795539033457246, "grad_norm": 3.66061165644885, "learning_rate": 9.994559071707663e-06, "loss": 0.1921, "step": 964 }, { "epoch": 0.04484200743494424, "grad_norm": 4.621016514105437, "learning_rate": 9.994523918746384e-06, "loss": 0.1809, "step": 965 }, { "epoch": 0.04488847583643123, "grad_norm": 4.2784177780338455, "learning_rate": 9.99448865265388e-06, "loss": 0.1509, "step": 966 }, { "epoch": 0.04493494423791822, "grad_norm": 2.354393831602962, "learning_rate": 9.99445327343095e-06, "loss": 0.1292, "step": 967 }, { "epoch": 0.04498141263940521, "grad_norm": 4.2768599176823345, "learning_rate": 9.9944177810784e-06, "loss": 0.1677, "step": 968 }, { "epoch": 0.04502788104089219, "grad_norm": 2.065536416701332, "learning_rate": 9.994382175597028e-06, "loss": 0.0961, "step": 969 }, { "epoch": 0.04507434944237918, "grad_norm": 1.9597630454174577, "learning_rate": 9.994346456987644e-06, "loss": 0.1368, "step": 970 }, { "epoch": 0.04512081784386617, "grad_norm": 4.479776803165555, "learning_rate": 9.994310625251057e-06, "loss": 0.2035, "step": 971 }, { "epoch": 0.04516728624535316, "grad_norm": 3.262011682692466, "learning_rate": 9.994274680388077e-06, "loss": 0.1725, "step": 972 }, { "epoch": 0.045213754646840146, "grad_norm": 2.3961354649074527, "learning_rate": 9.99423862239952e-06, "loss": 0.1685, "step": 973 }, { "epoch": 0.045260223048327136, "grad_norm": 2.193432022153209, "learning_rate": 9.994202451286202e-06, "loss": 0.2121, "step": 974 }, { "epoch": 0.04530669144981413, "grad_norm": 3.6818471685163847, "learning_rate": 9.994166167048941e-06, "loss": 0.1842, "step": 975 }, { "epoch": 0.04535315985130112, "grad_norm": 2.3348808820782048, "learning_rate": 9.99412976968856e-06, "loss": 0.266, "step": 976 }, { "epoch": 0.04539962825278811, "grad_norm": 3.357552079961166, "learning_rate": 9.994093259205886e-06, "loss": 0.204, "step": 977 }, { "epoch": 0.04544609665427509, "grad_norm": 9.114372200656863, "learning_rate": 9.99405663560174e-06, "loss": 0.2629, "step": 978 }, { "epoch": 0.04549256505576208, "grad_norm": 2.2301538294900953, "learning_rate": 9.994019898876957e-06, "loss": 0.108, "step": 979 }, { "epoch": 0.04553903345724907, "grad_norm": 2.3825498497689184, "learning_rate": 9.993983049032367e-06, "loss": 0.1417, "step": 980 }, { "epoch": 0.04558550185873606, "grad_norm": 2.9670175975918656, "learning_rate": 9.993946086068805e-06, "loss": 0.2149, "step": 981 }, { "epoch": 0.045631970260223045, "grad_norm": 3.9406661715767872, "learning_rate": 9.993909009987109e-06, "loss": 0.1738, "step": 982 }, { "epoch": 0.045678438661710036, "grad_norm": 1.8219284967955052, "learning_rate": 9.993871820788118e-06, "loss": 0.13, "step": 983 }, { "epoch": 0.045724907063197026, "grad_norm": 1.580337366741105, "learning_rate": 9.993834518472672e-06, "loss": 0.1839, "step": 984 }, { "epoch": 0.045771375464684017, "grad_norm": 4.110638710447729, "learning_rate": 9.993797103041618e-06, "loss": 0.2013, "step": 985 }, { "epoch": 0.04581784386617101, "grad_norm": 4.9359290988099485, "learning_rate": 9.993759574495806e-06, "loss": 0.195, "step": 986 }, { "epoch": 0.04586431226765799, "grad_norm": 1.3421603842406111, "learning_rate": 9.993721932836082e-06, "loss": 0.0794, "step": 987 }, { "epoch": 0.04591078066914498, "grad_norm": 3.669680564195783, "learning_rate": 9.993684178063302e-06, "loss": 0.1116, "step": 988 }, { "epoch": 0.04595724907063197, "grad_norm": 2.2609518315949297, "learning_rate": 9.99364631017832e-06, "loss": 0.1811, "step": 989 }, { "epoch": 0.04600371747211896, "grad_norm": 2.2630877401563425, "learning_rate": 9.993608329181991e-06, "loss": 0.1337, "step": 990 }, { "epoch": 0.046050185873605945, "grad_norm": 2.643333425399134, "learning_rate": 9.99357023507518e-06, "loss": 0.1171, "step": 991 }, { "epoch": 0.046096654275092935, "grad_norm": 1.907982169568737, "learning_rate": 9.993532027858745e-06, "loss": 0.1163, "step": 992 }, { "epoch": 0.046143122676579926, "grad_norm": 3.508591450742984, "learning_rate": 9.993493707533557e-06, "loss": 0.1271, "step": 993 }, { "epoch": 0.046189591078066916, "grad_norm": 2.4525817981803155, "learning_rate": 9.993455274100479e-06, "loss": 0.2048, "step": 994 }, { "epoch": 0.046236059479553906, "grad_norm": 2.8633835240575762, "learning_rate": 9.993416727560384e-06, "loss": 0.2025, "step": 995 }, { "epoch": 0.04628252788104089, "grad_norm": 3.364712680546293, "learning_rate": 9.993378067914144e-06, "loss": 0.1774, "step": 996 }, { "epoch": 0.04632899628252788, "grad_norm": 2.0994413322176593, "learning_rate": 9.993339295162637e-06, "loss": 0.1637, "step": 997 }, { "epoch": 0.04637546468401487, "grad_norm": 2.8893634916749034, "learning_rate": 9.993300409306738e-06, "loss": 0.1227, "step": 998 }, { "epoch": 0.04642193308550186, "grad_norm": 3.1563967550734002, "learning_rate": 9.99326141034733e-06, "loss": 0.1382, "step": 999 }, { "epoch": 0.046468401486988845, "grad_norm": 2.727423620635365, "learning_rate": 9.993222298285295e-06, "loss": 0.147, "step": 1000 }, { "epoch": 0.046514869888475835, "grad_norm": 2.839643809845794, "learning_rate": 9.993183073121521e-06, "loss": 0.1277, "step": 1001 }, { "epoch": 0.046561338289962825, "grad_norm": 2.16590818115136, "learning_rate": 9.993143734856895e-06, "loss": 0.1279, "step": 1002 }, { "epoch": 0.046607806691449816, "grad_norm": 2.1860880375700074, "learning_rate": 9.993104283492307e-06, "loss": 0.1616, "step": 1003 }, { "epoch": 0.046654275092936806, "grad_norm": 2.4022485427498985, "learning_rate": 9.993064719028653e-06, "loss": 0.2348, "step": 1004 }, { "epoch": 0.04670074349442379, "grad_norm": 1.9012164296838534, "learning_rate": 9.993025041466827e-06, "loss": 0.1327, "step": 1005 }, { "epoch": 0.04674721189591078, "grad_norm": 1.9180443981690065, "learning_rate": 9.99298525080773e-06, "loss": 0.1127, "step": 1006 }, { "epoch": 0.04679368029739777, "grad_norm": 2.8926016400364656, "learning_rate": 9.99294534705226e-06, "loss": 0.209, "step": 1007 }, { "epoch": 0.04684014869888476, "grad_norm": 1.8420514539598307, "learning_rate": 9.992905330201326e-06, "loss": 0.1727, "step": 1008 }, { "epoch": 0.046886617100371744, "grad_norm": 3.606640476347125, "learning_rate": 9.992865200255829e-06, "loss": 0.1854, "step": 1009 }, { "epoch": 0.046933085501858735, "grad_norm": 3.7523111664958337, "learning_rate": 9.99282495721668e-06, "loss": 0.1858, "step": 1010 }, { "epoch": 0.046979553903345725, "grad_norm": 2.159837136353221, "learning_rate": 9.992784601084793e-06, "loss": 0.0964, "step": 1011 }, { "epoch": 0.047026022304832715, "grad_norm": 2.147333277679544, "learning_rate": 9.992744131861079e-06, "loss": 0.1368, "step": 1012 }, { "epoch": 0.047072490706319706, "grad_norm": 5.0537618433147555, "learning_rate": 9.992703549546455e-06, "loss": 0.2488, "step": 1013 }, { "epoch": 0.04711895910780669, "grad_norm": 2.6372908727390016, "learning_rate": 9.992662854141842e-06, "loss": 0.1903, "step": 1014 }, { "epoch": 0.04716542750929368, "grad_norm": 2.3465108949106255, "learning_rate": 9.992622045648159e-06, "loss": 0.1447, "step": 1015 }, { "epoch": 0.04721189591078067, "grad_norm": 4.254718382669502, "learning_rate": 9.992581124066333e-06, "loss": 0.1844, "step": 1016 }, { "epoch": 0.04725836431226766, "grad_norm": 2.0350173539766336, "learning_rate": 9.99254008939729e-06, "loss": 0.1303, "step": 1017 }, { "epoch": 0.047304832713754644, "grad_norm": 2.1440059942013314, "learning_rate": 9.99249894164196e-06, "loss": 0.1323, "step": 1018 }, { "epoch": 0.047351301115241634, "grad_norm": 3.0869546973394635, "learning_rate": 9.992457680801273e-06, "loss": 0.1525, "step": 1019 }, { "epoch": 0.047397769516728624, "grad_norm": 2.995616234579955, "learning_rate": 9.992416306876166e-06, "loss": 0.1896, "step": 1020 }, { "epoch": 0.047444237918215615, "grad_norm": 2.4646761894764957, "learning_rate": 9.992374819867574e-06, "loss": 0.1583, "step": 1021 }, { "epoch": 0.047490706319702605, "grad_norm": 2.96621958117989, "learning_rate": 9.99233321977644e-06, "loss": 0.1131, "step": 1022 }, { "epoch": 0.04753717472118959, "grad_norm": 1.864327590328416, "learning_rate": 9.992291506603704e-06, "loss": 0.1703, "step": 1023 }, { "epoch": 0.04758364312267658, "grad_norm": 2.496943466669532, "learning_rate": 9.99224968035031e-06, "loss": 0.1567, "step": 1024 }, { "epoch": 0.04763011152416357, "grad_norm": 1.5347747085175452, "learning_rate": 9.992207741017206e-06, "loss": 0.136, "step": 1025 }, { "epoch": 0.04767657992565056, "grad_norm": 3.084377703877324, "learning_rate": 9.992165688605345e-06, "loss": 0.2647, "step": 1026 }, { "epoch": 0.04772304832713754, "grad_norm": 3.3767143524595062, "learning_rate": 9.992123523115673e-06, "loss": 0.2066, "step": 1027 }, { "epoch": 0.047769516728624534, "grad_norm": 2.01585846829879, "learning_rate": 9.992081244549152e-06, "loss": 0.1178, "step": 1028 }, { "epoch": 0.047815985130111524, "grad_norm": 20.65667233381447, "learning_rate": 9.992038852906735e-06, "loss": 0.1789, "step": 1029 }, { "epoch": 0.047862453531598514, "grad_norm": 6.435471167406935, "learning_rate": 9.991996348189387e-06, "loss": 0.1014, "step": 1030 }, { "epoch": 0.047908921933085505, "grad_norm": 2.032252942662383, "learning_rate": 9.991953730398066e-06, "loss": 0.1567, "step": 1031 }, { "epoch": 0.04795539033457249, "grad_norm": 1.4770778494532144, "learning_rate": 9.991910999533739e-06, "loss": 0.0851, "step": 1032 }, { "epoch": 0.04800185873605948, "grad_norm": 2.8672406596339535, "learning_rate": 9.991868155597373e-06, "loss": 0.1055, "step": 1033 }, { "epoch": 0.04804832713754647, "grad_norm": 3.7700420691345613, "learning_rate": 9.99182519858994e-06, "loss": 0.1901, "step": 1034 }, { "epoch": 0.04809479553903346, "grad_norm": 1.8955668688603435, "learning_rate": 9.991782128512414e-06, "loss": 0.1072, "step": 1035 }, { "epoch": 0.04814126394052044, "grad_norm": 2.661099032842143, "learning_rate": 9.991738945365768e-06, "loss": 0.1237, "step": 1036 }, { "epoch": 0.04818773234200743, "grad_norm": 3.4223128920398147, "learning_rate": 9.99169564915098e-06, "loss": 0.2253, "step": 1037 }, { "epoch": 0.048234200743494424, "grad_norm": 1.7130811683726308, "learning_rate": 9.991652239869033e-06, "loss": 0.1333, "step": 1038 }, { "epoch": 0.048280669144981414, "grad_norm": 2.4259077857766846, "learning_rate": 9.991608717520908e-06, "loss": 0.165, "step": 1039 }, { "epoch": 0.048327137546468404, "grad_norm": 2.3937527416327975, "learning_rate": 9.991565082107593e-06, "loss": 0.1438, "step": 1040 }, { "epoch": 0.04837360594795539, "grad_norm": 4.592015894436241, "learning_rate": 9.991521333630076e-06, "loss": 0.1518, "step": 1041 }, { "epoch": 0.04842007434944238, "grad_norm": 1.1885312375111352, "learning_rate": 9.991477472089345e-06, "loss": 0.0687, "step": 1042 }, { "epoch": 0.04846654275092937, "grad_norm": 1.872959674624761, "learning_rate": 9.9914334974864e-06, "loss": 0.1118, "step": 1043 }, { "epoch": 0.04851301115241636, "grad_norm": 3.626366002312335, "learning_rate": 9.99138940982223e-06, "loss": 0.1472, "step": 1044 }, { "epoch": 0.04855947955390334, "grad_norm": 2.40551222073491, "learning_rate": 9.991345209097835e-06, "loss": 0.1255, "step": 1045 }, { "epoch": 0.04860594795539033, "grad_norm": 2.0198412689683365, "learning_rate": 9.99130089531422e-06, "loss": 0.1655, "step": 1046 }, { "epoch": 0.04865241635687732, "grad_norm": 4.1038032155948585, "learning_rate": 9.991256468472388e-06, "loss": 0.1417, "step": 1047 }, { "epoch": 0.048698884758364314, "grad_norm": 5.252449451280142, "learning_rate": 9.991211928573342e-06, "loss": 0.1835, "step": 1048 }, { "epoch": 0.048745353159851304, "grad_norm": 1.599778239143478, "learning_rate": 9.991167275618092e-06, "loss": 0.1351, "step": 1049 }, { "epoch": 0.04879182156133829, "grad_norm": 3.7529779667896643, "learning_rate": 9.99112250960765e-06, "loss": 0.1358, "step": 1050 }, { "epoch": 0.04883828996282528, "grad_norm": 5.589422422183588, "learning_rate": 9.99107763054303e-06, "loss": 0.1914, "step": 1051 }, { "epoch": 0.04888475836431227, "grad_norm": 40.99037011307474, "learning_rate": 9.991032638425252e-06, "loss": 0.2018, "step": 1052 }, { "epoch": 0.04893122676579926, "grad_norm": 2.695148822885724, "learning_rate": 9.990987533255327e-06, "loss": 0.164, "step": 1053 }, { "epoch": 0.04897769516728624, "grad_norm": 3.8652656773678578, "learning_rate": 9.990942315034285e-06, "loss": 0.199, "step": 1054 }, { "epoch": 0.04902416356877323, "grad_norm": 3.9143789874133366, "learning_rate": 9.990896983763144e-06, "loss": 0.172, "step": 1055 }, { "epoch": 0.04907063197026022, "grad_norm": 2.407462732951649, "learning_rate": 9.990851539442935e-06, "loss": 0.1298, "step": 1056 }, { "epoch": 0.04911710037174721, "grad_norm": 2.5790940355050576, "learning_rate": 9.990805982074685e-06, "loss": 0.12, "step": 1057 }, { "epoch": 0.049163568773234204, "grad_norm": 1.9471323411587822, "learning_rate": 9.990760311659429e-06, "loss": 0.0771, "step": 1058 }, { "epoch": 0.04921003717472119, "grad_norm": 3.5709211454560164, "learning_rate": 9.990714528198196e-06, "loss": 0.0968, "step": 1059 }, { "epoch": 0.04925650557620818, "grad_norm": 7.190837845562485, "learning_rate": 9.990668631692029e-06, "loss": 0.1952, "step": 1060 }, { "epoch": 0.04930297397769517, "grad_norm": 2.7236694133292194, "learning_rate": 9.990622622141963e-06, "loss": 0.0962, "step": 1061 }, { "epoch": 0.04934944237918216, "grad_norm": 3.677657002153419, "learning_rate": 9.990576499549043e-06, "loss": 0.1326, "step": 1062 }, { "epoch": 0.04939591078066914, "grad_norm": 4.301656210633978, "learning_rate": 9.990530263914313e-06, "loss": 0.233, "step": 1063 }, { "epoch": 0.04944237918215613, "grad_norm": 2.363284316077123, "learning_rate": 9.99048391523882e-06, "loss": 0.0709, "step": 1064 }, { "epoch": 0.04948884758364312, "grad_norm": 7.638182361862872, "learning_rate": 9.990437453523613e-06, "loss": 0.2168, "step": 1065 }, { "epoch": 0.04953531598513011, "grad_norm": 5.279956157424289, "learning_rate": 9.990390878769747e-06, "loss": 0.0907, "step": 1066 }, { "epoch": 0.0495817843866171, "grad_norm": 3.805338760620107, "learning_rate": 9.990344190978274e-06, "loss": 0.1713, "step": 1067 }, { "epoch": 0.04962825278810409, "grad_norm": 2.6732292344115653, "learning_rate": 9.990297390150253e-06, "loss": 0.1887, "step": 1068 }, { "epoch": 0.04967472118959108, "grad_norm": 2.7483044242811694, "learning_rate": 9.990250476286745e-06, "loss": 0.1358, "step": 1069 }, { "epoch": 0.04972118959107807, "grad_norm": 1.8973488403863528, "learning_rate": 9.990203449388812e-06, "loss": 0.1156, "step": 1070 }, { "epoch": 0.04976765799256506, "grad_norm": 6.014219302039927, "learning_rate": 9.990156309457517e-06, "loss": 0.1176, "step": 1071 }, { "epoch": 0.04981412639405204, "grad_norm": 2.1472170861166315, "learning_rate": 9.990109056493929e-06, "loss": 0.1392, "step": 1072 }, { "epoch": 0.04986059479553903, "grad_norm": 3.2263449342862933, "learning_rate": 9.99006169049912e-06, "loss": 0.1824, "step": 1073 }, { "epoch": 0.04990706319702602, "grad_norm": 2.240461483815064, "learning_rate": 9.990014211474161e-06, "loss": 0.1323, "step": 1074 }, { "epoch": 0.04995353159851301, "grad_norm": 4.475553931953118, "learning_rate": 9.98996661942013e-06, "loss": 0.2678, "step": 1075 }, { "epoch": 0.05, "grad_norm": 1.9430264982250942, "learning_rate": 9.989918914338102e-06, "loss": 0.1177, "step": 1076 }, { "epoch": 0.050046468401486986, "grad_norm": 4.425757344395219, "learning_rate": 9.98987109622916e-06, "loss": 0.1344, "step": 1077 }, { "epoch": 0.05009293680297398, "grad_norm": 3.3990914015506006, "learning_rate": 9.989823165094384e-06, "loss": 0.216, "step": 1078 }, { "epoch": 0.05013940520446097, "grad_norm": 3.935792812013255, "learning_rate": 9.989775120934862e-06, "loss": 0.1212, "step": 1079 }, { "epoch": 0.05018587360594796, "grad_norm": 4.233915303801398, "learning_rate": 9.989726963751683e-06, "loss": 0.1734, "step": 1080 }, { "epoch": 0.05023234200743494, "grad_norm": 6.947101381439581, "learning_rate": 9.989678693545936e-06, "loss": 0.1807, "step": 1081 }, { "epoch": 0.05027881040892193, "grad_norm": 3.551858014771288, "learning_rate": 9.989630310318715e-06, "loss": 0.1099, "step": 1082 }, { "epoch": 0.05032527881040892, "grad_norm": 2.003774624494695, "learning_rate": 9.989581814071117e-06, "loss": 0.13, "step": 1083 }, { "epoch": 0.05037174721189591, "grad_norm": 6.813146180756814, "learning_rate": 9.989533204804237e-06, "loss": 0.1371, "step": 1084 }, { "epoch": 0.0504182156133829, "grad_norm": 5.241222914992616, "learning_rate": 9.98948448251918e-06, "loss": 0.1186, "step": 1085 }, { "epoch": 0.050464684014869886, "grad_norm": 7.189705485378345, "learning_rate": 9.989435647217049e-06, "loss": 0.1458, "step": 1086 }, { "epoch": 0.050511152416356876, "grad_norm": 1.6449120176884042, "learning_rate": 9.989386698898948e-06, "loss": 0.0952, "step": 1087 }, { "epoch": 0.05055762081784387, "grad_norm": 5.000450903922784, "learning_rate": 9.989337637565988e-06, "loss": 0.1401, "step": 1088 }, { "epoch": 0.05060408921933086, "grad_norm": 2.010617157409459, "learning_rate": 9.989288463219278e-06, "loss": 0.1535, "step": 1089 }, { "epoch": 0.05065055762081784, "grad_norm": 2.1731256373904233, "learning_rate": 9.989239175859934e-06, "loss": 0.1149, "step": 1090 }, { "epoch": 0.05069702602230483, "grad_norm": 5.525609926542439, "learning_rate": 9.989189775489072e-06, "loss": 0.1672, "step": 1091 }, { "epoch": 0.05074349442379182, "grad_norm": 2.7661391225565333, "learning_rate": 9.98914026210781e-06, "loss": 0.153, "step": 1092 }, { "epoch": 0.05078996282527881, "grad_norm": 2.771831977842605, "learning_rate": 9.989090635717272e-06, "loss": 0.1737, "step": 1093 }, { "epoch": 0.0508364312267658, "grad_norm": 342.12170419448944, "learning_rate": 9.989040896318577e-06, "loss": 0.1253, "step": 1094 }, { "epoch": 0.050882899628252785, "grad_norm": 5.0182585202985, "learning_rate": 9.988991043912857e-06, "loss": 0.1269, "step": 1095 }, { "epoch": 0.050929368029739776, "grad_norm": 12.761826552814513, "learning_rate": 9.988941078501237e-06, "loss": 0.1885, "step": 1096 }, { "epoch": 0.050975836431226766, "grad_norm": 13.4851131232319, "learning_rate": 9.988891000084853e-06, "loss": 0.094, "step": 1097 }, { "epoch": 0.05102230483271376, "grad_norm": 12.863794439137282, "learning_rate": 9.988840808664836e-06, "loss": 0.1291, "step": 1098 }, { "epoch": 0.05106877323420075, "grad_norm": 555.709678606607, "learning_rate": 9.988790504242325e-06, "loss": 0.1813, "step": 1099 }, { "epoch": 0.05111524163568773, "grad_norm": 24.213050756804066, "learning_rate": 9.988740086818457e-06, "loss": 0.3582, "step": 1100 }, { "epoch": 0.05116171003717472, "grad_norm": 28.6221838063133, "learning_rate": 9.988689556394377e-06, "loss": 0.2953, "step": 1101 }, { "epoch": 0.05120817843866171, "grad_norm": 31.076783517101568, "learning_rate": 9.988638912971227e-06, "loss": 0.3582, "step": 1102 }, { "epoch": 0.0512546468401487, "grad_norm": 82.71778491598948, "learning_rate": 9.988588156550154e-06, "loss": 0.373, "step": 1103 }, { "epoch": 0.051301115241635685, "grad_norm": 22.69896087116768, "learning_rate": 9.988537287132311e-06, "loss": 0.4046, "step": 1104 }, { "epoch": 0.051347583643122675, "grad_norm": 20.845029282234204, "learning_rate": 9.988486304718847e-06, "loss": 0.4015, "step": 1105 }, { "epoch": 0.051394052044609666, "grad_norm": 8.273363409641266, "learning_rate": 9.988435209310918e-06, "loss": 0.4437, "step": 1106 }, { "epoch": 0.051440520446096656, "grad_norm": 14.268647673711344, "learning_rate": 9.988384000909683e-06, "loss": 0.3382, "step": 1107 }, { "epoch": 0.051486988847583647, "grad_norm": 6.009939162832025, "learning_rate": 9.988332679516298e-06, "loss": 0.2641, "step": 1108 }, { "epoch": 0.05153345724907063, "grad_norm": 18.29362408299459, "learning_rate": 9.988281245131928e-06, "loss": 0.3531, "step": 1109 }, { "epoch": 0.05157992565055762, "grad_norm": 15.86874807982212, "learning_rate": 9.988229697757739e-06, "loss": 0.2759, "step": 1110 }, { "epoch": 0.05162639405204461, "grad_norm": 11.087543735387936, "learning_rate": 9.988178037394897e-06, "loss": 0.2954, "step": 1111 }, { "epoch": 0.0516728624535316, "grad_norm": 8.56173077680115, "learning_rate": 9.98812626404457e-06, "loss": 0.2839, "step": 1112 }, { "epoch": 0.051719330855018585, "grad_norm": 8.135207564457268, "learning_rate": 9.988074377707936e-06, "loss": 0.4439, "step": 1113 }, { "epoch": 0.051765799256505575, "grad_norm": 7.673813449816158, "learning_rate": 9.988022378386167e-06, "loss": 0.2875, "step": 1114 }, { "epoch": 0.051812267657992565, "grad_norm": 6.278918780669533, "learning_rate": 9.987970266080443e-06, "loss": 0.2908, "step": 1115 }, { "epoch": 0.051858736059479556, "grad_norm": 7.110293051610562, "learning_rate": 9.98791804079194e-06, "loss": 0.3358, "step": 1116 }, { "epoch": 0.051905204460966546, "grad_norm": 5.3378296676815555, "learning_rate": 9.987865702521844e-06, "loss": 0.2501, "step": 1117 }, { "epoch": 0.05195167286245353, "grad_norm": 7.867430025509763, "learning_rate": 9.987813251271341e-06, "loss": 0.2624, "step": 1118 }, { "epoch": 0.05199814126394052, "grad_norm": 9.228283768275405, "learning_rate": 9.987760687041617e-06, "loss": 0.2552, "step": 1119 }, { "epoch": 0.05204460966542751, "grad_norm": 4.4064018001102925, "learning_rate": 9.987708009833866e-06, "loss": 0.2762, "step": 1120 }, { "epoch": 0.0520910780669145, "grad_norm": 3.0926300116304475, "learning_rate": 9.987655219649277e-06, "loss": 0.1827, "step": 1121 }, { "epoch": 0.052137546468401484, "grad_norm": 4.776629978858264, "learning_rate": 9.987602316489049e-06, "loss": 0.3286, "step": 1122 }, { "epoch": 0.052184014869888475, "grad_norm": 5.454622624499336, "learning_rate": 9.98754930035438e-06, "loss": 0.3318, "step": 1123 }, { "epoch": 0.052230483271375465, "grad_norm": 4.403506216211925, "learning_rate": 9.98749617124647e-06, "loss": 0.2986, "step": 1124 }, { "epoch": 0.052276951672862455, "grad_norm": 6.476343569093524, "learning_rate": 9.98744292916652e-06, "loss": 0.2311, "step": 1125 }, { "epoch": 0.052323420074349446, "grad_norm": 5.289038239967672, "learning_rate": 9.98738957411574e-06, "loss": 0.2296, "step": 1126 }, { "epoch": 0.05236988847583643, "grad_norm": 5.327631873339283, "learning_rate": 9.987336106095337e-06, "loss": 0.3491, "step": 1127 }, { "epoch": 0.05241635687732342, "grad_norm": 12.99530825821171, "learning_rate": 9.987282525106523e-06, "loss": 0.3202, "step": 1128 }, { "epoch": 0.05246282527881041, "grad_norm": 2.8552040837830024, "learning_rate": 9.98722883115051e-06, "loss": 0.2195, "step": 1129 }, { "epoch": 0.0525092936802974, "grad_norm": 20.310679399248887, "learning_rate": 9.987175024228515e-06, "loss": 0.2928, "step": 1130 }, { "epoch": 0.052555762081784384, "grad_norm": 10.87639565600279, "learning_rate": 9.987121104341756e-06, "loss": 0.2383, "step": 1131 }, { "epoch": 0.052602230483271374, "grad_norm": 4.72686565328434, "learning_rate": 9.987067071491457e-06, "loss": 0.3495, "step": 1132 }, { "epoch": 0.052648698884758365, "grad_norm": 11.730518701923259, "learning_rate": 9.987012925678839e-06, "loss": 0.2736, "step": 1133 }, { "epoch": 0.052695167286245355, "grad_norm": 9.150892494768806, "learning_rate": 9.98695866690513e-06, "loss": 0.1999, "step": 1134 }, { "epoch": 0.052741635687732345, "grad_norm": 3.7944200322982153, "learning_rate": 9.986904295171556e-06, "loss": 0.1883, "step": 1135 }, { "epoch": 0.05278810408921933, "grad_norm": 7.580883675597001, "learning_rate": 9.986849810479354e-06, "loss": 0.1512, "step": 1136 }, { "epoch": 0.05283457249070632, "grad_norm": 5.605254156409237, "learning_rate": 9.986795212829754e-06, "loss": 0.152, "step": 1137 }, { "epoch": 0.05288104089219331, "grad_norm": 39.273130940635674, "learning_rate": 9.986740502223993e-06, "loss": 0.4497, "step": 1138 }, { "epoch": 0.0529275092936803, "grad_norm": 27.556771904304707, "learning_rate": 9.986685678663312e-06, "loss": 0.2291, "step": 1139 }, { "epoch": 0.05297397769516728, "grad_norm": 12.176729260823155, "learning_rate": 9.986630742148953e-06, "loss": 0.3273, "step": 1140 }, { "epoch": 0.053020446096654274, "grad_norm": 12.761228799247265, "learning_rate": 9.986575692682159e-06, "loss": 0.244, "step": 1141 }, { "epoch": 0.053066914498141264, "grad_norm": 6.150029105969377, "learning_rate": 9.986520530264177e-06, "loss": 0.2895, "step": 1142 }, { "epoch": 0.053113382899628254, "grad_norm": 4.695969618264955, "learning_rate": 9.986465254896257e-06, "loss": 0.2699, "step": 1143 }, { "epoch": 0.053159851301115245, "grad_norm": 16.682707785612642, "learning_rate": 9.98640986657965e-06, "loss": 0.3225, "step": 1144 }, { "epoch": 0.05320631970260223, "grad_norm": 6.0548032721928084, "learning_rate": 9.986354365315612e-06, "loss": 0.2895, "step": 1145 }, { "epoch": 0.05325278810408922, "grad_norm": 5.102123995260258, "learning_rate": 9.9862987511054e-06, "loss": 0.2491, "step": 1146 }, { "epoch": 0.05329925650557621, "grad_norm": 8.865751832705982, "learning_rate": 9.986243023950272e-06, "loss": 0.3474, "step": 1147 }, { "epoch": 0.0533457249070632, "grad_norm": 5.554834839045975, "learning_rate": 9.986187183851493e-06, "loss": 0.2851, "step": 1148 }, { "epoch": 0.05339219330855018, "grad_norm": 7.2417083812127405, "learning_rate": 9.986131230810326e-06, "loss": 0.2214, "step": 1149 }, { "epoch": 0.05343866171003717, "grad_norm": 5.528738637142486, "learning_rate": 9.98607516482804e-06, "loss": 0.1669, "step": 1150 }, { "epoch": 0.053485130111524164, "grad_norm": 7.422750297011825, "learning_rate": 9.986018985905901e-06, "loss": 0.2576, "step": 1151 }, { "epoch": 0.053531598513011154, "grad_norm": 4.338508803310501, "learning_rate": 9.985962694045185e-06, "loss": 0.2345, "step": 1152 }, { "epoch": 0.053578066914498144, "grad_norm": 3.0865291173257785, "learning_rate": 9.985906289247169e-06, "loss": 0.2609, "step": 1153 }, { "epoch": 0.05362453531598513, "grad_norm": 4.122697543803255, "learning_rate": 9.985849771513125e-06, "loss": 0.2065, "step": 1154 }, { "epoch": 0.05367100371747212, "grad_norm": 5.561229350552695, "learning_rate": 9.985793140844338e-06, "loss": 0.2481, "step": 1155 }, { "epoch": 0.05371747211895911, "grad_norm": 3.650568617350085, "learning_rate": 9.985736397242088e-06, "loss": 0.2213, "step": 1156 }, { "epoch": 0.0537639405204461, "grad_norm": 3.2421950345535158, "learning_rate": 9.985679540707662e-06, "loss": 0.251, "step": 1157 }, { "epoch": 0.05381040892193308, "grad_norm": 3.0269435648595633, "learning_rate": 9.985622571242347e-06, "loss": 0.1774, "step": 1158 }, { "epoch": 0.05385687732342007, "grad_norm": 5.2348855610218346, "learning_rate": 9.985565488847433e-06, "loss": 0.2845, "step": 1159 }, { "epoch": 0.05390334572490706, "grad_norm": 3.989298196763605, "learning_rate": 9.985508293524212e-06, "loss": 0.2227, "step": 1160 }, { "epoch": 0.053949814126394054, "grad_norm": 4.00295305773404, "learning_rate": 9.985450985273982e-06, "loss": 0.192, "step": 1161 }, { "epoch": 0.053996282527881044, "grad_norm": 22.76745270538593, "learning_rate": 9.98539356409804e-06, "loss": 0.2712, "step": 1162 }, { "epoch": 0.05404275092936803, "grad_norm": 3.0065808324872254, "learning_rate": 9.985336029997688e-06, "loss": 0.1712, "step": 1163 }, { "epoch": 0.05408921933085502, "grad_norm": 6.385968656535118, "learning_rate": 9.985278382974226e-06, "loss": 0.2705, "step": 1164 }, { "epoch": 0.05413568773234201, "grad_norm": 3.0923320057274686, "learning_rate": 9.985220623028964e-06, "loss": 0.2181, "step": 1165 }, { "epoch": 0.054182156133829, "grad_norm": 6.722467007142198, "learning_rate": 9.985162750163207e-06, "loss": 0.2169, "step": 1166 }, { "epoch": 0.05422862453531598, "grad_norm": 2.2361127856951977, "learning_rate": 9.985104764378266e-06, "loss": 0.1491, "step": 1167 }, { "epoch": 0.05427509293680297, "grad_norm": 56.807548090382085, "learning_rate": 9.985046665675457e-06, "loss": 0.2861, "step": 1168 }, { "epoch": 0.05432156133828996, "grad_norm": 29.354882234568613, "learning_rate": 9.984988454056093e-06, "loss": 0.1343, "step": 1169 }, { "epoch": 0.05436802973977695, "grad_norm": 3.527729518332858, "learning_rate": 9.984930129521496e-06, "loss": 0.144, "step": 1170 }, { "epoch": 0.054414498141263944, "grad_norm": 3.4371529229153497, "learning_rate": 9.984871692072983e-06, "loss": 0.1787, "step": 1171 }, { "epoch": 0.05446096654275093, "grad_norm": 3.8006449647701928, "learning_rate": 9.98481314171188e-06, "loss": 0.2241, "step": 1172 }, { "epoch": 0.05450743494423792, "grad_norm": 3.638739381253133, "learning_rate": 9.984754478439513e-06, "loss": 0.2811, "step": 1173 }, { "epoch": 0.05455390334572491, "grad_norm": 2.8199103190250963, "learning_rate": 9.98469570225721e-06, "loss": 0.2105, "step": 1174 }, { "epoch": 0.0546003717472119, "grad_norm": 3.113275727160309, "learning_rate": 9.984636813166304e-06, "loss": 0.1225, "step": 1175 }, { "epoch": 0.05464684014869888, "grad_norm": 3.428427343024679, "learning_rate": 9.984577811168129e-06, "loss": 0.2341, "step": 1176 }, { "epoch": 0.05469330855018587, "grad_norm": 8.621052892252589, "learning_rate": 9.984518696264018e-06, "loss": 0.2464, "step": 1177 }, { "epoch": 0.05473977695167286, "grad_norm": 1.9514302518328444, "learning_rate": 9.984459468455314e-06, "loss": 0.1496, "step": 1178 }, { "epoch": 0.05478624535315985, "grad_norm": 1.9729120985984585, "learning_rate": 9.984400127743357e-06, "loss": 0.2034, "step": 1179 }, { "epoch": 0.05483271375464684, "grad_norm": 4.072402433468757, "learning_rate": 9.984340674129492e-06, "loss": 0.255, "step": 1180 }, { "epoch": 0.05487918215613383, "grad_norm": 4.335432420365716, "learning_rate": 9.984281107615062e-06, "loss": 0.2413, "step": 1181 }, { "epoch": 0.05492565055762082, "grad_norm": 4.543916212361089, "learning_rate": 9.98422142820142e-06, "loss": 0.1389, "step": 1182 }, { "epoch": 0.05497211895910781, "grad_norm": 3.2810339398744737, "learning_rate": 9.984161635889917e-06, "loss": 0.165, "step": 1183 }, { "epoch": 0.0550185873605948, "grad_norm": 6.322629138183356, "learning_rate": 9.984101730681909e-06, "loss": 0.1281, "step": 1184 }, { "epoch": 0.05506505576208178, "grad_norm": 3.7317082678335294, "learning_rate": 9.984041712578749e-06, "loss": 0.1539, "step": 1185 }, { "epoch": 0.05511152416356877, "grad_norm": 5.030967451561428, "learning_rate": 9.9839815815818e-06, "loss": 0.2557, "step": 1186 }, { "epoch": 0.05515799256505576, "grad_norm": 3.3384799095429334, "learning_rate": 9.983921337692422e-06, "loss": 0.1797, "step": 1187 }, { "epoch": 0.05520446096654275, "grad_norm": 3.2954772243182324, "learning_rate": 9.98386098091198e-06, "loss": 0.1752, "step": 1188 }, { "epoch": 0.05525092936802974, "grad_norm": 3.6670673164592253, "learning_rate": 9.98380051124184e-06, "loss": 0.1874, "step": 1189 }, { "epoch": 0.055297397769516726, "grad_norm": 2.954986147369067, "learning_rate": 9.983739928683377e-06, "loss": 0.1537, "step": 1190 }, { "epoch": 0.05534386617100372, "grad_norm": 2.47767894295372, "learning_rate": 9.983679233237955e-06, "loss": 0.238, "step": 1191 }, { "epoch": 0.05539033457249071, "grad_norm": 3.2536432569520124, "learning_rate": 9.983618424906957e-06, "loss": 0.1815, "step": 1192 }, { "epoch": 0.0554368029739777, "grad_norm": 2.4255605793489043, "learning_rate": 9.983557503691755e-06, "loss": 0.2343, "step": 1193 }, { "epoch": 0.05548327137546468, "grad_norm": 3.7521062894853077, "learning_rate": 9.98349646959373e-06, "loss": 0.1682, "step": 1194 }, { "epoch": 0.05552973977695167, "grad_norm": 3.275789420666239, "learning_rate": 9.983435322614264e-06, "loss": 0.2274, "step": 1195 }, { "epoch": 0.05557620817843866, "grad_norm": 6.20687902255919, "learning_rate": 9.983374062754745e-06, "loss": 0.1373, "step": 1196 }, { "epoch": 0.05562267657992565, "grad_norm": 3.4534951397620746, "learning_rate": 9.983312690016557e-06, "loss": 0.2341, "step": 1197 }, { "epoch": 0.05566914498141264, "grad_norm": 1.5353321852098663, "learning_rate": 9.983251204401092e-06, "loss": 0.1086, "step": 1198 }, { "epoch": 0.055715613382899626, "grad_norm": 11.331172689998965, "learning_rate": 9.983189605909743e-06, "loss": 0.2359, "step": 1199 }, { "epoch": 0.055762081784386616, "grad_norm": 1.7972571589741329, "learning_rate": 9.983127894543904e-06, "loss": 0.1152, "step": 1200 }, { "epoch": 0.05580855018587361, "grad_norm": 2.1428386470302887, "learning_rate": 9.983066070304974e-06, "loss": 0.1616, "step": 1201 }, { "epoch": 0.0558550185873606, "grad_norm": 2.0926128655736402, "learning_rate": 9.983004133194354e-06, "loss": 0.2099, "step": 1202 }, { "epoch": 0.05590148698884758, "grad_norm": 5.957362110395402, "learning_rate": 9.982942083213444e-06, "loss": 0.2207, "step": 1203 }, { "epoch": 0.05594795539033457, "grad_norm": 4.746586247163681, "learning_rate": 9.982879920363652e-06, "loss": 0.301, "step": 1204 }, { "epoch": 0.05599442379182156, "grad_norm": 2.342607591413749, "learning_rate": 9.982817644646385e-06, "loss": 0.2046, "step": 1205 }, { "epoch": 0.05604089219330855, "grad_norm": 7.904865965471396, "learning_rate": 9.982755256063055e-06, "loss": 0.4215, "step": 1206 }, { "epoch": 0.05608736059479554, "grad_norm": 4.095910016574487, "learning_rate": 9.982692754615075e-06, "loss": 0.2112, "step": 1207 }, { "epoch": 0.056133828996282525, "grad_norm": 2.434403200810662, "learning_rate": 9.982630140303857e-06, "loss": 0.1423, "step": 1208 }, { "epoch": 0.056180297397769516, "grad_norm": 2.5114994463138567, "learning_rate": 9.982567413130824e-06, "loss": 0.2283, "step": 1209 }, { "epoch": 0.056226765799256506, "grad_norm": 2.5621031846731888, "learning_rate": 9.982504573097395e-06, "loss": 0.2006, "step": 1210 }, { "epoch": 0.0562732342007435, "grad_norm": 2.817256314587701, "learning_rate": 9.982441620204995e-06, "loss": 0.1518, "step": 1211 }, { "epoch": 0.05631970260223048, "grad_norm": 2.247304875639711, "learning_rate": 9.982378554455046e-06, "loss": 0.1558, "step": 1212 }, { "epoch": 0.05636617100371747, "grad_norm": 3.5975092016739687, "learning_rate": 9.982315375848978e-06, "loss": 0.1703, "step": 1213 }, { "epoch": 0.05641263940520446, "grad_norm": 3.849536516141187, "learning_rate": 9.982252084388227e-06, "loss": 0.1964, "step": 1214 }, { "epoch": 0.05645910780669145, "grad_norm": 3.8055049343408642, "learning_rate": 9.982188680074218e-06, "loss": 0.2307, "step": 1215 }, { "epoch": 0.05650557620817844, "grad_norm": 1.7283087190186974, "learning_rate": 9.982125162908396e-06, "loss": 0.1825, "step": 1216 }, { "epoch": 0.056552044609665425, "grad_norm": 1.5331554994079584, "learning_rate": 9.982061532892192e-06, "loss": 0.1411, "step": 1217 }, { "epoch": 0.056598513011152415, "grad_norm": 1.9730823695633084, "learning_rate": 9.981997790027053e-06, "loss": 0.1248, "step": 1218 }, { "epoch": 0.056644981412639406, "grad_norm": 2.997747460603835, "learning_rate": 9.98193393431442e-06, "loss": 0.2135, "step": 1219 }, { "epoch": 0.056691449814126396, "grad_norm": 3.5242444949719673, "learning_rate": 9.98186996575574e-06, "loss": 0.2512, "step": 1220 }, { "epoch": 0.05673791821561338, "grad_norm": 1.9500954740138885, "learning_rate": 9.98180588435246e-06, "loss": 0.1618, "step": 1221 }, { "epoch": 0.05678438661710037, "grad_norm": 3.0204624138637657, "learning_rate": 9.981741690106035e-06, "loss": 0.1838, "step": 1222 }, { "epoch": 0.05683085501858736, "grad_norm": 3.100497434051002, "learning_rate": 9.981677383017918e-06, "loss": 0.2029, "step": 1223 }, { "epoch": 0.05687732342007435, "grad_norm": 3.653658535755823, "learning_rate": 9.981612963089565e-06, "loss": 0.2267, "step": 1224 }, { "epoch": 0.05692379182156134, "grad_norm": 3.119637116925798, "learning_rate": 9.981548430322435e-06, "loss": 0.1648, "step": 1225 }, { "epoch": 0.056970260223048325, "grad_norm": 8.795932238851448, "learning_rate": 9.98148378471799e-06, "loss": 0.1905, "step": 1226 }, { "epoch": 0.057016728624535315, "grad_norm": 2.103452417495472, "learning_rate": 9.981419026277694e-06, "loss": 0.1797, "step": 1227 }, { "epoch": 0.057063197026022305, "grad_norm": 2.843799530464342, "learning_rate": 9.981354155003016e-06, "loss": 0.1229, "step": 1228 }, { "epoch": 0.057109665427509296, "grad_norm": 2.3535726035547766, "learning_rate": 9.981289170895421e-06, "loss": 0.1801, "step": 1229 }, { "epoch": 0.05715613382899628, "grad_norm": 3.514512081004711, "learning_rate": 9.981224073956386e-06, "loss": 0.2491, "step": 1230 }, { "epoch": 0.05720260223048327, "grad_norm": 3.929855266296105, "learning_rate": 9.98115886418738e-06, "loss": 0.1773, "step": 1231 }, { "epoch": 0.05724907063197026, "grad_norm": 4.349062883130451, "learning_rate": 9.981093541589888e-06, "loss": 0.2916, "step": 1232 }, { "epoch": 0.05729553903345725, "grad_norm": 1.413917814333972, "learning_rate": 9.981028106165381e-06, "loss": 0.1626, "step": 1233 }, { "epoch": 0.05734200743494424, "grad_norm": 4.662638230323737, "learning_rate": 9.980962557915346e-06, "loss": 0.1953, "step": 1234 }, { "epoch": 0.057388475836431224, "grad_norm": 2.9823845837937295, "learning_rate": 9.980896896841267e-06, "loss": 0.1922, "step": 1235 }, { "epoch": 0.057434944237918215, "grad_norm": 4.04101396848409, "learning_rate": 9.98083112294463e-06, "loss": 0.1746, "step": 1236 }, { "epoch": 0.057481412639405205, "grad_norm": 2.136672781641296, "learning_rate": 9.980765236226927e-06, "loss": 0.1871, "step": 1237 }, { "epoch": 0.057527881040892195, "grad_norm": 2.541392401220096, "learning_rate": 9.980699236689648e-06, "loss": 0.2044, "step": 1238 }, { "epoch": 0.05757434944237918, "grad_norm": 3.0646502202539683, "learning_rate": 9.98063312433429e-06, "loss": 0.2577, "step": 1239 }, { "epoch": 0.05762081784386617, "grad_norm": 2.929004184717667, "learning_rate": 9.98056689916235e-06, "loss": 0.2345, "step": 1240 }, { "epoch": 0.05766728624535316, "grad_norm": 1.6260756840164188, "learning_rate": 9.980500561175327e-06, "loss": 0.1568, "step": 1241 }, { "epoch": 0.05771375464684015, "grad_norm": 1.9763425787453948, "learning_rate": 9.980434110374725e-06, "loss": 0.1862, "step": 1242 }, { "epoch": 0.05776022304832714, "grad_norm": 3.5876831288805753, "learning_rate": 9.98036754676205e-06, "loss": 0.1775, "step": 1243 }, { "epoch": 0.057806691449814124, "grad_norm": 1.9389555879150875, "learning_rate": 9.980300870338805e-06, "loss": 0.1699, "step": 1244 }, { "epoch": 0.057853159851301114, "grad_norm": 2.099923796613005, "learning_rate": 9.980234081106506e-06, "loss": 0.2474, "step": 1245 }, { "epoch": 0.057899628252788105, "grad_norm": 5.059524276115075, "learning_rate": 9.980167179066663e-06, "loss": 0.2105, "step": 1246 }, { "epoch": 0.057946096654275095, "grad_norm": 2.2648647102901123, "learning_rate": 9.98010016422079e-06, "loss": 0.2075, "step": 1247 }, { "epoch": 0.05799256505576208, "grad_norm": 3.2483327469204566, "learning_rate": 9.980033036570408e-06, "loss": 0.1575, "step": 1248 }, { "epoch": 0.05803903345724907, "grad_norm": 2.2484813100772802, "learning_rate": 9.979965796117036e-06, "loss": 0.1382, "step": 1249 }, { "epoch": 0.05808550185873606, "grad_norm": 3.607255532491785, "learning_rate": 9.9798984428622e-06, "loss": 0.174, "step": 1250 }, { "epoch": 0.05813197026022305, "grad_norm": 3.0928719347692746, "learning_rate": 9.97983097680742e-06, "loss": 0.1735, "step": 1251 }, { "epoch": 0.05817843866171004, "grad_norm": 4.61184677672203, "learning_rate": 9.97976339795423e-06, "loss": 0.1382, "step": 1252 }, { "epoch": 0.05822490706319702, "grad_norm": 4.522084848687416, "learning_rate": 9.979695706304156e-06, "loss": 0.2285, "step": 1253 }, { "epoch": 0.058271375464684014, "grad_norm": 2.605921960636426, "learning_rate": 9.979627901858733e-06, "loss": 0.1821, "step": 1254 }, { "epoch": 0.058317843866171004, "grad_norm": 1.8980121674853112, "learning_rate": 9.979559984619499e-06, "loss": 0.1131, "step": 1255 }, { "epoch": 0.058364312267657995, "grad_norm": 2.877655934500002, "learning_rate": 9.979491954587988e-06, "loss": 0.2404, "step": 1256 }, { "epoch": 0.05841078066914498, "grad_norm": 3.266814556783174, "learning_rate": 9.979423811765746e-06, "loss": 0.145, "step": 1257 }, { "epoch": 0.05845724907063197, "grad_norm": 1.5347372958474517, "learning_rate": 9.979355556154311e-06, "loss": 0.0847, "step": 1258 }, { "epoch": 0.05850371747211896, "grad_norm": 2.112935410728985, "learning_rate": 9.979287187755233e-06, "loss": 0.1629, "step": 1259 }, { "epoch": 0.05855018587360595, "grad_norm": 3.27180688444499, "learning_rate": 9.97921870657006e-06, "loss": 0.2814, "step": 1260 }, { "epoch": 0.05859665427509294, "grad_norm": 2.6951093317023282, "learning_rate": 9.979150112600343e-06, "loss": 0.1935, "step": 1261 }, { "epoch": 0.05864312267657992, "grad_norm": 2.7593307914720637, "learning_rate": 9.979081405847634e-06, "loss": 0.1826, "step": 1262 }, { "epoch": 0.05868959107806691, "grad_norm": 2.151805638098466, "learning_rate": 9.979012586313494e-06, "loss": 0.1523, "step": 1263 }, { "epoch": 0.058736059479553904, "grad_norm": 2.25341360585361, "learning_rate": 9.978943653999476e-06, "loss": 0.1934, "step": 1264 }, { "epoch": 0.058782527881040894, "grad_norm": 2.04454473221903, "learning_rate": 9.978874608907145e-06, "loss": 0.2001, "step": 1265 }, { "epoch": 0.058828996282527884, "grad_norm": 4.132929370717234, "learning_rate": 9.978805451038064e-06, "loss": 0.1718, "step": 1266 }, { "epoch": 0.05887546468401487, "grad_norm": 3.6693714119683722, "learning_rate": 9.978736180393798e-06, "loss": 0.1331, "step": 1267 }, { "epoch": 0.05892193308550186, "grad_norm": 3.2885321764088546, "learning_rate": 9.978666796975918e-06, "loss": 0.1934, "step": 1268 }, { "epoch": 0.05896840148698885, "grad_norm": 2.721097105280326, "learning_rate": 9.978597300785995e-06, "loss": 0.093, "step": 1269 }, { "epoch": 0.05901486988847584, "grad_norm": 3.618468368521106, "learning_rate": 9.978527691825605e-06, "loss": 0.1635, "step": 1270 }, { "epoch": 0.05906133828996282, "grad_norm": 2.158091681812969, "learning_rate": 9.978457970096321e-06, "loss": 0.0966, "step": 1271 }, { "epoch": 0.05910780669144981, "grad_norm": 2.22052375239097, "learning_rate": 9.978388135599725e-06, "loss": 0.1109, "step": 1272 }, { "epoch": 0.0591542750929368, "grad_norm": 3.3496245539807066, "learning_rate": 9.9783181883374e-06, "loss": 0.1783, "step": 1273 }, { "epoch": 0.059200743494423794, "grad_norm": 3.7225885785321506, "learning_rate": 9.978248128310926e-06, "loss": 0.2521, "step": 1274 }, { "epoch": 0.059247211895910784, "grad_norm": 2.669669038844051, "learning_rate": 9.978177955521892e-06, "loss": 0.0974, "step": 1275 }, { "epoch": 0.05929368029739777, "grad_norm": 1.6174030523043668, "learning_rate": 9.978107669971888e-06, "loss": 0.152, "step": 1276 }, { "epoch": 0.05934014869888476, "grad_norm": 2.2346843934004035, "learning_rate": 9.978037271662507e-06, "loss": 0.1246, "step": 1277 }, { "epoch": 0.05938661710037175, "grad_norm": 5.959731967255024, "learning_rate": 9.977966760595343e-06, "loss": 0.1889, "step": 1278 }, { "epoch": 0.05943308550185874, "grad_norm": 1.8535663361411086, "learning_rate": 9.977896136771991e-06, "loss": 0.1027, "step": 1279 }, { "epoch": 0.05947955390334572, "grad_norm": 2.9957004365774518, "learning_rate": 9.977825400194055e-06, "loss": 0.1806, "step": 1280 }, { "epoch": 0.05952602230483271, "grad_norm": 1.6487258010977828, "learning_rate": 9.977754550863131e-06, "loss": 0.1336, "step": 1281 }, { "epoch": 0.0595724907063197, "grad_norm": 1.551017670745089, "learning_rate": 9.97768358878083e-06, "loss": 0.1148, "step": 1282 }, { "epoch": 0.05961895910780669, "grad_norm": 2.8611203053358354, "learning_rate": 9.977612513948756e-06, "loss": 0.1281, "step": 1283 }, { "epoch": 0.059665427509293684, "grad_norm": 3.3948095106368665, "learning_rate": 9.97754132636852e-06, "loss": 0.0929, "step": 1284 }, { "epoch": 0.05971189591078067, "grad_norm": 2.09383975512512, "learning_rate": 9.977470026041733e-06, "loss": 0.1798, "step": 1285 }, { "epoch": 0.05975836431226766, "grad_norm": 3.2120433700745683, "learning_rate": 9.977398612970014e-06, "loss": 0.1493, "step": 1286 }, { "epoch": 0.05980483271375465, "grad_norm": 2.5459317570351385, "learning_rate": 9.977327087154975e-06, "loss": 0.1584, "step": 1287 }, { "epoch": 0.05985130111524164, "grad_norm": 2.2622548416151225, "learning_rate": 9.97725544859824e-06, "loss": 0.1793, "step": 1288 }, { "epoch": 0.05989776951672862, "grad_norm": 11.252889986896461, "learning_rate": 9.97718369730143e-06, "loss": 0.2276, "step": 1289 }, { "epoch": 0.05994423791821561, "grad_norm": 3.10519345192006, "learning_rate": 9.977111833266171e-06, "loss": 0.139, "step": 1290 }, { "epoch": 0.0599907063197026, "grad_norm": 2.7730219921813744, "learning_rate": 9.977039856494091e-06, "loss": 0.1701, "step": 1291 }, { "epoch": 0.06003717472118959, "grad_norm": 2.7086778844145813, "learning_rate": 9.97696776698682e-06, "loss": 0.1402, "step": 1292 }, { "epoch": 0.06008364312267658, "grad_norm": 2.5893802039379934, "learning_rate": 9.976895564745993e-06, "loss": 0.1865, "step": 1293 }, { "epoch": 0.06013011152416357, "grad_norm": 4.084025407321682, "learning_rate": 9.97682324977324e-06, "loss": 0.2252, "step": 1294 }, { "epoch": 0.06017657992565056, "grad_norm": 3.136060646912661, "learning_rate": 9.976750822070204e-06, "loss": 0.1765, "step": 1295 }, { "epoch": 0.06022304832713755, "grad_norm": 1.8163739733153972, "learning_rate": 9.976678281638523e-06, "loss": 0.1165, "step": 1296 }, { "epoch": 0.06026951672862454, "grad_norm": 3.7048201620538848, "learning_rate": 9.976605628479843e-06, "loss": 0.154, "step": 1297 }, { "epoch": 0.06031598513011152, "grad_norm": 1.766196849887283, "learning_rate": 9.976532862595806e-06, "loss": 0.1451, "step": 1298 }, { "epoch": 0.06036245353159851, "grad_norm": 2.7320082386787496, "learning_rate": 9.976459983988063e-06, "loss": 0.1364, "step": 1299 }, { "epoch": 0.0604089219330855, "grad_norm": 2.289085778320409, "learning_rate": 9.976386992658262e-06, "loss": 0.1821, "step": 1300 }, { "epoch": 0.06045539033457249, "grad_norm": 1.4722262768599101, "learning_rate": 9.97631388860806e-06, "loss": 0.1548, "step": 1301 }, { "epoch": 0.06050185873605948, "grad_norm": 1.6161886970363788, "learning_rate": 9.97624067183911e-06, "loss": 0.1083, "step": 1302 }, { "epoch": 0.060548327137546466, "grad_norm": 3.7909827465995054, "learning_rate": 9.976167342353071e-06, "loss": 0.1508, "step": 1303 }, { "epoch": 0.06059479553903346, "grad_norm": 2.1089851643578075, "learning_rate": 9.976093900151605e-06, "loss": 0.1708, "step": 1304 }, { "epoch": 0.06064126394052045, "grad_norm": 2.473719457151189, "learning_rate": 9.976020345236374e-06, "loss": 0.127, "step": 1305 }, { "epoch": 0.06068773234200744, "grad_norm": 1.7939056274782463, "learning_rate": 9.975946677609047e-06, "loss": 0.1681, "step": 1306 }, { "epoch": 0.06073420074349442, "grad_norm": 2.7911319976618914, "learning_rate": 9.97587289727129e-06, "loss": 0.1966, "step": 1307 }, { "epoch": 0.06078066914498141, "grad_norm": 3.6745499430933886, "learning_rate": 9.975799004224774e-06, "loss": 0.1307, "step": 1308 }, { "epoch": 0.0608271375464684, "grad_norm": 2.5943201924631745, "learning_rate": 9.975724998471175e-06, "loss": 0.2131, "step": 1309 }, { "epoch": 0.06087360594795539, "grad_norm": 2.019866760815354, "learning_rate": 9.975650880012167e-06, "loss": 0.144, "step": 1310 }, { "epoch": 0.06092007434944238, "grad_norm": 3.268973956461927, "learning_rate": 9.975576648849429e-06, "loss": 0.2728, "step": 1311 }, { "epoch": 0.060966542750929366, "grad_norm": 1.284649213139499, "learning_rate": 9.975502304984643e-06, "loss": 0.0957, "step": 1312 }, { "epoch": 0.061013011152416356, "grad_norm": 2.1051709526333258, "learning_rate": 9.975427848419494e-06, "loss": 0.1407, "step": 1313 }, { "epoch": 0.06105947955390335, "grad_norm": 3.006522016364251, "learning_rate": 9.975353279155668e-06, "loss": 0.205, "step": 1314 }, { "epoch": 0.06110594795539034, "grad_norm": 2.1297290768597756, "learning_rate": 9.975278597194856e-06, "loss": 0.1477, "step": 1315 }, { "epoch": 0.06115241635687732, "grad_norm": 3.122707564519869, "learning_rate": 9.975203802538742e-06, "loss": 0.1963, "step": 1316 }, { "epoch": 0.06119888475836431, "grad_norm": 3.5967853903130127, "learning_rate": 9.97512889518903e-06, "loss": 0.2643, "step": 1317 }, { "epoch": 0.0612453531598513, "grad_norm": 5.052217348459915, "learning_rate": 9.975053875147411e-06, "loss": 0.121, "step": 1318 }, { "epoch": 0.06129182156133829, "grad_norm": 3.377977088338262, "learning_rate": 9.974978742415586e-06, "loss": 0.1836, "step": 1319 }, { "epoch": 0.06133828996282528, "grad_norm": 2.8372019268542257, "learning_rate": 9.974903496995255e-06, "loss": 0.2494, "step": 1320 }, { "epoch": 0.061384758364312265, "grad_norm": 2.908087145453302, "learning_rate": 9.974828138888124e-06, "loss": 0.1296, "step": 1321 }, { "epoch": 0.061431226765799256, "grad_norm": 2.379480149098715, "learning_rate": 9.974752668095902e-06, "loss": 0.1403, "step": 1322 }, { "epoch": 0.061477695167286246, "grad_norm": 4.427387403115998, "learning_rate": 9.974677084620294e-06, "loss": 0.1934, "step": 1323 }, { "epoch": 0.06152416356877324, "grad_norm": 2.525902737736701, "learning_rate": 9.974601388463015e-06, "loss": 0.1356, "step": 1324 }, { "epoch": 0.06157063197026022, "grad_norm": 3.7091698168856073, "learning_rate": 9.974525579625778e-06, "loss": 0.1694, "step": 1325 }, { "epoch": 0.06161710037174721, "grad_norm": 2.2931011773128764, "learning_rate": 9.974449658110302e-06, "loss": 0.1223, "step": 1326 }, { "epoch": 0.0616635687732342, "grad_norm": 4.418965774780323, "learning_rate": 9.974373623918306e-06, "loss": 0.1384, "step": 1327 }, { "epoch": 0.06171003717472119, "grad_norm": 1.773779249850325, "learning_rate": 9.974297477051511e-06, "loss": 0.0996, "step": 1328 }, { "epoch": 0.06175650557620818, "grad_norm": 1.9199511691834996, "learning_rate": 9.974221217511643e-06, "loss": 0.1055, "step": 1329 }, { "epoch": 0.061802973977695165, "grad_norm": 2.002042211907553, "learning_rate": 9.97414484530043e-06, "loss": 0.1404, "step": 1330 }, { "epoch": 0.061849442379182155, "grad_norm": 2.019536222114371, "learning_rate": 9.9740683604196e-06, "loss": 0.1606, "step": 1331 }, { "epoch": 0.061895910780669146, "grad_norm": 7.651269665839228, "learning_rate": 9.973991762870886e-06, "loss": 0.1523, "step": 1332 }, { "epoch": 0.061942379182156136, "grad_norm": 1.8548024412748998, "learning_rate": 9.973915052656026e-06, "loss": 0.2008, "step": 1333 }, { "epoch": 0.06198884758364312, "grad_norm": 2.1712783037418837, "learning_rate": 9.973838229776754e-06, "loss": 0.1834, "step": 1334 }, { "epoch": 0.06203531598513011, "grad_norm": 2.8864756361521673, "learning_rate": 9.97376129423481e-06, "loss": 0.0997, "step": 1335 }, { "epoch": 0.0620817843866171, "grad_norm": 6.424803362802397, "learning_rate": 9.97368424603194e-06, "loss": 0.1364, "step": 1336 }, { "epoch": 0.06212825278810409, "grad_norm": 7.750366284258469, "learning_rate": 9.973607085169886e-06, "loss": 0.1849, "step": 1337 }, { "epoch": 0.06217472118959108, "grad_norm": 5.747938269914111, "learning_rate": 9.973529811650395e-06, "loss": 0.0703, "step": 1338 }, { "epoch": 0.062221189591078065, "grad_norm": 3.42337377629049, "learning_rate": 9.973452425475223e-06, "loss": 0.1769, "step": 1339 }, { "epoch": 0.062267657992565055, "grad_norm": 14.08866451817147, "learning_rate": 9.973374926646117e-06, "loss": 0.1271, "step": 1340 }, { "epoch": 0.062314126394052045, "grad_norm": 6.2494792234971515, "learning_rate": 9.973297315164834e-06, "loss": 0.15, "step": 1341 }, { "epoch": 0.062360594795539036, "grad_norm": 2.9170204873169228, "learning_rate": 9.973219591033133e-06, "loss": 0.2114, "step": 1342 }, { "epoch": 0.06240706319702602, "grad_norm": 1.6745286234525356, "learning_rate": 9.973141754252775e-06, "loss": 0.1224, "step": 1343 }, { "epoch": 0.06245353159851301, "grad_norm": 2.722750760369558, "learning_rate": 9.97306380482552e-06, "loss": 0.175, "step": 1344 }, { "epoch": 0.0625, "grad_norm": 8.413498326529133, "learning_rate": 9.972985742753139e-06, "loss": 0.183, "step": 1345 }, { "epoch": 0.06254646840148699, "grad_norm": 2.919436001147968, "learning_rate": 9.972907568037392e-06, "loss": 0.159, "step": 1346 }, { "epoch": 0.06259293680297398, "grad_norm": 4.166596520434911, "learning_rate": 9.972829280680059e-06, "loss": 0.1087, "step": 1347 }, { "epoch": 0.06263940520446097, "grad_norm": 1.7219424457584698, "learning_rate": 9.972750880682909e-06, "loss": 0.0849, "step": 1348 }, { "epoch": 0.06268587360594796, "grad_norm": 1.7672769516755382, "learning_rate": 9.972672368047715e-06, "loss": 0.2011, "step": 1349 }, { "epoch": 0.06273234200743494, "grad_norm": 2.2188255894529822, "learning_rate": 9.97259374277626e-06, "loss": 0.144, "step": 1350 }, { "epoch": 0.06277881040892193, "grad_norm": 3.6587964633209733, "learning_rate": 9.972515004870325e-06, "loss": 0.1692, "step": 1351 }, { "epoch": 0.06282527881040892, "grad_norm": 1.7263635622431945, "learning_rate": 9.972436154331688e-06, "loss": 0.1265, "step": 1352 }, { "epoch": 0.06287174721189591, "grad_norm": 2.4868247852385146, "learning_rate": 9.972357191162141e-06, "loss": 0.142, "step": 1353 }, { "epoch": 0.0629182156133829, "grad_norm": 4.755074161719355, "learning_rate": 9.97227811536347e-06, "loss": 0.1038, "step": 1354 }, { "epoch": 0.06296468401486989, "grad_norm": 5.692429656277072, "learning_rate": 9.972198926937467e-06, "loss": 0.2354, "step": 1355 }, { "epoch": 0.06301115241635688, "grad_norm": 2.524775162514902, "learning_rate": 9.972119625885924e-06, "loss": 0.2174, "step": 1356 }, { "epoch": 0.06305762081784387, "grad_norm": 5.907050672116485, "learning_rate": 9.972040212210637e-06, "loss": 0.1265, "step": 1357 }, { "epoch": 0.06310408921933086, "grad_norm": 3.79260217796018, "learning_rate": 9.971960685913409e-06, "loss": 0.1472, "step": 1358 }, { "epoch": 0.06315055762081784, "grad_norm": 5.139496775428449, "learning_rate": 9.971881046996037e-06, "loss": 0.1647, "step": 1359 }, { "epoch": 0.06319702602230483, "grad_norm": 2.442051687103245, "learning_rate": 9.971801295460326e-06, "loss": 0.1234, "step": 1360 }, { "epoch": 0.06324349442379182, "grad_norm": 4.4982494352277405, "learning_rate": 9.971721431308085e-06, "loss": 0.1493, "step": 1361 }, { "epoch": 0.06328996282527881, "grad_norm": 1.8985246648124794, "learning_rate": 9.971641454541119e-06, "loss": 0.093, "step": 1362 }, { "epoch": 0.0633364312267658, "grad_norm": 6.606417642771048, "learning_rate": 9.971561365161243e-06, "loss": 0.2342, "step": 1363 }, { "epoch": 0.06338289962825279, "grad_norm": 3.6772056407752203, "learning_rate": 9.97148116317027e-06, "loss": 0.2553, "step": 1364 }, { "epoch": 0.06342936802973978, "grad_norm": 46.089160052463065, "learning_rate": 9.971400848570015e-06, "loss": 0.1474, "step": 1365 }, { "epoch": 0.06347583643122677, "grad_norm": 2.5728861678059163, "learning_rate": 9.971320421362298e-06, "loss": 0.1311, "step": 1366 }, { "epoch": 0.06352230483271376, "grad_norm": 3.0352255708153475, "learning_rate": 9.971239881548942e-06, "loss": 0.1754, "step": 1367 }, { "epoch": 0.06356877323420074, "grad_norm": 2.1101804560439343, "learning_rate": 9.97115922913177e-06, "loss": 0.1264, "step": 1368 }, { "epoch": 0.06361524163568773, "grad_norm": 1.8308459249450184, "learning_rate": 9.97107846411261e-06, "loss": 0.13, "step": 1369 }, { "epoch": 0.06366171003717472, "grad_norm": 2.198568710388905, "learning_rate": 9.97099758649329e-06, "loss": 0.1231, "step": 1370 }, { "epoch": 0.06370817843866171, "grad_norm": 3.177722237988848, "learning_rate": 9.970916596275643e-06, "loss": 0.168, "step": 1371 }, { "epoch": 0.0637546468401487, "grad_norm": 3.7778061254821855, "learning_rate": 9.970835493461505e-06, "loss": 0.3261, "step": 1372 }, { "epoch": 0.06380111524163569, "grad_norm": 2.6065568782784214, "learning_rate": 9.97075427805271e-06, "loss": 0.2078, "step": 1373 }, { "epoch": 0.06384758364312268, "grad_norm": 2.2417343104902376, "learning_rate": 9.9706729500511e-06, "loss": 0.1625, "step": 1374 }, { "epoch": 0.06389405204460967, "grad_norm": 2.271679406978466, "learning_rate": 9.970591509458514e-06, "loss": 0.1885, "step": 1375 }, { "epoch": 0.06394052044609666, "grad_norm": 2.5786554213741932, "learning_rate": 9.9705099562768e-06, "loss": 0.1075, "step": 1376 }, { "epoch": 0.06398698884758364, "grad_norm": 1.8659609651974352, "learning_rate": 9.970428290507805e-06, "loss": 0.0832, "step": 1377 }, { "epoch": 0.06403345724907063, "grad_norm": 2.153598740047861, "learning_rate": 9.970346512153377e-06, "loss": 0.1146, "step": 1378 }, { "epoch": 0.06407992565055762, "grad_norm": 2.6124392126838454, "learning_rate": 9.970264621215368e-06, "loss": 0.1428, "step": 1379 }, { "epoch": 0.06412639405204461, "grad_norm": 3.0929620996454488, "learning_rate": 9.970182617695637e-06, "loss": 0.2319, "step": 1380 }, { "epoch": 0.0641728624535316, "grad_norm": 1.76710188367996, "learning_rate": 9.970100501596038e-06, "loss": 0.1288, "step": 1381 }, { "epoch": 0.06421933085501859, "grad_norm": 1.6292534433084527, "learning_rate": 9.97001827291843e-06, "loss": 0.0796, "step": 1382 }, { "epoch": 0.06426579925650558, "grad_norm": 4.595510918193029, "learning_rate": 9.969935931664678e-06, "loss": 0.18, "step": 1383 }, { "epoch": 0.06431226765799257, "grad_norm": 1.8937871117398608, "learning_rate": 9.969853477836648e-06, "loss": 0.1142, "step": 1384 }, { "epoch": 0.06435873605947956, "grad_norm": 2.015093821922378, "learning_rate": 9.969770911436203e-06, "loss": 0.0781, "step": 1385 }, { "epoch": 0.06440520446096654, "grad_norm": 3.5224769281840684, "learning_rate": 9.96968823246522e-06, "loss": 0.1891, "step": 1386 }, { "epoch": 0.06445167286245353, "grad_norm": 3.801692759040775, "learning_rate": 9.969605440925564e-06, "loss": 0.15, "step": 1387 }, { "epoch": 0.06449814126394052, "grad_norm": 5.116870311246181, "learning_rate": 9.969522536819118e-06, "loss": 0.2222, "step": 1388 }, { "epoch": 0.06454460966542751, "grad_norm": 2.0484369792269552, "learning_rate": 9.969439520147754e-06, "loss": 0.1207, "step": 1389 }, { "epoch": 0.0645910780669145, "grad_norm": 3.0066761059163234, "learning_rate": 9.969356390913357e-06, "loss": 0.1868, "step": 1390 }, { "epoch": 0.06463754646840149, "grad_norm": 3.286063359899197, "learning_rate": 9.969273149117805e-06, "loss": 0.1762, "step": 1391 }, { "epoch": 0.06468401486988848, "grad_norm": 3.1363476570418034, "learning_rate": 9.969189794762987e-06, "loss": 0.1692, "step": 1392 }, { "epoch": 0.06473048327137547, "grad_norm": 3.3066966647923928, "learning_rate": 9.969106327850791e-06, "loss": 0.2154, "step": 1393 }, { "epoch": 0.06477695167286246, "grad_norm": 2.062244582561693, "learning_rate": 9.969022748383105e-06, "loss": 0.2153, "step": 1394 }, { "epoch": 0.06482342007434944, "grad_norm": 2.5946501475284474, "learning_rate": 9.968939056361823e-06, "loss": 0.2156, "step": 1395 }, { "epoch": 0.06486988847583643, "grad_norm": 2.2352399689123557, "learning_rate": 9.968855251788845e-06, "loss": 0.2082, "step": 1396 }, { "epoch": 0.06491635687732342, "grad_norm": 2.969171464520595, "learning_rate": 9.968771334666066e-06, "loss": 0.1567, "step": 1397 }, { "epoch": 0.06496282527881041, "grad_norm": 3.956024182030665, "learning_rate": 9.968687304995383e-06, "loss": 0.1403, "step": 1398 }, { "epoch": 0.0650092936802974, "grad_norm": 1.4421831279242405, "learning_rate": 9.968603162778705e-06, "loss": 0.1793, "step": 1399 }, { "epoch": 0.06505576208178439, "grad_norm": 2.6987243299925474, "learning_rate": 9.968518908017938e-06, "loss": 0.1729, "step": 1400 }, { "epoch": 0.06510223048327138, "grad_norm": 3.798603737841647, "learning_rate": 9.968434540714987e-06, "loss": 0.2091, "step": 1401 }, { "epoch": 0.06514869888475837, "grad_norm": 2.793891801307754, "learning_rate": 9.968350060871763e-06, "loss": 0.14, "step": 1402 }, { "epoch": 0.06519516728624536, "grad_norm": 2.0781088855967886, "learning_rate": 9.968265468490183e-06, "loss": 0.186, "step": 1403 }, { "epoch": 0.06524163568773234, "grad_norm": 1.3114272503304936, "learning_rate": 9.96818076357216e-06, "loss": 0.0883, "step": 1404 }, { "epoch": 0.06528810408921933, "grad_norm": 2.13358934709672, "learning_rate": 9.968095946119614e-06, "loss": 0.1692, "step": 1405 }, { "epoch": 0.06533457249070632, "grad_norm": 2.0324026732610796, "learning_rate": 9.968011016134467e-06, "loss": 0.152, "step": 1406 }, { "epoch": 0.0653810408921933, "grad_norm": 1.961888561672695, "learning_rate": 9.96792597361864e-06, "loss": 0.1178, "step": 1407 }, { "epoch": 0.0654275092936803, "grad_norm": 2.4526652822079233, "learning_rate": 9.967840818574062e-06, "loss": 0.1215, "step": 1408 }, { "epoch": 0.06547397769516729, "grad_norm": 3.2920574338076016, "learning_rate": 9.96775555100266e-06, "loss": 0.231, "step": 1409 }, { "epoch": 0.06552044609665428, "grad_norm": 3.819376191397615, "learning_rate": 9.967670170906369e-06, "loss": 0.2433, "step": 1410 }, { "epoch": 0.06556691449814127, "grad_norm": 2.7527132888907344, "learning_rate": 9.967584678287119e-06, "loss": 0.1267, "step": 1411 }, { "epoch": 0.06561338289962826, "grad_norm": 2.0871917549561485, "learning_rate": 9.967499073146846e-06, "loss": 0.1572, "step": 1412 }, { "epoch": 0.06565985130111524, "grad_norm": 4.849911730907106, "learning_rate": 9.967413355487491e-06, "loss": 0.2056, "step": 1413 }, { "epoch": 0.06570631970260223, "grad_norm": 1.8863680823882125, "learning_rate": 9.967327525310997e-06, "loss": 0.105, "step": 1414 }, { "epoch": 0.06575278810408922, "grad_norm": 1.815680416622233, "learning_rate": 9.967241582619306e-06, "loss": 0.147, "step": 1415 }, { "epoch": 0.0657992565055762, "grad_norm": 1.5741850983992762, "learning_rate": 9.967155527414365e-06, "loss": 0.1286, "step": 1416 }, { "epoch": 0.0658457249070632, "grad_norm": 1.6902755019593851, "learning_rate": 9.967069359698123e-06, "loss": 0.0997, "step": 1417 }, { "epoch": 0.06589219330855019, "grad_norm": 1.5457719985376757, "learning_rate": 9.966983079472534e-06, "loss": 0.1385, "step": 1418 }, { "epoch": 0.06593866171003718, "grad_norm": 1.8902246098733126, "learning_rate": 9.966896686739548e-06, "loss": 0.1211, "step": 1419 }, { "epoch": 0.06598513011152417, "grad_norm": 2.3482773289399814, "learning_rate": 9.966810181501127e-06, "loss": 0.1764, "step": 1420 }, { "epoch": 0.06603159851301116, "grad_norm": 3.4512340428155364, "learning_rate": 9.966723563759226e-06, "loss": 0.2454, "step": 1421 }, { "epoch": 0.06607806691449813, "grad_norm": 25.18102390824829, "learning_rate": 9.96663683351581e-06, "loss": 0.5044, "step": 1422 }, { "epoch": 0.06612453531598513, "grad_norm": 6.046687125440135, "learning_rate": 9.966549990772842e-06, "loss": 0.2441, "step": 1423 }, { "epoch": 0.06617100371747212, "grad_norm": 2.70236915374197, "learning_rate": 9.96646303553229e-06, "loss": 0.228, "step": 1424 }, { "epoch": 0.0662174721189591, "grad_norm": 2.701026486299467, "learning_rate": 9.966375967796122e-06, "loss": 0.2464, "step": 1425 }, { "epoch": 0.0662639405204461, "grad_norm": 2.802133730405134, "learning_rate": 9.966288787566313e-06, "loss": 0.2119, "step": 1426 }, { "epoch": 0.06631040892193309, "grad_norm": 4.95520992094801, "learning_rate": 9.966201494844836e-06, "loss": 0.3315, "step": 1427 }, { "epoch": 0.06635687732342008, "grad_norm": 2.4849116984551003, "learning_rate": 9.966114089633668e-06, "loss": 0.2343, "step": 1428 }, { "epoch": 0.06640334572490707, "grad_norm": 2.5651036158854965, "learning_rate": 9.966026571934788e-06, "loss": 0.2686, "step": 1429 }, { "epoch": 0.06644981412639406, "grad_norm": 2.195301817662217, "learning_rate": 9.96593894175018e-06, "loss": 0.2229, "step": 1430 }, { "epoch": 0.06649628252788103, "grad_norm": 1.8805170904684994, "learning_rate": 9.965851199081828e-06, "loss": 0.2075, "step": 1431 }, { "epoch": 0.06654275092936802, "grad_norm": 2.9095598326871777, "learning_rate": 9.96576334393172e-06, "loss": 0.2001, "step": 1432 }, { "epoch": 0.06658921933085502, "grad_norm": 2.078278559240246, "learning_rate": 9.965675376301846e-06, "loss": 0.1807, "step": 1433 }, { "epoch": 0.066635687732342, "grad_norm": 1.494747432187547, "learning_rate": 9.965587296194199e-06, "loss": 0.1549, "step": 1434 }, { "epoch": 0.066682156133829, "grad_norm": 2.161214544601994, "learning_rate": 9.965499103610775e-06, "loss": 0.1017, "step": 1435 }, { "epoch": 0.06672862453531599, "grad_norm": 1.9819548867533634, "learning_rate": 9.965410798553567e-06, "loss": 0.1755, "step": 1436 }, { "epoch": 0.06677509293680298, "grad_norm": 1.629342810698394, "learning_rate": 9.96532238102458e-06, "loss": 0.1313, "step": 1437 }, { "epoch": 0.06682156133828997, "grad_norm": 1.6611014219662552, "learning_rate": 9.965233851025816e-06, "loss": 0.1418, "step": 1438 }, { "epoch": 0.06686802973977696, "grad_norm": 2.709668392781272, "learning_rate": 9.965145208559277e-06, "loss": 0.193, "step": 1439 }, { "epoch": 0.06691449814126393, "grad_norm": 1.1584524438786734, "learning_rate": 9.965056453626975e-06, "loss": 0.0784, "step": 1440 }, { "epoch": 0.06696096654275092, "grad_norm": 1.6724251698110848, "learning_rate": 9.964967586230918e-06, "loss": 0.1329, "step": 1441 }, { "epoch": 0.06700743494423791, "grad_norm": 2.5859956049163664, "learning_rate": 9.964878606373121e-06, "loss": 0.1687, "step": 1442 }, { "epoch": 0.0670539033457249, "grad_norm": 2.728746843102671, "learning_rate": 9.964789514055596e-06, "loss": 0.1056, "step": 1443 }, { "epoch": 0.0671003717472119, "grad_norm": 2.546775437982862, "learning_rate": 9.964700309280367e-06, "loss": 0.2079, "step": 1444 }, { "epoch": 0.06714684014869889, "grad_norm": 1.9930384850828744, "learning_rate": 9.964610992049447e-06, "loss": 0.1576, "step": 1445 }, { "epoch": 0.06719330855018588, "grad_norm": 3.3061052014692396, "learning_rate": 9.964521562364864e-06, "loss": 0.075, "step": 1446 }, { "epoch": 0.06723977695167287, "grad_norm": 2.302922333281208, "learning_rate": 9.964432020228643e-06, "loss": 0.1577, "step": 1447 }, { "epoch": 0.06728624535315986, "grad_norm": 1.7598977322844531, "learning_rate": 9.964342365642812e-06, "loss": 0.1251, "step": 1448 }, { "epoch": 0.06733271375464683, "grad_norm": 2.2956949395521007, "learning_rate": 9.964252598609401e-06, "loss": 0.19, "step": 1449 }, { "epoch": 0.06737918215613382, "grad_norm": 1.3859681385685396, "learning_rate": 9.964162719130443e-06, "loss": 0.0646, "step": 1450 }, { "epoch": 0.06742565055762081, "grad_norm": 3.260563048387753, "learning_rate": 9.964072727207977e-06, "loss": 0.1482, "step": 1451 }, { "epoch": 0.0674721189591078, "grad_norm": 2.436372579272835, "learning_rate": 9.963982622844037e-06, "loss": 0.1974, "step": 1452 }, { "epoch": 0.0675185873605948, "grad_norm": 3.3949667185093446, "learning_rate": 9.963892406040665e-06, "loss": 0.1198, "step": 1453 }, { "epoch": 0.06756505576208179, "grad_norm": 1.5914403675880016, "learning_rate": 9.96380207679991e-06, "loss": 0.1092, "step": 1454 }, { "epoch": 0.06761152416356878, "grad_norm": 1.4407807381762037, "learning_rate": 9.963711635123809e-06, "loss": 0.1015, "step": 1455 }, { "epoch": 0.06765799256505577, "grad_norm": 3.0218689996526256, "learning_rate": 9.963621081014417e-06, "loss": 0.19, "step": 1456 }, { "epoch": 0.06770446096654276, "grad_norm": 2.700404073736944, "learning_rate": 9.963530414473784e-06, "loss": 0.2059, "step": 1457 }, { "epoch": 0.06775092936802973, "grad_norm": 2.273855722542229, "learning_rate": 9.963439635503963e-06, "loss": 0.0688, "step": 1458 }, { "epoch": 0.06779739776951672, "grad_norm": 2.4425224901139, "learning_rate": 9.96334874410701e-06, "loss": 0.1442, "step": 1459 }, { "epoch": 0.06784386617100371, "grad_norm": 2.597473441984526, "learning_rate": 9.963257740284984e-06, "loss": 0.1637, "step": 1460 }, { "epoch": 0.0678903345724907, "grad_norm": 2.6042319926213633, "learning_rate": 9.963166624039947e-06, "loss": 0.1616, "step": 1461 }, { "epoch": 0.0679368029739777, "grad_norm": 3.2392232612196796, "learning_rate": 9.963075395373963e-06, "loss": 0.2791, "step": 1462 }, { "epoch": 0.06798327137546469, "grad_norm": 2.8397480234067847, "learning_rate": 9.962984054289095e-06, "loss": 0.2472, "step": 1463 }, { "epoch": 0.06802973977695168, "grad_norm": 2.620701984300864, "learning_rate": 9.962892600787417e-06, "loss": 0.1543, "step": 1464 }, { "epoch": 0.06807620817843867, "grad_norm": 2.773232588435739, "learning_rate": 9.962801034870996e-06, "loss": 0.1853, "step": 1465 }, { "epoch": 0.06812267657992566, "grad_norm": 2.8334521121538083, "learning_rate": 9.962709356541912e-06, "loss": 0.1694, "step": 1466 }, { "epoch": 0.06816914498141263, "grad_norm": 1.6770643764766178, "learning_rate": 9.962617565802235e-06, "loss": 0.1213, "step": 1467 }, { "epoch": 0.06821561338289962, "grad_norm": 1.9803913302106748, "learning_rate": 9.962525662654048e-06, "loss": 0.1064, "step": 1468 }, { "epoch": 0.06826208178438661, "grad_norm": 1.927603338531629, "learning_rate": 9.962433647099433e-06, "loss": 0.2021, "step": 1469 }, { "epoch": 0.0683085501858736, "grad_norm": 1.2959741422787925, "learning_rate": 9.962341519140471e-06, "loss": 0.1347, "step": 1470 }, { "epoch": 0.0683550185873606, "grad_norm": 1.3051115705107192, "learning_rate": 9.962249278779253e-06, "loss": 0.1273, "step": 1471 }, { "epoch": 0.06840148698884758, "grad_norm": 2.2051599399122592, "learning_rate": 9.962156926017864e-06, "loss": 0.2097, "step": 1472 }, { "epoch": 0.06844795539033457, "grad_norm": 2.241976077035311, "learning_rate": 9.962064460858398e-06, "loss": 0.1942, "step": 1473 }, { "epoch": 0.06849442379182157, "grad_norm": 2.192122838299004, "learning_rate": 9.96197188330295e-06, "loss": 0.1119, "step": 1474 }, { "epoch": 0.06854089219330856, "grad_norm": 4.205737849122554, "learning_rate": 9.961879193353618e-06, "loss": 0.1307, "step": 1475 }, { "epoch": 0.06858736059479553, "grad_norm": 1.7073232399171203, "learning_rate": 9.961786391012497e-06, "loss": 0.1134, "step": 1476 }, { "epoch": 0.06863382899628252, "grad_norm": 1.8648003140660696, "learning_rate": 9.961693476281696e-06, "loss": 0.2146, "step": 1477 }, { "epoch": 0.06868029739776951, "grad_norm": 12.251932085770665, "learning_rate": 9.961600449163313e-06, "loss": 0.105, "step": 1478 }, { "epoch": 0.0687267657992565, "grad_norm": 2.8239000133866456, "learning_rate": 9.961507309659457e-06, "loss": 0.1707, "step": 1479 }, { "epoch": 0.0687732342007435, "grad_norm": 1.7594416571689844, "learning_rate": 9.96141405777224e-06, "loss": 0.1182, "step": 1480 }, { "epoch": 0.06881970260223048, "grad_norm": 1.4795167482549372, "learning_rate": 9.961320693503774e-06, "loss": 0.119, "step": 1481 }, { "epoch": 0.06886617100371747, "grad_norm": 3.1493363642557486, "learning_rate": 9.96122721685617e-06, "loss": 0.1115, "step": 1482 }, { "epoch": 0.06891263940520446, "grad_norm": 2.1940469370781717, "learning_rate": 9.96113362783155e-06, "loss": 0.1029, "step": 1483 }, { "epoch": 0.06895910780669146, "grad_norm": 1.450441470862169, "learning_rate": 9.961039926432032e-06, "loss": 0.0921, "step": 1484 }, { "epoch": 0.06900557620817843, "grad_norm": 2.584701559316731, "learning_rate": 9.960946112659736e-06, "loss": 0.1544, "step": 1485 }, { "epoch": 0.06905204460966542, "grad_norm": 2.584159220955798, "learning_rate": 9.96085218651679e-06, "loss": 0.2022, "step": 1486 }, { "epoch": 0.06909851301115241, "grad_norm": 2.643361940523115, "learning_rate": 9.960758148005323e-06, "loss": 0.1742, "step": 1487 }, { "epoch": 0.0691449814126394, "grad_norm": 3.481121757260837, "learning_rate": 9.96066399712746e-06, "loss": 0.1329, "step": 1488 }, { "epoch": 0.0691914498141264, "grad_norm": 1.4889394100212492, "learning_rate": 9.960569733885339e-06, "loss": 0.0949, "step": 1489 }, { "epoch": 0.06923791821561338, "grad_norm": 1.573450023161698, "learning_rate": 9.96047535828109e-06, "loss": 0.1249, "step": 1490 }, { "epoch": 0.06928438661710037, "grad_norm": 1.3646831966208242, "learning_rate": 9.960380870316855e-06, "loss": 0.113, "step": 1491 }, { "epoch": 0.06933085501858736, "grad_norm": 1.6771151911329834, "learning_rate": 9.960286269994773e-06, "loss": 0.1505, "step": 1492 }, { "epoch": 0.06937732342007435, "grad_norm": 2.0256529888411023, "learning_rate": 9.960191557316985e-06, "loss": 0.1426, "step": 1493 }, { "epoch": 0.06942379182156133, "grad_norm": 1.7185095817437175, "learning_rate": 9.960096732285638e-06, "loss": 0.1373, "step": 1494 }, { "epoch": 0.06947026022304832, "grad_norm": 1.5402935808157348, "learning_rate": 9.960001794902881e-06, "loss": 0.1327, "step": 1495 }, { "epoch": 0.06951672862453531, "grad_norm": 3.4049421735875884, "learning_rate": 9.959906745170861e-06, "loss": 0.183, "step": 1496 }, { "epoch": 0.0695631970260223, "grad_norm": 1.4231785637567358, "learning_rate": 9.959811583091734e-06, "loss": 0.1208, "step": 1497 }, { "epoch": 0.06960966542750929, "grad_norm": 6.8994515353759995, "learning_rate": 9.959716308667653e-06, "loss": 0.1971, "step": 1498 }, { "epoch": 0.06965613382899628, "grad_norm": 1.771604433692745, "learning_rate": 9.95962092190078e-06, "loss": 0.1135, "step": 1499 }, { "epoch": 0.06970260223048327, "grad_norm": 2.2309785562027087, "learning_rate": 9.959525422793272e-06, "loss": 0.1594, "step": 1500 }, { "epoch": 0.06974907063197026, "grad_norm": 13.9185131545327, "learning_rate": 9.959429811347294e-06, "loss": 0.188, "step": 1501 }, { "epoch": 0.06979553903345725, "grad_norm": 2.076586472020689, "learning_rate": 9.959334087565012e-06, "loss": 0.138, "step": 1502 }, { "epoch": 0.06984200743494423, "grad_norm": 2.56621892990018, "learning_rate": 9.959238251448591e-06, "loss": 0.1278, "step": 1503 }, { "epoch": 0.06988847583643122, "grad_norm": 1.58333414411127, "learning_rate": 9.959142303000205e-06, "loss": 0.1251, "step": 1504 }, { "epoch": 0.06993494423791821, "grad_norm": 2.1387736007433173, "learning_rate": 9.959046242222026e-06, "loss": 0.1252, "step": 1505 }, { "epoch": 0.0699814126394052, "grad_norm": 2.7909425492995763, "learning_rate": 9.95895006911623e-06, "loss": 0.1555, "step": 1506 }, { "epoch": 0.07002788104089219, "grad_norm": 3.3699296813556407, "learning_rate": 9.958853783684998e-06, "loss": 0.1514, "step": 1507 }, { "epoch": 0.07007434944237918, "grad_norm": 3.0623827411230207, "learning_rate": 9.958757385930507e-06, "loss": 0.1326, "step": 1508 }, { "epoch": 0.07012081784386617, "grad_norm": 2.2283557891348806, "learning_rate": 9.958660875854943e-06, "loss": 0.1584, "step": 1509 }, { "epoch": 0.07016728624535316, "grad_norm": 1.8390843898617475, "learning_rate": 9.958564253460489e-06, "loss": 0.132, "step": 1510 }, { "epoch": 0.07021375464684015, "grad_norm": 2.038042264058087, "learning_rate": 9.95846751874934e-06, "loss": 0.1259, "step": 1511 }, { "epoch": 0.07026022304832713, "grad_norm": 1.9369112504593429, "learning_rate": 9.95837067172368e-06, "loss": 0.1359, "step": 1512 }, { "epoch": 0.07030669144981412, "grad_norm": 3.589578722630958, "learning_rate": 9.958273712385707e-06, "loss": 0.12, "step": 1513 }, { "epoch": 0.07035315985130111, "grad_norm": 1.7197924547497607, "learning_rate": 9.958176640737614e-06, "loss": 0.1915, "step": 1514 }, { "epoch": 0.0703996282527881, "grad_norm": 3.780661597657724, "learning_rate": 9.958079456781603e-06, "loss": 0.2669, "step": 1515 }, { "epoch": 0.07044609665427509, "grad_norm": 1.839547614067626, "learning_rate": 9.957982160519875e-06, "loss": 0.1772, "step": 1516 }, { "epoch": 0.07049256505576208, "grad_norm": 1.9982893315098318, "learning_rate": 9.957884751954632e-06, "loss": 0.1125, "step": 1517 }, { "epoch": 0.07053903345724907, "grad_norm": 1.8217324231928207, "learning_rate": 9.95778723108808e-06, "loss": 0.1802, "step": 1518 }, { "epoch": 0.07058550185873606, "grad_norm": 2.5215000830626666, "learning_rate": 9.95768959792243e-06, "loss": 0.1458, "step": 1519 }, { "epoch": 0.07063197026022305, "grad_norm": 2.829633305926769, "learning_rate": 9.957591852459894e-06, "loss": 0.2256, "step": 1520 }, { "epoch": 0.07067843866171004, "grad_norm": 2.3416646611843923, "learning_rate": 9.957493994702684e-06, "loss": 0.1569, "step": 1521 }, { "epoch": 0.07072490706319702, "grad_norm": 2.5338938215236175, "learning_rate": 9.957396024653017e-06, "loss": 0.1545, "step": 1522 }, { "epoch": 0.07077137546468401, "grad_norm": 2.2298522190297296, "learning_rate": 9.957297942313114e-06, "loss": 0.1435, "step": 1523 }, { "epoch": 0.070817843866171, "grad_norm": 1.7564534388682975, "learning_rate": 9.957199747685194e-06, "loss": 0.1328, "step": 1524 }, { "epoch": 0.07086431226765799, "grad_norm": 1.748468318304615, "learning_rate": 9.957101440771481e-06, "loss": 0.192, "step": 1525 }, { "epoch": 0.07091078066914498, "grad_norm": 1.6349827341031982, "learning_rate": 9.957003021574205e-06, "loss": 0.1113, "step": 1526 }, { "epoch": 0.07095724907063197, "grad_norm": 1.418439527727253, "learning_rate": 9.956904490095593e-06, "loss": 0.1084, "step": 1527 }, { "epoch": 0.07100371747211896, "grad_norm": 2.1402374457672955, "learning_rate": 9.956805846337875e-06, "loss": 0.2018, "step": 1528 }, { "epoch": 0.07105018587360595, "grad_norm": 2.193652372928467, "learning_rate": 9.95670709030329e-06, "loss": 0.1587, "step": 1529 }, { "epoch": 0.07109665427509294, "grad_norm": 2.936004201967938, "learning_rate": 9.956608221994073e-06, "loss": 0.1891, "step": 1530 }, { "epoch": 0.07114312267657992, "grad_norm": 2.7353154130975423, "learning_rate": 9.956509241412462e-06, "loss": 0.2177, "step": 1531 }, { "epoch": 0.07118959107806691, "grad_norm": 1.5332704972150843, "learning_rate": 9.9564101485607e-06, "loss": 0.1117, "step": 1532 }, { "epoch": 0.0712360594795539, "grad_norm": 1.68181485148324, "learning_rate": 9.956310943441032e-06, "loss": 0.1207, "step": 1533 }, { "epoch": 0.07128252788104089, "grad_norm": 2.691843963027624, "learning_rate": 9.956211626055705e-06, "loss": 0.1884, "step": 1534 }, { "epoch": 0.07132899628252788, "grad_norm": 2.425215901384407, "learning_rate": 9.956112196406968e-06, "loss": 0.1615, "step": 1535 }, { "epoch": 0.07137546468401487, "grad_norm": 1.9401644054242406, "learning_rate": 9.956012654497073e-06, "loss": 0.1811, "step": 1536 }, { "epoch": 0.07142193308550186, "grad_norm": 2.0090814971516813, "learning_rate": 9.955913000328277e-06, "loss": 0.194, "step": 1537 }, { "epoch": 0.07146840148698885, "grad_norm": 2.6271534932825404, "learning_rate": 9.955813233902833e-06, "loss": 0.1187, "step": 1538 }, { "epoch": 0.07151486988847584, "grad_norm": 1.4414989260030147, "learning_rate": 9.955713355223005e-06, "loss": 0.166, "step": 1539 }, { "epoch": 0.07156133828996282, "grad_norm": 3.147694330522324, "learning_rate": 9.955613364291055e-06, "loss": 0.1612, "step": 1540 }, { "epoch": 0.07160780669144981, "grad_norm": 3.0897124166929815, "learning_rate": 9.955513261109245e-06, "loss": 0.1577, "step": 1541 }, { "epoch": 0.0716542750929368, "grad_norm": 4.095692628292289, "learning_rate": 9.955413045679843e-06, "loss": 0.1078, "step": 1542 }, { "epoch": 0.07170074349442379, "grad_norm": 2.0963466135428686, "learning_rate": 9.955312718005122e-06, "loss": 0.1714, "step": 1543 }, { "epoch": 0.07174721189591078, "grad_norm": 1.7209313857412551, "learning_rate": 9.955212278087351e-06, "loss": 0.1381, "step": 1544 }, { "epoch": 0.07179368029739777, "grad_norm": 1.782329495906389, "learning_rate": 9.955111725928807e-06, "loss": 0.1414, "step": 1545 }, { "epoch": 0.07184014869888476, "grad_norm": 2.0778911550637837, "learning_rate": 9.955011061531767e-06, "loss": 0.1102, "step": 1546 }, { "epoch": 0.07188661710037175, "grad_norm": 2.607115362294713, "learning_rate": 9.954910284898513e-06, "loss": 0.1618, "step": 1547 }, { "epoch": 0.07193308550185874, "grad_norm": 2.4343897642687073, "learning_rate": 9.954809396031325e-06, "loss": 0.1542, "step": 1548 }, { "epoch": 0.07197955390334572, "grad_norm": 2.61597825851295, "learning_rate": 9.954708394932488e-06, "loss": 0.2022, "step": 1549 }, { "epoch": 0.07202602230483271, "grad_norm": 8.684805146938993, "learning_rate": 9.954607281604294e-06, "loss": 0.1107, "step": 1550 }, { "epoch": 0.0720724907063197, "grad_norm": 1.7293466769762227, "learning_rate": 9.95450605604903e-06, "loss": 0.1127, "step": 1551 }, { "epoch": 0.07211895910780669, "grad_norm": 1.7198868779179086, "learning_rate": 9.954404718268986e-06, "loss": 0.1307, "step": 1552 }, { "epoch": 0.07216542750929368, "grad_norm": 1.3958418805961545, "learning_rate": 9.954303268266463e-06, "loss": 0.0928, "step": 1553 }, { "epoch": 0.07221189591078067, "grad_norm": 2.233938752719228, "learning_rate": 9.954201706043757e-06, "loss": 0.1095, "step": 1554 }, { "epoch": 0.07225836431226766, "grad_norm": 1.7832084043625789, "learning_rate": 9.954100031603167e-06, "loss": 0.1278, "step": 1555 }, { "epoch": 0.07230483271375465, "grad_norm": 13.68856496144824, "learning_rate": 9.953998244946998e-06, "loss": 0.1059, "step": 1556 }, { "epoch": 0.07235130111524164, "grad_norm": 3.7550380045935903, "learning_rate": 9.953896346077556e-06, "loss": 0.1845, "step": 1557 }, { "epoch": 0.07239776951672862, "grad_norm": 1.4942422870785386, "learning_rate": 9.953794334997146e-06, "loss": 0.1383, "step": 1558 }, { "epoch": 0.07244423791821561, "grad_norm": 3.4930892134621208, "learning_rate": 9.953692211708081e-06, "loss": 0.1847, "step": 1559 }, { "epoch": 0.0724907063197026, "grad_norm": 3.419014600324126, "learning_rate": 9.953589976212674e-06, "loss": 0.2814, "step": 1560 }, { "epoch": 0.07253717472118959, "grad_norm": 2.254678264708532, "learning_rate": 9.953487628513241e-06, "loss": 0.1077, "step": 1561 }, { "epoch": 0.07258364312267658, "grad_norm": 2.1848202742107294, "learning_rate": 9.953385168612099e-06, "loss": 0.1226, "step": 1562 }, { "epoch": 0.07263011152416357, "grad_norm": 2.8016804710945173, "learning_rate": 9.95328259651157e-06, "loss": 0.1279, "step": 1563 }, { "epoch": 0.07267657992565056, "grad_norm": 1.2048690897475867, "learning_rate": 9.953179912213976e-06, "loss": 0.0764, "step": 1564 }, { "epoch": 0.07272304832713755, "grad_norm": 1.8001143311990961, "learning_rate": 9.953077115721645e-06, "loss": 0.1038, "step": 1565 }, { "epoch": 0.07276951672862454, "grad_norm": 1.7428336998288583, "learning_rate": 9.952974207036902e-06, "loss": 0.1841, "step": 1566 }, { "epoch": 0.07281598513011152, "grad_norm": 7.701153130667044, "learning_rate": 9.952871186162083e-06, "loss": 0.2225, "step": 1567 }, { "epoch": 0.07286245353159851, "grad_norm": 2.1383455254392163, "learning_rate": 9.952768053099518e-06, "loss": 0.125, "step": 1568 }, { "epoch": 0.0729089219330855, "grad_norm": 3.9606444582769518, "learning_rate": 9.952664807851545e-06, "loss": 0.1005, "step": 1569 }, { "epoch": 0.07295539033457249, "grad_norm": 1.4800351237008276, "learning_rate": 9.952561450420498e-06, "loss": 0.1181, "step": 1570 }, { "epoch": 0.07300185873605948, "grad_norm": 2.2022027042450807, "learning_rate": 9.952457980808725e-06, "loss": 0.11, "step": 1571 }, { "epoch": 0.07304832713754647, "grad_norm": 2.2015393751857157, "learning_rate": 9.952354399018567e-06, "loss": 0.1436, "step": 1572 }, { "epoch": 0.07309479553903346, "grad_norm": 2.9936673837219194, "learning_rate": 9.952250705052368e-06, "loss": 0.1092, "step": 1573 }, { "epoch": 0.07314126394052045, "grad_norm": 9.365979440233614, "learning_rate": 9.952146898912479e-06, "loss": 0.1566, "step": 1574 }, { "epoch": 0.07318773234200744, "grad_norm": 2.9576535868331075, "learning_rate": 9.952042980601248e-06, "loss": 0.0999, "step": 1575 }, { "epoch": 0.07323420074349442, "grad_norm": 3.1282312768036515, "learning_rate": 9.951938950121034e-06, "loss": 0.1829, "step": 1576 }, { "epoch": 0.07328066914498141, "grad_norm": 2.9641318274917836, "learning_rate": 9.951834807474191e-06, "loss": 0.1933, "step": 1577 }, { "epoch": 0.0733271375464684, "grad_norm": 2.99522108493141, "learning_rate": 9.951730552663078e-06, "loss": 0.254, "step": 1578 }, { "epoch": 0.07337360594795539, "grad_norm": 2.340175299788912, "learning_rate": 9.951626185690055e-06, "loss": 0.2044, "step": 1579 }, { "epoch": 0.07342007434944238, "grad_norm": 2.03564431212108, "learning_rate": 9.95152170655749e-06, "loss": 0.1179, "step": 1580 }, { "epoch": 0.07346654275092937, "grad_norm": 2.069999415338539, "learning_rate": 9.951417115267744e-06, "loss": 0.146, "step": 1581 }, { "epoch": 0.07351301115241636, "grad_norm": 1.2514145142927702, "learning_rate": 9.95131241182319e-06, "loss": 0.1075, "step": 1582 }, { "epoch": 0.07355947955390335, "grad_norm": 3.058137120398173, "learning_rate": 9.9512075962262e-06, "loss": 0.2155, "step": 1583 }, { "epoch": 0.07360594795539034, "grad_norm": 1.9389290456893413, "learning_rate": 9.951102668479146e-06, "loss": 0.1571, "step": 1584 }, { "epoch": 0.07365241635687732, "grad_norm": 2.3689869223179763, "learning_rate": 9.950997628584404e-06, "loss": 0.1193, "step": 1585 }, { "epoch": 0.07369888475836431, "grad_norm": 2.1529344716415753, "learning_rate": 9.950892476544358e-06, "loss": 0.1403, "step": 1586 }, { "epoch": 0.0737453531598513, "grad_norm": 1.870931784953715, "learning_rate": 9.950787212361384e-06, "loss": 0.0934, "step": 1587 }, { "epoch": 0.07379182156133829, "grad_norm": 1.1117286211838073, "learning_rate": 9.950681836037871e-06, "loss": 0.0932, "step": 1588 }, { "epoch": 0.07383828996282528, "grad_norm": 2.495784293190887, "learning_rate": 9.950576347576203e-06, "loss": 0.1753, "step": 1589 }, { "epoch": 0.07388475836431227, "grad_norm": 1.8157442505972599, "learning_rate": 9.950470746978771e-06, "loss": 0.1379, "step": 1590 }, { "epoch": 0.07393122676579926, "grad_norm": 1.8163811183271836, "learning_rate": 9.950365034247966e-06, "loss": 0.1529, "step": 1591 }, { "epoch": 0.07397769516728625, "grad_norm": 1.4589256193589974, "learning_rate": 9.950259209386182e-06, "loss": 0.1006, "step": 1592 }, { "epoch": 0.07402416356877324, "grad_norm": 11.632217751748106, "learning_rate": 9.950153272395818e-06, "loss": 0.1653, "step": 1593 }, { "epoch": 0.07407063197026022, "grad_norm": 2.6508795517715096, "learning_rate": 9.95004722327927e-06, "loss": 0.1945, "step": 1594 }, { "epoch": 0.07411710037174721, "grad_norm": 3.085296166283109, "learning_rate": 9.949941062038946e-06, "loss": 0.1653, "step": 1595 }, { "epoch": 0.0741635687732342, "grad_norm": 1.3519627588308798, "learning_rate": 9.949834788677245e-06, "loss": 0.0933, "step": 1596 }, { "epoch": 0.07421003717472119, "grad_norm": 4.111600090836053, "learning_rate": 9.949728403196577e-06, "loss": 0.1612, "step": 1597 }, { "epoch": 0.07425650557620818, "grad_norm": 2.2141449010206204, "learning_rate": 9.949621905599351e-06, "loss": 0.1773, "step": 1598 }, { "epoch": 0.07430297397769517, "grad_norm": 1.942754335333842, "learning_rate": 9.94951529588798e-06, "loss": 0.107, "step": 1599 }, { "epoch": 0.07434944237918216, "grad_norm": 3.0772875084475118, "learning_rate": 9.949408574064878e-06, "loss": 0.1764, "step": 1600 }, { "epoch": 0.07439591078066915, "grad_norm": 1.481215516693712, "learning_rate": 9.949301740132462e-06, "loss": 0.099, "step": 1601 }, { "epoch": 0.07444237918215614, "grad_norm": 2.2493882793270608, "learning_rate": 9.949194794093154e-06, "loss": 0.0948, "step": 1602 }, { "epoch": 0.07448884758364312, "grad_norm": 1.5111940604066871, "learning_rate": 9.949087735949374e-06, "loss": 0.1439, "step": 1603 }, { "epoch": 0.07453531598513011, "grad_norm": 2.1888914383089473, "learning_rate": 9.948980565703547e-06, "loss": 0.1239, "step": 1604 }, { "epoch": 0.0745817843866171, "grad_norm": 2.621574386161162, "learning_rate": 9.948873283358103e-06, "loss": 0.1663, "step": 1605 }, { "epoch": 0.07462825278810409, "grad_norm": 2.025869495256636, "learning_rate": 9.94876588891547e-06, "loss": 0.1556, "step": 1606 }, { "epoch": 0.07467472118959108, "grad_norm": 1.52163733324169, "learning_rate": 9.948658382378083e-06, "loss": 0.0996, "step": 1607 }, { "epoch": 0.07472118959107807, "grad_norm": 2.3806329784630567, "learning_rate": 9.948550763748373e-06, "loss": 0.1375, "step": 1608 }, { "epoch": 0.07476765799256506, "grad_norm": 1.8957715830292197, "learning_rate": 9.94844303302878e-06, "loss": 0.1474, "step": 1609 }, { "epoch": 0.07481412639405205, "grad_norm": 1.4032633441479345, "learning_rate": 9.948335190221745e-06, "loss": 0.0953, "step": 1610 }, { "epoch": 0.07486059479553904, "grad_norm": 2.2212161834576682, "learning_rate": 9.948227235329708e-06, "loss": 0.1452, "step": 1611 }, { "epoch": 0.07490706319702602, "grad_norm": 2.5257474935075157, "learning_rate": 9.94811916835512e-06, "loss": 0.1885, "step": 1612 }, { "epoch": 0.074953531598513, "grad_norm": 2.390539420030108, "learning_rate": 9.94801098930042e-06, "loss": 0.2127, "step": 1613 }, { "epoch": 0.075, "grad_norm": 2.186157720488389, "learning_rate": 9.947902698168068e-06, "loss": 0.1391, "step": 1614 }, { "epoch": 0.07504646840148699, "grad_norm": 1.669514174718819, "learning_rate": 9.94779429496051e-06, "loss": 0.1416, "step": 1615 }, { "epoch": 0.07509293680297398, "grad_norm": 2.121584528579641, "learning_rate": 9.947685779680205e-06, "loss": 0.1615, "step": 1616 }, { "epoch": 0.07513940520446097, "grad_norm": 1.5404259559978852, "learning_rate": 9.947577152329609e-06, "loss": 0.1646, "step": 1617 }, { "epoch": 0.07518587360594796, "grad_norm": 0.9158918630365008, "learning_rate": 9.947468412911182e-06, "loss": 0.0703, "step": 1618 }, { "epoch": 0.07523234200743495, "grad_norm": 4.5758480939816, "learning_rate": 9.94735956142739e-06, "loss": 0.1662, "step": 1619 }, { "epoch": 0.07527881040892194, "grad_norm": 4.0409790389633455, "learning_rate": 9.947250597880696e-06, "loss": 0.2126, "step": 1620 }, { "epoch": 0.07532527881040892, "grad_norm": 1.6509196581934515, "learning_rate": 9.94714152227357e-06, "loss": 0.1321, "step": 1621 }, { "epoch": 0.0753717472118959, "grad_norm": 2.394254501098966, "learning_rate": 9.947032334608483e-06, "loss": 0.1729, "step": 1622 }, { "epoch": 0.0754182156133829, "grad_norm": 1.413851043169661, "learning_rate": 9.946923034887904e-06, "loss": 0.1551, "step": 1623 }, { "epoch": 0.07546468401486989, "grad_norm": 1.1513979004914923, "learning_rate": 9.946813623114312e-06, "loss": 0.1103, "step": 1624 }, { "epoch": 0.07551115241635688, "grad_norm": 1.6645161047250727, "learning_rate": 9.946704099290188e-06, "loss": 0.1161, "step": 1625 }, { "epoch": 0.07555762081784387, "grad_norm": 3.3515810958558236, "learning_rate": 9.946594463418006e-06, "loss": 0.1114, "step": 1626 }, { "epoch": 0.07560408921933086, "grad_norm": 2.495711432079152, "learning_rate": 9.946484715500255e-06, "loss": 0.1886, "step": 1627 }, { "epoch": 0.07565055762081785, "grad_norm": 1.677800217013468, "learning_rate": 9.94637485553942e-06, "loss": 0.1581, "step": 1628 }, { "epoch": 0.07569702602230484, "grad_norm": 1.8106193953850938, "learning_rate": 9.946264883537988e-06, "loss": 0.1091, "step": 1629 }, { "epoch": 0.07574349442379182, "grad_norm": 6.099560618225751, "learning_rate": 9.94615479949845e-06, "loss": 0.1332, "step": 1630 }, { "epoch": 0.0757899628252788, "grad_norm": 1.5823218290429306, "learning_rate": 9.9460446034233e-06, "loss": 0.1084, "step": 1631 }, { "epoch": 0.0758364312267658, "grad_norm": 2.34030958972813, "learning_rate": 9.945934295315034e-06, "loss": 0.1392, "step": 1632 }, { "epoch": 0.07588289962825279, "grad_norm": 2.0041782211587944, "learning_rate": 9.945823875176152e-06, "loss": 0.1317, "step": 1633 }, { "epoch": 0.07592936802973978, "grad_norm": 2.3770194350932754, "learning_rate": 9.945713343009154e-06, "loss": 0.1208, "step": 1634 }, { "epoch": 0.07597583643122677, "grad_norm": 1.606442613959843, "learning_rate": 9.945602698816541e-06, "loss": 0.1016, "step": 1635 }, { "epoch": 0.07602230483271376, "grad_norm": 1.5231335226539877, "learning_rate": 9.945491942600824e-06, "loss": 0.0831, "step": 1636 }, { "epoch": 0.07606877323420075, "grad_norm": 1.8174321381123897, "learning_rate": 9.945381074364508e-06, "loss": 0.1081, "step": 1637 }, { "epoch": 0.07611524163568774, "grad_norm": 2.103808969035877, "learning_rate": 9.945270094110106e-06, "loss": 0.1013, "step": 1638 }, { "epoch": 0.07616171003717472, "grad_norm": 1.568258425084078, "learning_rate": 9.945159001840132e-06, "loss": 0.1269, "step": 1639 }, { "epoch": 0.0762081784386617, "grad_norm": 2.13384003172337, "learning_rate": 9.945047797557103e-06, "loss": 0.1904, "step": 1640 }, { "epoch": 0.0762546468401487, "grad_norm": 3.9357504670773285, "learning_rate": 9.944936481263536e-06, "loss": 0.0924, "step": 1641 }, { "epoch": 0.07630111524163569, "grad_norm": 5.724275883224985, "learning_rate": 9.944825052961953e-06, "loss": 0.1879, "step": 1642 }, { "epoch": 0.07634758364312268, "grad_norm": 1.9550828401628113, "learning_rate": 9.944713512654877e-06, "loss": 0.1425, "step": 1643 }, { "epoch": 0.07639405204460967, "grad_norm": 3.15076177232949, "learning_rate": 9.944601860344838e-06, "loss": 0.1328, "step": 1644 }, { "epoch": 0.07644052044609666, "grad_norm": 1.86546007334719, "learning_rate": 9.944490096034361e-06, "loss": 0.1185, "step": 1645 }, { "epoch": 0.07648698884758365, "grad_norm": 2.0433257023570266, "learning_rate": 9.944378219725982e-06, "loss": 0.1454, "step": 1646 }, { "epoch": 0.07653345724907064, "grad_norm": 3.045840335079125, "learning_rate": 9.944266231422228e-06, "loss": 0.1458, "step": 1647 }, { "epoch": 0.07657992565055761, "grad_norm": 1.4307315637551679, "learning_rate": 9.944154131125643e-06, "loss": 0.0752, "step": 1648 }, { "epoch": 0.0766263940520446, "grad_norm": 2.087777591719846, "learning_rate": 9.944041918838763e-06, "loss": 0.1153, "step": 1649 }, { "epoch": 0.0766728624535316, "grad_norm": 1.58338720166235, "learning_rate": 9.943929594564131e-06, "loss": 0.1218, "step": 1650 }, { "epoch": 0.07671933085501859, "grad_norm": 2.262954368748188, "learning_rate": 9.943817158304289e-06, "loss": 0.1073, "step": 1651 }, { "epoch": 0.07676579925650558, "grad_norm": 2.7299203014328346, "learning_rate": 9.943704610061786e-06, "loss": 0.1528, "step": 1652 }, { "epoch": 0.07681226765799257, "grad_norm": 2.176664380437177, "learning_rate": 9.943591949839169e-06, "loss": 0.1766, "step": 1653 }, { "epoch": 0.07685873605947956, "grad_norm": 1.8445673336980757, "learning_rate": 9.943479177638992e-06, "loss": 0.1517, "step": 1654 }, { "epoch": 0.07690520446096655, "grad_norm": 2.846998204639872, "learning_rate": 9.943366293463808e-06, "loss": 0.0996, "step": 1655 }, { "epoch": 0.07695167286245354, "grad_norm": 2.4880078521446407, "learning_rate": 9.943253297316176e-06, "loss": 0.1117, "step": 1656 }, { "epoch": 0.07699814126394051, "grad_norm": 2.4739017014869886, "learning_rate": 9.943140189198653e-06, "loss": 0.1805, "step": 1657 }, { "epoch": 0.0770446096654275, "grad_norm": 2.954221838459905, "learning_rate": 9.943026969113803e-06, "loss": 0.1322, "step": 1658 }, { "epoch": 0.0770910780669145, "grad_norm": 1.763251438488054, "learning_rate": 9.94291363706419e-06, "loss": 0.1166, "step": 1659 }, { "epoch": 0.07713754646840149, "grad_norm": 1.6339579152794956, "learning_rate": 9.94280019305238e-06, "loss": 0.087, "step": 1660 }, { "epoch": 0.07718401486988848, "grad_norm": 2.67943667932037, "learning_rate": 9.942686637080944e-06, "loss": 0.2441, "step": 1661 }, { "epoch": 0.07723048327137547, "grad_norm": 2.982843885061948, "learning_rate": 9.942572969152452e-06, "loss": 0.1299, "step": 1662 }, { "epoch": 0.07727695167286246, "grad_norm": 2.926997257002127, "learning_rate": 9.942459189269483e-06, "loss": 0.1412, "step": 1663 }, { "epoch": 0.07732342007434945, "grad_norm": 1.341749826415154, "learning_rate": 9.942345297434611e-06, "loss": 0.1049, "step": 1664 }, { "epoch": 0.07736988847583644, "grad_norm": 1.6892914980958573, "learning_rate": 9.942231293650415e-06, "loss": 0.1133, "step": 1665 }, { "epoch": 0.07741635687732341, "grad_norm": 6.947686496579809, "learning_rate": 9.942117177919476e-06, "loss": 0.1553, "step": 1666 }, { "epoch": 0.0774628252788104, "grad_norm": 1.899260935149298, "learning_rate": 9.942002950244386e-06, "loss": 0.1074, "step": 1667 }, { "epoch": 0.0775092936802974, "grad_norm": 1.3968446443542257, "learning_rate": 9.941888610627726e-06, "loss": 0.105, "step": 1668 }, { "epoch": 0.07755576208178439, "grad_norm": 2.660357633661029, "learning_rate": 9.941774159072089e-06, "loss": 0.1971, "step": 1669 }, { "epoch": 0.07760223048327138, "grad_norm": 2.9879999093912954, "learning_rate": 9.941659595580064e-06, "loss": 0.1561, "step": 1670 }, { "epoch": 0.07764869888475837, "grad_norm": 2.1997111851020876, "learning_rate": 9.941544920154249e-06, "loss": 0.1099, "step": 1671 }, { "epoch": 0.07769516728624536, "grad_norm": 1.0969292407976954, "learning_rate": 9.94143013279724e-06, "loss": 0.1117, "step": 1672 }, { "epoch": 0.07774163568773235, "grad_norm": 2.7191998735258496, "learning_rate": 9.941315233511639e-06, "loss": 0.1582, "step": 1673 }, { "epoch": 0.07778810408921934, "grad_norm": 2.2953212567630885, "learning_rate": 9.941200222300049e-06, "loss": 0.1153, "step": 1674 }, { "epoch": 0.07783457249070631, "grad_norm": 2.080613458122971, "learning_rate": 9.94108509916507e-06, "loss": 0.1808, "step": 1675 }, { "epoch": 0.0778810408921933, "grad_norm": 2.055340097494846, "learning_rate": 9.940969864109315e-06, "loss": 0.1177, "step": 1676 }, { "epoch": 0.0779275092936803, "grad_norm": 2.452347644714282, "learning_rate": 9.940854517135392e-06, "loss": 0.2096, "step": 1677 }, { "epoch": 0.07797397769516728, "grad_norm": 2.997914829790539, "learning_rate": 9.940739058245914e-06, "loss": 0.1602, "step": 1678 }, { "epoch": 0.07802044609665428, "grad_norm": 2.2730205550788414, "learning_rate": 9.940623487443494e-06, "loss": 0.1416, "step": 1679 }, { "epoch": 0.07806691449814127, "grad_norm": 2.416868162626255, "learning_rate": 9.940507804730756e-06, "loss": 0.144, "step": 1680 }, { "epoch": 0.07811338289962826, "grad_norm": 2.7013456801886604, "learning_rate": 9.940392010110314e-06, "loss": 0.1224, "step": 1681 }, { "epoch": 0.07815985130111525, "grad_norm": 1.4922114427425828, "learning_rate": 9.940276103584794e-06, "loss": 0.1442, "step": 1682 }, { "epoch": 0.07820631970260224, "grad_norm": 1.4291702138859654, "learning_rate": 9.94016008515682e-06, "loss": 0.1176, "step": 1683 }, { "epoch": 0.07825278810408921, "grad_norm": 2.011316915258951, "learning_rate": 9.940043954829022e-06, "loss": 0.1719, "step": 1684 }, { "epoch": 0.0782992565055762, "grad_norm": 1.5438810831528795, "learning_rate": 9.93992771260403e-06, "loss": 0.0889, "step": 1685 }, { "epoch": 0.0783457249070632, "grad_norm": 2.9298521562338204, "learning_rate": 9.939811358484473e-06, "loss": 0.0662, "step": 1686 }, { "epoch": 0.07839219330855018, "grad_norm": 1.1534250623471936, "learning_rate": 9.939694892472991e-06, "loss": 0.1191, "step": 1687 }, { "epoch": 0.07843866171003717, "grad_norm": 1.5516137924072493, "learning_rate": 9.939578314572223e-06, "loss": 0.1197, "step": 1688 }, { "epoch": 0.07848513011152417, "grad_norm": 2.20117548686544, "learning_rate": 9.939461624784804e-06, "loss": 0.157, "step": 1689 }, { "epoch": 0.07853159851301116, "grad_norm": 1.7273304289458267, "learning_rate": 9.93934482311338e-06, "loss": 0.163, "step": 1690 }, { "epoch": 0.07857806691449815, "grad_norm": 2.2412992856860745, "learning_rate": 9.939227909560601e-06, "loss": 0.142, "step": 1691 }, { "epoch": 0.07862453531598514, "grad_norm": 2.171339472214295, "learning_rate": 9.93911088412911e-06, "loss": 0.1805, "step": 1692 }, { "epoch": 0.07867100371747211, "grad_norm": 1.6497102963931125, "learning_rate": 9.938993746821558e-06, "loss": 0.1217, "step": 1693 }, { "epoch": 0.0787174721189591, "grad_norm": 0.966337367617216, "learning_rate": 9.9388764976406e-06, "loss": 0.0635, "step": 1694 }, { "epoch": 0.0787639405204461, "grad_norm": 2.19198784266524, "learning_rate": 9.93875913658889e-06, "loss": 0.1383, "step": 1695 }, { "epoch": 0.07881040892193308, "grad_norm": 2.7966056293248096, "learning_rate": 9.93864166366909e-06, "loss": 0.1554, "step": 1696 }, { "epoch": 0.07885687732342007, "grad_norm": 1.3159706864571343, "learning_rate": 9.938524078883857e-06, "loss": 0.0984, "step": 1697 }, { "epoch": 0.07890334572490706, "grad_norm": 1.9418189885653268, "learning_rate": 9.938406382235856e-06, "loss": 0.1679, "step": 1698 }, { "epoch": 0.07894981412639406, "grad_norm": 2.2803201638878043, "learning_rate": 9.938288573727755e-06, "loss": 0.0983, "step": 1699 }, { "epoch": 0.07899628252788105, "grad_norm": 1.3156908522780575, "learning_rate": 9.938170653362219e-06, "loss": 0.1156, "step": 1700 }, { "epoch": 0.07904275092936804, "grad_norm": 2.192676937041939, "learning_rate": 9.938052621141918e-06, "loss": 0.1045, "step": 1701 }, { "epoch": 0.07908921933085501, "grad_norm": 1.8073671017730877, "learning_rate": 9.93793447706953e-06, "loss": 0.0977, "step": 1702 }, { "epoch": 0.079135687732342, "grad_norm": 1.650510355996378, "learning_rate": 9.93781622114773e-06, "loss": 0.1182, "step": 1703 }, { "epoch": 0.079182156133829, "grad_norm": 1.314985458473536, "learning_rate": 9.937697853379194e-06, "loss": 0.1264, "step": 1704 }, { "epoch": 0.07922862453531598, "grad_norm": 1.5902262467658692, "learning_rate": 9.937579373766605e-06, "loss": 0.1709, "step": 1705 }, { "epoch": 0.07927509293680297, "grad_norm": 2.1093892472762836, "learning_rate": 9.937460782312647e-06, "loss": 0.1553, "step": 1706 }, { "epoch": 0.07932156133828996, "grad_norm": 1.6995678319063978, "learning_rate": 9.937342079020006e-06, "loss": 0.1064, "step": 1707 }, { "epoch": 0.07936802973977695, "grad_norm": 1.4062952155972617, "learning_rate": 9.93722326389137e-06, "loss": 0.1126, "step": 1708 }, { "epoch": 0.07941449814126395, "grad_norm": 1.401937996764244, "learning_rate": 9.937104336929431e-06, "loss": 0.1253, "step": 1709 }, { "epoch": 0.07946096654275094, "grad_norm": 3.4987096178730264, "learning_rate": 9.93698529813688e-06, "loss": 0.2478, "step": 1710 }, { "epoch": 0.07950743494423791, "grad_norm": 1.0708382772236311, "learning_rate": 9.936866147516419e-06, "loss": 0.0809, "step": 1711 }, { "epoch": 0.0795539033457249, "grad_norm": 1.5106281302916948, "learning_rate": 9.936746885070742e-06, "loss": 0.0903, "step": 1712 }, { "epoch": 0.07960037174721189, "grad_norm": 2.4083752028419667, "learning_rate": 9.936627510802554e-06, "loss": 0.19, "step": 1713 }, { "epoch": 0.07964684014869888, "grad_norm": 1.2405271298463407, "learning_rate": 9.936508024714557e-06, "loss": 0.0937, "step": 1714 }, { "epoch": 0.07969330855018587, "grad_norm": 2.298354992158597, "learning_rate": 9.936388426809457e-06, "loss": 0.2119, "step": 1715 }, { "epoch": 0.07973977695167286, "grad_norm": 1.9328777093453888, "learning_rate": 9.936268717089962e-06, "loss": 0.1874, "step": 1716 }, { "epoch": 0.07978624535315985, "grad_norm": 1.7874962384682584, "learning_rate": 9.936148895558788e-06, "loss": 0.1208, "step": 1717 }, { "epoch": 0.07983271375464684, "grad_norm": 1.5480957593070304, "learning_rate": 9.936028962218645e-06, "loss": 0.1143, "step": 1718 }, { "epoch": 0.07987918215613383, "grad_norm": 2.391455737737542, "learning_rate": 9.935908917072253e-06, "loss": 0.1225, "step": 1719 }, { "epoch": 0.07992565055762081, "grad_norm": 2.582882743927252, "learning_rate": 9.935788760122326e-06, "loss": 0.1567, "step": 1720 }, { "epoch": 0.0799721189591078, "grad_norm": 1.6388056973632175, "learning_rate": 9.93566849137159e-06, "loss": 0.1365, "step": 1721 }, { "epoch": 0.08001858736059479, "grad_norm": 1.76058759308979, "learning_rate": 9.935548110822768e-06, "loss": 0.1189, "step": 1722 }, { "epoch": 0.08006505576208178, "grad_norm": 2.0878384554756804, "learning_rate": 9.935427618478586e-06, "loss": 0.1608, "step": 1723 }, { "epoch": 0.08011152416356877, "grad_norm": 1.926638122343311, "learning_rate": 9.935307014341776e-06, "loss": 0.1072, "step": 1724 }, { "epoch": 0.08015799256505576, "grad_norm": 2.384385281720817, "learning_rate": 9.935186298415065e-06, "loss": 0.1763, "step": 1725 }, { "epoch": 0.08020446096654275, "grad_norm": 2.6006663235407528, "learning_rate": 9.935065470701191e-06, "loss": 0.1702, "step": 1726 }, { "epoch": 0.08025092936802974, "grad_norm": 2.4248268607293153, "learning_rate": 9.934944531202892e-06, "loss": 0.1034, "step": 1727 }, { "epoch": 0.08029739776951673, "grad_norm": 2.042823794890811, "learning_rate": 9.934823479922905e-06, "loss": 0.206, "step": 1728 }, { "epoch": 0.08034386617100371, "grad_norm": 1.7235599209656416, "learning_rate": 9.93470231686397e-06, "loss": 0.1216, "step": 1729 }, { "epoch": 0.0803903345724907, "grad_norm": 2.5196914580162084, "learning_rate": 9.934581042028834e-06, "loss": 0.1881, "step": 1730 }, { "epoch": 0.08043680297397769, "grad_norm": 1.2328754388482905, "learning_rate": 9.934459655420246e-06, "loss": 0.0908, "step": 1731 }, { "epoch": 0.08048327137546468, "grad_norm": 1.5884274406041576, "learning_rate": 9.934338157040953e-06, "loss": 0.163, "step": 1732 }, { "epoch": 0.08052973977695167, "grad_norm": 1.8742763348517786, "learning_rate": 9.934216546893706e-06, "loss": 0.1114, "step": 1733 }, { "epoch": 0.08057620817843866, "grad_norm": 2.7261466091962014, "learning_rate": 9.934094824981263e-06, "loss": 0.1708, "step": 1734 }, { "epoch": 0.08062267657992565, "grad_norm": 1.2844263051936142, "learning_rate": 9.933972991306377e-06, "loss": 0.1017, "step": 1735 }, { "epoch": 0.08066914498141264, "grad_norm": 1.9974351287269214, "learning_rate": 9.933851045871812e-06, "loss": 0.1797, "step": 1736 }, { "epoch": 0.08071561338289963, "grad_norm": 1.8626233634043687, "learning_rate": 9.933728988680327e-06, "loss": 0.125, "step": 1737 }, { "epoch": 0.08076208178438661, "grad_norm": 1.6131667979099888, "learning_rate": 9.933606819734689e-06, "loss": 0.1075, "step": 1738 }, { "epoch": 0.0808085501858736, "grad_norm": 1.3886228393760887, "learning_rate": 9.933484539037661e-06, "loss": 0.1116, "step": 1739 }, { "epoch": 0.08085501858736059, "grad_norm": 2.180124880420894, "learning_rate": 9.933362146592018e-06, "loss": 0.2126, "step": 1740 }, { "epoch": 0.08090148698884758, "grad_norm": 1.6574998839850223, "learning_rate": 9.93323964240053e-06, "loss": 0.1195, "step": 1741 }, { "epoch": 0.08094795539033457, "grad_norm": 1.6210604378146332, "learning_rate": 9.933117026465971e-06, "loss": 0.1372, "step": 1742 }, { "epoch": 0.08099442379182156, "grad_norm": 1.7022721497305793, "learning_rate": 9.932994298791121e-06, "loss": 0.1946, "step": 1743 }, { "epoch": 0.08104089219330855, "grad_norm": 2.4275668096023635, "learning_rate": 9.932871459378756e-06, "loss": 0.1518, "step": 1744 }, { "epoch": 0.08108736059479554, "grad_norm": 1.9640475276443066, "learning_rate": 9.932748508231663e-06, "loss": 0.0916, "step": 1745 }, { "epoch": 0.08113382899628253, "grad_norm": 1.8548212514675846, "learning_rate": 9.932625445352624e-06, "loss": 0.1232, "step": 1746 }, { "epoch": 0.08118029739776951, "grad_norm": 1.345419699120516, "learning_rate": 9.932502270744426e-06, "loss": 0.0625, "step": 1747 }, { "epoch": 0.0812267657992565, "grad_norm": 3.2428196022614095, "learning_rate": 9.932378984409861e-06, "loss": 0.1978, "step": 1748 }, { "epoch": 0.08127323420074349, "grad_norm": 1.6833352588382833, "learning_rate": 9.93225558635172e-06, "loss": 0.0862, "step": 1749 }, { "epoch": 0.08131970260223048, "grad_norm": 1.8932398643967383, "learning_rate": 9.9321320765728e-06, "loss": 0.088, "step": 1750 }, { "epoch": 0.08136617100371747, "grad_norm": 1.6265083568033833, "learning_rate": 9.932008455075897e-06, "loss": 0.0442, "step": 1751 }, { "epoch": 0.08141263940520446, "grad_norm": 2.8727269243544766, "learning_rate": 9.931884721863812e-06, "loss": 0.1293, "step": 1752 }, { "epoch": 0.08145910780669145, "grad_norm": 2.874179779405711, "learning_rate": 9.931760876939347e-06, "loss": 0.1305, "step": 1753 }, { "epoch": 0.08150557620817844, "grad_norm": 3.143233586546413, "learning_rate": 9.931636920305307e-06, "loss": 0.1843, "step": 1754 }, { "epoch": 0.08155204460966543, "grad_norm": 2.8844798791240143, "learning_rate": 9.931512851964503e-06, "loss": 0.1855, "step": 1755 }, { "epoch": 0.08159851301115241, "grad_norm": 4.423735029494382, "learning_rate": 9.93138867191974e-06, "loss": 0.1478, "step": 1756 }, { "epoch": 0.0816449814126394, "grad_norm": 2.936086253606686, "learning_rate": 9.931264380173832e-06, "loss": 0.1301, "step": 1757 }, { "epoch": 0.08169144981412639, "grad_norm": 1.9476714177851346, "learning_rate": 9.9311399767296e-06, "loss": 0.1264, "step": 1758 }, { "epoch": 0.08173791821561338, "grad_norm": 1.2357760925628396, "learning_rate": 9.931015461589855e-06, "loss": 0.1119, "step": 1759 }, { "epoch": 0.08178438661710037, "grad_norm": 1.6592624931971323, "learning_rate": 9.93089083475742e-06, "loss": 0.0935, "step": 1760 }, { "epoch": 0.08183085501858736, "grad_norm": 2.8013539433331824, "learning_rate": 9.930766096235118e-06, "loss": 0.1827, "step": 1761 }, { "epoch": 0.08187732342007435, "grad_norm": 2.1947012221120716, "learning_rate": 9.930641246025775e-06, "loss": 0.1431, "step": 1762 }, { "epoch": 0.08192379182156134, "grad_norm": 5.7217800253616575, "learning_rate": 9.930516284132216e-06, "loss": 0.1989, "step": 1763 }, { "epoch": 0.08197026022304833, "grad_norm": 1.0646292679017868, "learning_rate": 9.930391210557277e-06, "loss": 0.1132, "step": 1764 }, { "epoch": 0.08201672862453531, "grad_norm": 1.2452778014230437, "learning_rate": 9.930266025303787e-06, "loss": 0.0724, "step": 1765 }, { "epoch": 0.0820631970260223, "grad_norm": 1.4339821044694292, "learning_rate": 9.930140728374583e-06, "loss": 0.1132, "step": 1766 }, { "epoch": 0.08210966542750929, "grad_norm": 1.7443214374252842, "learning_rate": 9.9300153197725e-06, "loss": 0.1126, "step": 1767 }, { "epoch": 0.08215613382899628, "grad_norm": 1.9058266607683965, "learning_rate": 9.929889799500383e-06, "loss": 0.1469, "step": 1768 }, { "epoch": 0.08220260223048327, "grad_norm": 1.6795176753323666, "learning_rate": 9.929764167561075e-06, "loss": 0.0751, "step": 1769 }, { "epoch": 0.08224907063197026, "grad_norm": 2.1136576342678115, "learning_rate": 9.929638423957418e-06, "loss": 0.1339, "step": 1770 }, { "epoch": 0.08229553903345725, "grad_norm": 1.917945998068925, "learning_rate": 9.929512568692262e-06, "loss": 0.1177, "step": 1771 }, { "epoch": 0.08234200743494424, "grad_norm": 3.551339162579776, "learning_rate": 9.92938660176846e-06, "loss": 0.2136, "step": 1772 }, { "epoch": 0.08238847583643123, "grad_norm": 2.885677500255099, "learning_rate": 9.929260523188862e-06, "loss": 0.2288, "step": 1773 }, { "epoch": 0.08243494423791821, "grad_norm": 2.014996961766141, "learning_rate": 9.929134332956326e-06, "loss": 0.1357, "step": 1774 }, { "epoch": 0.0824814126394052, "grad_norm": 1.289920833478595, "learning_rate": 9.92900803107371e-06, "loss": 0.1445, "step": 1775 }, { "epoch": 0.08252788104089219, "grad_norm": 1.3864681489941397, "learning_rate": 9.928881617543874e-06, "loss": 0.1466, "step": 1776 }, { "epoch": 0.08257434944237918, "grad_norm": 2.734450448865135, "learning_rate": 9.92875509236968e-06, "loss": 0.1323, "step": 1777 }, { "epoch": 0.08262081784386617, "grad_norm": 1.451506251076104, "learning_rate": 9.928628455553997e-06, "loss": 0.1589, "step": 1778 }, { "epoch": 0.08266728624535316, "grad_norm": 2.263017737484366, "learning_rate": 9.928501707099693e-06, "loss": 0.1066, "step": 1779 }, { "epoch": 0.08271375464684015, "grad_norm": 2.634037537968062, "learning_rate": 9.928374847009639e-06, "loss": 0.1313, "step": 1780 }, { "epoch": 0.08276022304832714, "grad_norm": 1.5228134835728244, "learning_rate": 9.928247875286705e-06, "loss": 0.0852, "step": 1781 }, { "epoch": 0.08280669144981413, "grad_norm": 2.2753453150592433, "learning_rate": 9.928120791933771e-06, "loss": 0.1775, "step": 1782 }, { "epoch": 0.08285315985130111, "grad_norm": 1.1051801636373266, "learning_rate": 9.927993596953715e-06, "loss": 0.0816, "step": 1783 }, { "epoch": 0.0828996282527881, "grad_norm": 1.5632323239629387, "learning_rate": 9.927866290349417e-06, "loss": 0.1181, "step": 1784 }, { "epoch": 0.08294609665427509, "grad_norm": 1.4526902335481322, "learning_rate": 9.92773887212376e-06, "loss": 0.0719, "step": 1785 }, { "epoch": 0.08299256505576208, "grad_norm": 1.5230587496989052, "learning_rate": 9.927611342279632e-06, "loss": 0.0991, "step": 1786 }, { "epoch": 0.08303903345724907, "grad_norm": 1.9721256665336908, "learning_rate": 9.92748370081992e-06, "loss": 0.1525, "step": 1787 }, { "epoch": 0.08308550185873606, "grad_norm": 1.9420333010303263, "learning_rate": 9.927355947747517e-06, "loss": 0.086, "step": 1788 }, { "epoch": 0.08313197026022305, "grad_norm": 1.7237324735784938, "learning_rate": 9.927228083065316e-06, "loss": 0.1155, "step": 1789 }, { "epoch": 0.08317843866171004, "grad_norm": 1.3183204631909207, "learning_rate": 9.927100106776213e-06, "loss": 0.1092, "step": 1790 }, { "epoch": 0.08322490706319703, "grad_norm": 1.7727151132097116, "learning_rate": 9.926972018883107e-06, "loss": 0.1468, "step": 1791 }, { "epoch": 0.08327137546468401, "grad_norm": 3.557254843163329, "learning_rate": 9.9268438193889e-06, "loss": 0.1936, "step": 1792 }, { "epoch": 0.083317843866171, "grad_norm": 1.858371650574894, "learning_rate": 9.926715508296495e-06, "loss": 0.1348, "step": 1793 }, { "epoch": 0.08336431226765799, "grad_norm": 2.881746069826127, "learning_rate": 9.926587085608797e-06, "loss": 0.1213, "step": 1794 }, { "epoch": 0.08341078066914498, "grad_norm": 1.1279026926634454, "learning_rate": 9.926458551328716e-06, "loss": 0.083, "step": 1795 }, { "epoch": 0.08345724907063197, "grad_norm": 2.5904411649404757, "learning_rate": 9.926329905459165e-06, "loss": 0.1371, "step": 1796 }, { "epoch": 0.08350371747211896, "grad_norm": 2.475013016800038, "learning_rate": 9.926201148003056e-06, "loss": 0.2044, "step": 1797 }, { "epoch": 0.08355018587360595, "grad_norm": 3.7075246195094134, "learning_rate": 9.926072278963306e-06, "loss": 0.1404, "step": 1798 }, { "epoch": 0.08359665427509294, "grad_norm": 1.9561204923178936, "learning_rate": 9.925943298342836e-06, "loss": 0.1427, "step": 1799 }, { "epoch": 0.08364312267657993, "grad_norm": 2.4471280329713014, "learning_rate": 9.925814206144566e-06, "loss": 0.1182, "step": 1800 }, { "epoch": 0.08368959107806691, "grad_norm": 1.3615844903951173, "learning_rate": 9.925685002371417e-06, "loss": 0.1541, "step": 1801 }, { "epoch": 0.0837360594795539, "grad_norm": 1.9628474725188758, "learning_rate": 9.92555568702632e-06, "loss": 0.1128, "step": 1802 }, { "epoch": 0.08378252788104089, "grad_norm": 1.5339887772729308, "learning_rate": 9.925426260112203e-06, "loss": 0.1162, "step": 1803 }, { "epoch": 0.08382899628252788, "grad_norm": 1.5425329677349404, "learning_rate": 9.925296721631998e-06, "loss": 0.1428, "step": 1804 }, { "epoch": 0.08387546468401487, "grad_norm": 1.1372309927248894, "learning_rate": 9.925167071588637e-06, "loss": 0.1244, "step": 1805 }, { "epoch": 0.08392193308550186, "grad_norm": 4.702117193656274, "learning_rate": 9.92503730998506e-06, "loss": 0.1127, "step": 1806 }, { "epoch": 0.08396840148698885, "grad_norm": 1.615460324853315, "learning_rate": 9.924907436824203e-06, "loss": 0.1462, "step": 1807 }, { "epoch": 0.08401486988847584, "grad_norm": 1.4333705075221903, "learning_rate": 9.924777452109008e-06, "loss": 0.116, "step": 1808 }, { "epoch": 0.08406133828996283, "grad_norm": 1.8999189284494014, "learning_rate": 9.924647355842422e-06, "loss": 0.1581, "step": 1809 }, { "epoch": 0.08410780669144981, "grad_norm": 1.4219357524813596, "learning_rate": 9.924517148027391e-06, "loss": 0.0954, "step": 1810 }, { "epoch": 0.0841542750929368, "grad_norm": 1.6178716426319053, "learning_rate": 9.924386828666863e-06, "loss": 0.1727, "step": 1811 }, { "epoch": 0.08420074349442379, "grad_norm": 2.121625720431479, "learning_rate": 9.924256397763793e-06, "loss": 0.1063, "step": 1812 }, { "epoch": 0.08424721189591078, "grad_norm": 1.7381069848390431, "learning_rate": 9.924125855321128e-06, "loss": 0.1355, "step": 1813 }, { "epoch": 0.08429368029739777, "grad_norm": 2.014940155580747, "learning_rate": 9.923995201341832e-06, "loss": 0.1413, "step": 1814 }, { "epoch": 0.08434014869888476, "grad_norm": 1.2490684669812522, "learning_rate": 9.923864435828863e-06, "loss": 0.0867, "step": 1815 }, { "epoch": 0.08438661710037175, "grad_norm": 1.1101753618893828, "learning_rate": 9.923733558785182e-06, "loss": 0.0956, "step": 1816 }, { "epoch": 0.08443308550185874, "grad_norm": 1.8679262515291277, "learning_rate": 9.923602570213752e-06, "loss": 0.1308, "step": 1817 }, { "epoch": 0.08447955390334573, "grad_norm": 2.8400666563809223, "learning_rate": 9.923471470117544e-06, "loss": 0.0957, "step": 1818 }, { "epoch": 0.08452602230483271, "grad_norm": 2.2315114158605076, "learning_rate": 9.923340258499525e-06, "loss": 0.2205, "step": 1819 }, { "epoch": 0.0845724907063197, "grad_norm": 1.7791867839934388, "learning_rate": 9.923208935362667e-06, "loss": 0.0804, "step": 1820 }, { "epoch": 0.08461895910780669, "grad_norm": 1.8427290533045266, "learning_rate": 9.923077500709946e-06, "loss": 0.088, "step": 1821 }, { "epoch": 0.08466542750929368, "grad_norm": 1.601275239239926, "learning_rate": 9.922945954544336e-06, "loss": 0.1554, "step": 1822 }, { "epoch": 0.08471189591078067, "grad_norm": 1.5006836998812692, "learning_rate": 9.92281429686882e-06, "loss": 0.0919, "step": 1823 }, { "epoch": 0.08475836431226766, "grad_norm": 12.876337053352577, "learning_rate": 9.92268252768638e-06, "loss": 0.1233, "step": 1824 }, { "epoch": 0.08480483271375465, "grad_norm": 1.3426873139131763, "learning_rate": 9.922550646999998e-06, "loss": 0.077, "step": 1825 }, { "epoch": 0.08485130111524164, "grad_norm": 1.5945170836852975, "learning_rate": 9.922418654812663e-06, "loss": 0.1073, "step": 1826 }, { "epoch": 0.08489776951672863, "grad_norm": 1.9578544914615255, "learning_rate": 9.922286551127366e-06, "loss": 0.2055, "step": 1827 }, { "epoch": 0.0849442379182156, "grad_norm": 2.980353408468959, "learning_rate": 9.922154335947098e-06, "loss": 0.1258, "step": 1828 }, { "epoch": 0.0849907063197026, "grad_norm": 1.6883821992821144, "learning_rate": 9.922022009274853e-06, "loss": 0.0934, "step": 1829 }, { "epoch": 0.08503717472118959, "grad_norm": 1.442130417019689, "learning_rate": 9.921889571113629e-06, "loss": 0.1174, "step": 1830 }, { "epoch": 0.08508364312267658, "grad_norm": 1.8515862926910973, "learning_rate": 9.921757021466427e-06, "loss": 0.13, "step": 1831 }, { "epoch": 0.08513011152416357, "grad_norm": 1.6594669445668733, "learning_rate": 9.921624360336248e-06, "loss": 0.1408, "step": 1832 }, { "epoch": 0.08517657992565056, "grad_norm": 1.5204443469893567, "learning_rate": 9.921491587726096e-06, "loss": 0.1123, "step": 1833 }, { "epoch": 0.08522304832713755, "grad_norm": 1.6636605918025598, "learning_rate": 9.921358703638982e-06, "loss": 0.1269, "step": 1834 }, { "epoch": 0.08526951672862454, "grad_norm": 2.202853041158122, "learning_rate": 9.921225708077911e-06, "loss": 0.1832, "step": 1835 }, { "epoch": 0.08531598513011153, "grad_norm": 1.597531647171719, "learning_rate": 9.921092601045901e-06, "loss": 0.0964, "step": 1836 }, { "epoch": 0.0853624535315985, "grad_norm": 1.3381440165474086, "learning_rate": 9.920959382545963e-06, "loss": 0.1268, "step": 1837 }, { "epoch": 0.0854089219330855, "grad_norm": 3.11112090956809, "learning_rate": 9.920826052581116e-06, "loss": 0.1312, "step": 1838 }, { "epoch": 0.08545539033457249, "grad_norm": 1.0224781282024373, "learning_rate": 9.92069261115438e-06, "loss": 0.106, "step": 1839 }, { "epoch": 0.08550185873605948, "grad_norm": 4.381088334540404, "learning_rate": 9.920559058268776e-06, "loss": 0.1252, "step": 1840 }, { "epoch": 0.08554832713754647, "grad_norm": 1.2025470182599316, "learning_rate": 9.920425393927332e-06, "loss": 0.0988, "step": 1841 }, { "epoch": 0.08559479553903346, "grad_norm": 1.3004053307605268, "learning_rate": 9.920291618133074e-06, "loss": 0.0757, "step": 1842 }, { "epoch": 0.08564126394052045, "grad_norm": 5.9137027820584676, "learning_rate": 9.920157730889032e-06, "loss": 0.169, "step": 1843 }, { "epoch": 0.08568773234200744, "grad_norm": 1.2912169919258303, "learning_rate": 9.920023732198239e-06, "loss": 0.1141, "step": 1844 }, { "epoch": 0.08573420074349443, "grad_norm": 1.964814966073968, "learning_rate": 9.91988962206373e-06, "loss": 0.1788, "step": 1845 }, { "epoch": 0.0857806691449814, "grad_norm": 1.3520477149426988, "learning_rate": 9.919755400488544e-06, "loss": 0.0853, "step": 1846 }, { "epoch": 0.0858271375464684, "grad_norm": 2.0638144083623153, "learning_rate": 9.91962106747572e-06, "loss": 0.1036, "step": 1847 }, { "epoch": 0.08587360594795539, "grad_norm": 1.8050560346127338, "learning_rate": 9.9194866230283e-06, "loss": 0.1026, "step": 1848 }, { "epoch": 0.08592007434944238, "grad_norm": 1.8378426110133133, "learning_rate": 9.91935206714933e-06, "loss": 0.0886, "step": 1849 }, { "epoch": 0.08596654275092937, "grad_norm": 2.748276437563327, "learning_rate": 9.91921739984186e-06, "loss": 0.1192, "step": 1850 }, { "epoch": 0.08601301115241636, "grad_norm": 1.8135461024274724, "learning_rate": 9.919082621108937e-06, "loss": 0.1477, "step": 1851 }, { "epoch": 0.08605947955390335, "grad_norm": 2.600464815165396, "learning_rate": 9.918947730953617e-06, "loss": 0.1564, "step": 1852 }, { "epoch": 0.08610594795539034, "grad_norm": 5.5610319042943415, "learning_rate": 9.918812729378953e-06, "loss": 0.2103, "step": 1853 }, { "epoch": 0.08615241635687733, "grad_norm": 1.9502903777382832, "learning_rate": 9.918677616388003e-06, "loss": 0.1082, "step": 1854 }, { "epoch": 0.08619888475836432, "grad_norm": 1.419810009514883, "learning_rate": 9.918542391983829e-06, "loss": 0.0877, "step": 1855 }, { "epoch": 0.0862453531598513, "grad_norm": 2.5624942254909606, "learning_rate": 9.918407056169493e-06, "loss": 0.1368, "step": 1856 }, { "epoch": 0.08629182156133829, "grad_norm": 2.068295091333275, "learning_rate": 9.918271608948061e-06, "loss": 0.083, "step": 1857 }, { "epoch": 0.08633828996282528, "grad_norm": 2.112271602309227, "learning_rate": 9.9181360503226e-06, "loss": 0.198, "step": 1858 }, { "epoch": 0.08638475836431227, "grad_norm": 2.9677739627775668, "learning_rate": 9.91800038029618e-06, "loss": 0.1671, "step": 1859 }, { "epoch": 0.08643122676579926, "grad_norm": 1.6140626143812662, "learning_rate": 9.917864598871877e-06, "loss": 0.1649, "step": 1860 }, { "epoch": 0.08647769516728625, "grad_norm": 0.918544919001674, "learning_rate": 9.917728706052765e-06, "loss": 0.0552, "step": 1861 }, { "epoch": 0.08652416356877324, "grad_norm": 2.5133791009813713, "learning_rate": 9.91759270184192e-06, "loss": 0.0898, "step": 1862 }, { "epoch": 0.08657063197026023, "grad_norm": 1.2660917767762527, "learning_rate": 9.917456586242427e-06, "loss": 0.0939, "step": 1863 }, { "epoch": 0.08661710037174722, "grad_norm": 2.6131163582981296, "learning_rate": 9.917320359257365e-06, "loss": 0.1303, "step": 1864 }, { "epoch": 0.0866635687732342, "grad_norm": 1.5552196172603616, "learning_rate": 9.917184020889823e-06, "loss": 0.0961, "step": 1865 }, { "epoch": 0.08671003717472119, "grad_norm": 1.7958015483325562, "learning_rate": 9.917047571142888e-06, "loss": 0.1322, "step": 1866 }, { "epoch": 0.08675650557620818, "grad_norm": 1.886210956619187, "learning_rate": 9.91691101001965e-06, "loss": 0.155, "step": 1867 }, { "epoch": 0.08680297397769517, "grad_norm": 1.9416508849461833, "learning_rate": 9.9167743375232e-06, "loss": 0.1241, "step": 1868 }, { "epoch": 0.08684944237918216, "grad_norm": 1.4677885036448886, "learning_rate": 9.91663755365664e-06, "loss": 0.139, "step": 1869 }, { "epoch": 0.08689591078066915, "grad_norm": 1.74082498099616, "learning_rate": 9.916500658423063e-06, "loss": 0.1321, "step": 1870 }, { "epoch": 0.08694237918215614, "grad_norm": 1.547126075546088, "learning_rate": 9.916363651825573e-06, "loss": 0.1554, "step": 1871 }, { "epoch": 0.08698884758364313, "grad_norm": 1.6983032295397034, "learning_rate": 9.91622653386727e-06, "loss": 0.1338, "step": 1872 }, { "epoch": 0.08703531598513012, "grad_norm": 2.028000640547693, "learning_rate": 9.916089304551262e-06, "loss": 0.1001, "step": 1873 }, { "epoch": 0.0870817843866171, "grad_norm": 1.6888416262640316, "learning_rate": 9.915951963880658e-06, "loss": 0.1108, "step": 1874 }, { "epoch": 0.08712825278810409, "grad_norm": 3.00892556360652, "learning_rate": 9.915814511858568e-06, "loss": 0.2429, "step": 1875 }, { "epoch": 0.08717472118959108, "grad_norm": 1.9324583220696097, "learning_rate": 9.915676948488106e-06, "loss": 0.127, "step": 1876 }, { "epoch": 0.08722118959107807, "grad_norm": 2.8192441348147543, "learning_rate": 9.915539273772387e-06, "loss": 0.0889, "step": 1877 }, { "epoch": 0.08726765799256506, "grad_norm": 2.320749676021971, "learning_rate": 9.915401487714528e-06, "loss": 0.1105, "step": 1878 }, { "epoch": 0.08731412639405205, "grad_norm": 2.1868611695096436, "learning_rate": 9.915263590317654e-06, "loss": 0.1471, "step": 1879 }, { "epoch": 0.08736059479553904, "grad_norm": 1.8314812277105263, "learning_rate": 9.915125581584886e-06, "loss": 0.0907, "step": 1880 }, { "epoch": 0.08740706319702603, "grad_norm": 2.1458913214557738, "learning_rate": 9.914987461519352e-06, "loss": 0.1042, "step": 1881 }, { "epoch": 0.08745353159851302, "grad_norm": 1.4436622696300407, "learning_rate": 9.914849230124177e-06, "loss": 0.1571, "step": 1882 }, { "epoch": 0.0875, "grad_norm": 3.661736959227894, "learning_rate": 9.914710887402494e-06, "loss": 0.1331, "step": 1883 }, { "epoch": 0.08754646840148698, "grad_norm": 2.260337053618473, "learning_rate": 9.91457243335744e-06, "loss": 0.1783, "step": 1884 }, { "epoch": 0.08759293680297398, "grad_norm": 2.628850351228511, "learning_rate": 9.914433867992144e-06, "loss": 0.171, "step": 1885 }, { "epoch": 0.08763940520446097, "grad_norm": 1.4383130953081076, "learning_rate": 9.91429519130975e-06, "loss": 0.1418, "step": 1886 }, { "epoch": 0.08768587360594796, "grad_norm": 2.049022899987123, "learning_rate": 9.914156403313398e-06, "loss": 0.1715, "step": 1887 }, { "epoch": 0.08773234200743495, "grad_norm": 1.8678779803195944, "learning_rate": 9.914017504006232e-06, "loss": 0.1502, "step": 1888 }, { "epoch": 0.08777881040892194, "grad_norm": 2.165664407979908, "learning_rate": 9.913878493391396e-06, "loss": 0.1417, "step": 1889 }, { "epoch": 0.08782527881040893, "grad_norm": 1.6355426513949625, "learning_rate": 9.913739371472042e-06, "loss": 0.1269, "step": 1890 }, { "epoch": 0.08787174721189592, "grad_norm": 1.8928956211567878, "learning_rate": 9.913600138251318e-06, "loss": 0.1182, "step": 1891 }, { "epoch": 0.0879182156133829, "grad_norm": 1.551571140398253, "learning_rate": 9.91346079373238e-06, "loss": 0.1048, "step": 1892 }, { "epoch": 0.08796468401486988, "grad_norm": 1.234486446147065, "learning_rate": 9.913321337918385e-06, "loss": 0.0638, "step": 1893 }, { "epoch": 0.08801115241635687, "grad_norm": 2.641293694105737, "learning_rate": 9.91318177081249e-06, "loss": 0.1488, "step": 1894 }, { "epoch": 0.08805762081784387, "grad_norm": 1.7729829045146703, "learning_rate": 9.913042092417855e-06, "loss": 0.1692, "step": 1895 }, { "epoch": 0.08810408921933086, "grad_norm": 1.2829227006642254, "learning_rate": 9.912902302737648e-06, "loss": 0.094, "step": 1896 }, { "epoch": 0.08815055762081785, "grad_norm": 2.136081828956918, "learning_rate": 9.91276240177503e-06, "loss": 0.1164, "step": 1897 }, { "epoch": 0.08819702602230484, "grad_norm": 1.6800567718844355, "learning_rate": 9.912622389533177e-06, "loss": 0.0627, "step": 1898 }, { "epoch": 0.08824349442379183, "grad_norm": 1.426404908430376, "learning_rate": 9.912482266015254e-06, "loss": 0.0771, "step": 1899 }, { "epoch": 0.08828996282527882, "grad_norm": 2.458936512125097, "learning_rate": 9.912342031224438e-06, "loss": 0.0688, "step": 1900 }, { "epoch": 0.0883364312267658, "grad_norm": 1.7772894582591814, "learning_rate": 9.912201685163905e-06, "loss": 0.1539, "step": 1901 }, { "epoch": 0.08838289962825278, "grad_norm": 1.2903355651361303, "learning_rate": 9.912061227836833e-06, "loss": 0.0761, "step": 1902 }, { "epoch": 0.08842936802973977, "grad_norm": 0.9834818801292873, "learning_rate": 9.911920659246406e-06, "loss": 0.0635, "step": 1903 }, { "epoch": 0.08847583643122676, "grad_norm": 2.5891088665314173, "learning_rate": 9.911779979395804e-06, "loss": 0.2, "step": 1904 }, { "epoch": 0.08852230483271376, "grad_norm": 1.0756566494479858, "learning_rate": 9.911639188288218e-06, "loss": 0.0602, "step": 1905 }, { "epoch": 0.08856877323420075, "grad_norm": 2.816471739941751, "learning_rate": 9.911498285926833e-06, "loss": 0.2204, "step": 1906 }, { "epoch": 0.08861524163568774, "grad_norm": 1.345891152459094, "learning_rate": 9.911357272314844e-06, "loss": 0.1018, "step": 1907 }, { "epoch": 0.08866171003717473, "grad_norm": 2.2640643859120533, "learning_rate": 9.911216147455444e-06, "loss": 0.0941, "step": 1908 }, { "epoch": 0.08870817843866172, "grad_norm": 1.632361274897586, "learning_rate": 9.911074911351828e-06, "loss": 0.1004, "step": 1909 }, { "epoch": 0.0887546468401487, "grad_norm": 1.9137795034615932, "learning_rate": 9.910933564007199e-06, "loss": 0.1461, "step": 1910 }, { "epoch": 0.08880111524163568, "grad_norm": 1.475606797311692, "learning_rate": 9.910792105424753e-06, "loss": 0.0955, "step": 1911 }, { "epoch": 0.08884758364312267, "grad_norm": 1.610617367176316, "learning_rate": 9.9106505356077e-06, "loss": 0.0998, "step": 1912 }, { "epoch": 0.08889405204460966, "grad_norm": 1.9212591959186325, "learning_rate": 9.910508854559243e-06, "loss": 0.1186, "step": 1913 }, { "epoch": 0.08894052044609665, "grad_norm": 4.2315942276339, "learning_rate": 9.91036706228259e-06, "loss": 0.1598, "step": 1914 }, { "epoch": 0.08898698884758365, "grad_norm": 1.885309824022624, "learning_rate": 9.910225158780958e-06, "loss": 0.0948, "step": 1915 }, { "epoch": 0.08903345724907064, "grad_norm": 3.050973834772147, "learning_rate": 9.910083144057557e-06, "loss": 0.1269, "step": 1916 }, { "epoch": 0.08907992565055763, "grad_norm": 2.081784211470069, "learning_rate": 9.909941018115604e-06, "loss": 0.1007, "step": 1917 }, { "epoch": 0.08912639405204462, "grad_norm": 1.8888782143883653, "learning_rate": 9.90979878095832e-06, "loss": 0.1044, "step": 1918 }, { "epoch": 0.08917286245353159, "grad_norm": 2.0191713803594493, "learning_rate": 9.909656432588927e-06, "loss": 0.1268, "step": 1919 }, { "epoch": 0.08921933085501858, "grad_norm": 2.1173976144374898, "learning_rate": 9.909513973010648e-06, "loss": 0.1882, "step": 1920 }, { "epoch": 0.08926579925650557, "grad_norm": 1.5381733705591587, "learning_rate": 9.909371402226712e-06, "loss": 0.1328, "step": 1921 }, { "epoch": 0.08931226765799256, "grad_norm": 1.228171933628276, "learning_rate": 9.909228720240343e-06, "loss": 0.164, "step": 1922 }, { "epoch": 0.08935873605947955, "grad_norm": 1.975720547226471, "learning_rate": 9.90908592705478e-06, "loss": 0.1169, "step": 1923 }, { "epoch": 0.08940520446096654, "grad_norm": 1.643146033718362, "learning_rate": 9.908943022673251e-06, "loss": 0.1639, "step": 1924 }, { "epoch": 0.08945167286245354, "grad_norm": 1.2575498843103834, "learning_rate": 9.908800007098997e-06, "loss": 0.071, "step": 1925 }, { "epoch": 0.08949814126394053, "grad_norm": 2.0678076219647097, "learning_rate": 9.908656880335259e-06, "loss": 0.0908, "step": 1926 }, { "epoch": 0.08954460966542752, "grad_norm": 1.8712063231524956, "learning_rate": 9.908513642385273e-06, "loss": 0.1501, "step": 1927 }, { "epoch": 0.08959107806691449, "grad_norm": 1.0333853709575405, "learning_rate": 9.90837029325229e-06, "loss": 0.0743, "step": 1928 }, { "epoch": 0.08963754646840148, "grad_norm": 2.239325345145664, "learning_rate": 9.908226832939551e-06, "loss": 0.1808, "step": 1929 }, { "epoch": 0.08968401486988847, "grad_norm": 1.729333387929573, "learning_rate": 9.90808326145031e-06, "loss": 0.0883, "step": 1930 }, { "epoch": 0.08973048327137546, "grad_norm": 1.532046196498878, "learning_rate": 9.907939578787817e-06, "loss": 0.1289, "step": 1931 }, { "epoch": 0.08977695167286245, "grad_norm": 2.097192862234557, "learning_rate": 9.907795784955327e-06, "loss": 0.1976, "step": 1932 }, { "epoch": 0.08982342007434944, "grad_norm": 2.946623389576736, "learning_rate": 9.907651879956097e-06, "loss": 0.136, "step": 1933 }, { "epoch": 0.08986988847583643, "grad_norm": 3.1449436593537667, "learning_rate": 9.907507863793388e-06, "loss": 0.2622, "step": 1934 }, { "epoch": 0.08991635687732343, "grad_norm": 1.6644129655373436, "learning_rate": 9.907363736470458e-06, "loss": 0.1609, "step": 1935 }, { "epoch": 0.08996282527881042, "grad_norm": 6.282699171526025, "learning_rate": 9.907219497990578e-06, "loss": 0.1416, "step": 1936 }, { "epoch": 0.09000929368029739, "grad_norm": 2.3721589431212635, "learning_rate": 9.907075148357011e-06, "loss": 0.1017, "step": 1937 }, { "epoch": 0.09005576208178438, "grad_norm": 1.219626811624255, "learning_rate": 9.906930687573024e-06, "loss": 0.0843, "step": 1938 }, { "epoch": 0.09010223048327137, "grad_norm": 0.9700574725461854, "learning_rate": 9.906786115641895e-06, "loss": 0.0719, "step": 1939 }, { "epoch": 0.09014869888475836, "grad_norm": 1.6311613591194225, "learning_rate": 9.906641432566897e-06, "loss": 0.1435, "step": 1940 }, { "epoch": 0.09019516728624535, "grad_norm": 2.204726271097203, "learning_rate": 9.906496638351305e-06, "loss": 0.112, "step": 1941 }, { "epoch": 0.09024163568773234, "grad_norm": 1.3158498365518647, "learning_rate": 9.9063517329984e-06, "loss": 0.0804, "step": 1942 }, { "epoch": 0.09028810408921933, "grad_norm": 1.3521616977296944, "learning_rate": 9.906206716511465e-06, "loss": 0.0839, "step": 1943 }, { "epoch": 0.09033457249070632, "grad_norm": 1.482988458988268, "learning_rate": 9.906061588893783e-06, "loss": 0.1254, "step": 1944 }, { "epoch": 0.09038104089219332, "grad_norm": 2.7033229368669978, "learning_rate": 9.905916350148645e-06, "loss": 0.1702, "step": 1945 }, { "epoch": 0.09042750929368029, "grad_norm": 1.6236315762298248, "learning_rate": 9.905771000279336e-06, "loss": 0.1015, "step": 1946 }, { "epoch": 0.09047397769516728, "grad_norm": 1.10422217876196, "learning_rate": 9.905625539289153e-06, "loss": 0.068, "step": 1947 }, { "epoch": 0.09052044609665427, "grad_norm": 2.451163350814501, "learning_rate": 9.905479967181385e-06, "loss": 0.2556, "step": 1948 }, { "epoch": 0.09056691449814126, "grad_norm": 1.8729280250569278, "learning_rate": 9.905334283959335e-06, "loss": 0.1451, "step": 1949 }, { "epoch": 0.09061338289962825, "grad_norm": 2.1385932577403968, "learning_rate": 9.905188489626301e-06, "loss": 0.1392, "step": 1950 }, { "epoch": 0.09065985130111524, "grad_norm": 2.575649348818046, "learning_rate": 9.905042584185584e-06, "loss": 0.1665, "step": 1951 }, { "epoch": 0.09070631970260223, "grad_norm": 1.7271099101339904, "learning_rate": 9.90489656764049e-06, "loss": 0.1266, "step": 1952 }, { "epoch": 0.09075278810408922, "grad_norm": 1.5138457480810892, "learning_rate": 9.904750439994328e-06, "loss": 0.0975, "step": 1953 }, { "epoch": 0.09079925650557621, "grad_norm": 1.963776599381811, "learning_rate": 9.904604201250404e-06, "loss": 0.1232, "step": 1954 }, { "epoch": 0.09084572490706319, "grad_norm": 1.3351436177935343, "learning_rate": 9.904457851412035e-06, "loss": 0.1659, "step": 1955 }, { "epoch": 0.09089219330855018, "grad_norm": 2.1617297484260125, "learning_rate": 9.904311390482532e-06, "loss": 0.2196, "step": 1956 }, { "epoch": 0.09093866171003717, "grad_norm": 2.3660986911259756, "learning_rate": 9.904164818465217e-06, "loss": 0.1688, "step": 1957 }, { "epoch": 0.09098513011152416, "grad_norm": 1.8751697959222122, "learning_rate": 9.904018135363405e-06, "loss": 0.1011, "step": 1958 }, { "epoch": 0.09103159851301115, "grad_norm": 2.427734362393494, "learning_rate": 9.90387134118042e-06, "loss": 0.1381, "step": 1959 }, { "epoch": 0.09107806691449814, "grad_norm": 2.571327650400592, "learning_rate": 9.90372443591959e-06, "loss": 0.142, "step": 1960 }, { "epoch": 0.09112453531598513, "grad_norm": 1.489406096692638, "learning_rate": 9.90357741958424e-06, "loss": 0.1246, "step": 1961 }, { "epoch": 0.09117100371747212, "grad_norm": 1.420243358254243, "learning_rate": 9.9034302921777e-06, "loss": 0.1062, "step": 1962 }, { "epoch": 0.09121747211895911, "grad_norm": 1.2844008183156148, "learning_rate": 9.903283053703302e-06, "loss": 0.0951, "step": 1963 }, { "epoch": 0.09126394052044609, "grad_norm": 1.5074954375504752, "learning_rate": 9.903135704164384e-06, "loss": 0.0736, "step": 1964 }, { "epoch": 0.09131040892193308, "grad_norm": 1.5662622437271665, "learning_rate": 9.902988243564282e-06, "loss": 0.166, "step": 1965 }, { "epoch": 0.09135687732342007, "grad_norm": 1.983331822401857, "learning_rate": 9.902840671906335e-06, "loss": 0.199, "step": 1966 }, { "epoch": 0.09140334572490706, "grad_norm": 1.3921188834173854, "learning_rate": 9.902692989193887e-06, "loss": 0.0967, "step": 1967 }, { "epoch": 0.09144981412639405, "grad_norm": 2.3749393020404064, "learning_rate": 9.902545195430283e-06, "loss": 0.1601, "step": 1968 }, { "epoch": 0.09149628252788104, "grad_norm": 2.4815023481444918, "learning_rate": 9.90239729061887e-06, "loss": 0.1093, "step": 1969 }, { "epoch": 0.09154275092936803, "grad_norm": 2.50338910531011, "learning_rate": 9.902249274763e-06, "loss": 0.1118, "step": 1970 }, { "epoch": 0.09158921933085502, "grad_norm": 2.2504733367639975, "learning_rate": 9.902101147866024e-06, "loss": 0.122, "step": 1971 }, { "epoch": 0.09163568773234201, "grad_norm": 6.183353095052249, "learning_rate": 9.901952909931296e-06, "loss": 0.1235, "step": 1972 }, { "epoch": 0.09168215613382899, "grad_norm": 1.9305785516432292, "learning_rate": 9.901804560962175e-06, "loss": 0.1446, "step": 1973 }, { "epoch": 0.09172862453531598, "grad_norm": 1.8736620535326565, "learning_rate": 9.901656100962024e-06, "loss": 0.0988, "step": 1974 }, { "epoch": 0.09177509293680297, "grad_norm": 2.271898987280245, "learning_rate": 9.901507529934202e-06, "loss": 0.124, "step": 1975 }, { "epoch": 0.09182156133828996, "grad_norm": 2.312640347217505, "learning_rate": 9.901358847882077e-06, "loss": 0.1435, "step": 1976 }, { "epoch": 0.09186802973977695, "grad_norm": 0.9001509828699842, "learning_rate": 9.901210054809015e-06, "loss": 0.0712, "step": 1977 }, { "epoch": 0.09191449814126394, "grad_norm": 2.4921017674506047, "learning_rate": 9.901061150718385e-06, "loss": 0.1072, "step": 1978 }, { "epoch": 0.09196096654275093, "grad_norm": 1.6896850027923638, "learning_rate": 9.900912135613564e-06, "loss": 0.1153, "step": 1979 }, { "epoch": 0.09200743494423792, "grad_norm": 1.4281039611824446, "learning_rate": 9.900763009497925e-06, "loss": 0.0939, "step": 1980 }, { "epoch": 0.09205390334572491, "grad_norm": 2.1191734571904575, "learning_rate": 9.900613772374845e-06, "loss": 0.1065, "step": 1981 }, { "epoch": 0.09210037174721189, "grad_norm": 1.8073244242490725, "learning_rate": 9.900464424247705e-06, "loss": 0.1272, "step": 1982 }, { "epoch": 0.09214684014869888, "grad_norm": 2.1252268954255458, "learning_rate": 9.900314965119888e-06, "loss": 0.1606, "step": 1983 }, { "epoch": 0.09219330855018587, "grad_norm": 1.7012861071190837, "learning_rate": 9.90016539499478e-06, "loss": 0.1109, "step": 1984 }, { "epoch": 0.09223977695167286, "grad_norm": 1.5620854677011662, "learning_rate": 9.90001571387577e-06, "loss": 0.1766, "step": 1985 }, { "epoch": 0.09228624535315985, "grad_norm": 2.4880194502157638, "learning_rate": 9.899865921766245e-06, "loss": 0.12, "step": 1986 }, { "epoch": 0.09233271375464684, "grad_norm": 1.652361199447832, "learning_rate": 9.8997160186696e-06, "loss": 0.1394, "step": 1987 }, { "epoch": 0.09237918215613383, "grad_norm": 1.26526223408925, "learning_rate": 9.89956600458923e-06, "loss": 0.1071, "step": 1988 }, { "epoch": 0.09242565055762082, "grad_norm": 3.658920834574547, "learning_rate": 9.899415879528534e-06, "loss": 0.0959, "step": 1989 }, { "epoch": 0.09247211895910781, "grad_norm": 2.0722329168962372, "learning_rate": 9.899265643490913e-06, "loss": 0.0935, "step": 1990 }, { "epoch": 0.09251858736059479, "grad_norm": 2.106914058632518, "learning_rate": 9.899115296479767e-06, "loss": 0.0985, "step": 1991 }, { "epoch": 0.09256505576208178, "grad_norm": 1.657784089403632, "learning_rate": 9.898964838498504e-06, "loss": 0.0784, "step": 1992 }, { "epoch": 0.09261152416356877, "grad_norm": 1.7530386123575554, "learning_rate": 9.89881426955053e-06, "loss": 0.1245, "step": 1993 }, { "epoch": 0.09265799256505576, "grad_norm": 2.711991718647239, "learning_rate": 9.89866358963926e-06, "loss": 0.1543, "step": 1994 }, { "epoch": 0.09270446096654275, "grad_norm": 2.241792662183988, "learning_rate": 9.8985127987681e-06, "loss": 0.1326, "step": 1995 }, { "epoch": 0.09275092936802974, "grad_norm": 1.438471481274521, "learning_rate": 9.898361896940472e-06, "loss": 0.0827, "step": 1996 }, { "epoch": 0.09279739776951673, "grad_norm": 2.8034634441806143, "learning_rate": 9.89821088415979e-06, "loss": 0.1366, "step": 1997 }, { "epoch": 0.09284386617100372, "grad_norm": 1.7303006767233198, "learning_rate": 9.898059760429479e-06, "loss": 0.1061, "step": 1998 }, { "epoch": 0.09289033457249071, "grad_norm": 2.372958713172228, "learning_rate": 9.897908525752956e-06, "loss": 0.0953, "step": 1999 }, { "epoch": 0.09293680297397769, "grad_norm": 2.3290386081562158, "learning_rate": 9.897757180133653e-06, "loss": 0.1059, "step": 2000 }, { "epoch": 0.09298327137546468, "grad_norm": 1.8370480955010788, "learning_rate": 9.897605723574992e-06, "loss": 0.1555, "step": 2001 }, { "epoch": 0.09302973977695167, "grad_norm": 1.4856768706008463, "learning_rate": 9.897454156080409e-06, "loss": 0.1109, "step": 2002 }, { "epoch": 0.09307620817843866, "grad_norm": 1.1389567360501218, "learning_rate": 9.897302477653334e-06, "loss": 0.1221, "step": 2003 }, { "epoch": 0.09312267657992565, "grad_norm": 2.2257348283213423, "learning_rate": 9.897150688297203e-06, "loss": 0.1758, "step": 2004 }, { "epoch": 0.09316914498141264, "grad_norm": 2.7122875831117264, "learning_rate": 9.896998788015457e-06, "loss": 0.174, "step": 2005 }, { "epoch": 0.09321561338289963, "grad_norm": 2.2249492711450283, "learning_rate": 9.896846776811533e-06, "loss": 0.1237, "step": 2006 }, { "epoch": 0.09326208178438662, "grad_norm": 2.889224058574102, "learning_rate": 9.896694654688875e-06, "loss": 0.0903, "step": 2007 }, { "epoch": 0.09330855018587361, "grad_norm": 2.4265322475982423, "learning_rate": 9.89654242165093e-06, "loss": 0.1546, "step": 2008 }, { "epoch": 0.09335501858736059, "grad_norm": 3.1511882827759528, "learning_rate": 9.896390077701146e-06, "loss": 0.142, "step": 2009 }, { "epoch": 0.09340148698884758, "grad_norm": 1.5864175667886462, "learning_rate": 9.896237622842972e-06, "loss": 0.1032, "step": 2010 }, { "epoch": 0.09344795539033457, "grad_norm": 1.3274384639178909, "learning_rate": 9.896085057079865e-06, "loss": 0.087, "step": 2011 }, { "epoch": 0.09349442379182156, "grad_norm": 2.246710440799586, "learning_rate": 9.895932380415277e-06, "loss": 0.1217, "step": 2012 }, { "epoch": 0.09354089219330855, "grad_norm": 2.3988364892404466, "learning_rate": 9.895779592852668e-06, "loss": 0.1388, "step": 2013 }, { "epoch": 0.09358736059479554, "grad_norm": 2.768150550820859, "learning_rate": 9.895626694395498e-06, "loss": 0.1908, "step": 2014 }, { "epoch": 0.09363382899628253, "grad_norm": 1.3800643824006247, "learning_rate": 9.89547368504723e-06, "loss": 0.1388, "step": 2015 }, { "epoch": 0.09368029739776952, "grad_norm": 1.2080285290381807, "learning_rate": 9.895320564811333e-06, "loss": 0.0754, "step": 2016 }, { "epoch": 0.09372676579925651, "grad_norm": 1.8749123179461642, "learning_rate": 9.895167333691272e-06, "loss": 0.1295, "step": 2017 }, { "epoch": 0.09377323420074349, "grad_norm": 1.9102528257988418, "learning_rate": 9.895013991690518e-06, "loss": 0.217, "step": 2018 }, { "epoch": 0.09381970260223048, "grad_norm": 1.5568983446915525, "learning_rate": 9.894860538812547e-06, "loss": 0.0948, "step": 2019 }, { "epoch": 0.09386617100371747, "grad_norm": 3.306196655843423, "learning_rate": 9.89470697506083e-06, "loss": 0.1274, "step": 2020 }, { "epoch": 0.09391263940520446, "grad_norm": 2.17338322885072, "learning_rate": 9.89455330043885e-06, "loss": 0.1445, "step": 2021 }, { "epoch": 0.09395910780669145, "grad_norm": 1.9557717746330188, "learning_rate": 9.894399514950087e-06, "loss": 0.1019, "step": 2022 }, { "epoch": 0.09400557620817844, "grad_norm": 1.4898644936835883, "learning_rate": 9.894245618598023e-06, "loss": 0.1281, "step": 2023 }, { "epoch": 0.09405204460966543, "grad_norm": 1.8061216122688082, "learning_rate": 9.894091611386145e-06, "loss": 0.1211, "step": 2024 }, { "epoch": 0.09409851301115242, "grad_norm": 1.3011378849705788, "learning_rate": 9.893937493317943e-06, "loss": 0.0703, "step": 2025 }, { "epoch": 0.09414498141263941, "grad_norm": 1.7035853005717454, "learning_rate": 9.893783264396903e-06, "loss": 0.0958, "step": 2026 }, { "epoch": 0.09419144981412639, "grad_norm": 1.3372247198801335, "learning_rate": 9.893628924626521e-06, "loss": 0.0825, "step": 2027 }, { "epoch": 0.09423791821561338, "grad_norm": 1.315293755453348, "learning_rate": 9.893474474010296e-06, "loss": 0.1218, "step": 2028 }, { "epoch": 0.09428438661710037, "grad_norm": 4.532749150308104, "learning_rate": 9.893319912551723e-06, "loss": 0.1799, "step": 2029 }, { "epoch": 0.09433085501858736, "grad_norm": 1.6510350041389583, "learning_rate": 9.893165240254304e-06, "loss": 0.1597, "step": 2030 }, { "epoch": 0.09437732342007435, "grad_norm": 3.19879140476608, "learning_rate": 9.893010457121541e-06, "loss": 0.1224, "step": 2031 }, { "epoch": 0.09442379182156134, "grad_norm": 3.6706234892245857, "learning_rate": 9.892855563156943e-06, "loss": 0.1996, "step": 2032 }, { "epoch": 0.09447026022304833, "grad_norm": 1.8471424789653799, "learning_rate": 9.892700558364017e-06, "loss": 0.1265, "step": 2033 }, { "epoch": 0.09451672862453532, "grad_norm": 2.245744791140102, "learning_rate": 9.892545442746272e-06, "loss": 0.1086, "step": 2034 }, { "epoch": 0.09456319702602231, "grad_norm": 1.8270716240465985, "learning_rate": 9.892390216307224e-06, "loss": 0.0676, "step": 2035 }, { "epoch": 0.09460966542750929, "grad_norm": 1.5895400934134276, "learning_rate": 9.892234879050389e-06, "loss": 0.1576, "step": 2036 }, { "epoch": 0.09465613382899628, "grad_norm": 1.4812959493761493, "learning_rate": 9.892079430979285e-06, "loss": 0.099, "step": 2037 }, { "epoch": 0.09470260223048327, "grad_norm": 1.893629871733479, "learning_rate": 9.891923872097434e-06, "loss": 0.1556, "step": 2038 }, { "epoch": 0.09474907063197026, "grad_norm": 2.319998695543005, "learning_rate": 9.891768202408357e-06, "loss": 0.1241, "step": 2039 }, { "epoch": 0.09479553903345725, "grad_norm": 1.4759929098440092, "learning_rate": 9.891612421915581e-06, "loss": 0.1039, "step": 2040 }, { "epoch": 0.09484200743494424, "grad_norm": 2.600191813713004, "learning_rate": 9.891456530622638e-06, "loss": 0.2276, "step": 2041 }, { "epoch": 0.09488847583643123, "grad_norm": 1.5104611890065933, "learning_rate": 9.891300528533055e-06, "loss": 0.1698, "step": 2042 }, { "epoch": 0.09493494423791822, "grad_norm": 0.9810607372510558, "learning_rate": 9.891144415650366e-06, "loss": 0.0778, "step": 2043 }, { "epoch": 0.09498141263940521, "grad_norm": 1.7986995125501615, "learning_rate": 9.890988191978111e-06, "loss": 0.1308, "step": 2044 }, { "epoch": 0.09502788104089219, "grad_norm": 1.1319420503908504, "learning_rate": 9.890831857519823e-06, "loss": 0.084, "step": 2045 }, { "epoch": 0.09507434944237918, "grad_norm": 3.326563694478116, "learning_rate": 9.890675412279047e-06, "loss": 0.0872, "step": 2046 }, { "epoch": 0.09512081784386617, "grad_norm": 1.9820565750821688, "learning_rate": 9.890518856259325e-06, "loss": 0.1934, "step": 2047 }, { "epoch": 0.09516728624535316, "grad_norm": 1.2870160124724384, "learning_rate": 9.890362189464205e-06, "loss": 0.1593, "step": 2048 }, { "epoch": 0.09521375464684015, "grad_norm": 2.953005606728357, "learning_rate": 9.890205411897234e-06, "loss": 0.1705, "step": 2049 }, { "epoch": 0.09526022304832714, "grad_norm": 1.6943449618300293, "learning_rate": 9.89004852356196e-06, "loss": 0.1276, "step": 2050 }, { "epoch": 0.09530669144981413, "grad_norm": 1.4810596338997517, "learning_rate": 9.889891524461946e-06, "loss": 0.0903, "step": 2051 }, { "epoch": 0.09535315985130112, "grad_norm": 1.3584917291467784, "learning_rate": 9.88973441460074e-06, "loss": 0.1107, "step": 2052 }, { "epoch": 0.09539962825278811, "grad_norm": 1.586490955166994, "learning_rate": 9.889577193981903e-06, "loss": 0.1098, "step": 2053 }, { "epoch": 0.09544609665427509, "grad_norm": 0.9928015486524383, "learning_rate": 9.889419862608995e-06, "loss": 0.0709, "step": 2054 }, { "epoch": 0.09549256505576208, "grad_norm": 1.3201941043428056, "learning_rate": 9.889262420485583e-06, "loss": 0.0794, "step": 2055 }, { "epoch": 0.09553903345724907, "grad_norm": 1.1939021024222052, "learning_rate": 9.889104867615232e-06, "loss": 0.0446, "step": 2056 }, { "epoch": 0.09558550185873606, "grad_norm": 2.1328949437688505, "learning_rate": 9.888947204001509e-06, "loss": 0.1201, "step": 2057 }, { "epoch": 0.09563197026022305, "grad_norm": 1.2895495477788874, "learning_rate": 9.888789429647985e-06, "loss": 0.091, "step": 2058 }, { "epoch": 0.09567843866171004, "grad_norm": 3.2786911911921517, "learning_rate": 9.888631544558236e-06, "loss": 0.1236, "step": 2059 }, { "epoch": 0.09572490706319703, "grad_norm": 1.4188259293171244, "learning_rate": 9.888473548735837e-06, "loss": 0.1585, "step": 2060 }, { "epoch": 0.09577137546468402, "grad_norm": 1.6016744536901588, "learning_rate": 9.88831544218437e-06, "loss": 0.112, "step": 2061 }, { "epoch": 0.09581784386617101, "grad_norm": 1.0895323077326131, "learning_rate": 9.88815722490741e-06, "loss": 0.0599, "step": 2062 }, { "epoch": 0.09586431226765799, "grad_norm": 1.6549667459500559, "learning_rate": 9.887998896908545e-06, "loss": 0.1032, "step": 2063 }, { "epoch": 0.09591078066914498, "grad_norm": 3.456078944535364, "learning_rate": 9.88784045819136e-06, "loss": 0.1852, "step": 2064 }, { "epoch": 0.09595724907063197, "grad_norm": 1.9355796371876248, "learning_rate": 9.887681908759445e-06, "loss": 0.1927, "step": 2065 }, { "epoch": 0.09600371747211896, "grad_norm": 1.507972534780926, "learning_rate": 9.88752324861639e-06, "loss": 0.1023, "step": 2066 }, { "epoch": 0.09605018587360595, "grad_norm": 2.593741491138483, "learning_rate": 9.88736447776579e-06, "loss": 0.1594, "step": 2067 }, { "epoch": 0.09609665427509294, "grad_norm": 2.6444910570083553, "learning_rate": 9.887205596211243e-06, "loss": 0.1199, "step": 2068 }, { "epoch": 0.09614312267657993, "grad_norm": 2.089680213707231, "learning_rate": 9.887046603956341e-06, "loss": 0.2095, "step": 2069 }, { "epoch": 0.09618959107806692, "grad_norm": 1.138278235239547, "learning_rate": 9.886887501004694e-06, "loss": 0.0948, "step": 2070 }, { "epoch": 0.09623605947955391, "grad_norm": 1.6687655288655214, "learning_rate": 9.8867282873599e-06, "loss": 0.125, "step": 2071 }, { "epoch": 0.09628252788104089, "grad_norm": 2.0096741822178945, "learning_rate": 9.886568963025567e-06, "loss": 0.0931, "step": 2072 }, { "epoch": 0.09632899628252788, "grad_norm": 1.2319967817883624, "learning_rate": 9.886409528005303e-06, "loss": 0.0735, "step": 2073 }, { "epoch": 0.09637546468401487, "grad_norm": 1.1141853865060003, "learning_rate": 9.88624998230272e-06, "loss": 0.1012, "step": 2074 }, { "epoch": 0.09642193308550186, "grad_norm": 1.5122038680077432, "learning_rate": 9.886090325921435e-06, "loss": 0.1494, "step": 2075 }, { "epoch": 0.09646840148698885, "grad_norm": 1.3429799006376153, "learning_rate": 9.88593055886506e-06, "loss": 0.1069, "step": 2076 }, { "epoch": 0.09651486988847584, "grad_norm": 1.1858273346257246, "learning_rate": 9.885770681137215e-06, "loss": 0.0869, "step": 2077 }, { "epoch": 0.09656133828996283, "grad_norm": 1.4846060858725407, "learning_rate": 9.885610692741523e-06, "loss": 0.1517, "step": 2078 }, { "epoch": 0.09660780669144982, "grad_norm": 2.2991157469463404, "learning_rate": 9.885450593681605e-06, "loss": 0.0682, "step": 2079 }, { "epoch": 0.09665427509293681, "grad_norm": 1.6729003193401963, "learning_rate": 9.88529038396109e-06, "loss": 0.039, "step": 2080 }, { "epoch": 0.09670074349442379, "grad_norm": 2.257137339447281, "learning_rate": 9.885130063583607e-06, "loss": 0.1635, "step": 2081 }, { "epoch": 0.09674721189591078, "grad_norm": 1.3289471494619547, "learning_rate": 9.884969632552787e-06, "loss": 0.1257, "step": 2082 }, { "epoch": 0.09679368029739777, "grad_norm": 1.0157833726883971, "learning_rate": 9.884809090872263e-06, "loss": 0.0761, "step": 2083 }, { "epoch": 0.09684014869888476, "grad_norm": 5.309213890079893, "learning_rate": 9.88464843854567e-06, "loss": 0.142, "step": 2084 }, { "epoch": 0.09688661710037175, "grad_norm": 4.315462507308983, "learning_rate": 9.884487675576651e-06, "loss": 0.2066, "step": 2085 }, { "epoch": 0.09693308550185874, "grad_norm": 2.1346744848796604, "learning_rate": 9.884326801968844e-06, "loss": 0.0876, "step": 2086 }, { "epoch": 0.09697955390334573, "grad_norm": 1.5634693297394988, "learning_rate": 9.884165817725896e-06, "loss": 0.1043, "step": 2087 }, { "epoch": 0.09702602230483272, "grad_norm": 4.068776035703123, "learning_rate": 9.88400472285145e-06, "loss": 0.1286, "step": 2088 }, { "epoch": 0.09707249070631971, "grad_norm": 1.8567909064653996, "learning_rate": 9.883843517349158e-06, "loss": 0.1107, "step": 2089 }, { "epoch": 0.09711895910780668, "grad_norm": 1.0818663717499355, "learning_rate": 9.883682201222669e-06, "loss": 0.0857, "step": 2090 }, { "epoch": 0.09716542750929368, "grad_norm": 1.1402802225095654, "learning_rate": 9.883520774475638e-06, "loss": 0.1214, "step": 2091 }, { "epoch": 0.09721189591078067, "grad_norm": 1.8821096149232721, "learning_rate": 9.883359237111724e-06, "loss": 0.1638, "step": 2092 }, { "epoch": 0.09725836431226766, "grad_norm": 2.7931791524556737, "learning_rate": 9.883197589134582e-06, "loss": 0.1477, "step": 2093 }, { "epoch": 0.09730483271375465, "grad_norm": 0.9341084172986429, "learning_rate": 9.883035830547876e-06, "loss": 0.081, "step": 2094 }, { "epoch": 0.09735130111524164, "grad_norm": 2.4432809254660435, "learning_rate": 9.882873961355267e-06, "loss": 0.1594, "step": 2095 }, { "epoch": 0.09739776951672863, "grad_norm": 1.5334545310432097, "learning_rate": 9.882711981560425e-06, "loss": 0.128, "step": 2096 }, { "epoch": 0.09744423791821562, "grad_norm": 4.609723846274513, "learning_rate": 9.882549891167018e-06, "loss": 0.12, "step": 2097 }, { "epoch": 0.09749070631970261, "grad_norm": 1.2759840420056516, "learning_rate": 9.882387690178717e-06, "loss": 0.1009, "step": 2098 }, { "epoch": 0.09753717472118958, "grad_norm": 1.7153261660294492, "learning_rate": 9.882225378599196e-06, "loss": 0.1229, "step": 2099 }, { "epoch": 0.09758364312267657, "grad_norm": 1.5970705228688689, "learning_rate": 9.882062956432132e-06, "loss": 0.1101, "step": 2100 }, { "epoch": 0.09763011152416357, "grad_norm": 2.432234715103208, "learning_rate": 9.881900423681204e-06, "loss": 0.0776, "step": 2101 }, { "epoch": 0.09767657992565056, "grad_norm": 1.218955815582429, "learning_rate": 9.88173778035009e-06, "loss": 0.1092, "step": 2102 }, { "epoch": 0.09772304832713755, "grad_norm": 1.474385044682836, "learning_rate": 9.881575026442482e-06, "loss": 0.0878, "step": 2103 }, { "epoch": 0.09776951672862454, "grad_norm": 2.3631110713933214, "learning_rate": 9.88141216196206e-06, "loss": 0.0987, "step": 2104 }, { "epoch": 0.09781598513011153, "grad_norm": 2.5515806289651586, "learning_rate": 9.881249186912514e-06, "loss": 0.1477, "step": 2105 }, { "epoch": 0.09786245353159852, "grad_norm": 1.816358534856804, "learning_rate": 9.881086101297539e-06, "loss": 0.1453, "step": 2106 }, { "epoch": 0.09790892193308551, "grad_norm": 1.13851015740811, "learning_rate": 9.880922905120824e-06, "loss": 0.0833, "step": 2107 }, { "epoch": 0.09795539033457248, "grad_norm": 3.1879536103457826, "learning_rate": 9.88075959838607e-06, "loss": 0.1745, "step": 2108 }, { "epoch": 0.09800185873605947, "grad_norm": 1.3912941948840685, "learning_rate": 9.880596181096973e-06, "loss": 0.0787, "step": 2109 }, { "epoch": 0.09804832713754646, "grad_norm": 2.9122566558812673, "learning_rate": 9.880432653257235e-06, "loss": 0.2316, "step": 2110 }, { "epoch": 0.09809479553903346, "grad_norm": 2.4108675781992073, "learning_rate": 9.880269014870562e-06, "loss": 0.1969, "step": 2111 }, { "epoch": 0.09814126394052045, "grad_norm": 1.5427511047139637, "learning_rate": 9.88010526594066e-06, "loss": 0.0854, "step": 2112 }, { "epoch": 0.09818773234200744, "grad_norm": 1.4177928121500312, "learning_rate": 9.879941406471237e-06, "loss": 0.0894, "step": 2113 }, { "epoch": 0.09823420074349443, "grad_norm": 1.6527347724952934, "learning_rate": 9.879777436466005e-06, "loss": 0.1339, "step": 2114 }, { "epoch": 0.09828066914498142, "grad_norm": 2.414446654990442, "learning_rate": 9.879613355928679e-06, "loss": 0.0906, "step": 2115 }, { "epoch": 0.09832713754646841, "grad_norm": 1.5529030385897837, "learning_rate": 9.879449164862974e-06, "loss": 0.1644, "step": 2116 }, { "epoch": 0.09837360594795538, "grad_norm": 1.8493451513442476, "learning_rate": 9.879284863272608e-06, "loss": 0.0974, "step": 2117 }, { "epoch": 0.09842007434944237, "grad_norm": 2.3833675093790374, "learning_rate": 9.879120451161308e-06, "loss": 0.1089, "step": 2118 }, { "epoch": 0.09846654275092936, "grad_norm": 1.6666366117919118, "learning_rate": 9.878955928532792e-06, "loss": 0.1486, "step": 2119 }, { "epoch": 0.09851301115241635, "grad_norm": 2.7849625207853226, "learning_rate": 9.87879129539079e-06, "loss": 0.1491, "step": 2120 }, { "epoch": 0.09855947955390335, "grad_norm": 2.23501106031851, "learning_rate": 9.87862655173903e-06, "loss": 0.1168, "step": 2121 }, { "epoch": 0.09860594795539034, "grad_norm": 2.260792799130659, "learning_rate": 9.878461697581245e-06, "loss": 0.1297, "step": 2122 }, { "epoch": 0.09865241635687733, "grad_norm": 2.8606971572619284, "learning_rate": 9.878296732921167e-06, "loss": 0.1474, "step": 2123 }, { "epoch": 0.09869888475836432, "grad_norm": 1.6557179105416275, "learning_rate": 9.878131657762535e-06, "loss": 0.1225, "step": 2124 }, { "epoch": 0.0987453531598513, "grad_norm": 1.9278729460469253, "learning_rate": 9.877966472109086e-06, "loss": 0.1006, "step": 2125 }, { "epoch": 0.09879182156133828, "grad_norm": 1.3987304012690784, "learning_rate": 9.877801175964562e-06, "loss": 0.0942, "step": 2126 }, { "epoch": 0.09883828996282527, "grad_norm": 1.5510441900317087, "learning_rate": 9.877635769332707e-06, "loss": 0.158, "step": 2127 }, { "epoch": 0.09888475836431226, "grad_norm": 2.3760905535168253, "learning_rate": 9.87747025221727e-06, "loss": 0.1692, "step": 2128 }, { "epoch": 0.09893122676579925, "grad_norm": 1.6028950056500089, "learning_rate": 9.877304624621998e-06, "loss": 0.1177, "step": 2129 }, { "epoch": 0.09897769516728624, "grad_norm": 1.4110752884616788, "learning_rate": 9.877138886550642e-06, "loss": 0.0942, "step": 2130 }, { "epoch": 0.09902416356877324, "grad_norm": 2.0068183288568604, "learning_rate": 9.876973038006958e-06, "loss": 0.1127, "step": 2131 }, { "epoch": 0.09907063197026023, "grad_norm": 1.1009603391804887, "learning_rate": 9.876807078994702e-06, "loss": 0.0651, "step": 2132 }, { "epoch": 0.09911710037174722, "grad_norm": 3.235770491287913, "learning_rate": 9.876641009517632e-06, "loss": 0.1822, "step": 2133 }, { "epoch": 0.0991635687732342, "grad_norm": 1.5516846642393263, "learning_rate": 9.87647482957951e-06, "loss": 0.1344, "step": 2134 }, { "epoch": 0.09921003717472118, "grad_norm": 2.0989049163437903, "learning_rate": 9.876308539184102e-06, "loss": 0.086, "step": 2135 }, { "epoch": 0.09925650557620817, "grad_norm": 2.021619936199057, "learning_rate": 9.876142138335173e-06, "loss": 0.1413, "step": 2136 }, { "epoch": 0.09930297397769516, "grad_norm": 1.4380021869125477, "learning_rate": 9.875975627036493e-06, "loss": 0.1103, "step": 2137 }, { "epoch": 0.09934944237918215, "grad_norm": 1.7399381174883786, "learning_rate": 9.875809005291832e-06, "loss": 0.1718, "step": 2138 }, { "epoch": 0.09939591078066914, "grad_norm": 1.7467690866993622, "learning_rate": 9.875642273104966e-06, "loss": 0.0995, "step": 2139 }, { "epoch": 0.09944237918215613, "grad_norm": 0.7118404995069734, "learning_rate": 9.875475430479671e-06, "loss": 0.0594, "step": 2140 }, { "epoch": 0.09948884758364313, "grad_norm": 1.8704665869735826, "learning_rate": 9.875308477419726e-06, "loss": 0.1323, "step": 2141 }, { "epoch": 0.09953531598513012, "grad_norm": 3.338991020638518, "learning_rate": 9.875141413928913e-06, "loss": 0.1166, "step": 2142 }, { "epoch": 0.0995817843866171, "grad_norm": 2.2054899415009106, "learning_rate": 9.874974240011016e-06, "loss": 0.1575, "step": 2143 }, { "epoch": 0.09962825278810408, "grad_norm": 2.039930524059767, "learning_rate": 9.87480695566982e-06, "loss": 0.1457, "step": 2144 }, { "epoch": 0.09967472118959107, "grad_norm": 1.3942936029742878, "learning_rate": 9.874639560909118e-06, "loss": 0.1081, "step": 2145 }, { "epoch": 0.09972118959107806, "grad_norm": 1.0817417596938976, "learning_rate": 9.8744720557327e-06, "loss": 0.07, "step": 2146 }, { "epoch": 0.09976765799256505, "grad_norm": 1.3691434518498389, "learning_rate": 9.874304440144358e-06, "loss": 0.0999, "step": 2147 }, { "epoch": 0.09981412639405204, "grad_norm": 1.3756310411409896, "learning_rate": 9.87413671414789e-06, "loss": 0.0886, "step": 2148 }, { "epoch": 0.09986059479553903, "grad_norm": 1.6237619999134965, "learning_rate": 9.873968877747097e-06, "loss": 0.102, "step": 2149 }, { "epoch": 0.09990706319702602, "grad_norm": 1.3154150151322472, "learning_rate": 9.873800930945778e-06, "loss": 0.0759, "step": 2150 }, { "epoch": 0.09995353159851302, "grad_norm": 1.8831017698285748, "learning_rate": 9.873632873747739e-06, "loss": 0.0976, "step": 2151 }, { "epoch": 0.1, "grad_norm": 1.70945610888348, "learning_rate": 9.873464706156785e-06, "loss": 0.1135, "step": 2152 }, { "epoch": 0.10004646840148698, "grad_norm": 1.6178643478420054, "learning_rate": 9.873296428176728e-06, "loss": 0.115, "step": 2153 }, { "epoch": 0.10009293680297397, "grad_norm": 1.3114643537737891, "learning_rate": 9.873128039811377e-06, "loss": 0.098, "step": 2154 }, { "epoch": 0.10013940520446096, "grad_norm": 2.9423816495200885, "learning_rate": 9.872959541064548e-06, "loss": 0.1253, "step": 2155 }, { "epoch": 0.10018587360594795, "grad_norm": 2.1364766758698632, "learning_rate": 9.872790931940056e-06, "loss": 0.1611, "step": 2156 }, { "epoch": 0.10023234200743494, "grad_norm": 0.8940884421749151, "learning_rate": 9.872622212441719e-06, "loss": 0.0502, "step": 2157 }, { "epoch": 0.10027881040892193, "grad_norm": 1.2874046565601374, "learning_rate": 9.872453382573361e-06, "loss": 0.0745, "step": 2158 }, { "epoch": 0.10032527881040892, "grad_norm": 1.505421264142253, "learning_rate": 9.872284442338808e-06, "loss": 0.0602, "step": 2159 }, { "epoch": 0.10037174721189591, "grad_norm": 1.4419277645313298, "learning_rate": 9.872115391741883e-06, "loss": 0.0805, "step": 2160 }, { "epoch": 0.1004182156133829, "grad_norm": 1.9590833391484044, "learning_rate": 9.871946230786417e-06, "loss": 0.0807, "step": 2161 }, { "epoch": 0.10046468401486988, "grad_norm": 1.8148742084107883, "learning_rate": 9.871776959476241e-06, "loss": 0.1131, "step": 2162 }, { "epoch": 0.10051115241635687, "grad_norm": 2.4607324993387203, "learning_rate": 9.871607577815188e-06, "loss": 0.0781, "step": 2163 }, { "epoch": 0.10055762081784386, "grad_norm": 2.278088500033675, "learning_rate": 9.871438085807098e-06, "loss": 0.139, "step": 2164 }, { "epoch": 0.10060408921933085, "grad_norm": 1.294605177720818, "learning_rate": 9.871268483455807e-06, "loss": 0.0901, "step": 2165 }, { "epoch": 0.10065055762081784, "grad_norm": 1.9798805084383493, "learning_rate": 9.87109877076516e-06, "loss": 0.1916, "step": 2166 }, { "epoch": 0.10069702602230483, "grad_norm": 2.2179693284574244, "learning_rate": 9.870928947738997e-06, "loss": 0.1632, "step": 2167 }, { "epoch": 0.10074349442379182, "grad_norm": 2.260275111801391, "learning_rate": 9.870759014381168e-06, "loss": 0.147, "step": 2168 }, { "epoch": 0.10078996282527881, "grad_norm": 1.5096663083868254, "learning_rate": 9.870588970695521e-06, "loss": 0.1198, "step": 2169 }, { "epoch": 0.1008364312267658, "grad_norm": 1.5092172348211366, "learning_rate": 9.870418816685905e-06, "loss": 0.1252, "step": 2170 }, { "epoch": 0.10088289962825278, "grad_norm": 1.1211700131101463, "learning_rate": 9.870248552356179e-06, "loss": 0.0783, "step": 2171 }, { "epoch": 0.10092936802973977, "grad_norm": 1.7332215920288156, "learning_rate": 9.870078177710198e-06, "loss": 0.1137, "step": 2172 }, { "epoch": 0.10097583643122676, "grad_norm": 1.5993734826135166, "learning_rate": 9.86990769275182e-06, "loss": 0.1266, "step": 2173 }, { "epoch": 0.10102230483271375, "grad_norm": 2.097616661006677, "learning_rate": 9.869737097484908e-06, "loss": 0.1274, "step": 2174 }, { "epoch": 0.10106877323420074, "grad_norm": 1.8824476913083272, "learning_rate": 9.869566391913324e-06, "loss": 0.1711, "step": 2175 }, { "epoch": 0.10111524163568773, "grad_norm": 2.3490717945112207, "learning_rate": 9.869395576040936e-06, "loss": 0.1118, "step": 2176 }, { "epoch": 0.10116171003717472, "grad_norm": 2.9652709245513513, "learning_rate": 9.869224649871613e-06, "loss": 0.0638, "step": 2177 }, { "epoch": 0.10120817843866171, "grad_norm": 2.568492364997396, "learning_rate": 9.869053613409229e-06, "loss": 0.1156, "step": 2178 }, { "epoch": 0.1012546468401487, "grad_norm": 1.305789552537695, "learning_rate": 9.868882466657654e-06, "loss": 0.0705, "step": 2179 }, { "epoch": 0.10130111524163568, "grad_norm": 1.266620455938361, "learning_rate": 9.868711209620769e-06, "loss": 0.1059, "step": 2180 }, { "epoch": 0.10134758364312267, "grad_norm": 2.1871982089502255, "learning_rate": 9.86853984230245e-06, "loss": 0.1358, "step": 2181 }, { "epoch": 0.10139405204460966, "grad_norm": 1.2668306656696906, "learning_rate": 9.868368364706577e-06, "loss": 0.0911, "step": 2182 }, { "epoch": 0.10144052044609665, "grad_norm": 4.451230083521672, "learning_rate": 9.86819677683704e-06, "loss": 0.0832, "step": 2183 }, { "epoch": 0.10148698884758364, "grad_norm": 3.119473003315454, "learning_rate": 9.86802507869772e-06, "loss": 0.1363, "step": 2184 }, { "epoch": 0.10153345724907063, "grad_norm": 3.9030703647527973, "learning_rate": 9.867853270292512e-06, "loss": 0.1315, "step": 2185 }, { "epoch": 0.10157992565055762, "grad_norm": 1.2706911343395342, "learning_rate": 9.8676813516253e-06, "loss": 0.0826, "step": 2186 }, { "epoch": 0.10162639405204461, "grad_norm": 1.7379740768896679, "learning_rate": 9.867509322699984e-06, "loss": 0.1077, "step": 2187 }, { "epoch": 0.1016728624535316, "grad_norm": 1.1471731565237877, "learning_rate": 9.867337183520458e-06, "loss": 0.0656, "step": 2188 }, { "epoch": 0.1017193308550186, "grad_norm": 1.733834110738573, "learning_rate": 9.867164934090623e-06, "loss": 0.0784, "step": 2189 }, { "epoch": 0.10176579925650557, "grad_norm": 3.9722285857298396, "learning_rate": 9.86699257441438e-06, "loss": 0.1045, "step": 2190 }, { "epoch": 0.10181226765799256, "grad_norm": 1.2560016330219743, "learning_rate": 9.86682010449563e-06, "loss": 0.0866, "step": 2191 }, { "epoch": 0.10185873605947955, "grad_norm": 1.5314996231049915, "learning_rate": 9.866647524338287e-06, "loss": 0.069, "step": 2192 }, { "epoch": 0.10190520446096654, "grad_norm": 2.083566922410472, "learning_rate": 9.866474833946253e-06, "loss": 0.1066, "step": 2193 }, { "epoch": 0.10195167286245353, "grad_norm": 1.47568516857474, "learning_rate": 9.866302033323441e-06, "loss": 0.1196, "step": 2194 }, { "epoch": 0.10199814126394052, "grad_norm": 1.1943861471505515, "learning_rate": 9.866129122473769e-06, "loss": 0.0575, "step": 2195 }, { "epoch": 0.10204460966542751, "grad_norm": 0.9326979374317051, "learning_rate": 9.865956101401148e-06, "loss": 0.0998, "step": 2196 }, { "epoch": 0.1020910780669145, "grad_norm": 1.5046606970682426, "learning_rate": 9.865782970109501e-06, "loss": 0.0903, "step": 2197 }, { "epoch": 0.1021375464684015, "grad_norm": 3.368942371879187, "learning_rate": 9.86560972860275e-06, "loss": 0.1627, "step": 2198 }, { "epoch": 0.10218401486988847, "grad_norm": 1.7536972060579168, "learning_rate": 9.865436376884817e-06, "loss": 0.0907, "step": 2199 }, { "epoch": 0.10223048327137546, "grad_norm": 1.5962253835774376, "learning_rate": 9.865262914959628e-06, "loss": 0.0966, "step": 2200 }, { "epoch": 0.10227695167286245, "grad_norm": 2.3225880710593247, "learning_rate": 9.865089342831115e-06, "loss": 0.0896, "step": 2201 }, { "epoch": 0.10232342007434944, "grad_norm": 1.5805679399556283, "learning_rate": 9.864915660503208e-06, "loss": 0.116, "step": 2202 }, { "epoch": 0.10236988847583643, "grad_norm": 1.162213233775755, "learning_rate": 9.86474186797984e-06, "loss": 0.0851, "step": 2203 }, { "epoch": 0.10241635687732342, "grad_norm": 2.1813425278874052, "learning_rate": 9.864567965264949e-06, "loss": 0.1056, "step": 2204 }, { "epoch": 0.10246282527881041, "grad_norm": 11.026289726819282, "learning_rate": 9.864393952362473e-06, "loss": 0.2226, "step": 2205 }, { "epoch": 0.1025092936802974, "grad_norm": 3.0183391872510357, "learning_rate": 9.864219829276355e-06, "loss": 0.1262, "step": 2206 }, { "epoch": 0.1025557620817844, "grad_norm": 2.8127599805009145, "learning_rate": 9.86404559601054e-06, "loss": 0.1714, "step": 2207 }, { "epoch": 0.10260223048327137, "grad_norm": 2.559100304031761, "learning_rate": 9.863871252568968e-06, "loss": 0.1221, "step": 2208 }, { "epoch": 0.10264869888475836, "grad_norm": 3.0700832769768196, "learning_rate": 9.863696798955597e-06, "loss": 0.1797, "step": 2209 }, { "epoch": 0.10269516728624535, "grad_norm": 3.0635280414142203, "learning_rate": 9.863522235174374e-06, "loss": 0.1107, "step": 2210 }, { "epoch": 0.10274163568773234, "grad_norm": 2.1546647050552092, "learning_rate": 9.863347561229253e-06, "loss": 0.196, "step": 2211 }, { "epoch": 0.10278810408921933, "grad_norm": 1.4576030014334094, "learning_rate": 9.863172777124193e-06, "loss": 0.1159, "step": 2212 }, { "epoch": 0.10283457249070632, "grad_norm": 1.3999265928671558, "learning_rate": 9.86299788286315e-06, "loss": 0.0967, "step": 2213 }, { "epoch": 0.10288104089219331, "grad_norm": 1.6299642101862524, "learning_rate": 9.862822878450085e-06, "loss": 0.0904, "step": 2214 }, { "epoch": 0.1029275092936803, "grad_norm": 2.810394753401393, "learning_rate": 9.862647763888967e-06, "loss": 0.1251, "step": 2215 }, { "epoch": 0.10297397769516729, "grad_norm": 3.235221573284403, "learning_rate": 9.862472539183757e-06, "loss": 0.1936, "step": 2216 }, { "epoch": 0.10302044609665427, "grad_norm": 2.093281387272465, "learning_rate": 9.862297204338427e-06, "loss": 0.1702, "step": 2217 }, { "epoch": 0.10306691449814126, "grad_norm": 2.261155503183074, "learning_rate": 9.862121759356947e-06, "loss": 0.0976, "step": 2218 }, { "epoch": 0.10311338289962825, "grad_norm": 1.5339214540742496, "learning_rate": 9.861946204243293e-06, "loss": 0.093, "step": 2219 }, { "epoch": 0.10315985130111524, "grad_norm": 3.1382258633082536, "learning_rate": 9.86177053900144e-06, "loss": 0.1718, "step": 2220 }, { "epoch": 0.10320631970260223, "grad_norm": 1.3289799128379094, "learning_rate": 9.861594763635367e-06, "loss": 0.1226, "step": 2221 }, { "epoch": 0.10325278810408922, "grad_norm": 1.805640866703044, "learning_rate": 9.861418878149056e-06, "loss": 0.1422, "step": 2222 }, { "epoch": 0.10329925650557621, "grad_norm": 2.1619026178981073, "learning_rate": 9.86124288254649e-06, "loss": 0.1554, "step": 2223 }, { "epoch": 0.1033457249070632, "grad_norm": 3.0987414153480177, "learning_rate": 9.861066776831658e-06, "loss": 0.1434, "step": 2224 }, { "epoch": 0.10339219330855019, "grad_norm": 2.353298132041161, "learning_rate": 9.860890561008547e-06, "loss": 0.1149, "step": 2225 }, { "epoch": 0.10343866171003717, "grad_norm": 1.6583718751080412, "learning_rate": 9.860714235081149e-06, "loss": 0.1041, "step": 2226 }, { "epoch": 0.10348513011152416, "grad_norm": 1.3648657241102327, "learning_rate": 9.860537799053458e-06, "loss": 0.1118, "step": 2227 }, { "epoch": 0.10353159851301115, "grad_norm": 1.3592530678147519, "learning_rate": 9.860361252929469e-06, "loss": 0.1119, "step": 2228 }, { "epoch": 0.10357806691449814, "grad_norm": 1.71217796723735, "learning_rate": 9.860184596713182e-06, "loss": 0.0991, "step": 2229 }, { "epoch": 0.10362453531598513, "grad_norm": 2.4310695376966143, "learning_rate": 9.8600078304086e-06, "loss": 0.1766, "step": 2230 }, { "epoch": 0.10367100371747212, "grad_norm": 2.1493642940373303, "learning_rate": 9.859830954019724e-06, "loss": 0.1074, "step": 2231 }, { "epoch": 0.10371747211895911, "grad_norm": 2.0149128952446476, "learning_rate": 9.859653967550565e-06, "loss": 0.101, "step": 2232 }, { "epoch": 0.1037639405204461, "grad_norm": 2.474721975039942, "learning_rate": 9.859476871005125e-06, "loss": 0.1082, "step": 2233 }, { "epoch": 0.10381040892193309, "grad_norm": 1.3908313356998776, "learning_rate": 9.859299664387422e-06, "loss": 0.0758, "step": 2234 }, { "epoch": 0.10385687732342007, "grad_norm": 2.2227003983762716, "learning_rate": 9.859122347701467e-06, "loss": 0.1229, "step": 2235 }, { "epoch": 0.10390334572490706, "grad_norm": 2.5916392855176844, "learning_rate": 9.858944920951277e-06, "loss": 0.1255, "step": 2236 }, { "epoch": 0.10394981412639405, "grad_norm": 3.8488581822840535, "learning_rate": 9.858767384140869e-06, "loss": 0.1289, "step": 2237 }, { "epoch": 0.10399628252788104, "grad_norm": 1.5582391274199674, "learning_rate": 9.858589737274266e-06, "loss": 0.0828, "step": 2238 }, { "epoch": 0.10404275092936803, "grad_norm": 2.03042330297753, "learning_rate": 9.858411980355493e-06, "loss": 0.0934, "step": 2239 }, { "epoch": 0.10408921933085502, "grad_norm": 1.9555688955967407, "learning_rate": 9.858234113388577e-06, "loss": 0.1348, "step": 2240 }, { "epoch": 0.10413568773234201, "grad_norm": 2.6394142990872442, "learning_rate": 9.858056136377542e-06, "loss": 0.0932, "step": 2241 }, { "epoch": 0.104182156133829, "grad_norm": 1.620812746703823, "learning_rate": 9.857878049326424e-06, "loss": 0.1394, "step": 2242 }, { "epoch": 0.10422862453531599, "grad_norm": 0.9526268654553142, "learning_rate": 9.857699852239255e-06, "loss": 0.0735, "step": 2243 }, { "epoch": 0.10427509293680297, "grad_norm": 1.0706701101727216, "learning_rate": 9.85752154512007e-06, "loss": 0.0737, "step": 2244 }, { "epoch": 0.10432156133828996, "grad_norm": 2.20971350004058, "learning_rate": 9.857343127972911e-06, "loss": 0.1651, "step": 2245 }, { "epoch": 0.10436802973977695, "grad_norm": 1.234897730141085, "learning_rate": 9.857164600801819e-06, "loss": 0.1006, "step": 2246 }, { "epoch": 0.10441449814126394, "grad_norm": 1.3103122344256317, "learning_rate": 9.856985963610835e-06, "loss": 0.0817, "step": 2247 }, { "epoch": 0.10446096654275093, "grad_norm": 1.3908253104696475, "learning_rate": 9.856807216404009e-06, "loss": 0.0633, "step": 2248 }, { "epoch": 0.10450743494423792, "grad_norm": 1.3257826372024424, "learning_rate": 9.856628359185387e-06, "loss": 0.0699, "step": 2249 }, { "epoch": 0.10455390334572491, "grad_norm": 1.2670865076618385, "learning_rate": 9.85644939195902e-06, "loss": 0.0956, "step": 2250 }, { "epoch": 0.1046003717472119, "grad_norm": 1.893709008885871, "learning_rate": 9.856270314728965e-06, "loss": 0.0867, "step": 2251 }, { "epoch": 0.10464684014869889, "grad_norm": 2.4620174025951234, "learning_rate": 9.856091127499274e-06, "loss": 0.1755, "step": 2252 }, { "epoch": 0.10469330855018587, "grad_norm": 1.2006970776148191, "learning_rate": 9.855911830274009e-06, "loss": 0.0894, "step": 2253 }, { "epoch": 0.10473977695167286, "grad_norm": 3.7526718745530028, "learning_rate": 9.85573242305723e-06, "loss": 0.1762, "step": 2254 }, { "epoch": 0.10478624535315985, "grad_norm": 1.9059086336421789, "learning_rate": 9.855552905853002e-06, "loss": 0.0709, "step": 2255 }, { "epoch": 0.10483271375464684, "grad_norm": 2.5438646044350777, "learning_rate": 9.85537327866539e-06, "loss": 0.0904, "step": 2256 }, { "epoch": 0.10487918215613383, "grad_norm": 3.580899462038051, "learning_rate": 9.855193541498462e-06, "loss": 0.1643, "step": 2257 }, { "epoch": 0.10492565055762082, "grad_norm": 1.532309273097261, "learning_rate": 9.855013694356292e-06, "loss": 0.1248, "step": 2258 }, { "epoch": 0.10497211895910781, "grad_norm": 1.2773013209846475, "learning_rate": 9.85483373724295e-06, "loss": 0.0709, "step": 2259 }, { "epoch": 0.1050185873605948, "grad_norm": 1.1481213333051394, "learning_rate": 9.854653670162516e-06, "loss": 0.0944, "step": 2260 }, { "epoch": 0.10506505576208179, "grad_norm": 3.484438814910281, "learning_rate": 9.854473493119066e-06, "loss": 0.0857, "step": 2261 }, { "epoch": 0.10511152416356877, "grad_norm": 1.9404046928638474, "learning_rate": 9.854293206116683e-06, "loss": 0.1058, "step": 2262 }, { "epoch": 0.10515799256505576, "grad_norm": 1.9345070528112969, "learning_rate": 9.854112809159448e-06, "loss": 0.1458, "step": 2263 }, { "epoch": 0.10520446096654275, "grad_norm": 2.418042021974656, "learning_rate": 9.85393230225145e-06, "loss": 0.1265, "step": 2264 }, { "epoch": 0.10525092936802974, "grad_norm": 2.8916113676836357, "learning_rate": 9.853751685396777e-06, "loss": 0.1482, "step": 2265 }, { "epoch": 0.10529739776951673, "grad_norm": 0.9610669825591752, "learning_rate": 9.85357095859952e-06, "loss": 0.0436, "step": 2266 }, { "epoch": 0.10534386617100372, "grad_norm": 1.9532956697134198, "learning_rate": 9.853390121863772e-06, "loss": 0.1947, "step": 2267 }, { "epoch": 0.10539033457249071, "grad_norm": 0.9507253353752719, "learning_rate": 9.85320917519363e-06, "loss": 0.0653, "step": 2268 }, { "epoch": 0.1054368029739777, "grad_norm": 2.2523510839692817, "learning_rate": 9.853028118593193e-06, "loss": 0.0936, "step": 2269 }, { "epoch": 0.10548327137546469, "grad_norm": 2.68932350662237, "learning_rate": 9.852846952066562e-06, "loss": 0.1215, "step": 2270 }, { "epoch": 0.10552973977695167, "grad_norm": 1.3270643946734069, "learning_rate": 9.852665675617837e-06, "loss": 0.0598, "step": 2271 }, { "epoch": 0.10557620817843866, "grad_norm": 1.4448038314747575, "learning_rate": 9.85248428925113e-06, "loss": 0.0922, "step": 2272 }, { "epoch": 0.10562267657992565, "grad_norm": 1.4752493800240059, "learning_rate": 9.852302792970545e-06, "loss": 0.075, "step": 2273 }, { "epoch": 0.10566914498141264, "grad_norm": 1.765281868344418, "learning_rate": 9.852121186780195e-06, "loss": 0.1088, "step": 2274 }, { "epoch": 0.10571561338289963, "grad_norm": 1.6694316458093954, "learning_rate": 9.851939470684192e-06, "loss": 0.0887, "step": 2275 }, { "epoch": 0.10576208178438662, "grad_norm": 1.5962983713402157, "learning_rate": 9.851757644686656e-06, "loss": 0.1096, "step": 2276 }, { "epoch": 0.10580855018587361, "grad_norm": 1.5319426996406138, "learning_rate": 9.851575708791702e-06, "loss": 0.0888, "step": 2277 }, { "epoch": 0.1058550185873606, "grad_norm": 1.7985858887053014, "learning_rate": 9.851393663003453e-06, "loss": 0.1191, "step": 2278 }, { "epoch": 0.10590148698884759, "grad_norm": 1.1554534359280386, "learning_rate": 9.85121150732603e-06, "loss": 0.0866, "step": 2279 }, { "epoch": 0.10594795539033457, "grad_norm": 1.183649999485869, "learning_rate": 9.851029241763561e-06, "loss": 0.0678, "step": 2280 }, { "epoch": 0.10599442379182156, "grad_norm": 2.2602181436099933, "learning_rate": 9.850846866320174e-06, "loss": 0.1074, "step": 2281 }, { "epoch": 0.10604089219330855, "grad_norm": 3.4652550943081106, "learning_rate": 9.850664380999999e-06, "loss": 0.1568, "step": 2282 }, { "epoch": 0.10608736059479554, "grad_norm": 1.57572764354307, "learning_rate": 9.850481785807173e-06, "loss": 0.1529, "step": 2283 }, { "epoch": 0.10613382899628253, "grad_norm": 1.4607971911471251, "learning_rate": 9.850299080745826e-06, "loss": 0.1112, "step": 2284 }, { "epoch": 0.10618029739776952, "grad_norm": 1.2484357011229228, "learning_rate": 9.850116265820104e-06, "loss": 0.07, "step": 2285 }, { "epoch": 0.10622676579925651, "grad_norm": 1.4618748577472545, "learning_rate": 9.84993334103414e-06, "loss": 0.1111, "step": 2286 }, { "epoch": 0.1062732342007435, "grad_norm": 2.3428109000510884, "learning_rate": 9.849750306392085e-06, "loss": 0.1614, "step": 2287 }, { "epoch": 0.10631970260223049, "grad_norm": 3.4795617796045355, "learning_rate": 9.849567161898079e-06, "loss": 0.0796, "step": 2288 }, { "epoch": 0.10636617100371747, "grad_norm": 1.9127312327386186, "learning_rate": 9.849383907556273e-06, "loss": 0.1431, "step": 2289 }, { "epoch": 0.10641263940520446, "grad_norm": 3.37009553045932, "learning_rate": 9.849200543370818e-06, "loss": 0.1327, "step": 2290 }, { "epoch": 0.10645910780669145, "grad_norm": 2.3806209466520016, "learning_rate": 9.849017069345867e-06, "loss": 0.1541, "step": 2291 }, { "epoch": 0.10650557620817844, "grad_norm": 1.866629014802557, "learning_rate": 9.848833485485577e-06, "loss": 0.153, "step": 2292 }, { "epoch": 0.10655204460966543, "grad_norm": 1.5613930749180274, "learning_rate": 9.848649791794105e-06, "loss": 0.0845, "step": 2293 }, { "epoch": 0.10659851301115242, "grad_norm": 1.7839331630815685, "learning_rate": 9.84846598827561e-06, "loss": 0.1033, "step": 2294 }, { "epoch": 0.10664498141263941, "grad_norm": 2.071876931910499, "learning_rate": 9.84828207493426e-06, "loss": 0.0827, "step": 2295 }, { "epoch": 0.1066914498141264, "grad_norm": 1.7189536875369433, "learning_rate": 9.848098051774216e-06, "loss": 0.0977, "step": 2296 }, { "epoch": 0.10673791821561339, "grad_norm": 1.5652333148985071, "learning_rate": 9.84791391879965e-06, "loss": 0.0862, "step": 2297 }, { "epoch": 0.10678438661710037, "grad_norm": 1.9781980788571585, "learning_rate": 9.84772967601473e-06, "loss": 0.0928, "step": 2298 }, { "epoch": 0.10683085501858736, "grad_norm": 1.087330502023449, "learning_rate": 9.847545323423634e-06, "loss": 0.1063, "step": 2299 }, { "epoch": 0.10687732342007435, "grad_norm": 1.7190243074099092, "learning_rate": 9.84736086103053e-06, "loss": 0.0986, "step": 2300 }, { "epoch": 0.10692379182156134, "grad_norm": 1.5542306726572255, "learning_rate": 9.847176288839604e-06, "loss": 0.1066, "step": 2301 }, { "epoch": 0.10697026022304833, "grad_norm": 0.9935243643650783, "learning_rate": 9.846991606855032e-06, "loss": 0.0795, "step": 2302 }, { "epoch": 0.10701672862453532, "grad_norm": 1.5737539278332404, "learning_rate": 9.846806815080998e-06, "loss": 0.1062, "step": 2303 }, { "epoch": 0.10706319702602231, "grad_norm": 1.9315512948611884, "learning_rate": 9.84662191352169e-06, "loss": 0.1486, "step": 2304 }, { "epoch": 0.1071096654275093, "grad_norm": 3.8070042535053368, "learning_rate": 9.846436902181295e-06, "loss": 0.1652, "step": 2305 }, { "epoch": 0.10715613382899629, "grad_norm": 1.8821292863665742, "learning_rate": 9.846251781064002e-06, "loss": 0.0876, "step": 2306 }, { "epoch": 0.10720260223048327, "grad_norm": 2.8678120972738146, "learning_rate": 9.846066550174007e-06, "loss": 0.1479, "step": 2307 }, { "epoch": 0.10724907063197026, "grad_norm": 4.9645669851452965, "learning_rate": 9.845881209515503e-06, "loss": 0.1363, "step": 2308 }, { "epoch": 0.10729553903345725, "grad_norm": 2.207842326585868, "learning_rate": 9.84569575909269e-06, "loss": 0.0792, "step": 2309 }, { "epoch": 0.10734200743494424, "grad_norm": 2.8153822852153123, "learning_rate": 9.845510198909768e-06, "loss": 0.0866, "step": 2310 }, { "epoch": 0.10738847583643123, "grad_norm": 4.022707144260597, "learning_rate": 9.845324528970942e-06, "loss": 0.1427, "step": 2311 }, { "epoch": 0.10743494423791822, "grad_norm": 1.5963173042268342, "learning_rate": 9.845138749280413e-06, "loss": 0.0964, "step": 2312 }, { "epoch": 0.10748141263940521, "grad_norm": 1.5836285158707717, "learning_rate": 9.844952859842391e-06, "loss": 0.1196, "step": 2313 }, { "epoch": 0.1075278810408922, "grad_norm": 2.176238506469598, "learning_rate": 9.84476686066109e-06, "loss": 0.0801, "step": 2314 }, { "epoch": 0.10757434944237919, "grad_norm": 1.3143320958918998, "learning_rate": 9.84458075174072e-06, "loss": 0.0816, "step": 2315 }, { "epoch": 0.10762081784386617, "grad_norm": 2.419769705709905, "learning_rate": 9.844394533085497e-06, "loss": 0.2028, "step": 2316 }, { "epoch": 0.10766728624535316, "grad_norm": 0.9272386045274515, "learning_rate": 9.844208204699638e-06, "loss": 0.0867, "step": 2317 }, { "epoch": 0.10771375464684015, "grad_norm": 1.8397034287205531, "learning_rate": 9.844021766587369e-06, "loss": 0.0896, "step": 2318 }, { "epoch": 0.10776022304832714, "grad_norm": 1.8516195347797573, "learning_rate": 9.843835218752904e-06, "loss": 0.0971, "step": 2319 }, { "epoch": 0.10780669144981413, "grad_norm": 1.3478181574913843, "learning_rate": 9.843648561200476e-06, "loss": 0.1137, "step": 2320 }, { "epoch": 0.10785315985130112, "grad_norm": 1.4611622499170913, "learning_rate": 9.843461793934308e-06, "loss": 0.1165, "step": 2321 }, { "epoch": 0.10789962825278811, "grad_norm": 2.4245728516408005, "learning_rate": 9.843274916958634e-06, "loss": 0.1004, "step": 2322 }, { "epoch": 0.1079460966542751, "grad_norm": 1.4601541122879247, "learning_rate": 9.843087930277685e-06, "loss": 0.0734, "step": 2323 }, { "epoch": 0.10799256505576209, "grad_norm": 2.8165541889033774, "learning_rate": 9.842900833895699e-06, "loss": 0.1458, "step": 2324 }, { "epoch": 0.10803903345724906, "grad_norm": 3.2185292403756094, "learning_rate": 9.842713627816911e-06, "loss": 0.1224, "step": 2325 }, { "epoch": 0.10808550185873605, "grad_norm": 1.4632692317897364, "learning_rate": 9.842526312045561e-06, "loss": 0.0926, "step": 2326 }, { "epoch": 0.10813197026022305, "grad_norm": 1.3985382532709574, "learning_rate": 9.842338886585896e-06, "loss": 0.091, "step": 2327 }, { "epoch": 0.10817843866171004, "grad_norm": 0.9698383138379524, "learning_rate": 9.842151351442157e-06, "loss": 0.0801, "step": 2328 }, { "epoch": 0.10822490706319703, "grad_norm": 1.7788726901071072, "learning_rate": 9.841963706618593e-06, "loss": 0.0679, "step": 2329 }, { "epoch": 0.10827137546468402, "grad_norm": 1.281516949457811, "learning_rate": 9.841775952119457e-06, "loss": 0.0533, "step": 2330 }, { "epoch": 0.108317843866171, "grad_norm": 1.4977014929011103, "learning_rate": 9.841588087949e-06, "loss": 0.0671, "step": 2331 }, { "epoch": 0.108364312267658, "grad_norm": 2.4274739109789403, "learning_rate": 9.841400114111474e-06, "loss": 0.1386, "step": 2332 }, { "epoch": 0.10841078066914499, "grad_norm": 1.437966907162882, "learning_rate": 9.841212030611143e-06, "loss": 0.0779, "step": 2333 }, { "epoch": 0.10845724907063196, "grad_norm": 4.493426267374619, "learning_rate": 9.841023837452263e-06, "loss": 0.0914, "step": 2334 }, { "epoch": 0.10850371747211895, "grad_norm": 1.750381348086529, "learning_rate": 9.840835534639097e-06, "loss": 0.0974, "step": 2335 }, { "epoch": 0.10855018587360594, "grad_norm": 1.1824927196732022, "learning_rate": 9.840647122175913e-06, "loss": 0.0716, "step": 2336 }, { "epoch": 0.10859665427509294, "grad_norm": 1.4624427878750967, "learning_rate": 9.840458600066976e-06, "loss": 0.0923, "step": 2337 }, { "epoch": 0.10864312267657993, "grad_norm": 2.524946749794118, "learning_rate": 9.840269968316557e-06, "loss": 0.1302, "step": 2338 }, { "epoch": 0.10868959107806692, "grad_norm": 2.094915344214921, "learning_rate": 9.84008122692893e-06, "loss": 0.1044, "step": 2339 }, { "epoch": 0.1087360594795539, "grad_norm": 1.4153066317780754, "learning_rate": 9.839892375908369e-06, "loss": 0.1436, "step": 2340 }, { "epoch": 0.1087825278810409, "grad_norm": 1.1498050530451491, "learning_rate": 9.839703415259152e-06, "loss": 0.0897, "step": 2341 }, { "epoch": 0.10882899628252789, "grad_norm": 1.2129106877883957, "learning_rate": 9.839514344985558e-06, "loss": 0.0773, "step": 2342 }, { "epoch": 0.10887546468401486, "grad_norm": 2.459761810831415, "learning_rate": 9.83932516509187e-06, "loss": 0.147, "step": 2343 }, { "epoch": 0.10892193308550185, "grad_norm": 3.0895475467156306, "learning_rate": 9.839135875582375e-06, "loss": 0.0903, "step": 2344 }, { "epoch": 0.10896840148698884, "grad_norm": 2.6270627893689436, "learning_rate": 9.83894647646136e-06, "loss": 0.1205, "step": 2345 }, { "epoch": 0.10901486988847583, "grad_norm": 1.2817690681505753, "learning_rate": 9.838756967733112e-06, "loss": 0.0838, "step": 2346 }, { "epoch": 0.10906133828996283, "grad_norm": 2.872899074107582, "learning_rate": 9.838567349401928e-06, "loss": 0.1406, "step": 2347 }, { "epoch": 0.10910780669144982, "grad_norm": 1.8317191621591418, "learning_rate": 9.838377621472101e-06, "loss": 0.0848, "step": 2348 }, { "epoch": 0.1091542750929368, "grad_norm": 1.2715540489448516, "learning_rate": 9.838187783947928e-06, "loss": 0.1036, "step": 2349 }, { "epoch": 0.1092007434944238, "grad_norm": 1.5481951393211533, "learning_rate": 9.837997836833708e-06, "loss": 0.1055, "step": 2350 }, { "epoch": 0.10924721189591079, "grad_norm": 1.7127408258652794, "learning_rate": 9.837807780133748e-06, "loss": 0.1171, "step": 2351 }, { "epoch": 0.10929368029739776, "grad_norm": 1.4627270200117266, "learning_rate": 9.837617613852348e-06, "loss": 0.1126, "step": 2352 }, { "epoch": 0.10934014869888475, "grad_norm": 1.3698494060376876, "learning_rate": 9.837427337993819e-06, "loss": 0.1291, "step": 2353 }, { "epoch": 0.10938661710037174, "grad_norm": 1.447394705686133, "learning_rate": 9.837236952562468e-06, "loss": 0.108, "step": 2354 }, { "epoch": 0.10943308550185873, "grad_norm": 1.414406489175215, "learning_rate": 9.837046457562609e-06, "loss": 0.154, "step": 2355 }, { "epoch": 0.10947955390334572, "grad_norm": 2.1397422794214824, "learning_rate": 9.836855852998559e-06, "loss": 0.1244, "step": 2356 }, { "epoch": 0.10952602230483272, "grad_norm": 1.6046898583883709, "learning_rate": 9.83666513887463e-06, "loss": 0.0891, "step": 2357 }, { "epoch": 0.1095724907063197, "grad_norm": 3.5201734267141775, "learning_rate": 9.836474315195148e-06, "loss": 0.1091, "step": 2358 }, { "epoch": 0.1096189591078067, "grad_norm": 1.0749047586112437, "learning_rate": 9.836283381964432e-06, "loss": 0.0844, "step": 2359 }, { "epoch": 0.10966542750929369, "grad_norm": 1.1751378555927945, "learning_rate": 9.836092339186805e-06, "loss": 0.0492, "step": 2360 }, { "epoch": 0.10971189591078066, "grad_norm": 2.8258845124068515, "learning_rate": 9.8359011868666e-06, "loss": 0.1465, "step": 2361 }, { "epoch": 0.10975836431226765, "grad_norm": 1.722162716600891, "learning_rate": 9.835709925008142e-06, "loss": 0.1783, "step": 2362 }, { "epoch": 0.10980483271375464, "grad_norm": 1.495290909434556, "learning_rate": 9.835518553615764e-06, "loss": 0.1022, "step": 2363 }, { "epoch": 0.10985130111524163, "grad_norm": 1.6271688531480526, "learning_rate": 9.835327072693802e-06, "loss": 0.1086, "step": 2364 }, { "epoch": 0.10989776951672862, "grad_norm": 1.6121800319636446, "learning_rate": 9.835135482246591e-06, "loss": 0.0936, "step": 2365 }, { "epoch": 0.10994423791821561, "grad_norm": 2.1223763993084224, "learning_rate": 9.834943782278475e-06, "loss": 0.0916, "step": 2366 }, { "epoch": 0.1099907063197026, "grad_norm": 1.2397383445597567, "learning_rate": 9.834751972793793e-06, "loss": 0.1038, "step": 2367 }, { "epoch": 0.1100371747211896, "grad_norm": 1.4425415573795777, "learning_rate": 9.834560053796888e-06, "loss": 0.1105, "step": 2368 }, { "epoch": 0.11008364312267659, "grad_norm": 2.1617506462106695, "learning_rate": 9.834368025292112e-06, "loss": 0.0719, "step": 2369 }, { "epoch": 0.11013011152416356, "grad_norm": 2.597783778262208, "learning_rate": 9.834175887283812e-06, "loss": 0.1301, "step": 2370 }, { "epoch": 0.11017657992565055, "grad_norm": 0.8116336696817391, "learning_rate": 9.83398363977634e-06, "loss": 0.0675, "step": 2371 }, { "epoch": 0.11022304832713754, "grad_norm": 2.244697928671715, "learning_rate": 9.833791282774051e-06, "loss": 0.1539, "step": 2372 }, { "epoch": 0.11026951672862453, "grad_norm": 2.4744618698513046, "learning_rate": 9.833598816281302e-06, "loss": 0.1312, "step": 2373 }, { "epoch": 0.11031598513011152, "grad_norm": 1.641074057543857, "learning_rate": 9.833406240302452e-06, "loss": 0.1139, "step": 2374 }, { "epoch": 0.11036245353159851, "grad_norm": 1.6716212222800373, "learning_rate": 9.833213554841864e-06, "loss": 0.1027, "step": 2375 }, { "epoch": 0.1104089219330855, "grad_norm": 1.7055079274061553, "learning_rate": 9.833020759903901e-06, "loss": 0.1378, "step": 2376 }, { "epoch": 0.1104553903345725, "grad_norm": 1.5130909877903926, "learning_rate": 9.832827855492932e-06, "loss": 0.1012, "step": 2377 }, { "epoch": 0.11050185873605949, "grad_norm": 2.475546814603583, "learning_rate": 9.832634841613327e-06, "loss": 0.1462, "step": 2378 }, { "epoch": 0.11054832713754646, "grad_norm": 1.4309111536657217, "learning_rate": 9.832441718269455e-06, "loss": 0.0921, "step": 2379 }, { "epoch": 0.11059479553903345, "grad_norm": 1.4179535840817319, "learning_rate": 9.832248485465692e-06, "loss": 0.0896, "step": 2380 }, { "epoch": 0.11064126394052044, "grad_norm": 2.574260287916275, "learning_rate": 9.832055143206416e-06, "loss": 0.1244, "step": 2381 }, { "epoch": 0.11068773234200743, "grad_norm": 1.7281716532707925, "learning_rate": 9.831861691496004e-06, "loss": 0.0917, "step": 2382 }, { "epoch": 0.11073420074349442, "grad_norm": 2.1023606816461324, "learning_rate": 9.83166813033884e-06, "loss": 0.1299, "step": 2383 }, { "epoch": 0.11078066914498141, "grad_norm": 3.0955935980961877, "learning_rate": 9.831474459739307e-06, "loss": 0.1741, "step": 2384 }, { "epoch": 0.1108271375464684, "grad_norm": 1.3105047931146616, "learning_rate": 9.831280679701794e-06, "loss": 0.1, "step": 2385 }, { "epoch": 0.1108736059479554, "grad_norm": 2.738961759846569, "learning_rate": 9.831086790230687e-06, "loss": 0.1029, "step": 2386 }, { "epoch": 0.11092007434944239, "grad_norm": 2.7758740127861214, "learning_rate": 9.830892791330381e-06, "loss": 0.1392, "step": 2387 }, { "epoch": 0.11096654275092936, "grad_norm": 1.4026623747165867, "learning_rate": 9.830698683005268e-06, "loss": 0.1169, "step": 2388 }, { "epoch": 0.11101301115241635, "grad_norm": 3.4349944051324828, "learning_rate": 9.830504465259746e-06, "loss": 0.1823, "step": 2389 }, { "epoch": 0.11105947955390334, "grad_norm": 1.1723914488058407, "learning_rate": 9.830310138098214e-06, "loss": 0.171, "step": 2390 }, { "epoch": 0.11110594795539033, "grad_norm": 1.6719030267404442, "learning_rate": 9.830115701525073e-06, "loss": 0.1075, "step": 2391 }, { "epoch": 0.11115241635687732, "grad_norm": 1.2273717894678096, "learning_rate": 9.829921155544729e-06, "loss": 0.104, "step": 2392 }, { "epoch": 0.11119888475836431, "grad_norm": 1.3051790296224817, "learning_rate": 9.829726500161584e-06, "loss": 0.0856, "step": 2393 }, { "epoch": 0.1112453531598513, "grad_norm": 1.8511666871482142, "learning_rate": 9.829531735380054e-06, "loss": 0.1656, "step": 2394 }, { "epoch": 0.1112918215613383, "grad_norm": 1.6423644377297737, "learning_rate": 9.829336861204546e-06, "loss": 0.0725, "step": 2395 }, { "epoch": 0.11133828996282528, "grad_norm": 2.1435822191103413, "learning_rate": 9.829141877639476e-06, "loss": 0.1557, "step": 2396 }, { "epoch": 0.11138475836431226, "grad_norm": 2.136351690515692, "learning_rate": 9.82894678468926e-06, "loss": 0.1166, "step": 2397 }, { "epoch": 0.11143122676579925, "grad_norm": 1.8866608053516554, "learning_rate": 9.828751582358315e-06, "loss": 0.1211, "step": 2398 }, { "epoch": 0.11147769516728624, "grad_norm": 0.9985963569764897, "learning_rate": 9.828556270651066e-06, "loss": 0.0868, "step": 2399 }, { "epoch": 0.11152416356877323, "grad_norm": 2.2790748310012265, "learning_rate": 9.828360849571935e-06, "loss": 0.1458, "step": 2400 }, { "epoch": 0.11157063197026022, "grad_norm": 2.235403572595463, "learning_rate": 9.82816531912535e-06, "loss": 0.0903, "step": 2401 }, { "epoch": 0.11161710037174721, "grad_norm": 0.8849819941598742, "learning_rate": 9.827969679315738e-06, "loss": 0.0565, "step": 2402 }, { "epoch": 0.1116635687732342, "grad_norm": 1.8235038943842123, "learning_rate": 9.827773930147532e-06, "loss": 0.0718, "step": 2403 }, { "epoch": 0.1117100371747212, "grad_norm": 2.179069995792522, "learning_rate": 9.827578071625164e-06, "loss": 0.0874, "step": 2404 }, { "epoch": 0.11175650557620818, "grad_norm": 1.1209502358262786, "learning_rate": 9.827382103753073e-06, "loss": 0.054, "step": 2405 }, { "epoch": 0.11180297397769516, "grad_norm": 1.381990809919877, "learning_rate": 9.827186026535697e-06, "loss": 0.0654, "step": 2406 }, { "epoch": 0.11184944237918215, "grad_norm": 1.5742423029188255, "learning_rate": 9.826989839977477e-06, "loss": 0.092, "step": 2407 }, { "epoch": 0.11189591078066914, "grad_norm": 0.9598017789957224, "learning_rate": 9.826793544082855e-06, "loss": 0.0668, "step": 2408 }, { "epoch": 0.11194237918215613, "grad_norm": 1.0420953885306536, "learning_rate": 9.826597138856281e-06, "loss": 0.0721, "step": 2409 }, { "epoch": 0.11198884758364312, "grad_norm": 1.853937824033759, "learning_rate": 9.826400624302201e-06, "loss": 0.1079, "step": 2410 }, { "epoch": 0.11203531598513011, "grad_norm": 2.0406548189434512, "learning_rate": 9.826204000425068e-06, "loss": 0.1451, "step": 2411 }, { "epoch": 0.1120817843866171, "grad_norm": 1.2840745526856838, "learning_rate": 9.826007267229333e-06, "loss": 0.0887, "step": 2412 }, { "epoch": 0.1121282527881041, "grad_norm": 1.2955885222701482, "learning_rate": 9.825810424719456e-06, "loss": 0.1609, "step": 2413 }, { "epoch": 0.11217472118959108, "grad_norm": 2.1376989690034933, "learning_rate": 9.825613472899894e-06, "loss": 0.0825, "step": 2414 }, { "epoch": 0.11222118959107806, "grad_norm": 4.4278799256053665, "learning_rate": 9.825416411775106e-06, "loss": 0.1492, "step": 2415 }, { "epoch": 0.11226765799256505, "grad_norm": 3.4162657791391915, "learning_rate": 9.825219241349559e-06, "loss": 0.1909, "step": 2416 }, { "epoch": 0.11231412639405204, "grad_norm": 1.0910773828076261, "learning_rate": 9.825021961627717e-06, "loss": 0.0671, "step": 2417 }, { "epoch": 0.11236059479553903, "grad_norm": 1.6743852339376026, "learning_rate": 9.82482457261405e-06, "loss": 0.0852, "step": 2418 }, { "epoch": 0.11240706319702602, "grad_norm": 1.8510270963636846, "learning_rate": 9.82462707431303e-06, "loss": 0.0797, "step": 2419 }, { "epoch": 0.11245353159851301, "grad_norm": 2.0954932101574055, "learning_rate": 9.824429466729126e-06, "loss": 0.126, "step": 2420 }, { "epoch": 0.1125, "grad_norm": 2.853233608421225, "learning_rate": 9.824231749866818e-06, "loss": 0.2044, "step": 2421 }, { "epoch": 0.112546468401487, "grad_norm": 1.3146860683780681, "learning_rate": 9.824033923730584e-06, "loss": 0.1084, "step": 2422 }, { "epoch": 0.11259293680297398, "grad_norm": 1.5197815502178749, "learning_rate": 9.823835988324904e-06, "loss": 0.0793, "step": 2423 }, { "epoch": 0.11263940520446096, "grad_norm": 1.240342723347162, "learning_rate": 9.823637943654261e-06, "loss": 0.0643, "step": 2424 }, { "epoch": 0.11268587360594795, "grad_norm": 1.9001480387623642, "learning_rate": 9.823439789723143e-06, "loss": 0.0977, "step": 2425 }, { "epoch": 0.11273234200743494, "grad_norm": 2.403319216860145, "learning_rate": 9.823241526536037e-06, "loss": 0.1296, "step": 2426 }, { "epoch": 0.11277881040892193, "grad_norm": 1.531051485359536, "learning_rate": 9.823043154097433e-06, "loss": 0.0788, "step": 2427 }, { "epoch": 0.11282527881040892, "grad_norm": 1.1155383379800246, "learning_rate": 9.822844672411827e-06, "loss": 0.1064, "step": 2428 }, { "epoch": 0.11287174721189591, "grad_norm": 1.6674077517795431, "learning_rate": 9.822646081483713e-06, "loss": 0.0882, "step": 2429 }, { "epoch": 0.1129182156133829, "grad_norm": 2.1033895029491805, "learning_rate": 9.82244738131759e-06, "loss": 0.1282, "step": 2430 }, { "epoch": 0.11296468401486989, "grad_norm": 1.1718972708162718, "learning_rate": 9.822248571917956e-06, "loss": 0.0951, "step": 2431 }, { "epoch": 0.11301115241635688, "grad_norm": 2.5228760303859237, "learning_rate": 9.822049653289318e-06, "loss": 0.1806, "step": 2432 }, { "epoch": 0.11305762081784386, "grad_norm": 4.5806807878806675, "learning_rate": 9.82185062543618e-06, "loss": 0.1641, "step": 2433 }, { "epoch": 0.11310408921933085, "grad_norm": 1.509683260076268, "learning_rate": 9.821651488363052e-06, "loss": 0.045, "step": 2434 }, { "epoch": 0.11315055762081784, "grad_norm": 1.2243317266884945, "learning_rate": 9.82145224207444e-06, "loss": 0.1417, "step": 2435 }, { "epoch": 0.11319702602230483, "grad_norm": 1.25121381006702, "learning_rate": 9.821252886574862e-06, "loss": 0.0758, "step": 2436 }, { "epoch": 0.11324349442379182, "grad_norm": 1.4851713676673943, "learning_rate": 9.821053421868832e-06, "loss": 0.0858, "step": 2437 }, { "epoch": 0.11328996282527881, "grad_norm": 2.0579953177440737, "learning_rate": 9.820853847960868e-06, "loss": 0.0984, "step": 2438 }, { "epoch": 0.1133364312267658, "grad_norm": 1.8534291060418728, "learning_rate": 9.82065416485549e-06, "loss": 0.1318, "step": 2439 }, { "epoch": 0.11338289962825279, "grad_norm": 1.725298014179234, "learning_rate": 9.820454372557221e-06, "loss": 0.0789, "step": 2440 }, { "epoch": 0.11342936802973978, "grad_norm": 0.983701091270338, "learning_rate": 9.820254471070588e-06, "loss": 0.0752, "step": 2441 }, { "epoch": 0.11347583643122676, "grad_norm": 4.015841589178946, "learning_rate": 9.820054460400117e-06, "loss": 0.0748, "step": 2442 }, { "epoch": 0.11352230483271375, "grad_norm": 1.7794646893670754, "learning_rate": 9.819854340550342e-06, "loss": 0.1256, "step": 2443 }, { "epoch": 0.11356877323420074, "grad_norm": 1.0248969193563364, "learning_rate": 9.81965411152579e-06, "loss": 0.1342, "step": 2444 }, { "epoch": 0.11361524163568773, "grad_norm": 1.5855525239407215, "learning_rate": 9.819453773331e-06, "loss": 0.1341, "step": 2445 }, { "epoch": 0.11366171003717472, "grad_norm": 2.605262795831239, "learning_rate": 9.819253325970512e-06, "loss": 0.0795, "step": 2446 }, { "epoch": 0.11370817843866171, "grad_norm": 2.805968428794113, "learning_rate": 9.819052769448862e-06, "loss": 0.115, "step": 2447 }, { "epoch": 0.1137546468401487, "grad_norm": 2.076338334526439, "learning_rate": 9.818852103770598e-06, "loss": 0.0992, "step": 2448 }, { "epoch": 0.11380111524163569, "grad_norm": 2.462276479160389, "learning_rate": 9.818651328940258e-06, "loss": 0.1153, "step": 2449 }, { "epoch": 0.11384758364312268, "grad_norm": 1.9331905838416825, "learning_rate": 9.818450444962397e-06, "loss": 0.0975, "step": 2450 }, { "epoch": 0.11389405204460966, "grad_norm": 1.7935040373202389, "learning_rate": 9.81824945184156e-06, "loss": 0.1044, "step": 2451 }, { "epoch": 0.11394052044609665, "grad_norm": 2.09113811718346, "learning_rate": 9.818048349582303e-06, "loss": 0.1012, "step": 2452 }, { "epoch": 0.11398698884758364, "grad_norm": 1.5180304577473045, "learning_rate": 9.81784713818918e-06, "loss": 0.1111, "step": 2453 }, { "epoch": 0.11403345724907063, "grad_norm": 1.582149609144018, "learning_rate": 9.817645817666747e-06, "loss": 0.1088, "step": 2454 }, { "epoch": 0.11407992565055762, "grad_norm": 1.1720152531429298, "learning_rate": 9.817444388019568e-06, "loss": 0.0781, "step": 2455 }, { "epoch": 0.11412639405204461, "grad_norm": 2.6112795635253936, "learning_rate": 9.817242849252202e-06, "loss": 0.0939, "step": 2456 }, { "epoch": 0.1141728624535316, "grad_norm": 2.1110211091037865, "learning_rate": 9.817041201369215e-06, "loss": 0.1633, "step": 2457 }, { "epoch": 0.11421933085501859, "grad_norm": 1.1206422175229869, "learning_rate": 9.816839444375176e-06, "loss": 0.0658, "step": 2458 }, { "epoch": 0.11426579925650558, "grad_norm": 1.3468906263744806, "learning_rate": 9.816637578274653e-06, "loss": 0.1031, "step": 2459 }, { "epoch": 0.11431226765799256, "grad_norm": 2.151434079931266, "learning_rate": 9.81643560307222e-06, "loss": 0.0616, "step": 2460 }, { "epoch": 0.11435873605947955, "grad_norm": 1.6639103503272508, "learning_rate": 9.816233518772452e-06, "loss": 0.0581, "step": 2461 }, { "epoch": 0.11440520446096654, "grad_norm": 9.683504769986357, "learning_rate": 9.816031325379924e-06, "loss": 0.1451, "step": 2462 }, { "epoch": 0.11445167286245353, "grad_norm": 2.784725964058443, "learning_rate": 9.81582902289922e-06, "loss": 0.1359, "step": 2463 }, { "epoch": 0.11449814126394052, "grad_norm": 1.5756567512240707, "learning_rate": 9.815626611334919e-06, "loss": 0.0865, "step": 2464 }, { "epoch": 0.11454460966542751, "grad_norm": 1.61905266202358, "learning_rate": 9.815424090691607e-06, "loss": 0.1071, "step": 2465 }, { "epoch": 0.1145910780669145, "grad_norm": 1.4989410399193157, "learning_rate": 9.815221460973871e-06, "loss": 0.0726, "step": 2466 }, { "epoch": 0.11463754646840149, "grad_norm": 2.277645948743509, "learning_rate": 9.8150187221863e-06, "loss": 0.0893, "step": 2467 }, { "epoch": 0.11468401486988848, "grad_norm": 2.527163228078553, "learning_rate": 9.81481587433349e-06, "loss": 0.136, "step": 2468 }, { "epoch": 0.11473048327137546, "grad_norm": 2.347693474763028, "learning_rate": 9.81461291742003e-06, "loss": 0.107, "step": 2469 }, { "epoch": 0.11477695167286245, "grad_norm": 2.206993451567654, "learning_rate": 9.814409851450522e-06, "loss": 0.0917, "step": 2470 }, { "epoch": 0.11482342007434944, "grad_norm": 1.5436417231983757, "learning_rate": 9.814206676429563e-06, "loss": 0.0907, "step": 2471 }, { "epoch": 0.11486988847583643, "grad_norm": 1.767529252935529, "learning_rate": 9.814003392361756e-06, "loss": 0.1012, "step": 2472 }, { "epoch": 0.11491635687732342, "grad_norm": 1.9462369424671557, "learning_rate": 9.813799999251705e-06, "loss": 0.1167, "step": 2473 }, { "epoch": 0.11496282527881041, "grad_norm": 1.8823949498943156, "learning_rate": 9.813596497104019e-06, "loss": 0.1108, "step": 2474 }, { "epoch": 0.1150092936802974, "grad_norm": 2.1218945558198916, "learning_rate": 9.813392885923306e-06, "loss": 0.1109, "step": 2475 }, { "epoch": 0.11505576208178439, "grad_norm": 1.4729793075707367, "learning_rate": 9.813189165714176e-06, "loss": 0.0629, "step": 2476 }, { "epoch": 0.11510223048327138, "grad_norm": 1.3811191670589622, "learning_rate": 9.812985336481248e-06, "loss": 0.0972, "step": 2477 }, { "epoch": 0.11514869888475836, "grad_norm": 3.7349838640147097, "learning_rate": 9.812781398229135e-06, "loss": 0.0834, "step": 2478 }, { "epoch": 0.11519516728624535, "grad_norm": 1.4118184109642562, "learning_rate": 9.812577350962459e-06, "loss": 0.0876, "step": 2479 }, { "epoch": 0.11524163568773234, "grad_norm": 2.199039916657604, "learning_rate": 9.81237319468584e-06, "loss": 0.1379, "step": 2480 }, { "epoch": 0.11528810408921933, "grad_norm": 2.1363459033291043, "learning_rate": 9.812168929403904e-06, "loss": 0.0886, "step": 2481 }, { "epoch": 0.11533457249070632, "grad_norm": 0.8762799724578475, "learning_rate": 9.811964555121275e-06, "loss": 0.0514, "step": 2482 }, { "epoch": 0.11538104089219331, "grad_norm": 1.4580844768987191, "learning_rate": 9.811760071842587e-06, "loss": 0.1447, "step": 2483 }, { "epoch": 0.1154275092936803, "grad_norm": 1.9174577847836165, "learning_rate": 9.811555479572468e-06, "loss": 0.0817, "step": 2484 }, { "epoch": 0.11547397769516729, "grad_norm": 1.58006636449947, "learning_rate": 9.811350778315553e-06, "loss": 0.1069, "step": 2485 }, { "epoch": 0.11552044609665428, "grad_norm": 2.589898875352749, "learning_rate": 9.811145968076481e-06, "loss": 0.138, "step": 2486 }, { "epoch": 0.11556691449814126, "grad_norm": 1.6524537898302978, "learning_rate": 9.810941048859886e-06, "loss": 0.1152, "step": 2487 }, { "epoch": 0.11561338289962825, "grad_norm": 1.5571560965833005, "learning_rate": 9.810736020670415e-06, "loss": 0.1079, "step": 2488 }, { "epoch": 0.11565985130111524, "grad_norm": 2.3521204560772784, "learning_rate": 9.81053088351271e-06, "loss": 0.1846, "step": 2489 }, { "epoch": 0.11570631970260223, "grad_norm": 1.4650850197029799, "learning_rate": 9.810325637391415e-06, "loss": 0.1081, "step": 2490 }, { "epoch": 0.11575278810408922, "grad_norm": 1.6661310950691421, "learning_rate": 9.810120282311185e-06, "loss": 0.0622, "step": 2491 }, { "epoch": 0.11579925650557621, "grad_norm": 1.3532810402344115, "learning_rate": 9.809914818276666e-06, "loss": 0.0838, "step": 2492 }, { "epoch": 0.1158457249070632, "grad_norm": 2.6662223090201342, "learning_rate": 9.809709245292515e-06, "loss": 0.1149, "step": 2493 }, { "epoch": 0.11589219330855019, "grad_norm": 3.7166943494497904, "learning_rate": 9.809503563363387e-06, "loss": 0.1291, "step": 2494 }, { "epoch": 0.11593866171003718, "grad_norm": 1.8420528694163727, "learning_rate": 9.809297772493939e-06, "loss": 0.1135, "step": 2495 }, { "epoch": 0.11598513011152416, "grad_norm": 1.769382838510112, "learning_rate": 9.809091872688838e-06, "loss": 0.155, "step": 2496 }, { "epoch": 0.11603159851301115, "grad_norm": 0.9595376434120432, "learning_rate": 9.808885863952745e-06, "loss": 0.0722, "step": 2497 }, { "epoch": 0.11607806691449814, "grad_norm": 2.100387490893197, "learning_rate": 9.808679746290324e-06, "loss": 0.0963, "step": 2498 }, { "epoch": 0.11612453531598513, "grad_norm": 1.1596483985557093, "learning_rate": 9.808473519706245e-06, "loss": 0.0567, "step": 2499 }, { "epoch": 0.11617100371747212, "grad_norm": 1.9345552386288964, "learning_rate": 9.808267184205182e-06, "loss": 0.1273, "step": 2500 }, { "epoch": 0.11621747211895911, "grad_norm": 2.670509397099432, "learning_rate": 9.808060739791806e-06, "loss": 0.1161, "step": 2501 }, { "epoch": 0.1162639405204461, "grad_norm": 2.214868948260286, "learning_rate": 9.807854186470794e-06, "loss": 0.1532, "step": 2502 }, { "epoch": 0.11631040892193309, "grad_norm": 1.9150552257937605, "learning_rate": 9.807647524246824e-06, "loss": 0.1166, "step": 2503 }, { "epoch": 0.11635687732342008, "grad_norm": 1.3486067741080738, "learning_rate": 9.807440753124578e-06, "loss": 0.1115, "step": 2504 }, { "epoch": 0.11640334572490706, "grad_norm": 1.3023849731810968, "learning_rate": 9.80723387310874e-06, "loss": 0.1113, "step": 2505 }, { "epoch": 0.11644981412639405, "grad_norm": 2.5115258696698324, "learning_rate": 9.807026884203994e-06, "loss": 0.0902, "step": 2506 }, { "epoch": 0.11649628252788104, "grad_norm": 1.5191662028843378, "learning_rate": 9.806819786415029e-06, "loss": 0.0564, "step": 2507 }, { "epoch": 0.11654275092936803, "grad_norm": 1.7446990831260243, "learning_rate": 9.806612579746539e-06, "loss": 0.1067, "step": 2508 }, { "epoch": 0.11658921933085502, "grad_norm": 1.8910147258035614, "learning_rate": 9.806405264203214e-06, "loss": 0.1396, "step": 2509 }, { "epoch": 0.11663568773234201, "grad_norm": 3.6700016298469103, "learning_rate": 9.806197839789751e-06, "loss": 0.1297, "step": 2510 }, { "epoch": 0.116682156133829, "grad_norm": 2.2940202864074575, "learning_rate": 9.805990306510849e-06, "loss": 0.1513, "step": 2511 }, { "epoch": 0.11672862453531599, "grad_norm": 2.511437383825135, "learning_rate": 9.805782664371206e-06, "loss": 0.1057, "step": 2512 }, { "epoch": 0.11677509293680298, "grad_norm": 1.1248431841405468, "learning_rate": 9.805574913375531e-06, "loss": 0.0583, "step": 2513 }, { "epoch": 0.11682156133828996, "grad_norm": 1.3391111396001616, "learning_rate": 9.805367053528526e-06, "loss": 0.0763, "step": 2514 }, { "epoch": 0.11686802973977695, "grad_norm": 1.2292127318970003, "learning_rate": 9.805159084834897e-06, "loss": 0.0951, "step": 2515 }, { "epoch": 0.11691449814126394, "grad_norm": 1.420736230077858, "learning_rate": 9.80495100729936e-06, "loss": 0.0792, "step": 2516 }, { "epoch": 0.11696096654275093, "grad_norm": 2.2307684317574896, "learning_rate": 9.804742820926625e-06, "loss": 0.1022, "step": 2517 }, { "epoch": 0.11700743494423792, "grad_norm": 1.4433090609616321, "learning_rate": 9.804534525721407e-06, "loss": 0.0974, "step": 2518 }, { "epoch": 0.11705390334572491, "grad_norm": 1.7160657547474707, "learning_rate": 9.804326121688426e-06, "loss": 0.1261, "step": 2519 }, { "epoch": 0.1171003717472119, "grad_norm": 2.096712638452986, "learning_rate": 9.804117608832402e-06, "loss": 0.1154, "step": 2520 }, { "epoch": 0.11714684014869889, "grad_norm": 1.9741454733206194, "learning_rate": 9.803908987158058e-06, "loss": 0.1065, "step": 2521 }, { "epoch": 0.11719330855018588, "grad_norm": 1.2097326513039897, "learning_rate": 9.80370025667012e-06, "loss": 0.0806, "step": 2522 }, { "epoch": 0.11723977695167287, "grad_norm": 1.1986614231864152, "learning_rate": 9.803491417373314e-06, "loss": 0.078, "step": 2523 }, { "epoch": 0.11728624535315985, "grad_norm": 1.7072864499201297, "learning_rate": 9.803282469272372e-06, "loss": 0.0744, "step": 2524 }, { "epoch": 0.11733271375464684, "grad_norm": 1.9490763357212046, "learning_rate": 9.803073412372028e-06, "loss": 0.1631, "step": 2525 }, { "epoch": 0.11737918215613383, "grad_norm": 1.783935684836179, "learning_rate": 9.802864246677016e-06, "loss": 0.1424, "step": 2526 }, { "epoch": 0.11742565055762082, "grad_norm": 1.4500827958685272, "learning_rate": 9.802654972192073e-06, "loss": 0.1069, "step": 2527 }, { "epoch": 0.11747211895910781, "grad_norm": 2.1280739690498494, "learning_rate": 9.802445588921941e-06, "loss": 0.0941, "step": 2528 }, { "epoch": 0.1175185873605948, "grad_norm": 1.4299311221463944, "learning_rate": 9.802236096871361e-06, "loss": 0.111, "step": 2529 }, { "epoch": 0.11756505576208179, "grad_norm": 2.0745152247195056, "learning_rate": 9.80202649604508e-06, "loss": 0.1354, "step": 2530 }, { "epoch": 0.11761152416356878, "grad_norm": 1.4329993973627286, "learning_rate": 9.801816786447844e-06, "loss": 0.0907, "step": 2531 }, { "epoch": 0.11765799256505577, "grad_norm": 2.4536491730555094, "learning_rate": 9.801606968084405e-06, "loss": 0.0831, "step": 2532 }, { "epoch": 0.11770446096654275, "grad_norm": 1.9653747981478589, "learning_rate": 9.801397040959512e-06, "loss": 0.1207, "step": 2533 }, { "epoch": 0.11775092936802974, "grad_norm": 2.4294953385994176, "learning_rate": 9.801187005077925e-06, "loss": 0.1327, "step": 2534 }, { "epoch": 0.11779739776951673, "grad_norm": 2.1072600440324893, "learning_rate": 9.8009768604444e-06, "loss": 0.1261, "step": 2535 }, { "epoch": 0.11784386617100372, "grad_norm": 1.3302699946907215, "learning_rate": 9.800766607063693e-06, "loss": 0.0945, "step": 2536 }, { "epoch": 0.11789033457249071, "grad_norm": 1.429808535156463, "learning_rate": 9.800556244940573e-06, "loss": 0.0917, "step": 2537 }, { "epoch": 0.1179368029739777, "grad_norm": 2.8165044290900005, "learning_rate": 9.8003457740798e-06, "loss": 0.2093, "step": 2538 }, { "epoch": 0.11798327137546469, "grad_norm": 2.973081566685356, "learning_rate": 9.800135194486143e-06, "loss": 0.1082, "step": 2539 }, { "epoch": 0.11802973977695168, "grad_norm": 2.0948925028638095, "learning_rate": 9.799924506164372e-06, "loss": 0.1162, "step": 2540 }, { "epoch": 0.11807620817843867, "grad_norm": 1.8286215516580435, "learning_rate": 9.799713709119259e-06, "loss": 0.0953, "step": 2541 }, { "epoch": 0.11812267657992565, "grad_norm": 2.187041743451464, "learning_rate": 9.799502803355579e-06, "loss": 0.1068, "step": 2542 }, { "epoch": 0.11816914498141264, "grad_norm": 1.323443336579329, "learning_rate": 9.79929178887811e-06, "loss": 0.1063, "step": 2543 }, { "epoch": 0.11821561338289963, "grad_norm": 1.7711353275207766, "learning_rate": 9.799080665691629e-06, "loss": 0.1038, "step": 2544 }, { "epoch": 0.11826208178438662, "grad_norm": 2.238186184433218, "learning_rate": 9.798869433800921e-06, "loss": 0.069, "step": 2545 }, { "epoch": 0.1183085501858736, "grad_norm": 1.4257896304372746, "learning_rate": 9.798658093210768e-06, "loss": 0.1028, "step": 2546 }, { "epoch": 0.1183550185873606, "grad_norm": 1.5858036326043996, "learning_rate": 9.798446643925961e-06, "loss": 0.0682, "step": 2547 }, { "epoch": 0.11840148698884759, "grad_norm": 1.9238972244937738, "learning_rate": 9.798235085951287e-06, "loss": 0.0933, "step": 2548 }, { "epoch": 0.11844795539033458, "grad_norm": 2.581304614540706, "learning_rate": 9.798023419291537e-06, "loss": 0.1565, "step": 2549 }, { "epoch": 0.11849442379182157, "grad_norm": 1.7336343030159338, "learning_rate": 9.797811643951507e-06, "loss": 0.0852, "step": 2550 }, { "epoch": 0.11854089219330854, "grad_norm": 1.406816792719141, "learning_rate": 9.797599759935995e-06, "loss": 0.1138, "step": 2551 }, { "epoch": 0.11858736059479554, "grad_norm": 1.442295016273511, "learning_rate": 9.797387767249797e-06, "loss": 0.1157, "step": 2552 }, { "epoch": 0.11863382899628253, "grad_norm": 1.8724826341972036, "learning_rate": 9.797175665897718e-06, "loss": 0.1, "step": 2553 }, { "epoch": 0.11868029739776952, "grad_norm": 2.04678774553546, "learning_rate": 9.79696345588456e-06, "loss": 0.1162, "step": 2554 }, { "epoch": 0.1187267657992565, "grad_norm": 3.0430936018956722, "learning_rate": 9.796751137215133e-06, "loss": 0.1109, "step": 2555 }, { "epoch": 0.1187732342007435, "grad_norm": 2.1681989114990303, "learning_rate": 9.796538709894243e-06, "loss": 0.1072, "step": 2556 }, { "epoch": 0.11881970260223049, "grad_norm": 1.3947437893095558, "learning_rate": 9.796326173926705e-06, "loss": 0.0927, "step": 2557 }, { "epoch": 0.11886617100371748, "grad_norm": 2.455616126798843, "learning_rate": 9.796113529317328e-06, "loss": 0.1309, "step": 2558 }, { "epoch": 0.11891263940520447, "grad_norm": 2.2881508673534916, "learning_rate": 9.795900776070933e-06, "loss": 0.1516, "step": 2559 }, { "epoch": 0.11895910780669144, "grad_norm": 1.1400469405904305, "learning_rate": 9.79568791419234e-06, "loss": 0.0548, "step": 2560 }, { "epoch": 0.11900557620817843, "grad_norm": 1.886368563607801, "learning_rate": 9.795474943686364e-06, "loss": 0.1494, "step": 2561 }, { "epoch": 0.11905204460966543, "grad_norm": 1.978471485385007, "learning_rate": 9.795261864557836e-06, "loss": 0.1439, "step": 2562 }, { "epoch": 0.11909851301115242, "grad_norm": 2.7762176125627582, "learning_rate": 9.79504867681158e-06, "loss": 0.1653, "step": 2563 }, { "epoch": 0.1191449814126394, "grad_norm": 1.800122920834153, "learning_rate": 9.794835380452425e-06, "loss": 0.1662, "step": 2564 }, { "epoch": 0.1191914498141264, "grad_norm": 1.5792071936039744, "learning_rate": 9.7946219754852e-06, "loss": 0.107, "step": 2565 }, { "epoch": 0.11923791821561339, "grad_norm": 1.883008828134637, "learning_rate": 9.794408461914745e-06, "loss": 0.0975, "step": 2566 }, { "epoch": 0.11928438661710038, "grad_norm": 2.047381358101899, "learning_rate": 9.79419483974589e-06, "loss": 0.1093, "step": 2567 }, { "epoch": 0.11933085501858737, "grad_norm": 1.517500555910082, "learning_rate": 9.793981108983475e-06, "loss": 0.1237, "step": 2568 }, { "epoch": 0.11937732342007434, "grad_norm": 2.091672903159208, "learning_rate": 9.793767269632343e-06, "loss": 0.1433, "step": 2569 }, { "epoch": 0.11942379182156133, "grad_norm": 1.122904742878996, "learning_rate": 9.793553321697337e-06, "loss": 0.0673, "step": 2570 }, { "epoch": 0.11947026022304832, "grad_norm": 2.5898459844405806, "learning_rate": 9.793339265183303e-06, "loss": 0.1637, "step": 2571 }, { "epoch": 0.11951672862453531, "grad_norm": 2.4222177634305346, "learning_rate": 9.79312510009509e-06, "loss": 0.1307, "step": 2572 }, { "epoch": 0.1195631970260223, "grad_norm": 1.252745630327469, "learning_rate": 9.79291082643755e-06, "loss": 0.1002, "step": 2573 }, { "epoch": 0.1196096654275093, "grad_norm": 1.0588906545334726, "learning_rate": 9.792696444215534e-06, "loss": 0.0566, "step": 2574 }, { "epoch": 0.11965613382899629, "grad_norm": 1.286549176405257, "learning_rate": 9.7924819534339e-06, "loss": 0.1322, "step": 2575 }, { "epoch": 0.11970260223048328, "grad_norm": 1.1220365337688507, "learning_rate": 9.792267354097505e-06, "loss": 0.0823, "step": 2576 }, { "epoch": 0.11974907063197027, "grad_norm": 1.0390874147456381, "learning_rate": 9.792052646211211e-06, "loss": 0.0622, "step": 2577 }, { "epoch": 0.11979553903345724, "grad_norm": 1.4585467238110508, "learning_rate": 9.791837829779882e-06, "loss": 0.0971, "step": 2578 }, { "epoch": 0.11984200743494423, "grad_norm": 1.3961107886228148, "learning_rate": 9.791622904808382e-06, "loss": 0.078, "step": 2579 }, { "epoch": 0.11988847583643122, "grad_norm": 1.5959334203305546, "learning_rate": 9.79140787130158e-06, "loss": 0.0941, "step": 2580 }, { "epoch": 0.11993494423791821, "grad_norm": 1.4015462974247086, "learning_rate": 9.791192729264346e-06, "loss": 0.0594, "step": 2581 }, { "epoch": 0.1199814126394052, "grad_norm": 1.9591360355555154, "learning_rate": 9.790977478701555e-06, "loss": 0.1279, "step": 2582 }, { "epoch": 0.1200278810408922, "grad_norm": 2.2906421229002563, "learning_rate": 9.790762119618083e-06, "loss": 0.1069, "step": 2583 }, { "epoch": 0.12007434944237919, "grad_norm": 1.7565040852447897, "learning_rate": 9.790546652018805e-06, "loss": 0.1212, "step": 2584 }, { "epoch": 0.12012081784386618, "grad_norm": 1.7734352627989747, "learning_rate": 9.790331075908604e-06, "loss": 0.1135, "step": 2585 }, { "epoch": 0.12016728624535317, "grad_norm": 1.7900857823339607, "learning_rate": 9.790115391292361e-06, "loss": 0.0864, "step": 2586 }, { "epoch": 0.12021375464684014, "grad_norm": 1.1332362242491518, "learning_rate": 9.789899598174965e-06, "loss": 0.0763, "step": 2587 }, { "epoch": 0.12026022304832713, "grad_norm": 1.8421876674862308, "learning_rate": 9.7896836965613e-06, "loss": 0.0981, "step": 2588 }, { "epoch": 0.12030669144981412, "grad_norm": 1.7150883498390308, "learning_rate": 9.78946768645626e-06, "loss": 0.1594, "step": 2589 }, { "epoch": 0.12035315985130111, "grad_norm": 2.4300540802664035, "learning_rate": 9.789251567864734e-06, "loss": 0.133, "step": 2590 }, { "epoch": 0.1203996282527881, "grad_norm": 2.0708420305080066, "learning_rate": 9.789035340791622e-06, "loss": 0.115, "step": 2591 }, { "epoch": 0.1204460966542751, "grad_norm": 1.0994678856659763, "learning_rate": 9.788819005241816e-06, "loss": 0.0868, "step": 2592 }, { "epoch": 0.12049256505576209, "grad_norm": 3.9089293602012596, "learning_rate": 9.788602561220221e-06, "loss": 0.1362, "step": 2593 }, { "epoch": 0.12053903345724908, "grad_norm": 1.4805391541069224, "learning_rate": 9.788386008731738e-06, "loss": 0.1041, "step": 2594 }, { "epoch": 0.12058550185873607, "grad_norm": 1.724998393167971, "learning_rate": 9.788169347781273e-06, "loss": 0.1215, "step": 2595 }, { "epoch": 0.12063197026022304, "grad_norm": 2.988062345000645, "learning_rate": 9.787952578373733e-06, "loss": 0.1213, "step": 2596 }, { "epoch": 0.12067843866171003, "grad_norm": 1.7429620967889081, "learning_rate": 9.787735700514027e-06, "loss": 0.1319, "step": 2597 }, { "epoch": 0.12072490706319702, "grad_norm": 1.9958285344456173, "learning_rate": 9.787518714207068e-06, "loss": 0.1062, "step": 2598 }, { "epoch": 0.12077137546468401, "grad_norm": 1.053405557477689, "learning_rate": 9.787301619457773e-06, "loss": 0.0635, "step": 2599 }, { "epoch": 0.120817843866171, "grad_norm": 2.3888711408849317, "learning_rate": 9.787084416271055e-06, "loss": 0.0911, "step": 2600 }, { "epoch": 0.120864312267658, "grad_norm": 2.564208827720921, "learning_rate": 9.78686710465184e-06, "loss": 0.102, "step": 2601 }, { "epoch": 0.12091078066914498, "grad_norm": 1.4252799041273634, "learning_rate": 9.786649684605045e-06, "loss": 0.1106, "step": 2602 }, { "epoch": 0.12095724907063198, "grad_norm": 0.899333845316022, "learning_rate": 9.786432156135597e-06, "loss": 0.0303, "step": 2603 }, { "epoch": 0.12100371747211897, "grad_norm": 2.779705212012576, "learning_rate": 9.786214519248423e-06, "loss": 0.112, "step": 2604 }, { "epoch": 0.12105018587360594, "grad_norm": 2.363679958315551, "learning_rate": 9.785996773948452e-06, "loss": 0.1485, "step": 2605 }, { "epoch": 0.12109665427509293, "grad_norm": 1.782326114367807, "learning_rate": 9.785778920240618e-06, "loss": 0.0807, "step": 2606 }, { "epoch": 0.12114312267657992, "grad_norm": 2.5216695856126945, "learning_rate": 9.785560958129854e-06, "loss": 0.1372, "step": 2607 }, { "epoch": 0.12118959107806691, "grad_norm": 1.9104526698046886, "learning_rate": 9.785342887621098e-06, "loss": 0.1334, "step": 2608 }, { "epoch": 0.1212360594795539, "grad_norm": 2.0102836823451207, "learning_rate": 9.785124708719288e-06, "loss": 0.1738, "step": 2609 }, { "epoch": 0.1212825278810409, "grad_norm": 2.706529077973414, "learning_rate": 9.784906421429369e-06, "loss": 0.0864, "step": 2610 }, { "epoch": 0.12132899628252788, "grad_norm": 1.727183038391371, "learning_rate": 9.784688025756283e-06, "loss": 0.1763, "step": 2611 }, { "epoch": 0.12137546468401487, "grad_norm": 1.1045734316313611, "learning_rate": 9.784469521704976e-06, "loss": 0.0884, "step": 2612 }, { "epoch": 0.12142193308550187, "grad_norm": 2.8738461593868534, "learning_rate": 9.784250909280401e-06, "loss": 0.1417, "step": 2613 }, { "epoch": 0.12146840148698884, "grad_norm": 1.1796051347828411, "learning_rate": 9.784032188487507e-06, "loss": 0.0767, "step": 2614 }, { "epoch": 0.12151486988847583, "grad_norm": 1.064558413948248, "learning_rate": 9.783813359331248e-06, "loss": 0.0534, "step": 2615 }, { "epoch": 0.12156133828996282, "grad_norm": 1.5827023789511345, "learning_rate": 9.78359442181658e-06, "loss": 0.064, "step": 2616 }, { "epoch": 0.12160780669144981, "grad_norm": 1.5866884405265802, "learning_rate": 9.783375375948465e-06, "loss": 0.1001, "step": 2617 }, { "epoch": 0.1216542750929368, "grad_norm": 1.2293103380741934, "learning_rate": 9.783156221731865e-06, "loss": 0.0884, "step": 2618 }, { "epoch": 0.1217007434944238, "grad_norm": 1.832424115311428, "learning_rate": 9.78293695917174e-06, "loss": 0.092, "step": 2619 }, { "epoch": 0.12174721189591078, "grad_norm": 1.6819236364913077, "learning_rate": 9.78271758827306e-06, "loss": 0.098, "step": 2620 }, { "epoch": 0.12179368029739777, "grad_norm": 1.9468757268241184, "learning_rate": 9.782498109040794e-06, "loss": 0.1035, "step": 2621 }, { "epoch": 0.12184014869888476, "grad_norm": 1.688549700630542, "learning_rate": 9.782278521479911e-06, "loss": 0.0911, "step": 2622 }, { "epoch": 0.12188661710037174, "grad_norm": 2.524626367235612, "learning_rate": 9.782058825595386e-06, "loss": 0.1177, "step": 2623 }, { "epoch": 0.12193308550185873, "grad_norm": 1.987886456879347, "learning_rate": 9.781839021392197e-06, "loss": 0.1075, "step": 2624 }, { "epoch": 0.12197955390334572, "grad_norm": 3.503327934813867, "learning_rate": 9.78161910887532e-06, "loss": 0.1651, "step": 2625 }, { "epoch": 0.12202602230483271, "grad_norm": 1.635000286927513, "learning_rate": 9.781399088049739e-06, "loss": 0.0938, "step": 2626 }, { "epoch": 0.1220724907063197, "grad_norm": 2.6626711761636632, "learning_rate": 9.781178958920435e-06, "loss": 0.1703, "step": 2627 }, { "epoch": 0.1221189591078067, "grad_norm": 1.7695005932120027, "learning_rate": 9.780958721492395e-06, "loss": 0.0954, "step": 2628 }, { "epoch": 0.12216542750929368, "grad_norm": 1.9202869628056514, "learning_rate": 9.78073837577061e-06, "loss": 0.1166, "step": 2629 }, { "epoch": 0.12221189591078067, "grad_norm": 1.6141684824134028, "learning_rate": 9.780517921760068e-06, "loss": 0.1113, "step": 2630 }, { "epoch": 0.12225836431226766, "grad_norm": 1.9758768034913177, "learning_rate": 9.780297359465765e-06, "loss": 0.1222, "step": 2631 }, { "epoch": 0.12230483271375464, "grad_norm": 2.0637951125246854, "learning_rate": 9.780076688892693e-06, "loss": 0.0976, "step": 2632 }, { "epoch": 0.12235130111524163, "grad_norm": 3.455668530434251, "learning_rate": 9.779855910045854e-06, "loss": 0.1073, "step": 2633 }, { "epoch": 0.12239776951672862, "grad_norm": 2.485388141660369, "learning_rate": 9.77963502293025e-06, "loss": 0.1092, "step": 2634 }, { "epoch": 0.12244423791821561, "grad_norm": 1.4618840395396235, "learning_rate": 9.77941402755088e-06, "loss": 0.1439, "step": 2635 }, { "epoch": 0.1224907063197026, "grad_norm": 1.6030469669166683, "learning_rate": 9.779192923912752e-06, "loss": 0.1152, "step": 2636 }, { "epoch": 0.12253717472118959, "grad_norm": 1.5518972557089343, "learning_rate": 9.778971712020876e-06, "loss": 0.1006, "step": 2637 }, { "epoch": 0.12258364312267658, "grad_norm": 2.101083425269295, "learning_rate": 9.778750391880259e-06, "loss": 0.1218, "step": 2638 }, { "epoch": 0.12263011152416357, "grad_norm": 1.7173818270718277, "learning_rate": 9.778528963495918e-06, "loss": 0.0737, "step": 2639 }, { "epoch": 0.12267657992565056, "grad_norm": 1.6638921580385762, "learning_rate": 9.778307426872865e-06, "loss": 0.1075, "step": 2640 }, { "epoch": 0.12272304832713754, "grad_norm": 0.9757300199408185, "learning_rate": 9.77808578201612e-06, "loss": 0.0744, "step": 2641 }, { "epoch": 0.12276951672862453, "grad_norm": 1.6828417234313977, "learning_rate": 9.777864028930705e-06, "loss": 0.0987, "step": 2642 }, { "epoch": 0.12281598513011152, "grad_norm": 2.404813375317855, "learning_rate": 9.77764216762164e-06, "loss": 0.1187, "step": 2643 }, { "epoch": 0.12286245353159851, "grad_norm": 2.2661552460220693, "learning_rate": 9.777420198093949e-06, "loss": 0.1135, "step": 2644 }, { "epoch": 0.1229089219330855, "grad_norm": 2.0083268175360374, "learning_rate": 9.777198120352667e-06, "loss": 0.1215, "step": 2645 }, { "epoch": 0.12295539033457249, "grad_norm": 1.16474860408027, "learning_rate": 9.776975934402817e-06, "loss": 0.0678, "step": 2646 }, { "epoch": 0.12300185873605948, "grad_norm": 1.3183392498771742, "learning_rate": 9.776753640249434e-06, "loss": 0.0921, "step": 2647 }, { "epoch": 0.12304832713754647, "grad_norm": 2.635893209731532, "learning_rate": 9.776531237897555e-06, "loss": 0.1163, "step": 2648 }, { "epoch": 0.12309479553903346, "grad_norm": 2.228787828256413, "learning_rate": 9.776308727352215e-06, "loss": 0.0774, "step": 2649 }, { "epoch": 0.12314126394052044, "grad_norm": 1.8995685696523283, "learning_rate": 9.776086108618456e-06, "loss": 0.1104, "step": 2650 }, { "epoch": 0.12318773234200743, "grad_norm": 1.2518491961198908, "learning_rate": 9.775863381701321e-06, "loss": 0.071, "step": 2651 }, { "epoch": 0.12323420074349442, "grad_norm": 1.427627317447455, "learning_rate": 9.775640546605851e-06, "loss": 0.1176, "step": 2652 }, { "epoch": 0.12328066914498141, "grad_norm": 5.861324223085836, "learning_rate": 9.775417603337099e-06, "loss": 0.1828, "step": 2653 }, { "epoch": 0.1233271375464684, "grad_norm": 2.4898191516815595, "learning_rate": 9.775194551900112e-06, "loss": 0.1308, "step": 2654 }, { "epoch": 0.12337360594795539, "grad_norm": 1.6431872380990573, "learning_rate": 9.774971392299941e-06, "loss": 0.0783, "step": 2655 }, { "epoch": 0.12342007434944238, "grad_norm": 0.7155238243861223, "learning_rate": 9.774748124541644e-06, "loss": 0.0413, "step": 2656 }, { "epoch": 0.12346654275092937, "grad_norm": 4.761229679980297, "learning_rate": 9.774524748630274e-06, "loss": 0.1049, "step": 2657 }, { "epoch": 0.12351301115241636, "grad_norm": 1.4619808441020887, "learning_rate": 9.774301264570895e-06, "loss": 0.0454, "step": 2658 }, { "epoch": 0.12355947955390334, "grad_norm": 1.892785698376174, "learning_rate": 9.774077672368569e-06, "loss": 0.106, "step": 2659 }, { "epoch": 0.12360594795539033, "grad_norm": 1.8661494293375132, "learning_rate": 9.773853972028357e-06, "loss": 0.1505, "step": 2660 }, { "epoch": 0.12365241635687732, "grad_norm": 1.536794433072491, "learning_rate": 9.77363016355533e-06, "loss": 0.1026, "step": 2661 }, { "epoch": 0.12369888475836431, "grad_norm": 1.4084023156321501, "learning_rate": 9.773406246954555e-06, "loss": 0.0802, "step": 2662 }, { "epoch": 0.1237453531598513, "grad_norm": 1.9339963318010893, "learning_rate": 9.773182222231104e-06, "loss": 0.1265, "step": 2663 }, { "epoch": 0.12379182156133829, "grad_norm": 2.4944954491902984, "learning_rate": 9.772958089390053e-06, "loss": 0.125, "step": 2664 }, { "epoch": 0.12383828996282528, "grad_norm": 1.7002265768246938, "learning_rate": 9.772733848436477e-06, "loss": 0.1014, "step": 2665 }, { "epoch": 0.12388475836431227, "grad_norm": 1.890673010637522, "learning_rate": 9.772509499375455e-06, "loss": 0.0769, "step": 2666 }, { "epoch": 0.12393122676579926, "grad_norm": 2.3084896791505463, "learning_rate": 9.77228504221207e-06, "loss": 0.1421, "step": 2667 }, { "epoch": 0.12397769516728624, "grad_norm": 1.3538069196290141, "learning_rate": 9.772060476951408e-06, "loss": 0.0882, "step": 2668 }, { "epoch": 0.12402416356877323, "grad_norm": 4.945588745540965, "learning_rate": 9.771835803598552e-06, "loss": 0.2019, "step": 2669 }, { "epoch": 0.12407063197026022, "grad_norm": 3.1957967109429943, "learning_rate": 9.771611022158595e-06, "loss": 0.2148, "step": 2670 }, { "epoch": 0.12411710037174721, "grad_norm": 1.7845635900351737, "learning_rate": 9.771386132636624e-06, "loss": 0.1517, "step": 2671 }, { "epoch": 0.1241635687732342, "grad_norm": 1.3465544761102957, "learning_rate": 9.771161135037737e-06, "loss": 0.0694, "step": 2672 }, { "epoch": 0.12421003717472119, "grad_norm": 2.1098244625847644, "learning_rate": 9.770936029367029e-06, "loss": 0.0774, "step": 2673 }, { "epoch": 0.12425650557620818, "grad_norm": 1.1272358766808637, "learning_rate": 9.770710815629596e-06, "loss": 0.0699, "step": 2674 }, { "epoch": 0.12430297397769517, "grad_norm": 1.8669897494891143, "learning_rate": 9.770485493830543e-06, "loss": 0.1743, "step": 2675 }, { "epoch": 0.12434944237918216, "grad_norm": 1.3714623262458434, "learning_rate": 9.770260063974975e-06, "loss": 0.0741, "step": 2676 }, { "epoch": 0.12439591078066914, "grad_norm": 0.9828209577034229, "learning_rate": 9.770034526067994e-06, "loss": 0.0414, "step": 2677 }, { "epoch": 0.12444237918215613, "grad_norm": 1.6961331800486905, "learning_rate": 9.769808880114711e-06, "loss": 0.0824, "step": 2678 }, { "epoch": 0.12448884758364312, "grad_norm": 3.0567377843315673, "learning_rate": 9.769583126120235e-06, "loss": 0.15, "step": 2679 }, { "epoch": 0.12453531598513011, "grad_norm": 0.7561557833856514, "learning_rate": 9.769357264089683e-06, "loss": 0.0756, "step": 2680 }, { "epoch": 0.1245817843866171, "grad_norm": 1.3832193682587963, "learning_rate": 9.76913129402817e-06, "loss": 0.0795, "step": 2681 }, { "epoch": 0.12462825278810409, "grad_norm": 1.7324044481266405, "learning_rate": 9.768905215940812e-06, "loss": 0.1113, "step": 2682 }, { "epoch": 0.12467472118959108, "grad_norm": 1.3831826396884577, "learning_rate": 9.768679029832734e-06, "loss": 0.0891, "step": 2683 }, { "epoch": 0.12472118959107807, "grad_norm": 2.9081752565320564, "learning_rate": 9.768452735709055e-06, "loss": 0.1273, "step": 2684 }, { "epoch": 0.12476765799256506, "grad_norm": 3.1144488255469533, "learning_rate": 9.768226333574904e-06, "loss": 0.1088, "step": 2685 }, { "epoch": 0.12481412639405204, "grad_norm": 1.2598653646250213, "learning_rate": 9.767999823435406e-06, "loss": 0.0642, "step": 2686 }, { "epoch": 0.12486059479553903, "grad_norm": 2.006092907723591, "learning_rate": 9.767773205295697e-06, "loss": 0.1472, "step": 2687 }, { "epoch": 0.12490706319702602, "grad_norm": 2.1450800369823475, "learning_rate": 9.767546479160904e-06, "loss": 0.1423, "step": 2688 }, { "epoch": 0.12495353159851301, "grad_norm": 1.291638397861412, "learning_rate": 9.767319645036166e-06, "loss": 0.061, "step": 2689 }, { "epoch": 0.125, "grad_norm": 1.8956103457029745, "learning_rate": 9.767092702926622e-06, "loss": 0.0929, "step": 2690 }, { "epoch": 0.12504646840148698, "grad_norm": 3.2766485837268964, "learning_rate": 9.766865652837409e-06, "loss": 0.1449, "step": 2691 }, { "epoch": 0.12509293680297398, "grad_norm": 2.363891014878214, "learning_rate": 9.766638494773676e-06, "loss": 0.1571, "step": 2692 }, { "epoch": 0.12513940520446096, "grad_norm": 2.348806335750741, "learning_rate": 9.76641122874056e-06, "loss": 0.0926, "step": 2693 }, { "epoch": 0.12518587360594796, "grad_norm": 1.1309011818292414, "learning_rate": 9.766183854743214e-06, "loss": 0.0659, "step": 2694 }, { "epoch": 0.12523234200743494, "grad_norm": 1.4957413674855353, "learning_rate": 9.765956372786788e-06, "loss": 0.0771, "step": 2695 }, { "epoch": 0.12527881040892194, "grad_norm": 2.2585230695601446, "learning_rate": 9.765728782876434e-06, "loss": 0.1442, "step": 2696 }, { "epoch": 0.12532527881040892, "grad_norm": 1.5439699690977182, "learning_rate": 9.765501085017308e-06, "loss": 0.0576, "step": 2697 }, { "epoch": 0.12537174721189592, "grad_norm": 2.2468731197330953, "learning_rate": 9.765273279214566e-06, "loss": 0.1605, "step": 2698 }, { "epoch": 0.1254182156133829, "grad_norm": 1.5252115820463874, "learning_rate": 9.765045365473368e-06, "loss": 0.1113, "step": 2699 }, { "epoch": 0.12546468401486988, "grad_norm": 1.433278183995905, "learning_rate": 9.764817343798879e-06, "loss": 0.0882, "step": 2700 }, { "epoch": 0.12551115241635688, "grad_norm": 1.6125918568776734, "learning_rate": 9.764589214196261e-06, "loss": 0.0992, "step": 2701 }, { "epoch": 0.12555762081784386, "grad_norm": 1.7479933665515082, "learning_rate": 9.764360976670684e-06, "loss": 0.1227, "step": 2702 }, { "epoch": 0.12560408921933086, "grad_norm": 1.0794347929672246, "learning_rate": 9.764132631227315e-06, "loss": 0.0786, "step": 2703 }, { "epoch": 0.12565055762081784, "grad_norm": 1.2062178572392404, "learning_rate": 9.763904177871328e-06, "loss": 0.0847, "step": 2704 }, { "epoch": 0.12569702602230484, "grad_norm": 2.323546246771655, "learning_rate": 9.7636756166079e-06, "loss": 0.0727, "step": 2705 }, { "epoch": 0.12574349442379182, "grad_norm": 1.3278235026321523, "learning_rate": 9.763446947442203e-06, "loss": 0.0656, "step": 2706 }, { "epoch": 0.12578996282527882, "grad_norm": 0.9534979707250278, "learning_rate": 9.763218170379418e-06, "loss": 0.0952, "step": 2707 }, { "epoch": 0.1258364312267658, "grad_norm": 1.9252852407502679, "learning_rate": 9.76298928542473e-06, "loss": 0.1497, "step": 2708 }, { "epoch": 0.12588289962825278, "grad_norm": 1.8534545700511396, "learning_rate": 9.762760292583321e-06, "loss": 0.0839, "step": 2709 }, { "epoch": 0.12592936802973978, "grad_norm": 1.0526581592966493, "learning_rate": 9.76253119186038e-06, "loss": 0.0621, "step": 2710 }, { "epoch": 0.12597583643122676, "grad_norm": 4.2196504330670335, "learning_rate": 9.762301983261096e-06, "loss": 0.1057, "step": 2711 }, { "epoch": 0.12602230483271376, "grad_norm": 3.432425570958725, "learning_rate": 9.762072666790658e-06, "loss": 0.1178, "step": 2712 }, { "epoch": 0.12606877323420074, "grad_norm": 1.7446046324686386, "learning_rate": 9.761843242454261e-06, "loss": 0.112, "step": 2713 }, { "epoch": 0.12611524163568774, "grad_norm": 3.2837215042508787, "learning_rate": 9.761613710257105e-06, "loss": 0.1087, "step": 2714 }, { "epoch": 0.12616171003717472, "grad_norm": 1.5197274924344977, "learning_rate": 9.761384070204387e-06, "loss": 0.086, "step": 2715 }, { "epoch": 0.12620817843866172, "grad_norm": 1.196672516677757, "learning_rate": 9.761154322301306e-06, "loss": 0.0654, "step": 2716 }, { "epoch": 0.1262546468401487, "grad_norm": 1.5988297559058122, "learning_rate": 9.760924466553071e-06, "loss": 0.1074, "step": 2717 }, { "epoch": 0.12630111524163568, "grad_norm": 1.2556972559763329, "learning_rate": 9.760694502964884e-06, "loss": 0.0607, "step": 2718 }, { "epoch": 0.12634758364312268, "grad_norm": 2.948733417345223, "learning_rate": 9.760464431541959e-06, "loss": 0.1295, "step": 2719 }, { "epoch": 0.12639405204460966, "grad_norm": 2.2566979116595176, "learning_rate": 9.760234252289502e-06, "loss": 0.1333, "step": 2720 }, { "epoch": 0.12644052044609666, "grad_norm": 1.6746943282026336, "learning_rate": 9.76000396521273e-06, "loss": 0.0645, "step": 2721 }, { "epoch": 0.12648698884758364, "grad_norm": 1.0720706906500688, "learning_rate": 9.759773570316857e-06, "loss": 0.0454, "step": 2722 }, { "epoch": 0.12653345724907064, "grad_norm": 2.1374636114340024, "learning_rate": 9.759543067607105e-06, "loss": 0.1045, "step": 2723 }, { "epoch": 0.12657992565055762, "grad_norm": 1.5855656490379961, "learning_rate": 9.759312457088692e-06, "loss": 0.0786, "step": 2724 }, { "epoch": 0.12662639405204462, "grad_norm": 3.2075348171883573, "learning_rate": 9.759081738766843e-06, "loss": 0.0713, "step": 2725 }, { "epoch": 0.1266728624535316, "grad_norm": 2.238856306144426, "learning_rate": 9.758850912646785e-06, "loss": 0.0899, "step": 2726 }, { "epoch": 0.12671933085501857, "grad_norm": 2.5439365065174933, "learning_rate": 9.758619978733743e-06, "loss": 0.1356, "step": 2727 }, { "epoch": 0.12676579925650558, "grad_norm": 2.8593971937517364, "learning_rate": 9.758388937032953e-06, "loss": 0.072, "step": 2728 }, { "epoch": 0.12681226765799256, "grad_norm": 1.8321749240942886, "learning_rate": 9.758157787549642e-06, "loss": 0.1198, "step": 2729 }, { "epoch": 0.12685873605947956, "grad_norm": 0.8064584307613741, "learning_rate": 9.75792653028905e-06, "loss": 0.0571, "step": 2730 }, { "epoch": 0.12690520446096654, "grad_norm": 2.3992388918638405, "learning_rate": 9.757695165256415e-06, "loss": 0.1379, "step": 2731 }, { "epoch": 0.12695167286245354, "grad_norm": 2.095163967092464, "learning_rate": 9.75746369245698e-06, "loss": 0.1082, "step": 2732 }, { "epoch": 0.12699814126394052, "grad_norm": 1.298217695728673, "learning_rate": 9.757232111895983e-06, "loss": 0.0794, "step": 2733 }, { "epoch": 0.12704460966542752, "grad_norm": 1.0567951713934705, "learning_rate": 9.757000423578671e-06, "loss": 0.0709, "step": 2734 }, { "epoch": 0.1270910780669145, "grad_norm": 1.7060739072434126, "learning_rate": 9.756768627510292e-06, "loss": 0.1017, "step": 2735 }, { "epoch": 0.12713754646840147, "grad_norm": 1.3996566174030958, "learning_rate": 9.7565367236961e-06, "loss": 0.0499, "step": 2736 }, { "epoch": 0.12718401486988848, "grad_norm": 1.6847915021961166, "learning_rate": 9.756304712141342e-06, "loss": 0.0979, "step": 2737 }, { "epoch": 0.12723048327137546, "grad_norm": 4.300673966101642, "learning_rate": 9.756072592851278e-06, "loss": 0.1192, "step": 2738 }, { "epoch": 0.12727695167286246, "grad_norm": 6.026428390245039, "learning_rate": 9.755840365831163e-06, "loss": 0.1307, "step": 2739 }, { "epoch": 0.12732342007434944, "grad_norm": 4.170941847305265, "learning_rate": 9.75560803108626e-06, "loss": 0.0913, "step": 2740 }, { "epoch": 0.12736988847583644, "grad_norm": 1.2460652666820533, "learning_rate": 9.75537558862183e-06, "loss": 0.1206, "step": 2741 }, { "epoch": 0.12741635687732342, "grad_norm": 16.652585991206074, "learning_rate": 9.755143038443136e-06, "loss": 0.1263, "step": 2742 }, { "epoch": 0.12746282527881042, "grad_norm": 2.3691537938432448, "learning_rate": 9.75491038055545e-06, "loss": 0.0727, "step": 2743 }, { "epoch": 0.1275092936802974, "grad_norm": 6.813800783533128, "learning_rate": 9.754677614964036e-06, "loss": 0.0693, "step": 2744 }, { "epoch": 0.12755576208178437, "grad_norm": 1.4446602703618256, "learning_rate": 9.754444741674173e-06, "loss": 0.0872, "step": 2745 }, { "epoch": 0.12760223048327138, "grad_norm": 2.3506014453468422, "learning_rate": 9.75421176069113e-06, "loss": 0.1653, "step": 2746 }, { "epoch": 0.12764869888475835, "grad_norm": 2.3703655051969124, "learning_rate": 9.75397867202019e-06, "loss": 0.1591, "step": 2747 }, { "epoch": 0.12769516728624536, "grad_norm": 2.83091636533188, "learning_rate": 9.753745475666628e-06, "loss": 0.0825, "step": 2748 }, { "epoch": 0.12774163568773234, "grad_norm": 1.7279095498381232, "learning_rate": 9.753512171635728e-06, "loss": 0.0634, "step": 2749 }, { "epoch": 0.12778810408921934, "grad_norm": 2.077284447083945, "learning_rate": 9.753278759932773e-06, "loss": 0.1649, "step": 2750 }, { "epoch": 0.12783457249070632, "grad_norm": 5.005506826188941, "learning_rate": 9.753045240563052e-06, "loss": 0.1338, "step": 2751 }, { "epoch": 0.12788104089219332, "grad_norm": 4.8642000782430594, "learning_rate": 9.752811613531855e-06, "loss": 0.0885, "step": 2752 }, { "epoch": 0.1279275092936803, "grad_norm": 4.401480757137353, "learning_rate": 9.752577878844472e-06, "loss": 0.1903, "step": 2753 }, { "epoch": 0.12797397769516727, "grad_norm": 1.9931315173013386, "learning_rate": 9.752344036506199e-06, "loss": 0.0699, "step": 2754 }, { "epoch": 0.12802044609665428, "grad_norm": 2.2425673211467876, "learning_rate": 9.75211008652233e-06, "loss": 0.1215, "step": 2755 }, { "epoch": 0.12806691449814125, "grad_norm": 12.13170802888408, "learning_rate": 9.751876028898165e-06, "loss": 0.1117, "step": 2756 }, { "epoch": 0.12811338289962826, "grad_norm": 1.6564538601296144, "learning_rate": 9.75164186363901e-06, "loss": 0.1505, "step": 2757 }, { "epoch": 0.12815985130111524, "grad_norm": 3.0591618332709065, "learning_rate": 9.751407590750163e-06, "loss": 0.123, "step": 2758 }, { "epoch": 0.12820631970260224, "grad_norm": 1.065771724302017, "learning_rate": 9.751173210236935e-06, "loss": 0.0524, "step": 2759 }, { "epoch": 0.12825278810408922, "grad_norm": 1.9775963674251824, "learning_rate": 9.750938722104633e-06, "loss": 0.1049, "step": 2760 }, { "epoch": 0.12829925650557622, "grad_norm": 4.0710484457414156, "learning_rate": 9.750704126358568e-06, "loss": 0.0958, "step": 2761 }, { "epoch": 0.1283457249070632, "grad_norm": 5.048768099715014, "learning_rate": 9.750469423004057e-06, "loss": 0.0965, "step": 2762 }, { "epoch": 0.12839219330855017, "grad_norm": 1.5719426084128478, "learning_rate": 9.75023461204641e-06, "loss": 0.1413, "step": 2763 }, { "epoch": 0.12843866171003718, "grad_norm": 2.6991345607545147, "learning_rate": 9.749999693490952e-06, "loss": 0.1754, "step": 2764 }, { "epoch": 0.12848513011152415, "grad_norm": 1.7222236764169552, "learning_rate": 9.749764667343001e-06, "loss": 0.0818, "step": 2765 }, { "epoch": 0.12853159851301116, "grad_norm": 3.2179599129450094, "learning_rate": 9.749529533607882e-06, "loss": 0.0897, "step": 2766 }, { "epoch": 0.12857806691449813, "grad_norm": 2.163341648068734, "learning_rate": 9.749294292290918e-06, "loss": 0.1374, "step": 2767 }, { "epoch": 0.12862453531598514, "grad_norm": 2.5354224625245485, "learning_rate": 9.749058943397442e-06, "loss": 0.0925, "step": 2768 }, { "epoch": 0.12867100371747212, "grad_norm": 1.460858505837479, "learning_rate": 9.748823486932782e-06, "loss": 0.0945, "step": 2769 }, { "epoch": 0.12871747211895912, "grad_norm": 2.3029573179747063, "learning_rate": 9.748587922902272e-06, "loss": 0.0922, "step": 2770 }, { "epoch": 0.1287639405204461, "grad_norm": 1.6430599433271011, "learning_rate": 9.748352251311248e-06, "loss": 0.0963, "step": 2771 }, { "epoch": 0.12881040892193307, "grad_norm": 1.1852008215361498, "learning_rate": 9.748116472165048e-06, "loss": 0.0922, "step": 2772 }, { "epoch": 0.12885687732342008, "grad_norm": 1.5527970043670145, "learning_rate": 9.747880585469011e-06, "loss": 0.121, "step": 2773 }, { "epoch": 0.12890334572490705, "grad_norm": 5.133661241450069, "learning_rate": 9.747644591228484e-06, "loss": 0.2133, "step": 2774 }, { "epoch": 0.12894981412639406, "grad_norm": 1.224258853532238, "learning_rate": 9.747408489448808e-06, "loss": 0.0879, "step": 2775 }, { "epoch": 0.12899628252788103, "grad_norm": 2.101699479980415, "learning_rate": 9.747172280135335e-06, "loss": 0.1629, "step": 2776 }, { "epoch": 0.12904275092936804, "grad_norm": 2.1034545537744607, "learning_rate": 9.746935963293412e-06, "loss": 0.1409, "step": 2777 }, { "epoch": 0.12908921933085502, "grad_norm": 2.2992180947235323, "learning_rate": 9.746699538928394e-06, "loss": 0.1, "step": 2778 }, { "epoch": 0.12913568773234202, "grad_norm": 3.6091541787323966, "learning_rate": 9.746463007045636e-06, "loss": 0.0955, "step": 2779 }, { "epoch": 0.129182156133829, "grad_norm": 5.732333291235302, "learning_rate": 9.746226367650495e-06, "loss": 0.1195, "step": 2780 }, { "epoch": 0.12922862453531597, "grad_norm": 1.578827646339806, "learning_rate": 9.745989620748332e-06, "loss": 0.0767, "step": 2781 }, { "epoch": 0.12927509293680298, "grad_norm": 2.2505619484711343, "learning_rate": 9.745752766344509e-06, "loss": 0.0844, "step": 2782 }, { "epoch": 0.12932156133828995, "grad_norm": 1.8866782265879112, "learning_rate": 9.74551580444439e-06, "loss": 0.1052, "step": 2783 }, { "epoch": 0.12936802973977696, "grad_norm": 2.256572749584586, "learning_rate": 9.745278735053345e-06, "loss": 0.1927, "step": 2784 }, { "epoch": 0.12941449814126393, "grad_norm": 1.9969599713915582, "learning_rate": 9.74504155817674e-06, "loss": 0.1377, "step": 2785 }, { "epoch": 0.12946096654275094, "grad_norm": 1.3953030091156478, "learning_rate": 9.744804273819952e-06, "loss": 0.1341, "step": 2786 }, { "epoch": 0.12950743494423791, "grad_norm": 1.4845863127288799, "learning_rate": 9.744566881988352e-06, "loss": 0.0965, "step": 2787 }, { "epoch": 0.12955390334572492, "grad_norm": 2.6576645852109717, "learning_rate": 9.744329382687319e-06, "loss": 0.1547, "step": 2788 }, { "epoch": 0.1296003717472119, "grad_norm": 2.1502540510018147, "learning_rate": 9.744091775922232e-06, "loss": 0.1167, "step": 2789 }, { "epoch": 0.12964684014869887, "grad_norm": 1.2132444967864207, "learning_rate": 9.743854061698474e-06, "loss": 0.0796, "step": 2790 }, { "epoch": 0.12969330855018588, "grad_norm": 3.034062171333819, "learning_rate": 9.743616240021427e-06, "loss": 0.0957, "step": 2791 }, { "epoch": 0.12973977695167285, "grad_norm": 2.055252034685132, "learning_rate": 9.743378310896481e-06, "loss": 0.1565, "step": 2792 }, { "epoch": 0.12978624535315986, "grad_norm": 2.819362329311172, "learning_rate": 9.743140274329024e-06, "loss": 0.1359, "step": 2793 }, { "epoch": 0.12983271375464683, "grad_norm": 2.792044547444526, "learning_rate": 9.742902130324447e-06, "loss": 0.0969, "step": 2794 }, { "epoch": 0.12987918215613384, "grad_norm": 1.7776104827254293, "learning_rate": 9.742663878888145e-06, "loss": 0.0861, "step": 2795 }, { "epoch": 0.12992565055762081, "grad_norm": 2.386713435427314, "learning_rate": 9.742425520025516e-06, "loss": 0.1198, "step": 2796 }, { "epoch": 0.12997211895910782, "grad_norm": 1.6885785758832708, "learning_rate": 9.742187053741956e-06, "loss": 0.0758, "step": 2797 }, { "epoch": 0.1300185873605948, "grad_norm": 1.9919432215522705, "learning_rate": 9.741948480042869e-06, "loss": 0.1054, "step": 2798 }, { "epoch": 0.13006505576208177, "grad_norm": 3.617373461998219, "learning_rate": 9.741709798933658e-06, "loss": 0.0761, "step": 2799 }, { "epoch": 0.13011152416356878, "grad_norm": 1.2360809452629824, "learning_rate": 9.741471010419729e-06, "loss": 0.0752, "step": 2800 }, { "epoch": 0.13015799256505575, "grad_norm": 2.4965546230106668, "learning_rate": 9.741232114506491e-06, "loss": 0.1764, "step": 2801 }, { "epoch": 0.13020446096654276, "grad_norm": 1.5863134689062166, "learning_rate": 9.740993111199356e-06, "loss": 0.0813, "step": 2802 }, { "epoch": 0.13025092936802973, "grad_norm": 5.060871553706864, "learning_rate": 9.740754000503738e-06, "loss": 0.0798, "step": 2803 }, { "epoch": 0.13029739776951674, "grad_norm": 2.7368553107670293, "learning_rate": 9.740514782425051e-06, "loss": 0.1459, "step": 2804 }, { "epoch": 0.1303438661710037, "grad_norm": 1.8147254688730352, "learning_rate": 9.740275456968716e-06, "loss": 0.1103, "step": 2805 }, { "epoch": 0.13039033457249072, "grad_norm": 6.207303503463998, "learning_rate": 9.740036024140151e-06, "loss": 0.1965, "step": 2806 }, { "epoch": 0.1304368029739777, "grad_norm": 1.5441203539162827, "learning_rate": 9.739796483944783e-06, "loss": 0.1078, "step": 2807 }, { "epoch": 0.13048327137546467, "grad_norm": 1.7618931176409527, "learning_rate": 9.739556836388037e-06, "loss": 0.0838, "step": 2808 }, { "epoch": 0.13052973977695168, "grad_norm": 3.35560480377524, "learning_rate": 9.739317081475338e-06, "loss": 0.0723, "step": 2809 }, { "epoch": 0.13057620817843865, "grad_norm": 2.1839784805395643, "learning_rate": 9.73907721921212e-06, "loss": 0.0973, "step": 2810 }, { "epoch": 0.13062267657992566, "grad_norm": 1.644298103429064, "learning_rate": 9.738837249603816e-06, "loss": 0.1603, "step": 2811 }, { "epoch": 0.13066914498141263, "grad_norm": 0.9176761050453891, "learning_rate": 9.73859717265586e-06, "loss": 0.0636, "step": 2812 }, { "epoch": 0.13071561338289964, "grad_norm": 3.2385037415048936, "learning_rate": 9.738356988373692e-06, "loss": 0.1124, "step": 2813 }, { "epoch": 0.1307620817843866, "grad_norm": 2.8023085550224156, "learning_rate": 9.73811669676275e-06, "loss": 0.1175, "step": 2814 }, { "epoch": 0.13080855018587362, "grad_norm": 1.4941310426004821, "learning_rate": 9.737876297828478e-06, "loss": 0.1275, "step": 2815 }, { "epoch": 0.1308550185873606, "grad_norm": 1.8191115798230701, "learning_rate": 9.737635791576322e-06, "loss": 0.0594, "step": 2816 }, { "epoch": 0.13090148698884757, "grad_norm": 2.9999268833894157, "learning_rate": 9.737395178011728e-06, "loss": 0.1206, "step": 2817 }, { "epoch": 0.13094795539033457, "grad_norm": 2.34010695252614, "learning_rate": 9.737154457140148e-06, "loss": 0.1213, "step": 2818 }, { "epoch": 0.13099442379182155, "grad_norm": 1.9967320956026673, "learning_rate": 9.736913628967033e-06, "loss": 0.1413, "step": 2819 }, { "epoch": 0.13104089219330856, "grad_norm": 2.2469994723859217, "learning_rate": 9.736672693497839e-06, "loss": 0.1275, "step": 2820 }, { "epoch": 0.13108736059479553, "grad_norm": 4.9976103834433365, "learning_rate": 9.736431650738022e-06, "loss": 0.0659, "step": 2821 }, { "epoch": 0.13113382899628254, "grad_norm": 2.302525057665293, "learning_rate": 9.736190500693046e-06, "loss": 0.1072, "step": 2822 }, { "epoch": 0.1311802973977695, "grad_norm": 2.2478968351399478, "learning_rate": 9.73594924336837e-06, "loss": 0.1585, "step": 2823 }, { "epoch": 0.13122676579925652, "grad_norm": 2.877341686124689, "learning_rate": 9.735707878769459e-06, "loss": 0.1523, "step": 2824 }, { "epoch": 0.1312732342007435, "grad_norm": 2.2131500300588884, "learning_rate": 9.735466406901779e-06, "loss": 0.1365, "step": 2825 }, { "epoch": 0.13131970260223047, "grad_norm": 1.652244637785827, "learning_rate": 9.7352248277708e-06, "loss": 0.1415, "step": 2826 }, { "epoch": 0.13136617100371747, "grad_norm": 1.5289424611698936, "learning_rate": 9.734983141381997e-06, "loss": 0.0705, "step": 2827 }, { "epoch": 0.13141263940520445, "grad_norm": 1.090635026473051, "learning_rate": 9.73474134774084e-06, "loss": 0.0924, "step": 2828 }, { "epoch": 0.13145910780669146, "grad_norm": 3.3053499087151486, "learning_rate": 9.734499446852813e-06, "loss": 0.1244, "step": 2829 }, { "epoch": 0.13150557620817843, "grad_norm": 1.0957770271552592, "learning_rate": 9.734257438723385e-06, "loss": 0.0543, "step": 2830 }, { "epoch": 0.13155204460966544, "grad_norm": 0.9874408215482414, "learning_rate": 9.734015323358046e-06, "loss": 0.0833, "step": 2831 }, { "epoch": 0.1315985130111524, "grad_norm": 1.983061779267227, "learning_rate": 9.733773100762278e-06, "loss": 0.0997, "step": 2832 }, { "epoch": 0.13164498141263942, "grad_norm": 2.1527672073400232, "learning_rate": 9.733530770941567e-06, "loss": 0.0957, "step": 2833 }, { "epoch": 0.1316914498141264, "grad_norm": 1.76016619536097, "learning_rate": 9.7332883339014e-06, "loss": 0.1125, "step": 2834 }, { "epoch": 0.13173791821561337, "grad_norm": 9.665798634481769, "learning_rate": 9.733045789647273e-06, "loss": 0.0606, "step": 2835 }, { "epoch": 0.13178438661710037, "grad_norm": 1.3295010921445238, "learning_rate": 9.732803138184676e-06, "loss": 0.0952, "step": 2836 }, { "epoch": 0.13183085501858735, "grad_norm": 1.3457996597151318, "learning_rate": 9.732560379519107e-06, "loss": 0.0813, "step": 2837 }, { "epoch": 0.13187732342007435, "grad_norm": 1.493780714330866, "learning_rate": 9.732317513656065e-06, "loss": 0.0623, "step": 2838 }, { "epoch": 0.13192379182156133, "grad_norm": 1.0504178537801778, "learning_rate": 9.73207454060105e-06, "loss": 0.0595, "step": 2839 }, { "epoch": 0.13197026022304834, "grad_norm": 1.3458033692333191, "learning_rate": 9.731831460359566e-06, "loss": 0.067, "step": 2840 }, { "epoch": 0.1320167286245353, "grad_norm": 3.8423715025206313, "learning_rate": 9.73158827293712e-06, "loss": 0.0961, "step": 2841 }, { "epoch": 0.13206319702602232, "grad_norm": 1.4380490570710724, "learning_rate": 9.731344978339219e-06, "loss": 0.0596, "step": 2842 }, { "epoch": 0.1321096654275093, "grad_norm": 1.5800902328939155, "learning_rate": 9.731101576571375e-06, "loss": 0.1609, "step": 2843 }, { "epoch": 0.13215613382899627, "grad_norm": 2.1438545346554605, "learning_rate": 9.730858067639102e-06, "loss": 0.1599, "step": 2844 }, { "epoch": 0.13220260223048327, "grad_norm": 3.2429590425204826, "learning_rate": 9.730614451547912e-06, "loss": 0.0972, "step": 2845 }, { "epoch": 0.13224907063197025, "grad_norm": 2.346552387439876, "learning_rate": 9.730370728303328e-06, "loss": 0.162, "step": 2846 }, { "epoch": 0.13229553903345725, "grad_norm": 1.7503871846912156, "learning_rate": 9.730126897910867e-06, "loss": 0.093, "step": 2847 }, { "epoch": 0.13234200743494423, "grad_norm": 3.1982858214538323, "learning_rate": 9.729882960376054e-06, "loss": 0.1039, "step": 2848 }, { "epoch": 0.13238847583643124, "grad_norm": 0.9867047538377649, "learning_rate": 9.729638915704415e-06, "loss": 0.0646, "step": 2849 }, { "epoch": 0.1324349442379182, "grad_norm": 1.0997764553414853, "learning_rate": 9.729394763901475e-06, "loss": 0.0861, "step": 2850 }, { "epoch": 0.13248141263940522, "grad_norm": 1.2311782982786332, "learning_rate": 9.729150504972766e-06, "loss": 0.0804, "step": 2851 }, { "epoch": 0.1325278810408922, "grad_norm": 2.146472814778547, "learning_rate": 9.728906138923823e-06, "loss": 0.1087, "step": 2852 }, { "epoch": 0.13257434944237917, "grad_norm": 1.0237385411392463, "learning_rate": 9.728661665760177e-06, "loss": 0.0392, "step": 2853 }, { "epoch": 0.13262081784386617, "grad_norm": 1.2757048255286634, "learning_rate": 9.728417085487367e-06, "loss": 0.0958, "step": 2854 }, { "epoch": 0.13266728624535315, "grad_norm": 1.175529020832777, "learning_rate": 9.728172398110935e-06, "loss": 0.1027, "step": 2855 }, { "epoch": 0.13271375464684015, "grad_norm": 2.272225645210569, "learning_rate": 9.72792760363642e-06, "loss": 0.1075, "step": 2856 }, { "epoch": 0.13276022304832713, "grad_norm": 1.8914623241706177, "learning_rate": 9.72768270206937e-06, "loss": 0.0712, "step": 2857 }, { "epoch": 0.13280669144981413, "grad_norm": 1.4656821425203381, "learning_rate": 9.727437693415332e-06, "loss": 0.141, "step": 2858 }, { "epoch": 0.1328531598513011, "grad_norm": 1.0688046752261564, "learning_rate": 9.727192577679852e-06, "loss": 0.0701, "step": 2859 }, { "epoch": 0.13289962825278812, "grad_norm": 1.625751674889809, "learning_rate": 9.726947354868487e-06, "loss": 0.0798, "step": 2860 }, { "epoch": 0.1329460966542751, "grad_norm": 1.9395147444523755, "learning_rate": 9.726702024986788e-06, "loss": 0.0728, "step": 2861 }, { "epoch": 0.13299256505576207, "grad_norm": 1.157626966390698, "learning_rate": 9.726456588040313e-06, "loss": 0.077, "step": 2862 }, { "epoch": 0.13303903345724907, "grad_norm": 3.0201060461031752, "learning_rate": 9.726211044034624e-06, "loss": 0.0897, "step": 2863 }, { "epoch": 0.13308550185873605, "grad_norm": 1.585653061402344, "learning_rate": 9.72596539297528e-06, "loss": 0.1383, "step": 2864 }, { "epoch": 0.13313197026022305, "grad_norm": 2.1016328923045546, "learning_rate": 9.725719634867844e-06, "loss": 0.0983, "step": 2865 }, { "epoch": 0.13317843866171003, "grad_norm": 1.2561219624255993, "learning_rate": 9.725473769717885e-06, "loss": 0.0464, "step": 2866 }, { "epoch": 0.13322490706319703, "grad_norm": 1.2177414622720322, "learning_rate": 9.725227797530973e-06, "loss": 0.071, "step": 2867 }, { "epoch": 0.133271375464684, "grad_norm": 1.247279113942268, "learning_rate": 9.724981718312678e-06, "loss": 0.0661, "step": 2868 }, { "epoch": 0.13331784386617102, "grad_norm": 2.22509832945013, "learning_rate": 9.724735532068575e-06, "loss": 0.1136, "step": 2869 }, { "epoch": 0.133364312267658, "grad_norm": 2.063662928389788, "learning_rate": 9.724489238804237e-06, "loss": 0.0564, "step": 2870 }, { "epoch": 0.13341078066914497, "grad_norm": 1.1066069863062444, "learning_rate": 9.724242838525246e-06, "loss": 0.0653, "step": 2871 }, { "epoch": 0.13345724907063197, "grad_norm": 1.5344036145646058, "learning_rate": 9.723996331237184e-06, "loss": 0.0799, "step": 2872 }, { "epoch": 0.13350371747211895, "grad_norm": 3.561372000787902, "learning_rate": 9.723749716945631e-06, "loss": 0.1482, "step": 2873 }, { "epoch": 0.13355018587360595, "grad_norm": 2.7408255492615443, "learning_rate": 9.723502995656177e-06, "loss": 0.1089, "step": 2874 }, { "epoch": 0.13359665427509293, "grad_norm": 1.2464304047186348, "learning_rate": 9.723256167374409e-06, "loss": 0.0553, "step": 2875 }, { "epoch": 0.13364312267657993, "grad_norm": 1.2950036201028101, "learning_rate": 9.723009232105918e-06, "loss": 0.0559, "step": 2876 }, { "epoch": 0.1336895910780669, "grad_norm": 12.510754155149993, "learning_rate": 9.722762189856295e-06, "loss": 0.1117, "step": 2877 }, { "epoch": 0.13373605947955391, "grad_norm": 3.6531371066856377, "learning_rate": 9.72251504063114e-06, "loss": 0.139, "step": 2878 }, { "epoch": 0.1337825278810409, "grad_norm": 1.857957403042265, "learning_rate": 9.722267784436045e-06, "loss": 0.0755, "step": 2879 }, { "epoch": 0.13382899628252787, "grad_norm": 2.1867307758475207, "learning_rate": 9.72202042127662e-06, "loss": 0.0763, "step": 2880 }, { "epoch": 0.13387546468401487, "grad_norm": 3.588639209081832, "learning_rate": 9.72177295115846e-06, "loss": 0.129, "step": 2881 }, { "epoch": 0.13392193308550185, "grad_norm": 1.8230393484106573, "learning_rate": 9.721525374087171e-06, "loss": 0.0964, "step": 2882 }, { "epoch": 0.13396840148698885, "grad_norm": 1.612438323385052, "learning_rate": 9.721277690068365e-06, "loss": 0.103, "step": 2883 }, { "epoch": 0.13401486988847583, "grad_norm": 2.15975101213991, "learning_rate": 9.721029899107652e-06, "loss": 0.0694, "step": 2884 }, { "epoch": 0.13406133828996283, "grad_norm": 4.047395945762943, "learning_rate": 9.72078200121064e-06, "loss": 0.1525, "step": 2885 }, { "epoch": 0.1341078066914498, "grad_norm": 1.2747124908099032, "learning_rate": 9.720533996382948e-06, "loss": 0.0838, "step": 2886 }, { "epoch": 0.13415427509293681, "grad_norm": 3.05135090193739, "learning_rate": 9.720285884630193e-06, "loss": 0.1541, "step": 2887 }, { "epoch": 0.1342007434944238, "grad_norm": 1.6177210935636592, "learning_rate": 9.720037665957996e-06, "loss": 0.085, "step": 2888 }, { "epoch": 0.13424721189591077, "grad_norm": 1.7957831825898158, "learning_rate": 9.719789340371977e-06, "loss": 0.0868, "step": 2889 }, { "epoch": 0.13429368029739777, "grad_norm": 1.9405942926203925, "learning_rate": 9.719540907877762e-06, "loss": 0.1106, "step": 2890 }, { "epoch": 0.13434014869888475, "grad_norm": 1.5617867309004425, "learning_rate": 9.719292368480977e-06, "loss": 0.1586, "step": 2891 }, { "epoch": 0.13438661710037175, "grad_norm": 0.9942630269141879, "learning_rate": 9.719043722187254e-06, "loss": 0.0526, "step": 2892 }, { "epoch": 0.13443308550185873, "grad_norm": 3.4154195878261393, "learning_rate": 9.718794969002225e-06, "loss": 0.1809, "step": 2893 }, { "epoch": 0.13447955390334573, "grad_norm": 1.7639147985336636, "learning_rate": 9.718546108931521e-06, "loss": 0.0697, "step": 2894 }, { "epoch": 0.1345260223048327, "grad_norm": 1.3155672971968697, "learning_rate": 9.718297141980784e-06, "loss": 0.0878, "step": 2895 }, { "epoch": 0.13457249070631971, "grad_norm": 1.2400398609549677, "learning_rate": 9.71804806815565e-06, "loss": 0.0824, "step": 2896 }, { "epoch": 0.1346189591078067, "grad_norm": 2.0608861403258802, "learning_rate": 9.71779888746176e-06, "loss": 0.1225, "step": 2897 }, { "epoch": 0.13466542750929367, "grad_norm": 1.8356968437651893, "learning_rate": 9.717549599904761e-06, "loss": 0.089, "step": 2898 }, { "epoch": 0.13471189591078067, "grad_norm": 1.9449027749742327, "learning_rate": 9.717300205490299e-06, "loss": 0.2178, "step": 2899 }, { "epoch": 0.13475836431226765, "grad_norm": 3.831745959705666, "learning_rate": 9.717050704224023e-06, "loss": 0.128, "step": 2900 }, { "epoch": 0.13480483271375465, "grad_norm": 1.0531181908730574, "learning_rate": 9.716801096111583e-06, "loss": 0.0798, "step": 2901 }, { "epoch": 0.13485130111524163, "grad_norm": 1.495996790220785, "learning_rate": 9.716551381158633e-06, "loss": 0.0833, "step": 2902 }, { "epoch": 0.13489776951672863, "grad_norm": 1.2569884320338949, "learning_rate": 9.716301559370831e-06, "loss": 0.0893, "step": 2903 }, { "epoch": 0.1349442379182156, "grad_norm": 0.9209688807865344, "learning_rate": 9.716051630753836e-06, "loss": 0.0723, "step": 2904 }, { "epoch": 0.1349907063197026, "grad_norm": 1.0562924449324043, "learning_rate": 9.715801595313306e-06, "loss": 0.0923, "step": 2905 }, { "epoch": 0.1350371747211896, "grad_norm": 1.3224300334399703, "learning_rate": 9.715551453054908e-06, "loss": 0.148, "step": 2906 }, { "epoch": 0.13508364312267657, "grad_norm": 1.412774550853623, "learning_rate": 9.715301203984305e-06, "loss": 0.1283, "step": 2907 }, { "epoch": 0.13513011152416357, "grad_norm": 2.1492251877384554, "learning_rate": 9.715050848107167e-06, "loss": 0.1656, "step": 2908 }, { "epoch": 0.13517657992565055, "grad_norm": 4.164195420273223, "learning_rate": 9.714800385429167e-06, "loss": 0.083, "step": 2909 }, { "epoch": 0.13522304832713755, "grad_norm": 1.2925037152900176, "learning_rate": 9.714549815955973e-06, "loss": 0.0918, "step": 2910 }, { "epoch": 0.13526951672862453, "grad_norm": 1.2410889494417494, "learning_rate": 9.714299139693267e-06, "loss": 0.1136, "step": 2911 }, { "epoch": 0.13531598513011153, "grad_norm": 0.9968495295853721, "learning_rate": 9.71404835664672e-06, "loss": 0.0646, "step": 2912 }, { "epoch": 0.1353624535315985, "grad_norm": 1.3064596570406082, "learning_rate": 9.713797466822021e-06, "loss": 0.1296, "step": 2913 }, { "epoch": 0.1354089219330855, "grad_norm": 1.7363755411940889, "learning_rate": 9.713546470224845e-06, "loss": 0.1203, "step": 2914 }, { "epoch": 0.1354553903345725, "grad_norm": 5.224992875284868, "learning_rate": 9.713295366860881e-06, "loss": 0.1059, "step": 2915 }, { "epoch": 0.13550185873605947, "grad_norm": 3.010378809910553, "learning_rate": 9.713044156735817e-06, "loss": 0.1285, "step": 2916 }, { "epoch": 0.13554832713754647, "grad_norm": 1.879671547376491, "learning_rate": 9.712792839855342e-06, "loss": 0.0878, "step": 2917 }, { "epoch": 0.13559479553903345, "grad_norm": 1.7093461037469488, "learning_rate": 9.712541416225149e-06, "loss": 0.0954, "step": 2918 }, { "epoch": 0.13564126394052045, "grad_norm": 4.869893794536843, "learning_rate": 9.712289885850932e-06, "loss": 0.1243, "step": 2919 }, { "epoch": 0.13568773234200743, "grad_norm": 2.1999373959004225, "learning_rate": 9.712038248738391e-06, "loss": 0.1103, "step": 2920 }, { "epoch": 0.13573420074349443, "grad_norm": 1.0786531660057603, "learning_rate": 9.711786504893223e-06, "loss": 0.0646, "step": 2921 }, { "epoch": 0.1357806691449814, "grad_norm": 1.3453086366068554, "learning_rate": 9.711534654321134e-06, "loss": 0.1145, "step": 2922 }, { "epoch": 0.1358271375464684, "grad_norm": 2.073998040789189, "learning_rate": 9.711282697027824e-06, "loss": 0.0882, "step": 2923 }, { "epoch": 0.1358736059479554, "grad_norm": 3.0261882454868565, "learning_rate": 9.711030633019004e-06, "loss": 0.1025, "step": 2924 }, { "epoch": 0.13592007434944237, "grad_norm": 3.761117382879081, "learning_rate": 9.710778462300381e-06, "loss": 0.1768, "step": 2925 }, { "epoch": 0.13596654275092937, "grad_norm": 2.375295928247699, "learning_rate": 9.710526184877667e-06, "loss": 0.1054, "step": 2926 }, { "epoch": 0.13601301115241635, "grad_norm": 1.994306380518375, "learning_rate": 9.710273800756578e-06, "loss": 0.1348, "step": 2927 }, { "epoch": 0.13605947955390335, "grad_norm": 1.327405559647067, "learning_rate": 9.710021309942831e-06, "loss": 0.0744, "step": 2928 }, { "epoch": 0.13610594795539033, "grad_norm": 1.1504871246724302, "learning_rate": 9.709768712442142e-06, "loss": 0.0792, "step": 2929 }, { "epoch": 0.13615241635687733, "grad_norm": 1.7269483020213643, "learning_rate": 9.709516008260236e-06, "loss": 0.1442, "step": 2930 }, { "epoch": 0.1361988847583643, "grad_norm": 1.3431735920604528, "learning_rate": 9.709263197402837e-06, "loss": 0.0877, "step": 2931 }, { "epoch": 0.1362453531598513, "grad_norm": 1.0016156386049653, "learning_rate": 9.709010279875668e-06, "loss": 0.0847, "step": 2932 }, { "epoch": 0.1362918215613383, "grad_norm": 1.9911669452150724, "learning_rate": 9.70875725568446e-06, "loss": 0.0971, "step": 2933 }, { "epoch": 0.13633828996282527, "grad_norm": 1.9937824438043663, "learning_rate": 9.708504124834946e-06, "loss": 0.1169, "step": 2934 }, { "epoch": 0.13638475836431227, "grad_norm": 1.4232440820865089, "learning_rate": 9.708250887332857e-06, "loss": 0.1211, "step": 2935 }, { "epoch": 0.13643122676579925, "grad_norm": 1.4889201971280932, "learning_rate": 9.707997543183929e-06, "loss": 0.121, "step": 2936 }, { "epoch": 0.13647769516728625, "grad_norm": 1.8602913154564649, "learning_rate": 9.707744092393904e-06, "loss": 0.0921, "step": 2937 }, { "epoch": 0.13652416356877323, "grad_norm": 1.313998727045992, "learning_rate": 9.707490534968517e-06, "loss": 0.0782, "step": 2938 }, { "epoch": 0.13657063197026023, "grad_norm": 2.2368625274805782, "learning_rate": 9.70723687091352e-06, "loss": 0.1892, "step": 2939 }, { "epoch": 0.1366171003717472, "grad_norm": 1.9061318115931147, "learning_rate": 9.706983100234649e-06, "loss": 0.1334, "step": 2940 }, { "epoch": 0.1366635687732342, "grad_norm": 3.075136610575473, "learning_rate": 9.706729222937658e-06, "loss": 0.1484, "step": 2941 }, { "epoch": 0.1367100371747212, "grad_norm": 0.9523055282467436, "learning_rate": 9.706475239028297e-06, "loss": 0.0738, "step": 2942 }, { "epoch": 0.13675650557620816, "grad_norm": 1.8675349346896521, "learning_rate": 9.706221148512317e-06, "loss": 0.1105, "step": 2943 }, { "epoch": 0.13680297397769517, "grad_norm": 2.222107753416408, "learning_rate": 9.705966951395477e-06, "loss": 0.0813, "step": 2944 }, { "epoch": 0.13684944237918215, "grad_norm": 1.8046275580539999, "learning_rate": 9.705712647683532e-06, "loss": 0.0744, "step": 2945 }, { "epoch": 0.13689591078066915, "grad_norm": 3.0733175586584416, "learning_rate": 9.705458237382241e-06, "loss": 0.1575, "step": 2946 }, { "epoch": 0.13694237918215613, "grad_norm": 1.7527246317022103, "learning_rate": 9.70520372049737e-06, "loss": 0.064, "step": 2947 }, { "epoch": 0.13698884758364313, "grad_norm": 2.933003896388898, "learning_rate": 9.704949097034683e-06, "loss": 0.1114, "step": 2948 }, { "epoch": 0.1370353159851301, "grad_norm": 1.5822544536879193, "learning_rate": 9.704694366999948e-06, "loss": 0.0991, "step": 2949 }, { "epoch": 0.1370817843866171, "grad_norm": 1.6194114633475112, "learning_rate": 9.70443953039893e-06, "loss": 0.0596, "step": 2950 }, { "epoch": 0.1371282527881041, "grad_norm": 1.857479240940131, "learning_rate": 9.70418458723741e-06, "loss": 0.1009, "step": 2951 }, { "epoch": 0.13717472118959106, "grad_norm": 1.5542202718440277, "learning_rate": 9.703929537521155e-06, "loss": 0.1041, "step": 2952 }, { "epoch": 0.13722118959107807, "grad_norm": 1.283044609201199, "learning_rate": 9.703674381255945e-06, "loss": 0.0756, "step": 2953 }, { "epoch": 0.13726765799256505, "grad_norm": 1.0952124748955256, "learning_rate": 9.70341911844756e-06, "loss": 0.0821, "step": 2954 }, { "epoch": 0.13731412639405205, "grad_norm": 1.1398796281271226, "learning_rate": 9.703163749101784e-06, "loss": 0.0629, "step": 2955 }, { "epoch": 0.13736059479553903, "grad_norm": 1.4227827459089155, "learning_rate": 9.702908273224397e-06, "loss": 0.0931, "step": 2956 }, { "epoch": 0.13740706319702603, "grad_norm": 2.15392093816361, "learning_rate": 9.702652690821188e-06, "loss": 0.0804, "step": 2957 }, { "epoch": 0.137453531598513, "grad_norm": 2.1089423793982918, "learning_rate": 9.702397001897947e-06, "loss": 0.0777, "step": 2958 }, { "epoch": 0.1375, "grad_norm": 1.8271300878174292, "learning_rate": 9.702141206460464e-06, "loss": 0.17, "step": 2959 }, { "epoch": 0.137546468401487, "grad_norm": 2.335359737858358, "learning_rate": 9.701885304514535e-06, "loss": 0.0923, "step": 2960 }, { "epoch": 0.13759293680297396, "grad_norm": 5.707091437990189, "learning_rate": 9.701629296065953e-06, "loss": 0.1481, "step": 2961 }, { "epoch": 0.13763940520446097, "grad_norm": 1.272386444635502, "learning_rate": 9.701373181120521e-06, "loss": 0.0868, "step": 2962 }, { "epoch": 0.13768587360594794, "grad_norm": 1.0427376845567693, "learning_rate": 9.701116959684038e-06, "loss": 0.0761, "step": 2963 }, { "epoch": 0.13773234200743495, "grad_norm": 1.3381563526271978, "learning_rate": 9.700860631762308e-06, "loss": 0.0825, "step": 2964 }, { "epoch": 0.13777881040892193, "grad_norm": 1.2664494300401135, "learning_rate": 9.700604197361136e-06, "loss": 0.0486, "step": 2965 }, { "epoch": 0.13782527881040893, "grad_norm": 0.8446981690997888, "learning_rate": 9.700347656486333e-06, "loss": 0.0496, "step": 2966 }, { "epoch": 0.1378717472118959, "grad_norm": 3.005393112369752, "learning_rate": 9.700091009143708e-06, "loss": 0.1742, "step": 2967 }, { "epoch": 0.1379182156133829, "grad_norm": 1.8925487734381794, "learning_rate": 9.699834255339074e-06, "loss": 0.1081, "step": 2968 }, { "epoch": 0.1379646840148699, "grad_norm": 1.188357389911714, "learning_rate": 9.699577395078249e-06, "loss": 0.0851, "step": 2969 }, { "epoch": 0.13801115241635686, "grad_norm": 1.2987773722690101, "learning_rate": 9.699320428367049e-06, "loss": 0.1054, "step": 2970 }, { "epoch": 0.13805762081784387, "grad_norm": 1.879989488755253, "learning_rate": 9.699063355211295e-06, "loss": 0.1031, "step": 2971 }, { "epoch": 0.13810408921933084, "grad_norm": 1.6395038589843327, "learning_rate": 9.69880617561681e-06, "loss": 0.1113, "step": 2972 }, { "epoch": 0.13815055762081785, "grad_norm": 1.865380935749818, "learning_rate": 9.698548889589422e-06, "loss": 0.0777, "step": 2973 }, { "epoch": 0.13819702602230483, "grad_norm": 1.353712272091713, "learning_rate": 9.698291497134955e-06, "loss": 0.0668, "step": 2974 }, { "epoch": 0.13824349442379183, "grad_norm": 1.0752431747976239, "learning_rate": 9.698033998259241e-06, "loss": 0.0865, "step": 2975 }, { "epoch": 0.1382899628252788, "grad_norm": 1.8521580540805442, "learning_rate": 9.697776392968112e-06, "loss": 0.1049, "step": 2976 }, { "epoch": 0.1383364312267658, "grad_norm": 1.9735926027490525, "learning_rate": 9.697518681267404e-06, "loss": 0.142, "step": 2977 }, { "epoch": 0.1383828996282528, "grad_norm": 1.28301934635926, "learning_rate": 9.697260863162953e-06, "loss": 0.0821, "step": 2978 }, { "epoch": 0.13842936802973976, "grad_norm": 2.03580103579119, "learning_rate": 9.6970029386606e-06, "loss": 0.1353, "step": 2979 }, { "epoch": 0.13847583643122677, "grad_norm": 1.8191444195348725, "learning_rate": 9.696744907766187e-06, "loss": 0.0885, "step": 2980 }, { "epoch": 0.13852230483271374, "grad_norm": 2.600936335020394, "learning_rate": 9.69648677048556e-06, "loss": 0.2081, "step": 2981 }, { "epoch": 0.13856877323420075, "grad_norm": 3.7572473925320673, "learning_rate": 9.696228526824563e-06, "loss": 0.086, "step": 2982 }, { "epoch": 0.13861524163568772, "grad_norm": 2.053518565077712, "learning_rate": 9.695970176789047e-06, "loss": 0.1415, "step": 2983 }, { "epoch": 0.13866171003717473, "grad_norm": 1.9405658203510507, "learning_rate": 9.695711720384865e-06, "loss": 0.0977, "step": 2984 }, { "epoch": 0.1387081784386617, "grad_norm": 1.789464937683371, "learning_rate": 9.69545315761787e-06, "loss": 0.0966, "step": 2985 }, { "epoch": 0.1387546468401487, "grad_norm": 2.2847809723785537, "learning_rate": 9.69519448849392e-06, "loss": 0.0969, "step": 2986 }, { "epoch": 0.1388011152416357, "grad_norm": 1.5908099137881342, "learning_rate": 9.694935713018872e-06, "loss": 0.1179, "step": 2987 }, { "epoch": 0.13884758364312266, "grad_norm": 1.0252494167541317, "learning_rate": 9.694676831198588e-06, "loss": 0.0586, "step": 2988 }, { "epoch": 0.13889405204460967, "grad_norm": 1.0281220887288356, "learning_rate": 9.694417843038934e-06, "loss": 0.079, "step": 2989 }, { "epoch": 0.13894052044609664, "grad_norm": 1.853200224771457, "learning_rate": 9.694158748545775e-06, "loss": 0.1097, "step": 2990 }, { "epoch": 0.13898698884758365, "grad_norm": 1.9543299833807035, "learning_rate": 9.69389954772498e-06, "loss": 0.0859, "step": 2991 }, { "epoch": 0.13903345724907062, "grad_norm": 1.5814084649452333, "learning_rate": 9.69364024058242e-06, "loss": 0.0755, "step": 2992 }, { "epoch": 0.13907992565055763, "grad_norm": 1.0721103866944888, "learning_rate": 9.693380827123968e-06, "loss": 0.0495, "step": 2993 }, { "epoch": 0.1391263940520446, "grad_norm": 1.849882108202489, "learning_rate": 9.693121307355501e-06, "loss": 0.1304, "step": 2994 }, { "epoch": 0.1391728624535316, "grad_norm": 2.6954907569492836, "learning_rate": 9.692861681282897e-06, "loss": 0.1009, "step": 2995 }, { "epoch": 0.13921933085501859, "grad_norm": 3.066353968045941, "learning_rate": 9.692601948912036e-06, "loss": 0.1477, "step": 2996 }, { "epoch": 0.13926579925650556, "grad_norm": 1.8238552644009973, "learning_rate": 9.692342110248802e-06, "loss": 0.0868, "step": 2997 }, { "epoch": 0.13931226765799257, "grad_norm": 2.366123843496833, "learning_rate": 9.692082165299082e-06, "loss": 0.1374, "step": 2998 }, { "epoch": 0.13935873605947954, "grad_norm": 1.251029786238576, "learning_rate": 9.691822114068763e-06, "loss": 0.1217, "step": 2999 }, { "epoch": 0.13940520446096655, "grad_norm": 1.5239839026219357, "learning_rate": 9.691561956563734e-06, "loss": 0.0835, "step": 3000 }, { "epoch": 0.13945167286245352, "grad_norm": 1.9868333176120812, "learning_rate": 9.691301692789887e-06, "loss": 0.1027, "step": 3001 }, { "epoch": 0.13949814126394053, "grad_norm": 1.1383377142558013, "learning_rate": 9.69104132275312e-06, "loss": 0.0776, "step": 3002 }, { "epoch": 0.1395446096654275, "grad_norm": 1.804475472320086, "learning_rate": 9.690780846459332e-06, "loss": 0.1262, "step": 3003 }, { "epoch": 0.1395910780669145, "grad_norm": 0.789196822962709, "learning_rate": 9.69052026391442e-06, "loss": 0.0544, "step": 3004 }, { "epoch": 0.13963754646840149, "grad_norm": 1.1103381528655731, "learning_rate": 9.690259575124287e-06, "loss": 0.1162, "step": 3005 }, { "epoch": 0.13968401486988846, "grad_norm": 2.083884377741759, "learning_rate": 9.689998780094839e-06, "loss": 0.0659, "step": 3006 }, { "epoch": 0.13973048327137547, "grad_norm": 1.5978786046652897, "learning_rate": 9.689737878831981e-06, "loss": 0.1117, "step": 3007 }, { "epoch": 0.13977695167286244, "grad_norm": 1.247563799185222, "learning_rate": 9.689476871341626e-06, "loss": 0.0677, "step": 3008 }, { "epoch": 0.13982342007434945, "grad_norm": 0.6707762948931311, "learning_rate": 9.689215757629682e-06, "loss": 0.0639, "step": 3009 }, { "epoch": 0.13986988847583642, "grad_norm": 2.474769594051131, "learning_rate": 9.688954537702067e-06, "loss": 0.0723, "step": 3010 }, { "epoch": 0.13991635687732343, "grad_norm": 1.2276984372237525, "learning_rate": 9.688693211564697e-06, "loss": 0.0904, "step": 3011 }, { "epoch": 0.1399628252788104, "grad_norm": 1.0091288718214269, "learning_rate": 9.688431779223491e-06, "loss": 0.0753, "step": 3012 }, { "epoch": 0.1400092936802974, "grad_norm": 5.608397217322183, "learning_rate": 9.688170240684369e-06, "loss": 0.1547, "step": 3013 }, { "epoch": 0.14005576208178439, "grad_norm": 1.6722345549497106, "learning_rate": 9.687908595953258e-06, "loss": 0.1063, "step": 3014 }, { "epoch": 0.14010223048327136, "grad_norm": 0.8794034902572375, "learning_rate": 9.687646845036083e-06, "loss": 0.0661, "step": 3015 }, { "epoch": 0.14014869888475837, "grad_norm": 2.6585076837830073, "learning_rate": 9.687384987938775e-06, "loss": 0.1501, "step": 3016 }, { "epoch": 0.14019516728624534, "grad_norm": 0.9810263433900627, "learning_rate": 9.687123024667262e-06, "loss": 0.0906, "step": 3017 }, { "epoch": 0.14024163568773235, "grad_norm": 1.8939656165022694, "learning_rate": 9.68686095522748e-06, "loss": 0.0871, "step": 3018 }, { "epoch": 0.14028810408921932, "grad_norm": 1.7994816074654902, "learning_rate": 9.686598779625365e-06, "loss": 0.0826, "step": 3019 }, { "epoch": 0.14033457249070633, "grad_norm": 1.4667302496648902, "learning_rate": 9.686336497866853e-06, "loss": 0.0786, "step": 3020 }, { "epoch": 0.1403810408921933, "grad_norm": 1.4876950762163441, "learning_rate": 9.68607410995789e-06, "loss": 0.1056, "step": 3021 }, { "epoch": 0.1404275092936803, "grad_norm": 1.916806000817334, "learning_rate": 9.685811615904415e-06, "loss": 0.1511, "step": 3022 }, { "epoch": 0.14047397769516728, "grad_norm": 1.7459065624796057, "learning_rate": 9.685549015712374e-06, "loss": 0.0887, "step": 3023 }, { "epoch": 0.14052044609665426, "grad_norm": 3.345840810803632, "learning_rate": 9.685286309387719e-06, "loss": 0.1359, "step": 3024 }, { "epoch": 0.14056691449814127, "grad_norm": 2.9520323883336492, "learning_rate": 9.685023496936396e-06, "loss": 0.1916, "step": 3025 }, { "epoch": 0.14061338289962824, "grad_norm": 1.4579417756837267, "learning_rate": 9.684760578364359e-06, "loss": 0.1239, "step": 3026 }, { "epoch": 0.14065985130111525, "grad_norm": 1.454125172506143, "learning_rate": 9.684497553677567e-06, "loss": 0.1045, "step": 3027 }, { "epoch": 0.14070631970260222, "grad_norm": 0.9854121650850461, "learning_rate": 9.684234422881972e-06, "loss": 0.0578, "step": 3028 }, { "epoch": 0.14075278810408923, "grad_norm": 4.67066566465014, "learning_rate": 9.683971185983539e-06, "loss": 0.0787, "step": 3029 }, { "epoch": 0.1407992565055762, "grad_norm": 6.674943694986556, "learning_rate": 9.683707842988229e-06, "loss": 0.1532, "step": 3030 }, { "epoch": 0.1408457249070632, "grad_norm": 1.2326585383598638, "learning_rate": 9.683444393902006e-06, "loss": 0.0586, "step": 3031 }, { "epoch": 0.14089219330855018, "grad_norm": 1.6398041920351853, "learning_rate": 9.68318083873084e-06, "loss": 0.0805, "step": 3032 }, { "epoch": 0.1409386617100372, "grad_norm": 1.4584134909858923, "learning_rate": 9.682917177480697e-06, "loss": 0.0648, "step": 3033 }, { "epoch": 0.14098513011152417, "grad_norm": 2.9220923251960595, "learning_rate": 9.682653410157552e-06, "loss": 0.1178, "step": 3034 }, { "epoch": 0.14103159851301114, "grad_norm": 2.4810262968338894, "learning_rate": 9.68238953676738e-06, "loss": 0.0985, "step": 3035 }, { "epoch": 0.14107806691449815, "grad_norm": 12.182300127408041, "learning_rate": 9.682125557316156e-06, "loss": 0.0709, "step": 3036 }, { "epoch": 0.14112453531598512, "grad_norm": 2.7694918033723166, "learning_rate": 9.681861471809858e-06, "loss": 0.1203, "step": 3037 }, { "epoch": 0.14117100371747213, "grad_norm": 1.8726611552930255, "learning_rate": 9.681597280254473e-06, "loss": 0.1024, "step": 3038 }, { "epoch": 0.1412174721189591, "grad_norm": 2.1751094902402213, "learning_rate": 9.681332982655983e-06, "loss": 0.0911, "step": 3039 }, { "epoch": 0.1412639405204461, "grad_norm": 2.1660704433220532, "learning_rate": 9.681068579020371e-06, "loss": 0.1467, "step": 3040 }, { "epoch": 0.14131040892193308, "grad_norm": 3.6822396815458465, "learning_rate": 9.680804069353631e-06, "loss": 0.1584, "step": 3041 }, { "epoch": 0.1413568773234201, "grad_norm": 5.587293549932772, "learning_rate": 9.680539453661751e-06, "loss": 0.1105, "step": 3042 }, { "epoch": 0.14140334572490706, "grad_norm": 1.2946177056608699, "learning_rate": 9.680274731950728e-06, "loss": 0.0699, "step": 3043 }, { "epoch": 0.14144981412639404, "grad_norm": 2.821609805410004, "learning_rate": 9.680009904226555e-06, "loss": 0.1661, "step": 3044 }, { "epoch": 0.14149628252788105, "grad_norm": 1.7433527537601323, "learning_rate": 9.679744970495233e-06, "loss": 0.0655, "step": 3045 }, { "epoch": 0.14154275092936802, "grad_norm": 1.8968569417045913, "learning_rate": 9.679479930762761e-06, "loss": 0.1102, "step": 3046 }, { "epoch": 0.14158921933085503, "grad_norm": 1.8728171223265084, "learning_rate": 9.679214785035144e-06, "loss": 0.1454, "step": 3047 }, { "epoch": 0.141635687732342, "grad_norm": 1.4590485342873745, "learning_rate": 9.678949533318387e-06, "loss": 0.1247, "step": 3048 }, { "epoch": 0.141682156133829, "grad_norm": 1.7102274399100725, "learning_rate": 9.6786841756185e-06, "loss": 0.0832, "step": 3049 }, { "epoch": 0.14172862453531598, "grad_norm": 1.7587075667056753, "learning_rate": 9.67841871194149e-06, "loss": 0.1008, "step": 3050 }, { "epoch": 0.141775092936803, "grad_norm": 1.8108893205243044, "learning_rate": 9.678153142293375e-06, "loss": 0.1137, "step": 3051 }, { "epoch": 0.14182156133828996, "grad_norm": 3.5274599051884357, "learning_rate": 9.677887466680165e-06, "loss": 0.1299, "step": 3052 }, { "epoch": 0.14186802973977694, "grad_norm": 2.8839517123706186, "learning_rate": 9.67762168510788e-06, "loss": 0.1335, "step": 3053 }, { "epoch": 0.14191449814126395, "grad_norm": 1.3810717002650552, "learning_rate": 9.677355797582544e-06, "loss": 0.0788, "step": 3054 }, { "epoch": 0.14196096654275092, "grad_norm": 2.813759861618201, "learning_rate": 9.677089804110173e-06, "loss": 0.1176, "step": 3055 }, { "epoch": 0.14200743494423793, "grad_norm": 1.520688094538695, "learning_rate": 9.676823704696799e-06, "loss": 0.083, "step": 3056 }, { "epoch": 0.1420539033457249, "grad_norm": 1.6013956049901221, "learning_rate": 9.676557499348442e-06, "loss": 0.1068, "step": 3057 }, { "epoch": 0.1421003717472119, "grad_norm": 2.8066263827005353, "learning_rate": 9.676291188071138e-06, "loss": 0.1747, "step": 3058 }, { "epoch": 0.14214684014869888, "grad_norm": 1.339341219822548, "learning_rate": 9.676024770870915e-06, "loss": 0.0842, "step": 3059 }, { "epoch": 0.1421933085501859, "grad_norm": 1.7666909375319149, "learning_rate": 9.675758247753813e-06, "loss": 0.1394, "step": 3060 }, { "epoch": 0.14223977695167286, "grad_norm": 1.8684974845941782, "learning_rate": 9.675491618725863e-06, "loss": 0.1118, "step": 3061 }, { "epoch": 0.14228624535315984, "grad_norm": 1.6744637164958676, "learning_rate": 9.675224883793106e-06, "loss": 0.0882, "step": 3062 }, { "epoch": 0.14233271375464684, "grad_norm": 1.7343351307585033, "learning_rate": 9.674958042961587e-06, "loss": 0.088, "step": 3063 }, { "epoch": 0.14237918215613382, "grad_norm": 1.1756962255747767, "learning_rate": 9.674691096237347e-06, "loss": 0.0801, "step": 3064 }, { "epoch": 0.14242565055762083, "grad_norm": 1.40044102690407, "learning_rate": 9.674424043626434e-06, "loss": 0.0928, "step": 3065 }, { "epoch": 0.1424721189591078, "grad_norm": 1.2977045806954004, "learning_rate": 9.674156885134896e-06, "loss": 0.0826, "step": 3066 }, { "epoch": 0.1425185873605948, "grad_norm": 1.0622597321243243, "learning_rate": 9.673889620768786e-06, "loss": 0.0955, "step": 3067 }, { "epoch": 0.14256505576208178, "grad_norm": 8.69878932186056, "learning_rate": 9.673622250534155e-06, "loss": 0.1432, "step": 3068 }, { "epoch": 0.1426115241635688, "grad_norm": 1.8932738917360166, "learning_rate": 9.673354774437064e-06, "loss": 0.1038, "step": 3069 }, { "epoch": 0.14265799256505576, "grad_norm": 1.3522860289893657, "learning_rate": 9.673087192483567e-06, "loss": 0.0862, "step": 3070 }, { "epoch": 0.14270446096654274, "grad_norm": 2.0486516458722956, "learning_rate": 9.672819504679727e-06, "loss": 0.1314, "step": 3071 }, { "epoch": 0.14275092936802974, "grad_norm": 2.3655107659448675, "learning_rate": 9.672551711031608e-06, "loss": 0.1708, "step": 3072 }, { "epoch": 0.14279739776951672, "grad_norm": 2.889017365926993, "learning_rate": 9.672283811545273e-06, "loss": 0.1393, "step": 3073 }, { "epoch": 0.14284386617100372, "grad_norm": 0.9944024782737089, "learning_rate": 9.672015806226794e-06, "loss": 0.0686, "step": 3074 }, { "epoch": 0.1428903345724907, "grad_norm": 1.7467464677445719, "learning_rate": 9.67174769508224e-06, "loss": 0.0836, "step": 3075 }, { "epoch": 0.1429368029739777, "grad_norm": 6.165833933721245, "learning_rate": 9.671479478117682e-06, "loss": 0.132, "step": 3076 }, { "epoch": 0.14298327137546468, "grad_norm": 2.326835529703187, "learning_rate": 9.671211155339197e-06, "loss": 0.0967, "step": 3077 }, { "epoch": 0.1430297397769517, "grad_norm": 2.3741147394092157, "learning_rate": 9.670942726752865e-06, "loss": 0.1255, "step": 3078 }, { "epoch": 0.14307620817843866, "grad_norm": 1.835547061030569, "learning_rate": 9.670674192364765e-06, "loss": 0.0708, "step": 3079 }, { "epoch": 0.14312267657992564, "grad_norm": 1.6004563399410687, "learning_rate": 9.670405552180977e-06, "loss": 0.0717, "step": 3080 }, { "epoch": 0.14316914498141264, "grad_norm": 2.0431648131387368, "learning_rate": 9.670136806207586e-06, "loss": 0.1151, "step": 3081 }, { "epoch": 0.14321561338289962, "grad_norm": 1.3448393419552038, "learning_rate": 9.669867954450684e-06, "loss": 0.0867, "step": 3082 }, { "epoch": 0.14326208178438662, "grad_norm": 1.5828193586680006, "learning_rate": 9.669598996916357e-06, "loss": 0.1078, "step": 3083 }, { "epoch": 0.1433085501858736, "grad_norm": 2.3058584772662334, "learning_rate": 9.669329933610695e-06, "loss": 0.1572, "step": 3084 }, { "epoch": 0.1433550185873606, "grad_norm": 1.7540221980504684, "learning_rate": 9.6690607645398e-06, "loss": 0.0582, "step": 3085 }, { "epoch": 0.14340148698884758, "grad_norm": 1.6416801597778692, "learning_rate": 9.668791489709764e-06, "loss": 0.1582, "step": 3086 }, { "epoch": 0.14344795539033459, "grad_norm": 1.7891277363144034, "learning_rate": 9.668522109126686e-06, "loss": 0.0683, "step": 3087 }, { "epoch": 0.14349442379182156, "grad_norm": 1.7109056731582457, "learning_rate": 9.668252622796667e-06, "loss": 0.1155, "step": 3088 }, { "epoch": 0.14354089219330854, "grad_norm": 1.0698139867537133, "learning_rate": 9.667983030725817e-06, "loss": 0.0796, "step": 3089 }, { "epoch": 0.14358736059479554, "grad_norm": 1.7681337931458931, "learning_rate": 9.667713332920235e-06, "loss": 0.1335, "step": 3090 }, { "epoch": 0.14363382899628252, "grad_norm": 1.0684080289949736, "learning_rate": 9.667443529386033e-06, "loss": 0.0594, "step": 3091 }, { "epoch": 0.14368029739776952, "grad_norm": 2.3585705487323474, "learning_rate": 9.667173620129325e-06, "loss": 0.1262, "step": 3092 }, { "epoch": 0.1437267657992565, "grad_norm": 1.6302106143705335, "learning_rate": 9.66690360515622e-06, "loss": 0.0796, "step": 3093 }, { "epoch": 0.1437732342007435, "grad_norm": 1.9897656983291683, "learning_rate": 9.666633484472839e-06, "loss": 0.0876, "step": 3094 }, { "epoch": 0.14381970260223048, "grad_norm": 3.3568746736161565, "learning_rate": 9.666363258085296e-06, "loss": 0.1109, "step": 3095 }, { "epoch": 0.14386617100371749, "grad_norm": 1.7902697712138942, "learning_rate": 9.666092925999715e-06, "loss": 0.0739, "step": 3096 }, { "epoch": 0.14391263940520446, "grad_norm": 2.0741230148290195, "learning_rate": 9.665822488222218e-06, "loss": 0.1145, "step": 3097 }, { "epoch": 0.14395910780669144, "grad_norm": 1.8949448425421616, "learning_rate": 9.66555194475893e-06, "loss": 0.1161, "step": 3098 }, { "epoch": 0.14400557620817844, "grad_norm": 3.805801165924311, "learning_rate": 9.665281295615983e-06, "loss": 0.1132, "step": 3099 }, { "epoch": 0.14405204460966542, "grad_norm": 1.6799629714212703, "learning_rate": 9.665010540799502e-06, "loss": 0.0901, "step": 3100 }, { "epoch": 0.14409851301115242, "grad_norm": 2.9371422279145816, "learning_rate": 9.664739680315622e-06, "loss": 0.1021, "step": 3101 }, { "epoch": 0.1441449814126394, "grad_norm": 2.4415093427803325, "learning_rate": 9.66446871417048e-06, "loss": 0.1057, "step": 3102 }, { "epoch": 0.1441914498141264, "grad_norm": 2.2795362200935, "learning_rate": 9.664197642370212e-06, "loss": 0.1132, "step": 3103 }, { "epoch": 0.14423791821561338, "grad_norm": 3.273211430052109, "learning_rate": 9.663926464920959e-06, "loss": 0.1183, "step": 3104 }, { "epoch": 0.14428438661710039, "grad_norm": 2.7191535333834023, "learning_rate": 9.663655181828862e-06, "loss": 0.124, "step": 3105 }, { "epoch": 0.14433085501858736, "grad_norm": 1.2665532579789798, "learning_rate": 9.663383793100067e-06, "loss": 0.0815, "step": 3106 }, { "epoch": 0.14437732342007434, "grad_norm": 1.3557733011604471, "learning_rate": 9.663112298740721e-06, "loss": 0.1028, "step": 3107 }, { "epoch": 0.14442379182156134, "grad_norm": 1.9540409644886998, "learning_rate": 9.662840698756973e-06, "loss": 0.0944, "step": 3108 }, { "epoch": 0.14447026022304832, "grad_norm": 1.9120938686698636, "learning_rate": 9.662568993154977e-06, "loss": 0.1328, "step": 3109 }, { "epoch": 0.14451672862453532, "grad_norm": 1.3111372138069017, "learning_rate": 9.662297181940886e-06, "loss": 0.1455, "step": 3110 }, { "epoch": 0.1445631970260223, "grad_norm": 2.3406203563022085, "learning_rate": 9.662025265120855e-06, "loss": 0.107, "step": 3111 }, { "epoch": 0.1446096654275093, "grad_norm": 3.223355689309769, "learning_rate": 9.661753242701046e-06, "loss": 0.143, "step": 3112 }, { "epoch": 0.14465613382899628, "grad_norm": 1.6626467789384516, "learning_rate": 9.661481114687621e-06, "loss": 0.1108, "step": 3113 }, { "epoch": 0.14470260223048328, "grad_norm": 1.9011471451085955, "learning_rate": 9.661208881086742e-06, "loss": 0.1023, "step": 3114 }, { "epoch": 0.14474907063197026, "grad_norm": 1.8725499093792983, "learning_rate": 9.660936541904574e-06, "loss": 0.0996, "step": 3115 }, { "epoch": 0.14479553903345724, "grad_norm": 2.292939135713636, "learning_rate": 9.66066409714729e-06, "loss": 0.1237, "step": 3116 }, { "epoch": 0.14484200743494424, "grad_norm": 1.2487757436472096, "learning_rate": 9.660391546821057e-06, "loss": 0.0691, "step": 3117 }, { "epoch": 0.14488847583643122, "grad_norm": 1.7424150595407832, "learning_rate": 9.660118890932053e-06, "loss": 0.0802, "step": 3118 }, { "epoch": 0.14493494423791822, "grad_norm": 1.9852490394146542, "learning_rate": 9.65984612948645e-06, "loss": 0.0958, "step": 3119 }, { "epoch": 0.1449814126394052, "grad_norm": 2.04895078104183, "learning_rate": 9.659573262490427e-06, "loss": 0.1539, "step": 3120 }, { "epoch": 0.1450278810408922, "grad_norm": 1.830311627733948, "learning_rate": 9.659300289950167e-06, "loss": 0.0733, "step": 3121 }, { "epoch": 0.14507434944237918, "grad_norm": 1.1223219535675606, "learning_rate": 9.65902721187185e-06, "loss": 0.0634, "step": 3122 }, { "epoch": 0.14512081784386618, "grad_norm": 0.9968692899451117, "learning_rate": 9.658754028261665e-06, "loss": 0.0526, "step": 3123 }, { "epoch": 0.14516728624535316, "grad_norm": 1.511221966643235, "learning_rate": 9.658480739125798e-06, "loss": 0.0807, "step": 3124 }, { "epoch": 0.14521375464684014, "grad_norm": 2.7551806576818074, "learning_rate": 9.658207344470437e-06, "loss": 0.1302, "step": 3125 }, { "epoch": 0.14526022304832714, "grad_norm": 1.639085114369281, "learning_rate": 9.657933844301779e-06, "loss": 0.1221, "step": 3126 }, { "epoch": 0.14530669144981412, "grad_norm": 2.8945029488251923, "learning_rate": 9.657660238626016e-06, "loss": 0.0686, "step": 3127 }, { "epoch": 0.14535315985130112, "grad_norm": 2.377760674120945, "learning_rate": 9.657386527449348e-06, "loss": 0.1081, "step": 3128 }, { "epoch": 0.1453996282527881, "grad_norm": 2.930747041580269, "learning_rate": 9.657112710777971e-06, "loss": 0.1111, "step": 3129 }, { "epoch": 0.1454460966542751, "grad_norm": 2.050363266847825, "learning_rate": 9.65683878861809e-06, "loss": 0.086, "step": 3130 }, { "epoch": 0.14549256505576208, "grad_norm": 1.143951375361423, "learning_rate": 9.65656476097591e-06, "loss": 0.069, "step": 3131 }, { "epoch": 0.14553903345724908, "grad_norm": 1.8628051997338897, "learning_rate": 9.656290627857638e-06, "loss": 0.1414, "step": 3132 }, { "epoch": 0.14558550185873606, "grad_norm": 1.103303067289649, "learning_rate": 9.656016389269482e-06, "loss": 0.1219, "step": 3133 }, { "epoch": 0.14563197026022304, "grad_norm": 2.3649676038554617, "learning_rate": 9.655742045217654e-06, "loss": 0.0898, "step": 3134 }, { "epoch": 0.14567843866171004, "grad_norm": 2.992983537458644, "learning_rate": 9.65546759570837e-06, "loss": 0.151, "step": 3135 }, { "epoch": 0.14572490706319702, "grad_norm": 1.0050967015048073, "learning_rate": 9.655193040747844e-06, "loss": 0.057, "step": 3136 }, { "epoch": 0.14577137546468402, "grad_norm": 1.888455455617989, "learning_rate": 9.654918380342295e-06, "loss": 0.0541, "step": 3137 }, { "epoch": 0.145817843866171, "grad_norm": 1.963870609972864, "learning_rate": 9.654643614497948e-06, "loss": 0.1167, "step": 3138 }, { "epoch": 0.145864312267658, "grad_norm": 1.0140340224695699, "learning_rate": 9.654368743221022e-06, "loss": 0.0882, "step": 3139 }, { "epoch": 0.14591078066914498, "grad_norm": 1.61178317534085, "learning_rate": 9.654093766517746e-06, "loss": 0.1088, "step": 3140 }, { "epoch": 0.14595724907063198, "grad_norm": 1.3842955406442334, "learning_rate": 9.653818684394347e-06, "loss": 0.099, "step": 3141 }, { "epoch": 0.14600371747211896, "grad_norm": 1.0951595254587478, "learning_rate": 9.653543496857058e-06, "loss": 0.0431, "step": 3142 }, { "epoch": 0.14605018587360594, "grad_norm": 1.6695126777640943, "learning_rate": 9.653268203912112e-06, "loss": 0.0932, "step": 3143 }, { "epoch": 0.14609665427509294, "grad_norm": 1.5826484512450687, "learning_rate": 9.652992805565742e-06, "loss": 0.1108, "step": 3144 }, { "epoch": 0.14614312267657992, "grad_norm": 1.8222326493674785, "learning_rate": 9.652717301824188e-06, "loss": 0.1072, "step": 3145 }, { "epoch": 0.14618959107806692, "grad_norm": 2.199709504661357, "learning_rate": 9.652441692693691e-06, "loss": 0.109, "step": 3146 }, { "epoch": 0.1462360594795539, "grad_norm": 1.0773128256101205, "learning_rate": 9.652165978180491e-06, "loss": 0.0744, "step": 3147 }, { "epoch": 0.1462825278810409, "grad_norm": 1.8950979205192728, "learning_rate": 9.651890158290838e-06, "loss": 0.0747, "step": 3148 }, { "epoch": 0.14632899628252788, "grad_norm": 1.674338639338766, "learning_rate": 9.651614233030977e-06, "loss": 0.0684, "step": 3149 }, { "epoch": 0.14637546468401488, "grad_norm": 1.5905795092903576, "learning_rate": 9.651338202407155e-06, "loss": 0.0986, "step": 3150 }, { "epoch": 0.14642193308550186, "grad_norm": 3.8638975272698888, "learning_rate": 9.651062066425629e-06, "loss": 0.1112, "step": 3151 }, { "epoch": 0.14646840148698884, "grad_norm": 0.9978937786255769, "learning_rate": 9.650785825092652e-06, "loss": 0.0772, "step": 3152 }, { "epoch": 0.14651486988847584, "grad_norm": 1.7697796583902756, "learning_rate": 9.650509478414483e-06, "loss": 0.0908, "step": 3153 }, { "epoch": 0.14656133828996282, "grad_norm": 1.0820004308901998, "learning_rate": 9.650233026397377e-06, "loss": 0.0658, "step": 3154 }, { "epoch": 0.14660780669144982, "grad_norm": 1.8936229369106017, "learning_rate": 9.649956469047598e-06, "loss": 0.0914, "step": 3155 }, { "epoch": 0.1466542750929368, "grad_norm": 4.769040297930651, "learning_rate": 9.649679806371413e-06, "loss": 0.2037, "step": 3156 }, { "epoch": 0.1467007434944238, "grad_norm": 1.1698581956506269, "learning_rate": 9.649403038375086e-06, "loss": 0.0832, "step": 3157 }, { "epoch": 0.14674721189591078, "grad_norm": 2.10248737345191, "learning_rate": 9.649126165064888e-06, "loss": 0.0663, "step": 3158 }, { "epoch": 0.14679368029739778, "grad_norm": 1.7319924680654823, "learning_rate": 9.648849186447089e-06, "loss": 0.0943, "step": 3159 }, { "epoch": 0.14684014869888476, "grad_norm": 0.8076828979586175, "learning_rate": 9.648572102527962e-06, "loss": 0.0596, "step": 3160 }, { "epoch": 0.14688661710037174, "grad_norm": 1.4545132386977906, "learning_rate": 9.648294913313784e-06, "loss": 0.0747, "step": 3161 }, { "epoch": 0.14693308550185874, "grad_norm": 2.6001021690883475, "learning_rate": 9.648017618810834e-06, "loss": 0.1624, "step": 3162 }, { "epoch": 0.14697955390334572, "grad_norm": 0.7673973870577375, "learning_rate": 9.647740219025395e-06, "loss": 0.0537, "step": 3163 }, { "epoch": 0.14702602230483272, "grad_norm": 2.651284240330332, "learning_rate": 9.647462713963747e-06, "loss": 0.1041, "step": 3164 }, { "epoch": 0.1470724907063197, "grad_norm": 2.3263322712099885, "learning_rate": 9.647185103632178e-06, "loss": 0.1283, "step": 3165 }, { "epoch": 0.1471189591078067, "grad_norm": 10.473505773521383, "learning_rate": 9.646907388036972e-06, "loss": 0.0986, "step": 3166 }, { "epoch": 0.14716542750929368, "grad_norm": 2.2214921814527058, "learning_rate": 9.646629567184426e-06, "loss": 0.1237, "step": 3167 }, { "epoch": 0.14721189591078068, "grad_norm": 1.0948475044311816, "learning_rate": 9.646351641080829e-06, "loss": 0.0635, "step": 3168 }, { "epoch": 0.14725836431226766, "grad_norm": 1.2101363669123917, "learning_rate": 9.646073609732478e-06, "loss": 0.1076, "step": 3169 }, { "epoch": 0.14730483271375464, "grad_norm": 1.4951500171790952, "learning_rate": 9.645795473145669e-06, "loss": 0.1486, "step": 3170 }, { "epoch": 0.14735130111524164, "grad_norm": 1.631026140411837, "learning_rate": 9.645517231326703e-06, "loss": 0.0767, "step": 3171 }, { "epoch": 0.14739776951672862, "grad_norm": 1.3956304343278496, "learning_rate": 9.645238884281883e-06, "loss": 0.113, "step": 3172 }, { "epoch": 0.14744423791821562, "grad_norm": 0.944997256611341, "learning_rate": 9.644960432017512e-06, "loss": 0.0643, "step": 3173 }, { "epoch": 0.1474907063197026, "grad_norm": 1.565624296223789, "learning_rate": 9.644681874539899e-06, "loss": 0.0634, "step": 3174 }, { "epoch": 0.1475371747211896, "grad_norm": 1.4391545917714403, "learning_rate": 9.644403211855353e-06, "loss": 0.0974, "step": 3175 }, { "epoch": 0.14758364312267658, "grad_norm": 1.3676042720309896, "learning_rate": 9.644124443970187e-06, "loss": 0.0826, "step": 3176 }, { "epoch": 0.14763011152416358, "grad_norm": 1.8627662428859082, "learning_rate": 9.643845570890713e-06, "loss": 0.0843, "step": 3177 }, { "epoch": 0.14767657992565056, "grad_norm": 0.8254443970203527, "learning_rate": 9.643566592623251e-06, "loss": 0.0612, "step": 3178 }, { "epoch": 0.14772304832713753, "grad_norm": 1.39899029048143, "learning_rate": 9.643287509174118e-06, "loss": 0.0911, "step": 3179 }, { "epoch": 0.14776951672862454, "grad_norm": 1.7098829600192231, "learning_rate": 9.643008320549637e-06, "loss": 0.081, "step": 3180 }, { "epoch": 0.14781598513011152, "grad_norm": 2.829891709216028, "learning_rate": 9.642729026756129e-06, "loss": 0.0819, "step": 3181 }, { "epoch": 0.14786245353159852, "grad_norm": 1.6250910166846766, "learning_rate": 9.642449627799921e-06, "loss": 0.0876, "step": 3182 }, { "epoch": 0.1479089219330855, "grad_norm": 1.7603233675495793, "learning_rate": 9.642170123687344e-06, "loss": 0.0746, "step": 3183 }, { "epoch": 0.1479553903345725, "grad_norm": 0.8487456915188352, "learning_rate": 9.64189051442473e-06, "loss": 0.0707, "step": 3184 }, { "epoch": 0.14800185873605948, "grad_norm": 1.824881593846444, "learning_rate": 9.641610800018409e-06, "loss": 0.1084, "step": 3185 }, { "epoch": 0.14804832713754648, "grad_norm": 2.5649735147192523, "learning_rate": 9.641330980474717e-06, "loss": 0.0986, "step": 3186 }, { "epoch": 0.14809479553903346, "grad_norm": 6.5690172103495135, "learning_rate": 9.641051055799994e-06, "loss": 0.1108, "step": 3187 }, { "epoch": 0.14814126394052043, "grad_norm": 1.4461016173403995, "learning_rate": 9.64077102600058e-06, "loss": 0.1391, "step": 3188 }, { "epoch": 0.14818773234200744, "grad_norm": 1.9900367605613116, "learning_rate": 9.640490891082817e-06, "loss": 0.1223, "step": 3189 }, { "epoch": 0.14823420074349442, "grad_norm": 2.8331317170054753, "learning_rate": 9.640210651053054e-06, "loss": 0.1325, "step": 3190 }, { "epoch": 0.14828066914498142, "grad_norm": 1.381638694109125, "learning_rate": 9.639930305917632e-06, "loss": 0.0982, "step": 3191 }, { "epoch": 0.1483271375464684, "grad_norm": 1.625283114992538, "learning_rate": 9.63964985568291e-06, "loss": 0.1616, "step": 3192 }, { "epoch": 0.1483736059479554, "grad_norm": 4.020777420750549, "learning_rate": 9.639369300355232e-06, "loss": 0.0708, "step": 3193 }, { "epoch": 0.14842007434944238, "grad_norm": 2.373410170941948, "learning_rate": 9.639088639940958e-06, "loss": 0.1068, "step": 3194 }, { "epoch": 0.14846654275092938, "grad_norm": 1.4983382253841957, "learning_rate": 9.638807874446443e-06, "loss": 0.1696, "step": 3195 }, { "epoch": 0.14851301115241636, "grad_norm": 2.5620761531533813, "learning_rate": 9.638527003878049e-06, "loss": 0.0838, "step": 3196 }, { "epoch": 0.14855947955390333, "grad_norm": 1.1276852852012489, "learning_rate": 9.638246028242135e-06, "loss": 0.0731, "step": 3197 }, { "epoch": 0.14860594795539034, "grad_norm": 1.3176140484986605, "learning_rate": 9.637964947545066e-06, "loss": 0.0969, "step": 3198 }, { "epoch": 0.14865241635687731, "grad_norm": 9.57451048236997, "learning_rate": 9.63768376179321e-06, "loss": 0.0779, "step": 3199 }, { "epoch": 0.14869888475836432, "grad_norm": 2.7429995911355847, "learning_rate": 9.637402470992937e-06, "loss": 0.0664, "step": 3200 }, { "epoch": 0.1487453531598513, "grad_norm": 4.507657063689263, "learning_rate": 9.637121075150617e-06, "loss": 0.0686, "step": 3201 }, { "epoch": 0.1487918215613383, "grad_norm": 1.06152397719085, "learning_rate": 9.636839574272623e-06, "loss": 0.0616, "step": 3202 }, { "epoch": 0.14883828996282528, "grad_norm": 1.9071460698548288, "learning_rate": 9.636557968365334e-06, "loss": 0.0941, "step": 3203 }, { "epoch": 0.14888475836431228, "grad_norm": 3.418072979771253, "learning_rate": 9.636276257435126e-06, "loss": 0.1272, "step": 3204 }, { "epoch": 0.14893122676579926, "grad_norm": 1.2351588409105654, "learning_rate": 9.63599444148838e-06, "loss": 0.1078, "step": 3205 }, { "epoch": 0.14897769516728623, "grad_norm": 1.8319087055847372, "learning_rate": 9.635712520531483e-06, "loss": 0.0788, "step": 3206 }, { "epoch": 0.14902416356877324, "grad_norm": 2.505801461812259, "learning_rate": 9.635430494570818e-06, "loss": 0.0795, "step": 3207 }, { "epoch": 0.14907063197026021, "grad_norm": 2.7744418222373413, "learning_rate": 9.635148363612774e-06, "loss": 0.0928, "step": 3208 }, { "epoch": 0.14911710037174722, "grad_norm": 1.462639730898746, "learning_rate": 9.634866127663739e-06, "loss": 0.0903, "step": 3209 }, { "epoch": 0.1491635687732342, "grad_norm": 2.064253628577882, "learning_rate": 9.63458378673011e-06, "loss": 0.1621, "step": 3210 }, { "epoch": 0.1492100371747212, "grad_norm": 1.0371307199545163, "learning_rate": 9.63430134081828e-06, "loss": 0.0554, "step": 3211 }, { "epoch": 0.14925650557620818, "grad_norm": 1.2358646602942351, "learning_rate": 9.634018789934647e-06, "loss": 0.0755, "step": 3212 }, { "epoch": 0.14930297397769518, "grad_norm": 4.009192986149783, "learning_rate": 9.633736134085611e-06, "loss": 0.1365, "step": 3213 }, { "epoch": 0.14934944237918216, "grad_norm": 2.297047338160307, "learning_rate": 9.633453373277576e-06, "loss": 0.062, "step": 3214 }, { "epoch": 0.14939591078066913, "grad_norm": 1.9556735587653848, "learning_rate": 9.633170507516942e-06, "loss": 0.1195, "step": 3215 }, { "epoch": 0.14944237918215614, "grad_norm": 1.3524467952862678, "learning_rate": 9.632887536810122e-06, "loss": 0.0826, "step": 3216 }, { "epoch": 0.14948884758364311, "grad_norm": 1.3171988549517208, "learning_rate": 9.632604461163525e-06, "loss": 0.1005, "step": 3217 }, { "epoch": 0.14953531598513012, "grad_norm": 1.3605573751601736, "learning_rate": 9.632321280583557e-06, "loss": 0.0801, "step": 3218 }, { "epoch": 0.1495817843866171, "grad_norm": 1.2753982738639453, "learning_rate": 9.632037995076639e-06, "loss": 0.0758, "step": 3219 }, { "epoch": 0.1496282527881041, "grad_norm": 1.9184661923026165, "learning_rate": 9.631754604649186e-06, "loss": 0.0906, "step": 3220 }, { "epoch": 0.14967472118959108, "grad_norm": 2.0442968629123266, "learning_rate": 9.631471109307615e-06, "loss": 0.0988, "step": 3221 }, { "epoch": 0.14972118959107808, "grad_norm": 2.905185031572473, "learning_rate": 9.63118750905835e-06, "loss": 0.139, "step": 3222 }, { "epoch": 0.14976765799256506, "grad_norm": 1.83263266793738, "learning_rate": 9.630903803907813e-06, "loss": 0.1545, "step": 3223 }, { "epoch": 0.14981412639405203, "grad_norm": 5.877124860073619, "learning_rate": 9.630619993862432e-06, "loss": 0.1281, "step": 3224 }, { "epoch": 0.14986059479553904, "grad_norm": 0.8146148846247796, "learning_rate": 9.630336078928632e-06, "loss": 0.0375, "step": 3225 }, { "epoch": 0.149907063197026, "grad_norm": 2.2450391429087677, "learning_rate": 9.630052059112849e-06, "loss": 0.073, "step": 3226 }, { "epoch": 0.14995353159851302, "grad_norm": 4.469787522033551, "learning_rate": 9.629767934421513e-06, "loss": 0.1668, "step": 3227 }, { "epoch": 0.15, "grad_norm": 23.025746931585964, "learning_rate": 9.62948370486106e-06, "loss": 0.0648, "step": 3228 }, { "epoch": 0.150046468401487, "grad_norm": 2.1296739136336305, "learning_rate": 9.629199370437928e-06, "loss": 0.1177, "step": 3229 }, { "epoch": 0.15009293680297398, "grad_norm": 4.691109204428548, "learning_rate": 9.628914931158559e-06, "loss": 0.1269, "step": 3230 }, { "epoch": 0.15013940520446098, "grad_norm": 1.1948680816438169, "learning_rate": 9.628630387029396e-06, "loss": 0.0604, "step": 3231 }, { "epoch": 0.15018587360594796, "grad_norm": 1.7887719047303683, "learning_rate": 9.628345738056882e-06, "loss": 0.0804, "step": 3232 }, { "epoch": 0.15023234200743493, "grad_norm": 1.2903699939483895, "learning_rate": 9.628060984247466e-06, "loss": 0.0984, "step": 3233 }, { "epoch": 0.15027881040892194, "grad_norm": 5.031818379102015, "learning_rate": 9.627776125607597e-06, "loss": 0.1394, "step": 3234 }, { "epoch": 0.1503252788104089, "grad_norm": 1.287062579806243, "learning_rate": 9.62749116214373e-06, "loss": 0.08, "step": 3235 }, { "epoch": 0.15037174721189592, "grad_norm": 1.1290867653137888, "learning_rate": 9.627206093862316e-06, "loss": 0.0577, "step": 3236 }, { "epoch": 0.1504182156133829, "grad_norm": 4.559262482675835, "learning_rate": 9.626920920769815e-06, "loss": 0.1221, "step": 3237 }, { "epoch": 0.1504646840148699, "grad_norm": 2.416451815602755, "learning_rate": 9.626635642872685e-06, "loss": 0.0878, "step": 3238 }, { "epoch": 0.15051115241635687, "grad_norm": 8.32173698691152, "learning_rate": 9.626350260177388e-06, "loss": 0.0887, "step": 3239 }, { "epoch": 0.15055762081784388, "grad_norm": 5.166928710958925, "learning_rate": 9.62606477269039e-06, "loss": 0.0782, "step": 3240 }, { "epoch": 0.15060408921933086, "grad_norm": 3.006173508625177, "learning_rate": 9.625779180418153e-06, "loss": 0.1208, "step": 3241 }, { "epoch": 0.15065055762081783, "grad_norm": 1.8228936525448147, "learning_rate": 9.625493483367151e-06, "loss": 0.1065, "step": 3242 }, { "epoch": 0.15069702602230484, "grad_norm": 1.201925042823866, "learning_rate": 9.625207681543853e-06, "loss": 0.0528, "step": 3243 }, { "epoch": 0.1507434944237918, "grad_norm": 1.9061626558771154, "learning_rate": 9.624921774954734e-06, "loss": 0.0956, "step": 3244 }, { "epoch": 0.15078996282527882, "grad_norm": 2.0015891659664913, "learning_rate": 9.624635763606268e-06, "loss": 0.083, "step": 3245 }, { "epoch": 0.1508364312267658, "grad_norm": 2.263505763856773, "learning_rate": 9.624349647504936e-06, "loss": 0.1397, "step": 3246 }, { "epoch": 0.1508828996282528, "grad_norm": 4.863786657521326, "learning_rate": 9.624063426657216e-06, "loss": 0.1017, "step": 3247 }, { "epoch": 0.15092936802973977, "grad_norm": 1.282957557088409, "learning_rate": 9.623777101069594e-06, "loss": 0.1011, "step": 3248 }, { "epoch": 0.15097583643122678, "grad_norm": 1.7886243852533918, "learning_rate": 9.623490670748551e-06, "loss": 0.1035, "step": 3249 }, { "epoch": 0.15102230483271376, "grad_norm": 2.886397764535757, "learning_rate": 9.623204135700583e-06, "loss": 0.1041, "step": 3250 }, { "epoch": 0.15106877323420073, "grad_norm": 1.503432571634903, "learning_rate": 9.622917495932172e-06, "loss": 0.0866, "step": 3251 }, { "epoch": 0.15111524163568774, "grad_norm": 1.9330243329925547, "learning_rate": 9.622630751449815e-06, "loss": 0.0939, "step": 3252 }, { "epoch": 0.1511617100371747, "grad_norm": 52.77744259556171, "learning_rate": 9.622343902260007e-06, "loss": 0.0551, "step": 3253 }, { "epoch": 0.15120817843866172, "grad_norm": 2.3054355116091187, "learning_rate": 9.622056948369244e-06, "loss": 0.1145, "step": 3254 }, { "epoch": 0.1512546468401487, "grad_norm": 0.9759288797484473, "learning_rate": 9.621769889784027e-06, "loss": 0.0635, "step": 3255 }, { "epoch": 0.1513011152416357, "grad_norm": 1.6345749458281513, "learning_rate": 9.621482726510857e-06, "loss": 0.1064, "step": 3256 }, { "epoch": 0.15134758364312267, "grad_norm": 2.7536828632043457, "learning_rate": 9.62119545855624e-06, "loss": 0.1167, "step": 3257 }, { "epoch": 0.15139405204460968, "grad_norm": 3.1000719348817447, "learning_rate": 9.620908085926681e-06, "loss": 0.0918, "step": 3258 }, { "epoch": 0.15144052044609665, "grad_norm": 1.9597838288918232, "learning_rate": 9.62062060862869e-06, "loss": 0.0824, "step": 3259 }, { "epoch": 0.15148698884758363, "grad_norm": 1.3907411650098642, "learning_rate": 9.620333026668782e-06, "loss": 0.0746, "step": 3260 }, { "epoch": 0.15153345724907064, "grad_norm": 1.7095619412869971, "learning_rate": 9.620045340053465e-06, "loss": 0.1276, "step": 3261 }, { "epoch": 0.1515799256505576, "grad_norm": 1.8043604338679284, "learning_rate": 9.619757548789261e-06, "loss": 0.1439, "step": 3262 }, { "epoch": 0.15162639405204462, "grad_norm": 3.1398880431934257, "learning_rate": 9.619469652882684e-06, "loss": 0.1247, "step": 3263 }, { "epoch": 0.1516728624535316, "grad_norm": 4.35680135531049, "learning_rate": 9.61918165234026e-06, "loss": 0.1212, "step": 3264 }, { "epoch": 0.1517193308550186, "grad_norm": 2.264555169908362, "learning_rate": 9.618893547168508e-06, "loss": 0.0908, "step": 3265 }, { "epoch": 0.15176579925650557, "grad_norm": 5.680987936369406, "learning_rate": 9.618605337373956e-06, "loss": 0.2743, "step": 3266 }, { "epoch": 0.15181226765799258, "grad_norm": 1.598582612656462, "learning_rate": 9.618317022963133e-06, "loss": 0.2071, "step": 3267 }, { "epoch": 0.15185873605947955, "grad_norm": 1.4238447740745956, "learning_rate": 9.618028603942569e-06, "loss": 0.0844, "step": 3268 }, { "epoch": 0.15190520446096653, "grad_norm": 1.733840547466856, "learning_rate": 9.617740080318795e-06, "loss": 0.0951, "step": 3269 }, { "epoch": 0.15195167286245354, "grad_norm": 3.0392110891459363, "learning_rate": 9.61745145209835e-06, "loss": 0.1189, "step": 3270 }, { "epoch": 0.1519981412639405, "grad_norm": 1.7010399468419901, "learning_rate": 9.61716271928777e-06, "loss": 0.0911, "step": 3271 }, { "epoch": 0.15204460966542752, "grad_norm": 1.5069462839714713, "learning_rate": 9.616873881893593e-06, "loss": 0.0716, "step": 3272 }, { "epoch": 0.1520910780669145, "grad_norm": 1.7524062471492714, "learning_rate": 9.616584939922365e-06, "loss": 0.1089, "step": 3273 }, { "epoch": 0.1521375464684015, "grad_norm": 2.130469712890724, "learning_rate": 9.616295893380628e-06, "loss": 0.0864, "step": 3274 }, { "epoch": 0.15218401486988847, "grad_norm": 1.9569512968690543, "learning_rate": 9.61600674227493e-06, "loss": 0.1165, "step": 3275 }, { "epoch": 0.15223048327137548, "grad_norm": 4.7521911108977815, "learning_rate": 9.615717486611822e-06, "loss": 0.1277, "step": 3276 }, { "epoch": 0.15227695167286245, "grad_norm": 2.1162324304083397, "learning_rate": 9.615428126397856e-06, "loss": 0.1184, "step": 3277 }, { "epoch": 0.15232342007434943, "grad_norm": 2.0105649663435368, "learning_rate": 9.615138661639584e-06, "loss": 0.0795, "step": 3278 }, { "epoch": 0.15236988847583643, "grad_norm": 1.828001722199491, "learning_rate": 9.614849092343565e-06, "loss": 0.1056, "step": 3279 }, { "epoch": 0.1524163568773234, "grad_norm": 1.2806345533875727, "learning_rate": 9.614559418516356e-06, "loss": 0.0857, "step": 3280 }, { "epoch": 0.15246282527881042, "grad_norm": 3.167248716972215, "learning_rate": 9.61426964016452e-06, "loss": 0.079, "step": 3281 }, { "epoch": 0.1525092936802974, "grad_norm": 1.057025693563354, "learning_rate": 9.61397975729462e-06, "loss": 0.0856, "step": 3282 }, { "epoch": 0.1525557620817844, "grad_norm": 2.9883763800140484, "learning_rate": 9.613689769913222e-06, "loss": 0.1437, "step": 3283 }, { "epoch": 0.15260223048327137, "grad_norm": 2.64102234291588, "learning_rate": 9.613399678026895e-06, "loss": 0.1685, "step": 3284 }, { "epoch": 0.15264869888475838, "grad_norm": 1.8015664055987926, "learning_rate": 9.61310948164221e-06, "loss": 0.1363, "step": 3285 }, { "epoch": 0.15269516728624535, "grad_norm": 1.9816431429260821, "learning_rate": 9.612819180765741e-06, "loss": 0.1336, "step": 3286 }, { "epoch": 0.15274163568773233, "grad_norm": 2.609536936051634, "learning_rate": 9.61252877540406e-06, "loss": 0.0935, "step": 3287 }, { "epoch": 0.15278810408921933, "grad_norm": 1.2586461707726022, "learning_rate": 9.61223826556375e-06, "loss": 0.0606, "step": 3288 }, { "epoch": 0.1528345724907063, "grad_norm": 1.1682212172670923, "learning_rate": 9.611947651251387e-06, "loss": 0.0602, "step": 3289 }, { "epoch": 0.15288104089219332, "grad_norm": 1.7325735834179248, "learning_rate": 9.611656932473558e-06, "loss": 0.0906, "step": 3290 }, { "epoch": 0.1529275092936803, "grad_norm": 1.6370693725230272, "learning_rate": 9.611366109236845e-06, "loss": 0.1449, "step": 3291 }, { "epoch": 0.1529739776951673, "grad_norm": 1.813080031003281, "learning_rate": 9.611075181547836e-06, "loss": 0.1126, "step": 3292 }, { "epoch": 0.15302044609665427, "grad_norm": 2.025427807006145, "learning_rate": 9.610784149413121e-06, "loss": 0.0742, "step": 3293 }, { "epoch": 0.15306691449814128, "grad_norm": 2.8678673104967465, "learning_rate": 9.610493012839293e-06, "loss": 0.0905, "step": 3294 }, { "epoch": 0.15311338289962825, "grad_norm": 3.0939002582817903, "learning_rate": 9.610201771832945e-06, "loss": 0.158, "step": 3295 }, { "epoch": 0.15315985130111523, "grad_norm": 1.5954525123819154, "learning_rate": 9.609910426400675e-06, "loss": 0.0955, "step": 3296 }, { "epoch": 0.15320631970260223, "grad_norm": 1.6679691109520958, "learning_rate": 9.609618976549082e-06, "loss": 0.1135, "step": 3297 }, { "epoch": 0.1532527881040892, "grad_norm": 2.464148168490007, "learning_rate": 9.609327422284766e-06, "loss": 0.0647, "step": 3298 }, { "epoch": 0.15329925650557621, "grad_norm": 1.4194280264450638, "learning_rate": 9.609035763614335e-06, "loss": 0.1208, "step": 3299 }, { "epoch": 0.1533457249070632, "grad_norm": 1.4604001906457773, "learning_rate": 9.608744000544392e-06, "loss": 0.1033, "step": 3300 }, { "epoch": 0.1533921933085502, "grad_norm": 1.235431464369887, "learning_rate": 9.608452133081548e-06, "loss": 0.0729, "step": 3301 }, { "epoch": 0.15343866171003717, "grad_norm": 1.8948100204104195, "learning_rate": 9.608160161232411e-06, "loss": 0.1095, "step": 3302 }, { "epoch": 0.15348513011152418, "grad_norm": 1.894567322897057, "learning_rate": 9.607868085003597e-06, "loss": 0.1046, "step": 3303 }, { "epoch": 0.15353159851301115, "grad_norm": 2.2742121883784687, "learning_rate": 9.607575904401722e-06, "loss": 0.101, "step": 3304 }, { "epoch": 0.15357806691449813, "grad_norm": 1.4675351826954102, "learning_rate": 9.607283619433402e-06, "loss": 0.1001, "step": 3305 }, { "epoch": 0.15362453531598513, "grad_norm": 2.3365816628062506, "learning_rate": 9.60699123010526e-06, "loss": 0.113, "step": 3306 }, { "epoch": 0.1536710037174721, "grad_norm": 2.888403521433208, "learning_rate": 9.60669873642392e-06, "loss": 0.0924, "step": 3307 }, { "epoch": 0.15371747211895911, "grad_norm": 2.0296743936234227, "learning_rate": 9.606406138396e-06, "loss": 0.1199, "step": 3308 }, { "epoch": 0.1537639405204461, "grad_norm": 0.8963634056069044, "learning_rate": 9.606113436028138e-06, "loss": 0.0783, "step": 3309 }, { "epoch": 0.1538104089219331, "grad_norm": 1.542015417081811, "learning_rate": 9.605820629326955e-06, "loss": 0.1061, "step": 3310 }, { "epoch": 0.15385687732342007, "grad_norm": 1.113855709956564, "learning_rate": 9.605527718299088e-06, "loss": 0.0577, "step": 3311 }, { "epoch": 0.15390334572490708, "grad_norm": 1.8816910399522855, "learning_rate": 9.605234702951173e-06, "loss": 0.0964, "step": 3312 }, { "epoch": 0.15394981412639405, "grad_norm": 2.8069798147916636, "learning_rate": 9.604941583289843e-06, "loss": 0.0942, "step": 3313 }, { "epoch": 0.15399628252788103, "grad_norm": 1.6144688090363017, "learning_rate": 9.60464835932174e-06, "loss": 0.1273, "step": 3314 }, { "epoch": 0.15404275092936803, "grad_norm": 1.997560880296864, "learning_rate": 9.604355031053504e-06, "loss": 0.0732, "step": 3315 }, { "epoch": 0.154089219330855, "grad_norm": 2.696759716824837, "learning_rate": 9.604061598491785e-06, "loss": 0.0965, "step": 3316 }, { "epoch": 0.154135687732342, "grad_norm": 1.3578312560761967, "learning_rate": 9.603768061643221e-06, "loss": 0.0715, "step": 3317 }, { "epoch": 0.154182156133829, "grad_norm": 2.223989668572733, "learning_rate": 9.603474420514466e-06, "loss": 0.0914, "step": 3318 }, { "epoch": 0.154228624535316, "grad_norm": 0.8765772313604272, "learning_rate": 9.603180675112171e-06, "loss": 0.0428, "step": 3319 }, { "epoch": 0.15427509293680297, "grad_norm": 2.2944346083447678, "learning_rate": 9.602886825442988e-06, "loss": 0.082, "step": 3320 }, { "epoch": 0.15432156133828998, "grad_norm": 2.9140986409787297, "learning_rate": 9.602592871513573e-06, "loss": 0.1512, "step": 3321 }, { "epoch": 0.15436802973977695, "grad_norm": 2.1973268555648797, "learning_rate": 9.602298813330588e-06, "loss": 0.1267, "step": 3322 }, { "epoch": 0.15441449814126393, "grad_norm": 1.4199247008777567, "learning_rate": 9.602004650900691e-06, "loss": 0.1096, "step": 3323 }, { "epoch": 0.15446096654275093, "grad_norm": 3.337238542332666, "learning_rate": 9.601710384230542e-06, "loss": 0.0917, "step": 3324 }, { "epoch": 0.1545074349442379, "grad_norm": 1.4067014002714322, "learning_rate": 9.601416013326811e-06, "loss": 0.062, "step": 3325 }, { "epoch": 0.1545539033457249, "grad_norm": 2.1121938718729045, "learning_rate": 9.601121538196167e-06, "loss": 0.0767, "step": 3326 }, { "epoch": 0.1546003717472119, "grad_norm": 2.4876213361806117, "learning_rate": 9.600826958845275e-06, "loss": 0.1102, "step": 3327 }, { "epoch": 0.1546468401486989, "grad_norm": 1.5334827852406863, "learning_rate": 9.60053227528081e-06, "loss": 0.0825, "step": 3328 }, { "epoch": 0.15469330855018587, "grad_norm": 4.988467242980157, "learning_rate": 9.600237487509448e-06, "loss": 0.1512, "step": 3329 }, { "epoch": 0.15473977695167287, "grad_norm": 1.0500396598197486, "learning_rate": 9.599942595537865e-06, "loss": 0.0784, "step": 3330 }, { "epoch": 0.15478624535315985, "grad_norm": 1.444531237256978, "learning_rate": 9.599647599372742e-06, "loss": 0.0869, "step": 3331 }, { "epoch": 0.15483271375464683, "grad_norm": 1.4036764419057286, "learning_rate": 9.599352499020757e-06, "loss": 0.1315, "step": 3332 }, { "epoch": 0.15487918215613383, "grad_norm": 0.8830325374062288, "learning_rate": 9.599057294488599e-06, "loss": 0.0616, "step": 3333 }, { "epoch": 0.1549256505576208, "grad_norm": 1.9588166255480262, "learning_rate": 9.598761985782951e-06, "loss": 0.1064, "step": 3334 }, { "epoch": 0.1549721189591078, "grad_norm": 2.2912130496218386, "learning_rate": 9.598466572910507e-06, "loss": 0.1229, "step": 3335 }, { "epoch": 0.1550185873605948, "grad_norm": 2.468327449390649, "learning_rate": 9.598171055877953e-06, "loss": 0.1532, "step": 3336 }, { "epoch": 0.1550650557620818, "grad_norm": 1.3132762283002009, "learning_rate": 9.597875434691986e-06, "loss": 0.0783, "step": 3337 }, { "epoch": 0.15511152416356877, "grad_norm": 0.958670766919676, "learning_rate": 9.597579709359302e-06, "loss": 0.074, "step": 3338 }, { "epoch": 0.15515799256505577, "grad_norm": 3.9058180166012106, "learning_rate": 9.597283879886596e-06, "loss": 0.1233, "step": 3339 }, { "epoch": 0.15520446096654275, "grad_norm": 1.6143936697079218, "learning_rate": 9.596987946280574e-06, "loss": 0.1399, "step": 3340 }, { "epoch": 0.15525092936802973, "grad_norm": 1.6180057058696635, "learning_rate": 9.596691908547937e-06, "loss": 0.0904, "step": 3341 }, { "epoch": 0.15529739776951673, "grad_norm": 1.2621305396777052, "learning_rate": 9.596395766695389e-06, "loss": 0.0842, "step": 3342 }, { "epoch": 0.1553438661710037, "grad_norm": 1.822269814597055, "learning_rate": 9.59609952072964e-06, "loss": 0.1283, "step": 3343 }, { "epoch": 0.1553903345724907, "grad_norm": 1.597203091917046, "learning_rate": 9.595803170657398e-06, "loss": 0.1004, "step": 3344 }, { "epoch": 0.1554368029739777, "grad_norm": 1.5567953029691144, "learning_rate": 9.59550671648538e-06, "loss": 0.0654, "step": 3345 }, { "epoch": 0.1554832713754647, "grad_norm": 2.00354492719125, "learning_rate": 9.595210158220296e-06, "loss": 0.0811, "step": 3346 }, { "epoch": 0.15552973977695167, "grad_norm": 0.8886831851655702, "learning_rate": 9.594913495868867e-06, "loss": 0.106, "step": 3347 }, { "epoch": 0.15557620817843867, "grad_norm": 1.737896004426921, "learning_rate": 9.59461672943781e-06, "loss": 0.1128, "step": 3348 }, { "epoch": 0.15562267657992565, "grad_norm": 1.0608347868034886, "learning_rate": 9.594319858933847e-06, "loss": 0.1063, "step": 3349 }, { "epoch": 0.15566914498141263, "grad_norm": 2.22442601696837, "learning_rate": 9.594022884363707e-06, "loss": 0.1263, "step": 3350 }, { "epoch": 0.15571561338289963, "grad_norm": 1.6565364647040026, "learning_rate": 9.593725805734112e-06, "loss": 0.102, "step": 3351 }, { "epoch": 0.1557620817843866, "grad_norm": 1.5759131025838642, "learning_rate": 9.593428623051793e-06, "loss": 0.0805, "step": 3352 }, { "epoch": 0.1558085501858736, "grad_norm": 1.7689677595334148, "learning_rate": 9.593131336323481e-06, "loss": 0.1199, "step": 3353 }, { "epoch": 0.1558550185873606, "grad_norm": 1.4932119851725894, "learning_rate": 9.59283394555591e-06, "loss": 0.0971, "step": 3354 }, { "epoch": 0.1559014869888476, "grad_norm": 4.029425835916088, "learning_rate": 9.592536450755817e-06, "loss": 0.0814, "step": 3355 }, { "epoch": 0.15594795539033457, "grad_norm": 2.0791013906328395, "learning_rate": 9.592238851929938e-06, "loss": 0.1057, "step": 3356 }, { "epoch": 0.15599442379182157, "grad_norm": 2.382787118079704, "learning_rate": 9.591941149085018e-06, "loss": 0.0694, "step": 3357 }, { "epoch": 0.15604089219330855, "grad_norm": 1.4664581708091866, "learning_rate": 9.591643342227796e-06, "loss": 0.0838, "step": 3358 }, { "epoch": 0.15608736059479553, "grad_norm": 1.9632687903314252, "learning_rate": 9.59134543136502e-06, "loss": 0.0825, "step": 3359 }, { "epoch": 0.15613382899628253, "grad_norm": 2.3338321952136702, "learning_rate": 9.59104741650344e-06, "loss": 0.1532, "step": 3360 }, { "epoch": 0.1561802973977695, "grad_norm": 3.826525464891822, "learning_rate": 9.590749297649802e-06, "loss": 0.1125, "step": 3361 }, { "epoch": 0.1562267657992565, "grad_norm": 2.1886142729240254, "learning_rate": 9.59045107481086e-06, "loss": 0.0834, "step": 3362 }, { "epoch": 0.1562732342007435, "grad_norm": 1.4232995528614352, "learning_rate": 9.590152747993371e-06, "loss": 0.1593, "step": 3363 }, { "epoch": 0.1563197026022305, "grad_norm": 1.8913812501176077, "learning_rate": 9.589854317204091e-06, "loss": 0.1271, "step": 3364 }, { "epoch": 0.15636617100371747, "grad_norm": 0.9315409736869987, "learning_rate": 9.58955578244978e-06, "loss": 0.0469, "step": 3365 }, { "epoch": 0.15641263940520447, "grad_norm": 1.7818598088747455, "learning_rate": 9.5892571437372e-06, "loss": 0.0887, "step": 3366 }, { "epoch": 0.15645910780669145, "grad_norm": 1.3780142878728556, "learning_rate": 9.588958401073117e-06, "loss": 0.1037, "step": 3367 }, { "epoch": 0.15650557620817843, "grad_norm": 1.4152777456666628, "learning_rate": 9.588659554464294e-06, "loss": 0.1012, "step": 3368 }, { "epoch": 0.15655204460966543, "grad_norm": 1.825014735407179, "learning_rate": 9.588360603917505e-06, "loss": 0.1159, "step": 3369 }, { "epoch": 0.1565985130111524, "grad_norm": 3.9700583893178245, "learning_rate": 9.588061549439518e-06, "loss": 0.0783, "step": 3370 }, { "epoch": 0.1566449814126394, "grad_norm": 1.5127304468613192, "learning_rate": 9.587762391037108e-06, "loss": 0.1021, "step": 3371 }, { "epoch": 0.1566914498141264, "grad_norm": 2.607636153955958, "learning_rate": 9.58746312871705e-06, "loss": 0.1196, "step": 3372 }, { "epoch": 0.1567379182156134, "grad_norm": 2.0277733340935638, "learning_rate": 9.587163762486126e-06, "loss": 0.0726, "step": 3373 }, { "epoch": 0.15678438661710037, "grad_norm": 1.1747189220737808, "learning_rate": 9.586864292351114e-06, "loss": 0.076, "step": 3374 }, { "epoch": 0.15683085501858737, "grad_norm": 1.2615266709962636, "learning_rate": 9.586564718318799e-06, "loss": 0.1112, "step": 3375 }, { "epoch": 0.15687732342007435, "grad_norm": 2.245511276066009, "learning_rate": 9.586265040395964e-06, "loss": 0.0869, "step": 3376 }, { "epoch": 0.15692379182156133, "grad_norm": 2.0475004272012134, "learning_rate": 9.5859652585894e-06, "loss": 0.1427, "step": 3377 }, { "epoch": 0.15697026022304833, "grad_norm": 14.174592628323689, "learning_rate": 9.585665372905897e-06, "loss": 0.0848, "step": 3378 }, { "epoch": 0.1570167286245353, "grad_norm": 1.9389614798637975, "learning_rate": 9.585365383352245e-06, "loss": 0.0693, "step": 3379 }, { "epoch": 0.1570631970260223, "grad_norm": 1.6834330119890006, "learning_rate": 9.585065289935243e-06, "loss": 0.0672, "step": 3380 }, { "epoch": 0.1571096654275093, "grad_norm": 1.7811836936289311, "learning_rate": 9.584765092661687e-06, "loss": 0.0877, "step": 3381 }, { "epoch": 0.1571561338289963, "grad_norm": 1.4122629233738027, "learning_rate": 9.584464791538375e-06, "loss": 0.1277, "step": 3382 }, { "epoch": 0.15720260223048327, "grad_norm": 23.64594218078356, "learning_rate": 9.58416438657211e-06, "loss": 0.1561, "step": 3383 }, { "epoch": 0.15724907063197027, "grad_norm": 1.8371728966751808, "learning_rate": 9.583863877769696e-06, "loss": 0.0944, "step": 3384 }, { "epoch": 0.15729553903345725, "grad_norm": 1.7759998775262864, "learning_rate": 9.583563265137943e-06, "loss": 0.1241, "step": 3385 }, { "epoch": 0.15734200743494423, "grad_norm": 1.5207349790308242, "learning_rate": 9.583262548683657e-06, "loss": 0.0793, "step": 3386 }, { "epoch": 0.15738847583643123, "grad_norm": 1.4514570977853622, "learning_rate": 9.582961728413652e-06, "loss": 0.0717, "step": 3387 }, { "epoch": 0.1574349442379182, "grad_norm": 2.9163642525720594, "learning_rate": 9.582660804334738e-06, "loss": 0.2002, "step": 3388 }, { "epoch": 0.1574814126394052, "grad_norm": 2.21288950413991, "learning_rate": 9.582359776453733e-06, "loss": 0.1355, "step": 3389 }, { "epoch": 0.1575278810408922, "grad_norm": 1.5108126358585212, "learning_rate": 9.582058644777459e-06, "loss": 0.0982, "step": 3390 }, { "epoch": 0.1575743494423792, "grad_norm": 2.235150344551734, "learning_rate": 9.581757409312733e-06, "loss": 0.121, "step": 3391 }, { "epoch": 0.15762081784386617, "grad_norm": 2.0737108250123297, "learning_rate": 9.58145607006638e-06, "loss": 0.0826, "step": 3392 }, { "epoch": 0.15766728624535317, "grad_norm": 1.562389229780867, "learning_rate": 9.581154627045224e-06, "loss": 0.1364, "step": 3393 }, { "epoch": 0.15771375464684015, "grad_norm": 4.062180470303124, "learning_rate": 9.580853080256094e-06, "loss": 0.1129, "step": 3394 }, { "epoch": 0.15776022304832713, "grad_norm": 3.10576819640901, "learning_rate": 9.580551429705822e-06, "loss": 0.114, "step": 3395 }, { "epoch": 0.15780669144981413, "grad_norm": 1.5986225550028654, "learning_rate": 9.580249675401239e-06, "loss": 0.0773, "step": 3396 }, { "epoch": 0.1578531598513011, "grad_norm": 1.4872158649383413, "learning_rate": 9.579947817349177e-06, "loss": 0.0476, "step": 3397 }, { "epoch": 0.1578996282527881, "grad_norm": 1.3013298500529942, "learning_rate": 9.579645855556481e-06, "loss": 0.0671, "step": 3398 }, { "epoch": 0.1579460966542751, "grad_norm": 4.110907132692198, "learning_rate": 9.579343790029984e-06, "loss": 0.1195, "step": 3399 }, { "epoch": 0.1579925650557621, "grad_norm": 2.265877843979651, "learning_rate": 9.57904162077653e-06, "loss": 0.1191, "step": 3400 }, { "epoch": 0.15803903345724907, "grad_norm": 1.5620124172028458, "learning_rate": 9.578739347802965e-06, "loss": 0.0699, "step": 3401 }, { "epoch": 0.15808550185873607, "grad_norm": 13.68251288933193, "learning_rate": 9.578436971116137e-06, "loss": 0.1015, "step": 3402 }, { "epoch": 0.15813197026022305, "grad_norm": 1.7847548707974505, "learning_rate": 9.578134490722889e-06, "loss": 0.0971, "step": 3403 }, { "epoch": 0.15817843866171002, "grad_norm": 1.351410882152051, "learning_rate": 9.577831906630077e-06, "loss": 0.0885, "step": 3404 }, { "epoch": 0.15822490706319703, "grad_norm": 2.4796324999164536, "learning_rate": 9.577529218844557e-06, "loss": 0.1196, "step": 3405 }, { "epoch": 0.158271375464684, "grad_norm": 2.2424349184805856, "learning_rate": 9.577226427373179e-06, "loss": 0.1135, "step": 3406 }, { "epoch": 0.158317843866171, "grad_norm": 1.5746281708847851, "learning_rate": 9.576923532222808e-06, "loss": 0.1154, "step": 3407 }, { "epoch": 0.158364312267658, "grad_norm": 1.8759818769073817, "learning_rate": 9.5766205334003e-06, "loss": 0.1023, "step": 3408 }, { "epoch": 0.158410780669145, "grad_norm": 7.133567538387627, "learning_rate": 9.576317430912521e-06, "loss": 0.1014, "step": 3409 }, { "epoch": 0.15845724907063197, "grad_norm": 1.295288865748648, "learning_rate": 9.576014224766335e-06, "loss": 0.0813, "step": 3410 }, { "epoch": 0.15850371747211897, "grad_norm": 1.721118336963438, "learning_rate": 9.57571091496861e-06, "loss": 0.0916, "step": 3411 }, { "epoch": 0.15855018587360595, "grad_norm": 1.5118428841995364, "learning_rate": 9.575407501526218e-06, "loss": 0.0991, "step": 3412 }, { "epoch": 0.15859665427509292, "grad_norm": 1.3943718070362057, "learning_rate": 9.57510398444603e-06, "loss": 0.0875, "step": 3413 }, { "epoch": 0.15864312267657993, "grad_norm": 1.6015747532812863, "learning_rate": 9.574800363734921e-06, "loss": 0.0956, "step": 3414 }, { "epoch": 0.1586895910780669, "grad_norm": 1.9117679746795917, "learning_rate": 9.574496639399771e-06, "loss": 0.1004, "step": 3415 }, { "epoch": 0.1587360594795539, "grad_norm": 1.7927504739241218, "learning_rate": 9.574192811447457e-06, "loss": 0.1525, "step": 3416 }, { "epoch": 0.15878252788104089, "grad_norm": 1.9622373163079347, "learning_rate": 9.573888879884861e-06, "loss": 0.0926, "step": 3417 }, { "epoch": 0.1588289962825279, "grad_norm": 1.5473543850993974, "learning_rate": 9.573584844718869e-06, "loss": 0.0774, "step": 3418 }, { "epoch": 0.15887546468401487, "grad_norm": 3.2879060309726675, "learning_rate": 9.573280705956365e-06, "loss": 0.128, "step": 3419 }, { "epoch": 0.15892193308550187, "grad_norm": 1.3398670408395923, "learning_rate": 9.57297646360424e-06, "loss": 0.1224, "step": 3420 }, { "epoch": 0.15896840148698885, "grad_norm": 2.742659279325866, "learning_rate": 9.572672117669384e-06, "loss": 0.1027, "step": 3421 }, { "epoch": 0.15901486988847582, "grad_norm": 2.4405581392137705, "learning_rate": 9.572367668158694e-06, "loss": 0.1017, "step": 3422 }, { "epoch": 0.15906133828996283, "grad_norm": 1.591079683858807, "learning_rate": 9.572063115079063e-06, "loss": 0.0932, "step": 3423 }, { "epoch": 0.1591078066914498, "grad_norm": 2.386228942032387, "learning_rate": 9.57175845843739e-06, "loss": 0.0822, "step": 3424 }, { "epoch": 0.1591542750929368, "grad_norm": 1.4342882125643697, "learning_rate": 9.571453698240578e-06, "loss": 0.0758, "step": 3425 }, { "epoch": 0.15920074349442379, "grad_norm": 1.3190264404767749, "learning_rate": 9.571148834495527e-06, "loss": 0.0837, "step": 3426 }, { "epoch": 0.1592472118959108, "grad_norm": 1.9111463170850524, "learning_rate": 9.570843867209144e-06, "loss": 0.0804, "step": 3427 }, { "epoch": 0.15929368029739777, "grad_norm": 0.9190668161737571, "learning_rate": 9.570538796388337e-06, "loss": 0.0549, "step": 3428 }, { "epoch": 0.15934014869888477, "grad_norm": 0.9591199726638346, "learning_rate": 9.570233622040014e-06, "loss": 0.059, "step": 3429 }, { "epoch": 0.15938661710037175, "grad_norm": 0.9466715124921281, "learning_rate": 9.569928344171094e-06, "loss": 0.0518, "step": 3430 }, { "epoch": 0.15943308550185872, "grad_norm": 2.3651588527730034, "learning_rate": 9.569622962788483e-06, "loss": 0.1299, "step": 3431 }, { "epoch": 0.15947955390334573, "grad_norm": 1.876502976381504, "learning_rate": 9.569317477899104e-06, "loss": 0.0805, "step": 3432 }, { "epoch": 0.1595260223048327, "grad_norm": 1.6173275217588183, "learning_rate": 9.569011889509874e-06, "loss": 0.0798, "step": 3433 }, { "epoch": 0.1595724907063197, "grad_norm": 1.7328003616436385, "learning_rate": 9.568706197627717e-06, "loss": 0.071, "step": 3434 }, { "epoch": 0.15961895910780668, "grad_norm": 3.366102106791454, "learning_rate": 9.568400402259555e-06, "loss": 0.1173, "step": 3435 }, { "epoch": 0.1596654275092937, "grad_norm": 8.092008726720024, "learning_rate": 9.568094503412317e-06, "loss": 0.0703, "step": 3436 }, { "epoch": 0.15971189591078067, "grad_norm": 1.9508044217653326, "learning_rate": 9.56778850109293e-06, "loss": 0.1137, "step": 3437 }, { "epoch": 0.15975836431226767, "grad_norm": 1.6158975928139288, "learning_rate": 9.567482395308327e-06, "loss": 0.0946, "step": 3438 }, { "epoch": 0.15980483271375465, "grad_norm": 1.4517678013460087, "learning_rate": 9.567176186065439e-06, "loss": 0.0725, "step": 3439 }, { "epoch": 0.15985130111524162, "grad_norm": 2.0746196379168356, "learning_rate": 9.566869873371205e-06, "loss": 0.0975, "step": 3440 }, { "epoch": 0.15989776951672863, "grad_norm": 1.7045763262924405, "learning_rate": 9.56656345723256e-06, "loss": 0.0885, "step": 3441 }, { "epoch": 0.1599442379182156, "grad_norm": 3.5718876910871895, "learning_rate": 9.566256937656448e-06, "loss": 0.1779, "step": 3442 }, { "epoch": 0.1599907063197026, "grad_norm": 3.203331198762576, "learning_rate": 9.56595031464981e-06, "loss": 0.0733, "step": 3443 }, { "epoch": 0.16003717472118958, "grad_norm": 1.799771440138235, "learning_rate": 9.565643588219592e-06, "loss": 0.0924, "step": 3444 }, { "epoch": 0.1600836431226766, "grad_norm": 2.412380134503881, "learning_rate": 9.565336758372741e-06, "loss": 0.096, "step": 3445 }, { "epoch": 0.16013011152416357, "grad_norm": 4.944625978259024, "learning_rate": 9.565029825116207e-06, "loss": 0.1137, "step": 3446 }, { "epoch": 0.16017657992565057, "grad_norm": 2.7336251564829306, "learning_rate": 9.564722788456943e-06, "loss": 0.1017, "step": 3447 }, { "epoch": 0.16022304832713755, "grad_norm": 1.293235982608772, "learning_rate": 9.564415648401906e-06, "loss": 0.0842, "step": 3448 }, { "epoch": 0.16026951672862452, "grad_norm": 4.573105746649488, "learning_rate": 9.564108404958049e-06, "loss": 0.1354, "step": 3449 }, { "epoch": 0.16031598513011153, "grad_norm": 1.3552336882852842, "learning_rate": 9.563801058132331e-06, "loss": 0.0709, "step": 3450 }, { "epoch": 0.1603624535315985, "grad_norm": 4.801207459391128, "learning_rate": 9.563493607931718e-06, "loss": 0.0782, "step": 3451 }, { "epoch": 0.1604089219330855, "grad_norm": 2.111181732497248, "learning_rate": 9.563186054363171e-06, "loss": 0.0577, "step": 3452 }, { "epoch": 0.16045539033457248, "grad_norm": 2.671715085847613, "learning_rate": 9.562878397433659e-06, "loss": 0.1255, "step": 3453 }, { "epoch": 0.1605018587360595, "grad_norm": 1.188153039422751, "learning_rate": 9.562570637150147e-06, "loss": 0.0817, "step": 3454 }, { "epoch": 0.16054832713754646, "grad_norm": 1.398557008942073, "learning_rate": 9.562262773519608e-06, "loss": 0.0993, "step": 3455 }, { "epoch": 0.16059479553903347, "grad_norm": 2.501733413530959, "learning_rate": 9.561954806549016e-06, "loss": 0.1202, "step": 3456 }, { "epoch": 0.16064126394052045, "grad_norm": 1.9524017254992307, "learning_rate": 9.561646736245347e-06, "loss": 0.0851, "step": 3457 }, { "epoch": 0.16068773234200742, "grad_norm": 4.671219031843179, "learning_rate": 9.561338562615577e-06, "loss": 0.1321, "step": 3458 }, { "epoch": 0.16073420074349443, "grad_norm": 1.1646560795743277, "learning_rate": 9.561030285666689e-06, "loss": 0.1304, "step": 3459 }, { "epoch": 0.1607806691449814, "grad_norm": 2.713311412045806, "learning_rate": 9.560721905405663e-06, "loss": 0.1277, "step": 3460 }, { "epoch": 0.1608271375464684, "grad_norm": 1.6490223504802601, "learning_rate": 9.560413421839487e-06, "loss": 0.0995, "step": 3461 }, { "epoch": 0.16087360594795538, "grad_norm": 2.1930623139280674, "learning_rate": 9.560104834975148e-06, "loss": 0.1074, "step": 3462 }, { "epoch": 0.1609200743494424, "grad_norm": 1.0172512139684515, "learning_rate": 9.559796144819632e-06, "loss": 0.0601, "step": 3463 }, { "epoch": 0.16096654275092936, "grad_norm": 1.2423314457877053, "learning_rate": 9.559487351379938e-06, "loss": 0.1412, "step": 3464 }, { "epoch": 0.16101301115241637, "grad_norm": 1.6255304015094036, "learning_rate": 9.559178454663054e-06, "loss": 0.0981, "step": 3465 }, { "epoch": 0.16105947955390335, "grad_norm": 2.004257204332794, "learning_rate": 9.558869454675981e-06, "loss": 0.1472, "step": 3466 }, { "epoch": 0.16110594795539032, "grad_norm": 1.9245498203388631, "learning_rate": 9.558560351425717e-06, "loss": 0.1122, "step": 3467 }, { "epoch": 0.16115241635687733, "grad_norm": 1.2454678195268203, "learning_rate": 9.558251144919262e-06, "loss": 0.0782, "step": 3468 }, { "epoch": 0.1611988847583643, "grad_norm": 2.209495863665773, "learning_rate": 9.557941835163623e-06, "loss": 0.0818, "step": 3469 }, { "epoch": 0.1612453531598513, "grad_norm": 2.2909455308255926, "learning_rate": 9.5576324221658e-06, "loss": 0.07, "step": 3470 }, { "epoch": 0.16129182156133828, "grad_norm": 1.2119961878335648, "learning_rate": 9.55732290593281e-06, "loss": 0.0611, "step": 3471 }, { "epoch": 0.1613382899628253, "grad_norm": 0.8298201114968299, "learning_rate": 9.55701328647166e-06, "loss": 0.0409, "step": 3472 }, { "epoch": 0.16138475836431226, "grad_norm": 1.1582989438589897, "learning_rate": 9.55670356378936e-06, "loss": 0.0924, "step": 3473 }, { "epoch": 0.16143122676579927, "grad_norm": 1.6031314591076238, "learning_rate": 9.556393737892931e-06, "loss": 0.0813, "step": 3474 }, { "epoch": 0.16147769516728624, "grad_norm": 2.157014385303675, "learning_rate": 9.556083808789388e-06, "loss": 0.0814, "step": 3475 }, { "epoch": 0.16152416356877322, "grad_norm": 1.8632095804925828, "learning_rate": 9.555773776485749e-06, "loss": 0.081, "step": 3476 }, { "epoch": 0.16157063197026023, "grad_norm": 3.9836899565628734, "learning_rate": 9.555463640989041e-06, "loss": 0.063, "step": 3477 }, { "epoch": 0.1616171003717472, "grad_norm": 1.108387794553639, "learning_rate": 9.555153402306287e-06, "loss": 0.0902, "step": 3478 }, { "epoch": 0.1616635687732342, "grad_norm": 1.4892933844128782, "learning_rate": 9.554843060444515e-06, "loss": 0.0741, "step": 3479 }, { "epoch": 0.16171003717472118, "grad_norm": 1.1621415913138864, "learning_rate": 9.554532615410752e-06, "loss": 0.0693, "step": 3480 }, { "epoch": 0.1617565055762082, "grad_norm": 1.3995361182822472, "learning_rate": 9.554222067212031e-06, "loss": 0.0742, "step": 3481 }, { "epoch": 0.16180297397769516, "grad_norm": 1.4299874129621388, "learning_rate": 9.55391141585539e-06, "loss": 0.1129, "step": 3482 }, { "epoch": 0.16184944237918217, "grad_norm": 1.857186955393293, "learning_rate": 9.553600661347861e-06, "loss": 0.1023, "step": 3483 }, { "epoch": 0.16189591078066914, "grad_norm": 1.2609142144268355, "learning_rate": 9.553289803696484e-06, "loss": 0.0581, "step": 3484 }, { "epoch": 0.16194237918215612, "grad_norm": 2.993433107207769, "learning_rate": 9.5529788429083e-06, "loss": 0.1307, "step": 3485 }, { "epoch": 0.16198884758364313, "grad_norm": 10.181089923879021, "learning_rate": 9.552667778990354e-06, "loss": 0.1253, "step": 3486 }, { "epoch": 0.1620353159851301, "grad_norm": 1.8681042181311456, "learning_rate": 9.552356611949693e-06, "loss": 0.1021, "step": 3487 }, { "epoch": 0.1620817843866171, "grad_norm": 1.7710772986260042, "learning_rate": 9.552045341793362e-06, "loss": 0.1198, "step": 3488 }, { "epoch": 0.16212825278810408, "grad_norm": 2.9524468677502886, "learning_rate": 9.551733968528413e-06, "loss": 0.1001, "step": 3489 }, { "epoch": 0.1621747211895911, "grad_norm": 2.9913122977574944, "learning_rate": 9.551422492161899e-06, "loss": 0.0849, "step": 3490 }, { "epoch": 0.16222118959107806, "grad_norm": 1.5035731724519377, "learning_rate": 9.551110912700875e-06, "loss": 0.1946, "step": 3491 }, { "epoch": 0.16226765799256507, "grad_norm": 1.9207245064662013, "learning_rate": 9.550799230152398e-06, "loss": 0.0956, "step": 3492 }, { "epoch": 0.16231412639405204, "grad_norm": 1.3558373501274712, "learning_rate": 9.550487444523532e-06, "loss": 0.0903, "step": 3493 }, { "epoch": 0.16236059479553902, "grad_norm": 3.1943139139991614, "learning_rate": 9.550175555821333e-06, "loss": 0.2037, "step": 3494 }, { "epoch": 0.16240706319702602, "grad_norm": 1.9460166254122686, "learning_rate": 9.54986356405287e-06, "loss": 0.1114, "step": 3495 }, { "epoch": 0.162453531598513, "grad_norm": 1.8122533391522595, "learning_rate": 9.54955146922521e-06, "loss": 0.0816, "step": 3496 }, { "epoch": 0.1625, "grad_norm": 1.1092666137372291, "learning_rate": 9.549239271345419e-06, "loss": 0.0581, "step": 3497 }, { "epoch": 0.16254646840148698, "grad_norm": 2.0050178730906447, "learning_rate": 9.548926970420571e-06, "loss": 0.1072, "step": 3498 }, { "epoch": 0.162592936802974, "grad_norm": 1.1522168747368637, "learning_rate": 9.54861456645774e-06, "loss": 0.0724, "step": 3499 }, { "epoch": 0.16263940520446096, "grad_norm": 2.156825998871078, "learning_rate": 9.548302059464e-06, "loss": 0.0993, "step": 3500 }, { "epoch": 0.16268587360594797, "grad_norm": 2.3998804493696184, "learning_rate": 9.547989449446433e-06, "loss": 0.2073, "step": 3501 }, { "epoch": 0.16273234200743494, "grad_norm": 1.3860430218532074, "learning_rate": 9.547676736412118e-06, "loss": 0.0903, "step": 3502 }, { "epoch": 0.16277881040892192, "grad_norm": 2.3086819530495375, "learning_rate": 9.547363920368138e-06, "loss": 0.1874, "step": 3503 }, { "epoch": 0.16282527881040892, "grad_norm": 1.6945630644391898, "learning_rate": 9.54705100132158e-06, "loss": 0.1029, "step": 3504 }, { "epoch": 0.1628717472118959, "grad_norm": 2.2666522321569778, "learning_rate": 9.54673797927953e-06, "loss": 0.1168, "step": 3505 }, { "epoch": 0.1629182156133829, "grad_norm": 2.318409599637791, "learning_rate": 9.54642485424908e-06, "loss": 0.1145, "step": 3506 }, { "epoch": 0.16296468401486988, "grad_norm": 1.2375302097927312, "learning_rate": 9.546111626237323e-06, "loss": 0.0757, "step": 3507 }, { "epoch": 0.16301115241635689, "grad_norm": 1.616674698685485, "learning_rate": 9.54579829525135e-06, "loss": 0.0735, "step": 3508 }, { "epoch": 0.16305762081784386, "grad_norm": 2.25710321684957, "learning_rate": 9.545484861298264e-06, "loss": 0.0667, "step": 3509 }, { "epoch": 0.16310408921933087, "grad_norm": 1.655715187902915, "learning_rate": 9.54517132438516e-06, "loss": 0.1266, "step": 3510 }, { "epoch": 0.16315055762081784, "grad_norm": 1.2275081154778693, "learning_rate": 9.544857684519144e-06, "loss": 0.0586, "step": 3511 }, { "epoch": 0.16319702602230482, "grad_norm": 3.058492204157249, "learning_rate": 9.544543941707316e-06, "loss": 0.0982, "step": 3512 }, { "epoch": 0.16324349442379182, "grad_norm": 1.1474854221640498, "learning_rate": 9.544230095956786e-06, "loss": 0.0282, "step": 3513 }, { "epoch": 0.1632899628252788, "grad_norm": 0.863867115951796, "learning_rate": 9.543916147274663e-06, "loss": 0.0547, "step": 3514 }, { "epoch": 0.1633364312267658, "grad_norm": 1.4072260096680393, "learning_rate": 9.543602095668055e-06, "loss": 0.0661, "step": 3515 }, { "epoch": 0.16338289962825278, "grad_norm": 2.6499065927288696, "learning_rate": 9.543287941144078e-06, "loss": 0.1545, "step": 3516 }, { "epoch": 0.16342936802973979, "grad_norm": 2.430904284890492, "learning_rate": 9.542973683709846e-06, "loss": 0.1516, "step": 3517 }, { "epoch": 0.16347583643122676, "grad_norm": 3.1038080229575615, "learning_rate": 9.54265932337248e-06, "loss": 0.0933, "step": 3518 }, { "epoch": 0.16352230483271377, "grad_norm": 0.9879316025287432, "learning_rate": 9.542344860139099e-06, "loss": 0.079, "step": 3519 }, { "epoch": 0.16356877323420074, "grad_norm": 5.324094813433074, "learning_rate": 9.542030294016827e-06, "loss": 0.0383, "step": 3520 }, { "epoch": 0.16361524163568772, "grad_norm": 1.2476390659876622, "learning_rate": 9.541715625012786e-06, "loss": 0.1048, "step": 3521 }, { "epoch": 0.16366171003717472, "grad_norm": 1.7405972278040533, "learning_rate": 9.541400853134108e-06, "loss": 0.1518, "step": 3522 }, { "epoch": 0.1637081784386617, "grad_norm": 1.2255375274373639, "learning_rate": 9.541085978387921e-06, "loss": 0.1176, "step": 3523 }, { "epoch": 0.1637546468401487, "grad_norm": 1.6566292632918171, "learning_rate": 9.540771000781357e-06, "loss": 0.0764, "step": 3524 }, { "epoch": 0.16380111524163568, "grad_norm": 1.9239238046810545, "learning_rate": 9.54045592032155e-06, "loss": 0.1247, "step": 3525 }, { "epoch": 0.16384758364312269, "grad_norm": 1.8907172500126634, "learning_rate": 9.540140737015639e-06, "loss": 0.1013, "step": 3526 }, { "epoch": 0.16389405204460966, "grad_norm": 2.4029704810048806, "learning_rate": 9.53982545087076e-06, "loss": 0.0905, "step": 3527 }, { "epoch": 0.16394052044609667, "grad_norm": 1.7686701492929746, "learning_rate": 9.539510061894058e-06, "loss": 0.1005, "step": 3528 }, { "epoch": 0.16398698884758364, "grad_norm": 1.3932550692383947, "learning_rate": 9.539194570092676e-06, "loss": 0.0975, "step": 3529 }, { "epoch": 0.16403345724907062, "grad_norm": 1.110435503320776, "learning_rate": 9.538878975473758e-06, "loss": 0.0947, "step": 3530 }, { "epoch": 0.16407992565055762, "grad_norm": 1.0172313841684735, "learning_rate": 9.538563278044454e-06, "loss": 0.0375, "step": 3531 }, { "epoch": 0.1641263940520446, "grad_norm": 0.986231027896432, "learning_rate": 9.538247477811916e-06, "loss": 0.0868, "step": 3532 }, { "epoch": 0.1641728624535316, "grad_norm": 1.57377083067201, "learning_rate": 9.537931574783296e-06, "loss": 0.1234, "step": 3533 }, { "epoch": 0.16421933085501858, "grad_norm": 1.7154323829554565, "learning_rate": 9.537615568965749e-06, "loss": 0.0915, "step": 3534 }, { "epoch": 0.16426579925650558, "grad_norm": 2.0031443553415187, "learning_rate": 9.537299460366434e-06, "loss": 0.0965, "step": 3535 }, { "epoch": 0.16431226765799256, "grad_norm": 1.409931542099597, "learning_rate": 9.536983248992511e-06, "loss": 0.0774, "step": 3536 }, { "epoch": 0.16435873605947957, "grad_norm": 1.0396456631504696, "learning_rate": 9.536666934851141e-06, "loss": 0.0853, "step": 3537 }, { "epoch": 0.16440520446096654, "grad_norm": 2.1034168604859667, "learning_rate": 9.536350517949492e-06, "loss": 0.0813, "step": 3538 }, { "epoch": 0.16445167286245352, "grad_norm": 3.134115827746775, "learning_rate": 9.536033998294729e-06, "loss": 0.1202, "step": 3539 }, { "epoch": 0.16449814126394052, "grad_norm": 1.394029837543705, "learning_rate": 9.535717375894022e-06, "loss": 0.0823, "step": 3540 }, { "epoch": 0.1645446096654275, "grad_norm": 1.804509892443864, "learning_rate": 9.535400650754543e-06, "loss": 0.0895, "step": 3541 }, { "epoch": 0.1645910780669145, "grad_norm": 1.6325963734994193, "learning_rate": 9.535083822883465e-06, "loss": 0.1086, "step": 3542 }, { "epoch": 0.16463754646840148, "grad_norm": 1.208640474890346, "learning_rate": 9.534766892287965e-06, "loss": 0.0664, "step": 3543 }, { "epoch": 0.16468401486988848, "grad_norm": 1.9764532842721925, "learning_rate": 9.534449858975223e-06, "loss": 0.1396, "step": 3544 }, { "epoch": 0.16473048327137546, "grad_norm": 1.3949129678143628, "learning_rate": 9.534132722952419e-06, "loss": 0.1046, "step": 3545 }, { "epoch": 0.16477695167286247, "grad_norm": 2.465633807787964, "learning_rate": 9.533815484226736e-06, "loss": 0.1199, "step": 3546 }, { "epoch": 0.16482342007434944, "grad_norm": 0.46159624913545827, "learning_rate": 9.533498142805361e-06, "loss": 0.0257, "step": 3547 }, { "epoch": 0.16486988847583642, "grad_norm": 0.8317032831185656, "learning_rate": 9.533180698695483e-06, "loss": 0.0521, "step": 3548 }, { "epoch": 0.16491635687732342, "grad_norm": 1.9738630433335385, "learning_rate": 9.53286315190429e-06, "loss": 0.0904, "step": 3549 }, { "epoch": 0.1649628252788104, "grad_norm": 1.9887929382319334, "learning_rate": 9.532545502438974e-06, "loss": 0.1177, "step": 3550 }, { "epoch": 0.1650092936802974, "grad_norm": 2.1569094716739334, "learning_rate": 9.532227750306735e-06, "loss": 0.0968, "step": 3551 }, { "epoch": 0.16505576208178438, "grad_norm": 0.9386702536734473, "learning_rate": 9.531909895514766e-06, "loss": 0.0582, "step": 3552 }, { "epoch": 0.16510223048327138, "grad_norm": 1.086555272082903, "learning_rate": 9.531591938070269e-06, "loss": 0.1091, "step": 3553 }, { "epoch": 0.16514869888475836, "grad_norm": 1.1603265306558976, "learning_rate": 9.531273877980444e-06, "loss": 0.0775, "step": 3554 }, { "epoch": 0.16519516728624536, "grad_norm": 0.7320993551817502, "learning_rate": 9.530955715252497e-06, "loss": 0.0612, "step": 3555 }, { "epoch": 0.16524163568773234, "grad_norm": 0.6490843856496058, "learning_rate": 9.530637449893635e-06, "loss": 0.0467, "step": 3556 }, { "epoch": 0.16528810408921932, "grad_norm": 2.206943247023249, "learning_rate": 9.530319081911065e-06, "loss": 0.0947, "step": 3557 }, { "epoch": 0.16533457249070632, "grad_norm": 1.6125281865710244, "learning_rate": 9.530000611312e-06, "loss": 0.0763, "step": 3558 }, { "epoch": 0.1653810408921933, "grad_norm": 2.4890349462810475, "learning_rate": 9.529682038103654e-06, "loss": 0.118, "step": 3559 }, { "epoch": 0.1654275092936803, "grad_norm": 2.4465865639966173, "learning_rate": 9.529363362293242e-06, "loss": 0.1362, "step": 3560 }, { "epoch": 0.16547397769516728, "grad_norm": 1.3876828558476693, "learning_rate": 9.529044583887983e-06, "loss": 0.0863, "step": 3561 }, { "epoch": 0.16552044609665428, "grad_norm": 2.5416028525907577, "learning_rate": 9.528725702895098e-06, "loss": 0.1092, "step": 3562 }, { "epoch": 0.16556691449814126, "grad_norm": 2.0883840099070836, "learning_rate": 9.528406719321807e-06, "loss": 0.0891, "step": 3563 }, { "epoch": 0.16561338289962826, "grad_norm": 0.9186170112949678, "learning_rate": 9.52808763317534e-06, "loss": 0.0829, "step": 3564 }, { "epoch": 0.16565985130111524, "grad_norm": 1.3073105617899943, "learning_rate": 9.527768444462922e-06, "loss": 0.0728, "step": 3565 }, { "epoch": 0.16570631970260222, "grad_norm": 1.121114393714625, "learning_rate": 9.527449153191784e-06, "loss": 0.0624, "step": 3566 }, { "epoch": 0.16575278810408922, "grad_norm": 1.8388320158783766, "learning_rate": 9.527129759369157e-06, "loss": 0.1073, "step": 3567 }, { "epoch": 0.1657992565055762, "grad_norm": 1.0222752157529396, "learning_rate": 9.526810263002276e-06, "loss": 0.079, "step": 3568 }, { "epoch": 0.1658457249070632, "grad_norm": 0.9510215043396717, "learning_rate": 9.526490664098378e-06, "loss": 0.0635, "step": 3569 }, { "epoch": 0.16589219330855018, "grad_norm": 0.8640262805905119, "learning_rate": 9.526170962664703e-06, "loss": 0.0626, "step": 3570 }, { "epoch": 0.16593866171003718, "grad_norm": 1.9745583710486283, "learning_rate": 9.525851158708493e-06, "loss": 0.1563, "step": 3571 }, { "epoch": 0.16598513011152416, "grad_norm": 1.753949321922813, "learning_rate": 9.525531252236988e-06, "loss": 0.0918, "step": 3572 }, { "epoch": 0.16603159851301116, "grad_norm": 1.3219634000845302, "learning_rate": 9.525211243257439e-06, "loss": 0.0728, "step": 3573 }, { "epoch": 0.16607806691449814, "grad_norm": 1.6783798747203418, "learning_rate": 9.524891131777093e-06, "loss": 0.1271, "step": 3574 }, { "epoch": 0.16612453531598512, "grad_norm": 1.501498245234326, "learning_rate": 9.5245709178032e-06, "loss": 0.0742, "step": 3575 }, { "epoch": 0.16617100371747212, "grad_norm": 1.4213324329948207, "learning_rate": 9.524250601343012e-06, "loss": 0.0467, "step": 3576 }, { "epoch": 0.1662174721189591, "grad_norm": 1.3029605228024064, "learning_rate": 9.523930182403789e-06, "loss": 0.0833, "step": 3577 }, { "epoch": 0.1662639405204461, "grad_norm": 2.0333358514304445, "learning_rate": 9.523609660992785e-06, "loss": 0.0938, "step": 3578 }, { "epoch": 0.16631040892193308, "grad_norm": 1.442649190053285, "learning_rate": 9.52328903711726e-06, "loss": 0.079, "step": 3579 }, { "epoch": 0.16635687732342008, "grad_norm": 1.8372146302205106, "learning_rate": 9.52296831078448e-06, "loss": 0.1224, "step": 3580 }, { "epoch": 0.16640334572490706, "grad_norm": 2.6664915297131384, "learning_rate": 9.522647482001706e-06, "loss": 0.09, "step": 3581 }, { "epoch": 0.16644981412639406, "grad_norm": 2.8385853130293364, "learning_rate": 9.522326550776206e-06, "loss": 0.1887, "step": 3582 }, { "epoch": 0.16649628252788104, "grad_norm": 0.7651197067436292, "learning_rate": 9.52200551711525e-06, "loss": 0.0442, "step": 3583 }, { "epoch": 0.16654275092936802, "grad_norm": 2.227552580295956, "learning_rate": 9.521684381026111e-06, "loss": 0.1324, "step": 3584 }, { "epoch": 0.16658921933085502, "grad_norm": 0.9303569790759534, "learning_rate": 9.52136314251606e-06, "loss": 0.0441, "step": 3585 }, { "epoch": 0.166635687732342, "grad_norm": 1.3330348322768069, "learning_rate": 9.521041801592377e-06, "loss": 0.0673, "step": 3586 }, { "epoch": 0.166682156133829, "grad_norm": 2.4235088049760996, "learning_rate": 9.520720358262338e-06, "loss": 0.1014, "step": 3587 }, { "epoch": 0.16672862453531598, "grad_norm": 1.5617564420522738, "learning_rate": 9.520398812533225e-06, "loss": 0.1064, "step": 3588 }, { "epoch": 0.16677509293680298, "grad_norm": 1.8498611851776257, "learning_rate": 9.52007716441232e-06, "loss": 0.0659, "step": 3589 }, { "epoch": 0.16682156133828996, "grad_norm": 2.0562083963963897, "learning_rate": 9.519755413906912e-06, "loss": 0.1128, "step": 3590 }, { "epoch": 0.16686802973977696, "grad_norm": 0.9627538142332747, "learning_rate": 9.519433561024284e-06, "loss": 0.1088, "step": 3591 }, { "epoch": 0.16691449814126394, "grad_norm": 2.5851752567865764, "learning_rate": 9.519111605771733e-06, "loss": 0.1754, "step": 3592 }, { "epoch": 0.16696096654275092, "grad_norm": 1.5848509552414884, "learning_rate": 9.518789548156546e-06, "loss": 0.1705, "step": 3593 }, { "epoch": 0.16700743494423792, "grad_norm": 1.407557599789012, "learning_rate": 9.51846738818602e-06, "loss": 0.0893, "step": 3594 }, { "epoch": 0.1670539033457249, "grad_norm": 1.577395549062968, "learning_rate": 9.518145125867452e-06, "loss": 0.0838, "step": 3595 }, { "epoch": 0.1671003717472119, "grad_norm": 0.9792920014462748, "learning_rate": 9.517822761208143e-06, "loss": 0.0645, "step": 3596 }, { "epoch": 0.16714684014869888, "grad_norm": 1.3884497081669878, "learning_rate": 9.517500294215392e-06, "loss": 0.086, "step": 3597 }, { "epoch": 0.16719330855018588, "grad_norm": 1.8563208623056984, "learning_rate": 9.517177724896505e-06, "loss": 0.1226, "step": 3598 }, { "epoch": 0.16723977695167286, "grad_norm": 1.222624812230828, "learning_rate": 9.516855053258786e-06, "loss": 0.0694, "step": 3599 }, { "epoch": 0.16728624535315986, "grad_norm": 1.4071591202925318, "learning_rate": 9.51653227930955e-06, "loss": 0.0875, "step": 3600 }, { "epoch": 0.16733271375464684, "grad_norm": 2.076824492288387, "learning_rate": 9.516209403056103e-06, "loss": 0.1317, "step": 3601 }, { "epoch": 0.16737918215613382, "grad_norm": 2.77843629398638, "learning_rate": 9.515886424505759e-06, "loss": 0.1109, "step": 3602 }, { "epoch": 0.16742565055762082, "grad_norm": 11.706346150738627, "learning_rate": 9.515563343665835e-06, "loss": 0.1015, "step": 3603 }, { "epoch": 0.1674721189591078, "grad_norm": 2.3639377663496313, "learning_rate": 9.515240160543648e-06, "loss": 0.1205, "step": 3604 }, { "epoch": 0.1675185873605948, "grad_norm": 1.4708986793797925, "learning_rate": 9.514916875146521e-06, "loss": 0.0776, "step": 3605 }, { "epoch": 0.16756505576208178, "grad_norm": 1.593807435370857, "learning_rate": 9.514593487481774e-06, "loss": 0.0812, "step": 3606 }, { "epoch": 0.16761152416356878, "grad_norm": 1.662106562352161, "learning_rate": 9.514269997556733e-06, "loss": 0.0784, "step": 3607 }, { "epoch": 0.16765799256505576, "grad_norm": 1.6491276857271926, "learning_rate": 9.513946405378725e-06, "loss": 0.1344, "step": 3608 }, { "epoch": 0.16770446096654276, "grad_norm": 8.475147144984284, "learning_rate": 9.51362271095508e-06, "loss": 0.0847, "step": 3609 }, { "epoch": 0.16775092936802974, "grad_norm": 3.251660348292261, "learning_rate": 9.51329891429313e-06, "loss": 0.1544, "step": 3610 }, { "epoch": 0.16779739776951672, "grad_norm": 1.9872689967739912, "learning_rate": 9.51297501540021e-06, "loss": 0.0643, "step": 3611 }, { "epoch": 0.16784386617100372, "grad_norm": 4.250885930962193, "learning_rate": 9.512651014283656e-06, "loss": 0.1079, "step": 3612 }, { "epoch": 0.1678903345724907, "grad_norm": 1.8951036801418033, "learning_rate": 9.512326910950807e-06, "loss": 0.0967, "step": 3613 }, { "epoch": 0.1679368029739777, "grad_norm": 1.7291517197951294, "learning_rate": 9.512002705409003e-06, "loss": 0.1074, "step": 3614 }, { "epoch": 0.16798327137546468, "grad_norm": 4.288912404614347, "learning_rate": 9.511678397665591e-06, "loss": 0.107, "step": 3615 }, { "epoch": 0.16802973977695168, "grad_norm": 2.336584912842698, "learning_rate": 9.511353987727913e-06, "loss": 0.0752, "step": 3616 }, { "epoch": 0.16807620817843866, "grad_norm": 1.5605800753165746, "learning_rate": 9.511029475603322e-06, "loss": 0.073, "step": 3617 }, { "epoch": 0.16812267657992566, "grad_norm": 3.351578664873546, "learning_rate": 9.510704861299162e-06, "loss": 0.1195, "step": 3618 }, { "epoch": 0.16816914498141264, "grad_norm": 2.2071989462862294, "learning_rate": 9.510380144822791e-06, "loss": 0.0825, "step": 3619 }, { "epoch": 0.16821561338289961, "grad_norm": 2.020444180224223, "learning_rate": 9.510055326181564e-06, "loss": 0.1019, "step": 3620 }, { "epoch": 0.16826208178438662, "grad_norm": 6.470590548301697, "learning_rate": 9.509730405382835e-06, "loss": 0.1509, "step": 3621 }, { "epoch": 0.1683085501858736, "grad_norm": 1.8774813265590302, "learning_rate": 9.509405382433967e-06, "loss": 0.0948, "step": 3622 }, { "epoch": 0.1683550185873606, "grad_norm": 1.749971295807511, "learning_rate": 9.509080257342322e-06, "loss": 0.087, "step": 3623 }, { "epoch": 0.16840148698884758, "grad_norm": 2.8518293695587946, "learning_rate": 9.508755030115262e-06, "loss": 0.0906, "step": 3624 }, { "epoch": 0.16844795539033458, "grad_norm": 5.770996908700038, "learning_rate": 9.508429700760155e-06, "loss": 0.0795, "step": 3625 }, { "epoch": 0.16849442379182156, "grad_norm": 1.3285814992479164, "learning_rate": 9.50810426928437e-06, "loss": 0.1481, "step": 3626 }, { "epoch": 0.16854089219330856, "grad_norm": 2.7372869914502878, "learning_rate": 9.50777873569528e-06, "loss": 0.1277, "step": 3627 }, { "epoch": 0.16858736059479554, "grad_norm": 3.6805980972429437, "learning_rate": 9.507453100000257e-06, "loss": 0.1639, "step": 3628 }, { "epoch": 0.16863382899628251, "grad_norm": 1.2857878274851833, "learning_rate": 9.507127362206677e-06, "loss": 0.0645, "step": 3629 }, { "epoch": 0.16868029739776952, "grad_norm": 1.6556979769627949, "learning_rate": 9.506801522321918e-06, "loss": 0.082, "step": 3630 }, { "epoch": 0.1687267657992565, "grad_norm": 3.949747800788824, "learning_rate": 9.506475580353361e-06, "loss": 0.1435, "step": 3631 }, { "epoch": 0.1687732342007435, "grad_norm": 1.2428026707261286, "learning_rate": 9.506149536308392e-06, "loss": 0.0959, "step": 3632 }, { "epoch": 0.16881970260223048, "grad_norm": 1.2587123771514224, "learning_rate": 9.50582339019439e-06, "loss": 0.1467, "step": 3633 }, { "epoch": 0.16886617100371748, "grad_norm": 1.7287444553754288, "learning_rate": 9.505497142018747e-06, "loss": 0.1364, "step": 3634 }, { "epoch": 0.16891263940520446, "grad_norm": 1.208335858718173, "learning_rate": 9.505170791788852e-06, "loss": 0.0817, "step": 3635 }, { "epoch": 0.16895910780669146, "grad_norm": 2.3862372001184227, "learning_rate": 9.504844339512096e-06, "loss": 0.1024, "step": 3636 }, { "epoch": 0.16900557620817844, "grad_norm": 1.5451922864843866, "learning_rate": 9.504517785195876e-06, "loss": 0.0902, "step": 3637 }, { "epoch": 0.16905204460966541, "grad_norm": 4.506889784944055, "learning_rate": 9.504191128847586e-06, "loss": 0.1179, "step": 3638 }, { "epoch": 0.16909851301115242, "grad_norm": 1.780801195800774, "learning_rate": 9.503864370474627e-06, "loss": 0.0687, "step": 3639 }, { "epoch": 0.1691449814126394, "grad_norm": 3.0737511302772456, "learning_rate": 9.503537510084399e-06, "loss": 0.0772, "step": 3640 }, { "epoch": 0.1691914498141264, "grad_norm": 2.7311672999598238, "learning_rate": 9.503210547684307e-06, "loss": 0.1189, "step": 3641 }, { "epoch": 0.16923791821561338, "grad_norm": 2.535412682409822, "learning_rate": 9.502883483281757e-06, "loss": 0.0639, "step": 3642 }, { "epoch": 0.16928438661710038, "grad_norm": 2.7755670925460376, "learning_rate": 9.502556316884157e-06, "loss": 0.0916, "step": 3643 }, { "epoch": 0.16933085501858736, "grad_norm": 2.948696484602881, "learning_rate": 9.502229048498917e-06, "loss": 0.1444, "step": 3644 }, { "epoch": 0.16937732342007436, "grad_norm": 1.70817139862751, "learning_rate": 9.501901678133452e-06, "loss": 0.1018, "step": 3645 }, { "epoch": 0.16942379182156134, "grad_norm": 3.2411103614065153, "learning_rate": 9.501574205795173e-06, "loss": 0.0993, "step": 3646 }, { "epoch": 0.1694702602230483, "grad_norm": 1.6972825884106704, "learning_rate": 9.501246631491503e-06, "loss": 0.1039, "step": 3647 }, { "epoch": 0.16951672862453532, "grad_norm": 1.681741296413171, "learning_rate": 9.500918955229858e-06, "loss": 0.0714, "step": 3648 }, { "epoch": 0.1695631970260223, "grad_norm": 1.471686905027883, "learning_rate": 9.500591177017662e-06, "loss": 0.0991, "step": 3649 }, { "epoch": 0.1696096654275093, "grad_norm": 0.7679502798461956, "learning_rate": 9.500263296862339e-06, "loss": 0.0595, "step": 3650 }, { "epoch": 0.16965613382899628, "grad_norm": 2.253546133477494, "learning_rate": 9.499935314771315e-06, "loss": 0.1459, "step": 3651 }, { "epoch": 0.16970260223048328, "grad_norm": 1.9639474230693714, "learning_rate": 9.499607230752021e-06, "loss": 0.1453, "step": 3652 }, { "epoch": 0.16974907063197026, "grad_norm": 1.9761820054966257, "learning_rate": 9.499279044811888e-06, "loss": 0.0969, "step": 3653 }, { "epoch": 0.16979553903345726, "grad_norm": 1.9966182705345865, "learning_rate": 9.498950756958347e-06, "loss": 0.0878, "step": 3654 }, { "epoch": 0.16984200743494424, "grad_norm": 1.2227017764665287, "learning_rate": 9.498622367198838e-06, "loss": 0.0863, "step": 3655 }, { "epoch": 0.1698884758364312, "grad_norm": 1.5794163369104384, "learning_rate": 9.498293875540798e-06, "loss": 0.0901, "step": 3656 }, { "epoch": 0.16993494423791822, "grad_norm": 1.5514787671774344, "learning_rate": 9.497965281991666e-06, "loss": 0.0875, "step": 3657 }, { "epoch": 0.1699814126394052, "grad_norm": 1.3291304303947635, "learning_rate": 9.497636586558888e-06, "loss": 0.0702, "step": 3658 }, { "epoch": 0.1700278810408922, "grad_norm": 2.665215704345848, "learning_rate": 9.497307789249908e-06, "loss": 0.1436, "step": 3659 }, { "epoch": 0.17007434944237917, "grad_norm": 2.143918621947528, "learning_rate": 9.49697889007217e-06, "loss": 0.0641, "step": 3660 }, { "epoch": 0.17012081784386618, "grad_norm": 1.5418557168724418, "learning_rate": 9.49664988903313e-06, "loss": 0.106, "step": 3661 }, { "epoch": 0.17016728624535316, "grad_norm": 1.0375782361396682, "learning_rate": 9.496320786140235e-06, "loss": 0.0717, "step": 3662 }, { "epoch": 0.17021375464684016, "grad_norm": 1.8096490347135366, "learning_rate": 9.495991581400946e-06, "loss": 0.1183, "step": 3663 }, { "epoch": 0.17026022304832714, "grad_norm": 1.90925918130328, "learning_rate": 9.495662274822712e-06, "loss": 0.0809, "step": 3664 }, { "epoch": 0.1703066914498141, "grad_norm": 2.4002083334467246, "learning_rate": 9.495332866412998e-06, "loss": 0.1782, "step": 3665 }, { "epoch": 0.17035315985130112, "grad_norm": 1.999766489456687, "learning_rate": 9.495003356179261e-06, "loss": 0.0748, "step": 3666 }, { "epoch": 0.1703996282527881, "grad_norm": 1.941703830109646, "learning_rate": 9.49467374412897e-06, "loss": 0.1132, "step": 3667 }, { "epoch": 0.1704460966542751, "grad_norm": 1.322382465287209, "learning_rate": 9.494344030269587e-06, "loss": 0.0883, "step": 3668 }, { "epoch": 0.17049256505576207, "grad_norm": 3.1001383941627365, "learning_rate": 9.49401421460858e-06, "loss": 0.094, "step": 3669 }, { "epoch": 0.17053903345724908, "grad_norm": 2.205456645018428, "learning_rate": 9.493684297153423e-06, "loss": 0.0963, "step": 3670 }, { "epoch": 0.17058550185873605, "grad_norm": 2.406824474280715, "learning_rate": 9.493354277911587e-06, "loss": 0.1023, "step": 3671 }, { "epoch": 0.17063197026022306, "grad_norm": 0.9405917762997434, "learning_rate": 9.493024156890545e-06, "loss": 0.0725, "step": 3672 }, { "epoch": 0.17067843866171004, "grad_norm": 1.2661175835680278, "learning_rate": 9.492693934097779e-06, "loss": 0.0834, "step": 3673 }, { "epoch": 0.170724907063197, "grad_norm": 2.412566646497669, "learning_rate": 9.492363609540767e-06, "loss": 0.1025, "step": 3674 }, { "epoch": 0.17077137546468402, "grad_norm": 0.6518625845260305, "learning_rate": 9.492033183226993e-06, "loss": 0.0593, "step": 3675 }, { "epoch": 0.170817843866171, "grad_norm": 1.9053537825096023, "learning_rate": 9.491702655163937e-06, "loss": 0.0997, "step": 3676 }, { "epoch": 0.170864312267658, "grad_norm": 4.0586059615004055, "learning_rate": 9.491372025359088e-06, "loss": 0.1821, "step": 3677 }, { "epoch": 0.17091078066914497, "grad_norm": 1.1187519461382491, "learning_rate": 9.491041293819936e-06, "loss": 0.0765, "step": 3678 }, { "epoch": 0.17095724907063198, "grad_norm": 1.9101811439129075, "learning_rate": 9.490710460553973e-06, "loss": 0.1401, "step": 3679 }, { "epoch": 0.17100371747211895, "grad_norm": 0.822416244129031, "learning_rate": 9.490379525568691e-06, "loss": 0.0744, "step": 3680 }, { "epoch": 0.17105018587360596, "grad_norm": 1.8333953308138795, "learning_rate": 9.490048488871586e-06, "loss": 0.1659, "step": 3681 }, { "epoch": 0.17109665427509294, "grad_norm": 7.399002878031296, "learning_rate": 9.48971735047016e-06, "loss": 0.0559, "step": 3682 }, { "epoch": 0.1711431226765799, "grad_norm": 1.7443644436369659, "learning_rate": 9.489386110371908e-06, "loss": 0.0782, "step": 3683 }, { "epoch": 0.17118959107806692, "grad_norm": 2.8257069183916768, "learning_rate": 9.489054768584336e-06, "loss": 0.1271, "step": 3684 }, { "epoch": 0.1712360594795539, "grad_norm": 1.5627875237684954, "learning_rate": 9.488723325114948e-06, "loss": 0.111, "step": 3685 }, { "epoch": 0.1712825278810409, "grad_norm": 2.127471332252338, "learning_rate": 9.488391779971252e-06, "loss": 0.122, "step": 3686 }, { "epoch": 0.17132899628252787, "grad_norm": 2.6761041903103346, "learning_rate": 9.488060133160761e-06, "loss": 0.1414, "step": 3687 }, { "epoch": 0.17137546468401488, "grad_norm": 1.1960071259011058, "learning_rate": 9.487728384690984e-06, "loss": 0.1072, "step": 3688 }, { "epoch": 0.17142193308550185, "grad_norm": 1.1750209406722762, "learning_rate": 9.487396534569436e-06, "loss": 0.0742, "step": 3689 }, { "epoch": 0.17146840148698886, "grad_norm": 2.0282715001826195, "learning_rate": 9.48706458280363e-06, "loss": 0.0894, "step": 3690 }, { "epoch": 0.17151486988847583, "grad_norm": 1.8993209249406307, "learning_rate": 9.486732529401092e-06, "loss": 0.1041, "step": 3691 }, { "epoch": 0.1715613382899628, "grad_norm": 2.2062917464293053, "learning_rate": 9.48640037436934e-06, "loss": 0.0907, "step": 3692 }, { "epoch": 0.17160780669144982, "grad_norm": 2.381794769674824, "learning_rate": 9.486068117715897e-06, "loss": 0.1946, "step": 3693 }, { "epoch": 0.1716542750929368, "grad_norm": 1.807761092926781, "learning_rate": 9.485735759448289e-06, "loss": 0.059, "step": 3694 }, { "epoch": 0.1717007434944238, "grad_norm": 1.5381133623948213, "learning_rate": 9.485403299574044e-06, "loss": 0.1028, "step": 3695 }, { "epoch": 0.17174721189591077, "grad_norm": 1.6337188302436185, "learning_rate": 9.485070738100695e-06, "loss": 0.0923, "step": 3696 }, { "epoch": 0.17179368029739778, "grad_norm": 1.8521673216860257, "learning_rate": 9.484738075035774e-06, "loss": 0.1208, "step": 3697 }, { "epoch": 0.17184014869888475, "grad_norm": 4.083285328067909, "learning_rate": 9.484405310386815e-06, "loss": 0.0761, "step": 3698 }, { "epoch": 0.17188661710037176, "grad_norm": 1.2823865577803932, "learning_rate": 9.484072444161355e-06, "loss": 0.0655, "step": 3699 }, { "epoch": 0.17193308550185873, "grad_norm": 2.1005687827190473, "learning_rate": 9.483739476366936e-06, "loss": 0.0715, "step": 3700 }, { "epoch": 0.17197955390334574, "grad_norm": 1.9446058093652605, "learning_rate": 9.483406407011098e-06, "loss": 0.102, "step": 3701 }, { "epoch": 0.17202602230483272, "grad_norm": 1.320841453195344, "learning_rate": 9.483073236101386e-06, "loss": 0.0804, "step": 3702 }, { "epoch": 0.1720724907063197, "grad_norm": 1.654918864766356, "learning_rate": 9.482739963645347e-06, "loss": 0.0712, "step": 3703 }, { "epoch": 0.1721189591078067, "grad_norm": 1.2999961516103022, "learning_rate": 9.48240658965053e-06, "loss": 0.0431, "step": 3704 }, { "epoch": 0.17216542750929367, "grad_norm": 2.2263871027803543, "learning_rate": 9.482073114124486e-06, "loss": 0.1175, "step": 3705 }, { "epoch": 0.17221189591078068, "grad_norm": 1.9311072571236416, "learning_rate": 9.481739537074768e-06, "loss": 0.0962, "step": 3706 }, { "epoch": 0.17225836431226765, "grad_norm": 2.106374766273244, "learning_rate": 9.481405858508935e-06, "loss": 0.1478, "step": 3707 }, { "epoch": 0.17230483271375466, "grad_norm": 1.752545954005802, "learning_rate": 9.481072078434541e-06, "loss": 0.1122, "step": 3708 }, { "epoch": 0.17235130111524163, "grad_norm": 1.9966642293211128, "learning_rate": 9.480738196859147e-06, "loss": 0.145, "step": 3709 }, { "epoch": 0.17239776951672864, "grad_norm": 2.211952087737294, "learning_rate": 9.480404213790319e-06, "loss": 0.0949, "step": 3710 }, { "epoch": 0.17244423791821561, "grad_norm": 2.838672156733417, "learning_rate": 9.480070129235619e-06, "loss": 0.1365, "step": 3711 }, { "epoch": 0.1724907063197026, "grad_norm": 1.3759404310136738, "learning_rate": 9.479735943202614e-06, "loss": 0.1064, "step": 3712 }, { "epoch": 0.1725371747211896, "grad_norm": 6.889671992256581, "learning_rate": 9.47940165569888e-06, "loss": 0.0926, "step": 3713 }, { "epoch": 0.17258364312267657, "grad_norm": 3.7516920515402186, "learning_rate": 9.479067266731978e-06, "loss": 0.1073, "step": 3714 }, { "epoch": 0.17263011152416358, "grad_norm": 1.271226451258576, "learning_rate": 9.478732776309491e-06, "loss": 0.0823, "step": 3715 }, { "epoch": 0.17267657992565055, "grad_norm": 1.2357565631958491, "learning_rate": 9.478398184438993e-06, "loss": 0.0738, "step": 3716 }, { "epoch": 0.17272304832713756, "grad_norm": 1.6981253336753308, "learning_rate": 9.478063491128062e-06, "loss": 0.1159, "step": 3717 }, { "epoch": 0.17276951672862453, "grad_norm": 1.641011443523856, "learning_rate": 9.477728696384281e-06, "loss": 0.1197, "step": 3718 }, { "epoch": 0.17281598513011154, "grad_norm": 0.709535096732507, "learning_rate": 9.477393800215232e-06, "loss": 0.0337, "step": 3719 }, { "epoch": 0.17286245353159851, "grad_norm": 0.8089963124292457, "learning_rate": 9.4770588026285e-06, "loss": 0.0551, "step": 3720 }, { "epoch": 0.1729089219330855, "grad_norm": 1.3203768116420342, "learning_rate": 9.476723703631675e-06, "loss": 0.0776, "step": 3721 }, { "epoch": 0.1729553903345725, "grad_norm": 1.3964111447062528, "learning_rate": 9.476388503232345e-06, "loss": 0.036, "step": 3722 }, { "epoch": 0.17300185873605947, "grad_norm": 1.748217312520398, "learning_rate": 9.476053201438106e-06, "loss": 0.1258, "step": 3723 }, { "epoch": 0.17304832713754648, "grad_norm": 0.7967217491949222, "learning_rate": 9.475717798256551e-06, "loss": 0.0603, "step": 3724 }, { "epoch": 0.17309479553903345, "grad_norm": 6.872396686828788, "learning_rate": 9.475382293695276e-06, "loss": 0.1168, "step": 3725 }, { "epoch": 0.17314126394052046, "grad_norm": 1.003553420553196, "learning_rate": 9.475046687761884e-06, "loss": 0.0686, "step": 3726 }, { "epoch": 0.17318773234200743, "grad_norm": 2.8203685862854955, "learning_rate": 9.474710980463972e-06, "loss": 0.122, "step": 3727 }, { "epoch": 0.17323420074349444, "grad_norm": 1.8412456778020312, "learning_rate": 9.474375171809148e-06, "loss": 0.1173, "step": 3728 }, { "epoch": 0.17328066914498141, "grad_norm": 1.7502398344965027, "learning_rate": 9.474039261805017e-06, "loss": 0.1078, "step": 3729 }, { "epoch": 0.1733271375464684, "grad_norm": 3.6231687220826094, "learning_rate": 9.473703250459188e-06, "loss": 0.1762, "step": 3730 }, { "epoch": 0.1733736059479554, "grad_norm": 1.2823655267859064, "learning_rate": 9.47336713777927e-06, "loss": 0.1116, "step": 3731 }, { "epoch": 0.17342007434944237, "grad_norm": 1.641110395011722, "learning_rate": 9.473030923772881e-06, "loss": 0.0937, "step": 3732 }, { "epoch": 0.17346654275092938, "grad_norm": 2.0301914181451965, "learning_rate": 9.472694608447632e-06, "loss": 0.1759, "step": 3733 }, { "epoch": 0.17351301115241635, "grad_norm": 1.0846585650433211, "learning_rate": 9.472358191811144e-06, "loss": 0.0533, "step": 3734 }, { "epoch": 0.17355947955390336, "grad_norm": 5.368124916701292, "learning_rate": 9.472021673871035e-06, "loss": 0.1525, "step": 3735 }, { "epoch": 0.17360594795539033, "grad_norm": 2.122539971702295, "learning_rate": 9.471685054634928e-06, "loss": 0.0656, "step": 3736 }, { "epoch": 0.17365241635687734, "grad_norm": 3.714773894580092, "learning_rate": 9.471348334110448e-06, "loss": 0.0947, "step": 3737 }, { "epoch": 0.1736988847583643, "grad_norm": 1.123062668650319, "learning_rate": 9.471011512305223e-06, "loss": 0.0798, "step": 3738 }, { "epoch": 0.1737453531598513, "grad_norm": 1.1385542731080631, "learning_rate": 9.470674589226882e-06, "loss": 0.0722, "step": 3739 }, { "epoch": 0.1737918215613383, "grad_norm": 1.4853398127552628, "learning_rate": 9.470337564883056e-06, "loss": 0.0784, "step": 3740 }, { "epoch": 0.17383828996282527, "grad_norm": 1.7651252464020766, "learning_rate": 9.470000439281379e-06, "loss": 0.1208, "step": 3741 }, { "epoch": 0.17388475836431228, "grad_norm": 1.5531861272260452, "learning_rate": 9.469663212429487e-06, "loss": 0.1143, "step": 3742 }, { "epoch": 0.17393122676579925, "grad_norm": 2.1133320031386567, "learning_rate": 9.46932588433502e-06, "loss": 0.0719, "step": 3743 }, { "epoch": 0.17397769516728626, "grad_norm": 1.580007107587017, "learning_rate": 9.468988455005617e-06, "loss": 0.0818, "step": 3744 }, { "epoch": 0.17402416356877323, "grad_norm": 1.4018943745868808, "learning_rate": 9.468650924448924e-06, "loss": 0.0757, "step": 3745 }, { "epoch": 0.17407063197026024, "grad_norm": 3.045760679025062, "learning_rate": 9.468313292672581e-06, "loss": 0.1243, "step": 3746 }, { "epoch": 0.1741171003717472, "grad_norm": 1.684154353862931, "learning_rate": 9.467975559684241e-06, "loss": 0.115, "step": 3747 }, { "epoch": 0.1741635687732342, "grad_norm": 2.0360832597835623, "learning_rate": 9.467637725491551e-06, "loss": 0.1165, "step": 3748 }, { "epoch": 0.1742100371747212, "grad_norm": 1.0088659129149111, "learning_rate": 9.467299790102166e-06, "loss": 0.0858, "step": 3749 }, { "epoch": 0.17425650557620817, "grad_norm": 2.4972803635901473, "learning_rate": 9.46696175352374e-06, "loss": 0.0779, "step": 3750 }, { "epoch": 0.17430297397769517, "grad_norm": 2.3260785649242792, "learning_rate": 9.466623615763926e-06, "loss": 0.0752, "step": 3751 }, { "epoch": 0.17434944237918215, "grad_norm": 1.1807086730809087, "learning_rate": 9.466285376830388e-06, "loss": 0.056, "step": 3752 }, { "epoch": 0.17439591078066916, "grad_norm": 1.0218754436879658, "learning_rate": 9.465947036730785e-06, "loss": 0.073, "step": 3753 }, { "epoch": 0.17444237918215613, "grad_norm": 1.2608037727681936, "learning_rate": 9.465608595472781e-06, "loss": 0.0649, "step": 3754 }, { "epoch": 0.17448884758364314, "grad_norm": 0.9286598787155298, "learning_rate": 9.465270053064041e-06, "loss": 0.0455, "step": 3755 }, { "epoch": 0.1745353159851301, "grad_norm": 2.6390618219810906, "learning_rate": 9.464931409512237e-06, "loss": 0.1544, "step": 3756 }, { "epoch": 0.1745817843866171, "grad_norm": 1.553468699183106, "learning_rate": 9.464592664825036e-06, "loss": 0.0894, "step": 3757 }, { "epoch": 0.1746282527881041, "grad_norm": 0.9633504301890339, "learning_rate": 9.464253819010112e-06, "loss": 0.0528, "step": 3758 }, { "epoch": 0.17467472118959107, "grad_norm": 1.340831523518138, "learning_rate": 9.463914872075142e-06, "loss": 0.1077, "step": 3759 }, { "epoch": 0.17472118959107807, "grad_norm": 1.6028849040640276, "learning_rate": 9.463575824027802e-06, "loss": 0.0989, "step": 3760 }, { "epoch": 0.17476765799256505, "grad_norm": 2.6396932585153707, "learning_rate": 9.46323667487577e-06, "loss": 0.1011, "step": 3761 }, { "epoch": 0.17481412639405206, "grad_norm": 1.226866184658463, "learning_rate": 9.462897424626731e-06, "loss": 0.0861, "step": 3762 }, { "epoch": 0.17486059479553903, "grad_norm": 2.783219035400105, "learning_rate": 9.462558073288369e-06, "loss": 0.1088, "step": 3763 }, { "epoch": 0.17490706319702604, "grad_norm": 2.143359414464587, "learning_rate": 9.462218620868369e-06, "loss": 0.1028, "step": 3764 }, { "epoch": 0.174953531598513, "grad_norm": 1.7615803413616975, "learning_rate": 9.46187906737442e-06, "loss": 0.1066, "step": 3765 }, { "epoch": 0.175, "grad_norm": 1.0847127524019873, "learning_rate": 9.461539412814213e-06, "loss": 0.0757, "step": 3766 }, { "epoch": 0.175046468401487, "grad_norm": 1.4659764275496392, "learning_rate": 9.461199657195446e-06, "loss": 0.1242, "step": 3767 }, { "epoch": 0.17509293680297397, "grad_norm": 3.1096021089696047, "learning_rate": 9.46085980052581e-06, "loss": 0.1027, "step": 3768 }, { "epoch": 0.17513940520446097, "grad_norm": 1.3880507757125737, "learning_rate": 9.460519842813004e-06, "loss": 0.1068, "step": 3769 }, { "epoch": 0.17518587360594795, "grad_norm": 2.162940950567449, "learning_rate": 9.460179784064729e-06, "loss": 0.1373, "step": 3770 }, { "epoch": 0.17523234200743495, "grad_norm": 0.811928589205812, "learning_rate": 9.459839624288689e-06, "loss": 0.0484, "step": 3771 }, { "epoch": 0.17527881040892193, "grad_norm": 2.607330632589359, "learning_rate": 9.459499363492585e-06, "loss": 0.1013, "step": 3772 }, { "epoch": 0.17532527881040894, "grad_norm": 1.8737854118998074, "learning_rate": 9.459159001684129e-06, "loss": 0.1088, "step": 3773 }, { "epoch": 0.1753717472118959, "grad_norm": 2.144305282049012, "learning_rate": 9.458818538871029e-06, "loss": 0.1127, "step": 3774 }, { "epoch": 0.1754182156133829, "grad_norm": 1.5913109586632828, "learning_rate": 9.458477975060993e-06, "loss": 0.0823, "step": 3775 }, { "epoch": 0.1754646840148699, "grad_norm": 2.4398738938205624, "learning_rate": 9.458137310261741e-06, "loss": 0.077, "step": 3776 }, { "epoch": 0.17551115241635687, "grad_norm": 0.9252721730788673, "learning_rate": 9.457796544480986e-06, "loss": 0.109, "step": 3777 }, { "epoch": 0.17555762081784387, "grad_norm": 1.6967949159572617, "learning_rate": 9.457455677726447e-06, "loss": 0.0714, "step": 3778 }, { "epoch": 0.17560408921933085, "grad_norm": 2.170793837080294, "learning_rate": 9.457114710005849e-06, "loss": 0.0756, "step": 3779 }, { "epoch": 0.17565055762081785, "grad_norm": 2.1068886588004654, "learning_rate": 9.456773641326908e-06, "loss": 0.0895, "step": 3780 }, { "epoch": 0.17569702602230483, "grad_norm": 1.4051695106344806, "learning_rate": 9.456432471697355e-06, "loss": 0.0631, "step": 3781 }, { "epoch": 0.17574349442379184, "grad_norm": 14.654149120051294, "learning_rate": 9.456091201124915e-06, "loss": 0.1437, "step": 3782 }, { "epoch": 0.1757899628252788, "grad_norm": 1.0311304394844696, "learning_rate": 9.45574982961732e-06, "loss": 0.1089, "step": 3783 }, { "epoch": 0.1758364312267658, "grad_norm": 1.583596655942238, "learning_rate": 9.455408357182303e-06, "loss": 0.1302, "step": 3784 }, { "epoch": 0.1758828996282528, "grad_norm": 1.6628017278691287, "learning_rate": 9.455066783827595e-06, "loss": 0.0749, "step": 3785 }, { "epoch": 0.17592936802973977, "grad_norm": 1.6075425022316026, "learning_rate": 9.454725109560938e-06, "loss": 0.1837, "step": 3786 }, { "epoch": 0.17597583643122677, "grad_norm": 1.2151597817860809, "learning_rate": 9.454383334390067e-06, "loss": 0.0909, "step": 3787 }, { "epoch": 0.17602230483271375, "grad_norm": 2.2762985264192044, "learning_rate": 9.454041458322726e-06, "loss": 0.1328, "step": 3788 }, { "epoch": 0.17606877323420075, "grad_norm": 2.672473225595897, "learning_rate": 9.453699481366658e-06, "loss": 0.0826, "step": 3789 }, { "epoch": 0.17611524163568773, "grad_norm": 1.9310574003180168, "learning_rate": 9.453357403529609e-06, "loss": 0.0545, "step": 3790 }, { "epoch": 0.17616171003717473, "grad_norm": 1.20125679773326, "learning_rate": 9.45301522481933e-06, "loss": 0.0626, "step": 3791 }, { "epoch": 0.1762081784386617, "grad_norm": 1.6379627257945215, "learning_rate": 9.452672945243567e-06, "loss": 0.1121, "step": 3792 }, { "epoch": 0.1762546468401487, "grad_norm": 1.6931338886115872, "learning_rate": 9.452330564810075e-06, "loss": 0.082, "step": 3793 }, { "epoch": 0.1763011152416357, "grad_norm": 1.0956165618151767, "learning_rate": 9.451988083526611e-06, "loss": 0.0714, "step": 3794 }, { "epoch": 0.17634758364312267, "grad_norm": 2.4985459574564945, "learning_rate": 9.45164550140093e-06, "loss": 0.1198, "step": 3795 }, { "epoch": 0.17639405204460967, "grad_norm": 1.9162016974582023, "learning_rate": 9.451302818440795e-06, "loss": 0.1024, "step": 3796 }, { "epoch": 0.17644052044609665, "grad_norm": 3.75217463571071, "learning_rate": 9.450960034653965e-06, "loss": 0.1538, "step": 3797 }, { "epoch": 0.17648698884758365, "grad_norm": 1.2516596010752121, "learning_rate": 9.450617150048208e-06, "loss": 0.0848, "step": 3798 }, { "epoch": 0.17653345724907063, "grad_norm": 1.6451095581539898, "learning_rate": 9.450274164631284e-06, "loss": 0.1246, "step": 3799 }, { "epoch": 0.17657992565055763, "grad_norm": 1.2106982626797171, "learning_rate": 9.44993107841097e-06, "loss": 0.1292, "step": 3800 }, { "epoch": 0.1766263940520446, "grad_norm": 1.2742753084553327, "learning_rate": 9.449587891395032e-06, "loss": 0.1218, "step": 3801 }, { "epoch": 0.1766728624535316, "grad_norm": 0.928851328661604, "learning_rate": 9.449244603591246e-06, "loss": 0.1324, "step": 3802 }, { "epoch": 0.1767193308550186, "grad_norm": 2.1808020979587948, "learning_rate": 9.448901215007385e-06, "loss": 0.109, "step": 3803 }, { "epoch": 0.17676579925650557, "grad_norm": 8.413335504902118, "learning_rate": 9.44855772565123e-06, "loss": 0.0929, "step": 3804 }, { "epoch": 0.17681226765799257, "grad_norm": 1.6893200013190317, "learning_rate": 9.448214135530563e-06, "loss": 0.0637, "step": 3805 }, { "epoch": 0.17685873605947955, "grad_norm": 1.3576590179272963, "learning_rate": 9.44787044465316e-06, "loss": 0.1168, "step": 3806 }, { "epoch": 0.17690520446096655, "grad_norm": 1.259346077437816, "learning_rate": 9.447526653026813e-06, "loss": 0.0748, "step": 3807 }, { "epoch": 0.17695167286245353, "grad_norm": 0.7694595435443164, "learning_rate": 9.447182760659306e-06, "loss": 0.0742, "step": 3808 }, { "epoch": 0.17699814126394053, "grad_norm": 0.9736181801349162, "learning_rate": 9.446838767558429e-06, "loss": 0.0762, "step": 3809 }, { "epoch": 0.1770446096654275, "grad_norm": 1.251506264596705, "learning_rate": 9.446494673731973e-06, "loss": 0.0728, "step": 3810 }, { "epoch": 0.1770910780669145, "grad_norm": 1.6519626638595213, "learning_rate": 9.446150479187734e-06, "loss": 0.0538, "step": 3811 }, { "epoch": 0.1771375464684015, "grad_norm": 1.4976988837234257, "learning_rate": 9.445806183933506e-06, "loss": 0.0859, "step": 3812 }, { "epoch": 0.17718401486988847, "grad_norm": 0.9031672949706221, "learning_rate": 9.445461787977088e-06, "loss": 0.0642, "step": 3813 }, { "epoch": 0.17723048327137547, "grad_norm": 1.2123649662970521, "learning_rate": 9.445117291326283e-06, "loss": 0.0632, "step": 3814 }, { "epoch": 0.17727695167286245, "grad_norm": 1.6050789310407847, "learning_rate": 9.444772693988893e-06, "loss": 0.1024, "step": 3815 }, { "epoch": 0.17732342007434945, "grad_norm": 2.4259499424845643, "learning_rate": 9.444427995972723e-06, "loss": 0.0875, "step": 3816 }, { "epoch": 0.17736988847583643, "grad_norm": 4.839679511882178, "learning_rate": 9.444083197285581e-06, "loss": 0.1333, "step": 3817 }, { "epoch": 0.17741635687732343, "grad_norm": 2.251486968769517, "learning_rate": 9.443738297935279e-06, "loss": 0.069, "step": 3818 }, { "epoch": 0.1774628252788104, "grad_norm": 0.6413247220309964, "learning_rate": 9.443393297929627e-06, "loss": 0.0338, "step": 3819 }, { "epoch": 0.1775092936802974, "grad_norm": 1.0696317523822023, "learning_rate": 9.443048197276439e-06, "loss": 0.0754, "step": 3820 }, { "epoch": 0.1775557620817844, "grad_norm": 2.4712125125159976, "learning_rate": 9.442702995983533e-06, "loss": 0.0703, "step": 3821 }, { "epoch": 0.17760223048327137, "grad_norm": 5.506085945694551, "learning_rate": 9.44235769405873e-06, "loss": 0.1507, "step": 3822 }, { "epoch": 0.17764869888475837, "grad_norm": 1.9667366804420239, "learning_rate": 9.442012291509848e-06, "loss": 0.0919, "step": 3823 }, { "epoch": 0.17769516728624535, "grad_norm": 3.1264820680786367, "learning_rate": 9.441666788344713e-06, "loss": 0.1171, "step": 3824 }, { "epoch": 0.17774163568773235, "grad_norm": 1.6742308607171106, "learning_rate": 9.44132118457115e-06, "loss": 0.0733, "step": 3825 }, { "epoch": 0.17778810408921933, "grad_norm": 3.876549827492081, "learning_rate": 9.44097548019699e-06, "loss": 0.0791, "step": 3826 }, { "epoch": 0.17783457249070633, "grad_norm": 1.5219986739489866, "learning_rate": 9.440629675230058e-06, "loss": 0.0674, "step": 3827 }, { "epoch": 0.1778810408921933, "grad_norm": 3.441996534647065, "learning_rate": 9.440283769678192e-06, "loss": 0.0795, "step": 3828 }, { "epoch": 0.17792750929368029, "grad_norm": 1.7703916393935912, "learning_rate": 9.439937763549225e-06, "loss": 0.0585, "step": 3829 }, { "epoch": 0.1779739776951673, "grad_norm": 1.8378065948600955, "learning_rate": 9.439591656850995e-06, "loss": 0.0958, "step": 3830 }, { "epoch": 0.17802044609665427, "grad_norm": 1.4915048907018966, "learning_rate": 9.439245449591342e-06, "loss": 0.1094, "step": 3831 }, { "epoch": 0.17806691449814127, "grad_norm": 3.2436783777943567, "learning_rate": 9.438899141778105e-06, "loss": 0.186, "step": 3832 }, { "epoch": 0.17811338289962825, "grad_norm": 1.8185944235740585, "learning_rate": 9.438552733419132e-06, "loss": 0.1117, "step": 3833 }, { "epoch": 0.17815985130111525, "grad_norm": 1.7325756893744664, "learning_rate": 9.43820622452227e-06, "loss": 0.0999, "step": 3834 }, { "epoch": 0.17820631970260223, "grad_norm": 2.190152931911186, "learning_rate": 9.437859615095365e-06, "loss": 0.1201, "step": 3835 }, { "epoch": 0.17825278810408923, "grad_norm": 0.7523811206707087, "learning_rate": 9.437512905146267e-06, "loss": 0.0652, "step": 3836 }, { "epoch": 0.1782992565055762, "grad_norm": 0.8868467554976639, "learning_rate": 9.437166094682832e-06, "loss": 0.0707, "step": 3837 }, { "epoch": 0.17834572490706319, "grad_norm": 0.9133482781080089, "learning_rate": 9.436819183712916e-06, "loss": 0.0563, "step": 3838 }, { "epoch": 0.1783921933085502, "grad_norm": 1.4034388444944084, "learning_rate": 9.436472172244374e-06, "loss": 0.0613, "step": 3839 }, { "epoch": 0.17843866171003717, "grad_norm": 2.1377410280235027, "learning_rate": 9.436125060285069e-06, "loss": 0.0765, "step": 3840 }, { "epoch": 0.17848513011152417, "grad_norm": 3.4747696466317906, "learning_rate": 9.435777847842862e-06, "loss": 0.147, "step": 3841 }, { "epoch": 0.17853159851301115, "grad_norm": 1.6525992779969767, "learning_rate": 9.435430534925621e-06, "loss": 0.0929, "step": 3842 }, { "epoch": 0.17857806691449815, "grad_norm": 1.5213246820551407, "learning_rate": 9.435083121541207e-06, "loss": 0.1018, "step": 3843 }, { "epoch": 0.17862453531598513, "grad_norm": 2.3175605719151577, "learning_rate": 9.434735607697493e-06, "loss": 0.1418, "step": 3844 }, { "epoch": 0.17867100371747213, "grad_norm": 1.3656231529785936, "learning_rate": 9.434387993402351e-06, "loss": 0.0702, "step": 3845 }, { "epoch": 0.1787174721189591, "grad_norm": 2.2710980371382328, "learning_rate": 9.434040278663652e-06, "loss": 0.1079, "step": 3846 }, { "epoch": 0.17876394052044609, "grad_norm": 1.8593768446179229, "learning_rate": 9.433692463489277e-06, "loss": 0.1291, "step": 3847 }, { "epoch": 0.1788104089219331, "grad_norm": 0.8029417049280673, "learning_rate": 9.433344547887097e-06, "loss": 0.0553, "step": 3848 }, { "epoch": 0.17885687732342007, "grad_norm": 1.073477512915749, "learning_rate": 9.432996531865001e-06, "loss": 0.0808, "step": 3849 }, { "epoch": 0.17890334572490707, "grad_norm": 1.3442414175087023, "learning_rate": 9.432648415430868e-06, "loss": 0.0623, "step": 3850 }, { "epoch": 0.17894981412639405, "grad_norm": 1.6064330681001464, "learning_rate": 9.43230019859258e-06, "loss": 0.0948, "step": 3851 }, { "epoch": 0.17899628252788105, "grad_norm": 2.3543768074107443, "learning_rate": 9.431951881358029e-06, "loss": 0.1198, "step": 3852 }, { "epoch": 0.17904275092936803, "grad_norm": 2.157549424667796, "learning_rate": 9.431603463735103e-06, "loss": 0.0949, "step": 3853 }, { "epoch": 0.17908921933085503, "grad_norm": 1.4243240768086818, "learning_rate": 9.431254945731696e-06, "loss": 0.0738, "step": 3854 }, { "epoch": 0.179135687732342, "grad_norm": 2.7187316350040356, "learning_rate": 9.4309063273557e-06, "loss": 0.1428, "step": 3855 }, { "epoch": 0.17918215613382898, "grad_norm": 1.5810737271843853, "learning_rate": 9.43055760861501e-06, "loss": 0.0798, "step": 3856 }, { "epoch": 0.179228624535316, "grad_norm": 1.2768637044684066, "learning_rate": 9.43020878951753e-06, "loss": 0.1249, "step": 3857 }, { "epoch": 0.17927509293680297, "grad_norm": 1.8487724389032032, "learning_rate": 9.429859870071156e-06, "loss": 0.0813, "step": 3858 }, { "epoch": 0.17932156133828997, "grad_norm": 1.5138929552250513, "learning_rate": 9.429510850283794e-06, "loss": 0.1117, "step": 3859 }, { "epoch": 0.17936802973977695, "grad_norm": 1.4883675534538408, "learning_rate": 9.429161730163352e-06, "loss": 0.0938, "step": 3860 }, { "epoch": 0.17941449814126395, "grad_norm": 1.8364828986870863, "learning_rate": 9.428812509717732e-06, "loss": 0.0872, "step": 3861 }, { "epoch": 0.17946096654275093, "grad_norm": 2.60778788769935, "learning_rate": 9.428463188954848e-06, "loss": 0.1191, "step": 3862 }, { "epoch": 0.17950743494423793, "grad_norm": 3.2589410419900067, "learning_rate": 9.428113767882613e-06, "loss": 0.0749, "step": 3863 }, { "epoch": 0.1795539033457249, "grad_norm": 1.3800259706104983, "learning_rate": 9.42776424650894e-06, "loss": 0.0628, "step": 3864 }, { "epoch": 0.17960037174721188, "grad_norm": 0.9532165995511271, "learning_rate": 9.427414624841747e-06, "loss": 0.0651, "step": 3865 }, { "epoch": 0.1796468401486989, "grad_norm": 1.3317663746109154, "learning_rate": 9.427064902888952e-06, "loss": 0.0911, "step": 3866 }, { "epoch": 0.17969330855018587, "grad_norm": 1.8004258144166376, "learning_rate": 9.426715080658479e-06, "loss": 0.1028, "step": 3867 }, { "epoch": 0.17973977695167287, "grad_norm": 1.9672366458773232, "learning_rate": 9.426365158158249e-06, "loss": 0.0949, "step": 3868 }, { "epoch": 0.17978624535315985, "grad_norm": 2.1345898867689903, "learning_rate": 9.42601513539619e-06, "loss": 0.0756, "step": 3869 }, { "epoch": 0.17983271375464685, "grad_norm": 2.2509182965207852, "learning_rate": 9.425665012380231e-06, "loss": 0.134, "step": 3870 }, { "epoch": 0.17987918215613383, "grad_norm": 0.9635403830225902, "learning_rate": 9.4253147891183e-06, "loss": 0.0503, "step": 3871 }, { "epoch": 0.17992565055762083, "grad_norm": 1.0418412459561304, "learning_rate": 9.424964465618333e-06, "loss": 0.1255, "step": 3872 }, { "epoch": 0.1799721189591078, "grad_norm": 1.4128346928354352, "learning_rate": 9.424614041888262e-06, "loss": 0.0698, "step": 3873 }, { "epoch": 0.18001858736059478, "grad_norm": 1.5383379227261398, "learning_rate": 9.424263517936028e-06, "loss": 0.0905, "step": 3874 }, { "epoch": 0.1800650557620818, "grad_norm": 1.2361618320580017, "learning_rate": 9.423912893769566e-06, "loss": 0.1082, "step": 3875 }, { "epoch": 0.18011152416356876, "grad_norm": 1.582006398701244, "learning_rate": 9.423562169396823e-06, "loss": 0.1119, "step": 3876 }, { "epoch": 0.18015799256505577, "grad_norm": 1.264604345600858, "learning_rate": 9.42321134482574e-06, "loss": 0.091, "step": 3877 }, { "epoch": 0.18020446096654275, "grad_norm": 1.436622345595359, "learning_rate": 9.422860420064265e-06, "loss": 0.0865, "step": 3878 }, { "epoch": 0.18025092936802975, "grad_norm": 4.388374504071646, "learning_rate": 9.422509395120348e-06, "loss": 0.0664, "step": 3879 }, { "epoch": 0.18029739776951673, "grad_norm": 2.5794332571870457, "learning_rate": 9.422158270001938e-06, "loss": 0.1275, "step": 3880 }, { "epoch": 0.18034386617100373, "grad_norm": 1.803342853992001, "learning_rate": 9.421807044716988e-06, "loss": 0.0702, "step": 3881 }, { "epoch": 0.1803903345724907, "grad_norm": 1.993788032654306, "learning_rate": 9.421455719273456e-06, "loss": 0.1042, "step": 3882 }, { "epoch": 0.18043680297397768, "grad_norm": 2.662129216042366, "learning_rate": 9.421104293679297e-06, "loss": 0.1129, "step": 3883 }, { "epoch": 0.1804832713754647, "grad_norm": 1.917874459113562, "learning_rate": 9.420752767942474e-06, "loss": 0.1633, "step": 3884 }, { "epoch": 0.18052973977695166, "grad_norm": 2.0492544897947322, "learning_rate": 9.420401142070946e-06, "loss": 0.1304, "step": 3885 }, { "epoch": 0.18057620817843867, "grad_norm": 2.4825693812219587, "learning_rate": 9.42004941607268e-06, "loss": 0.0959, "step": 3886 }, { "epoch": 0.18062267657992565, "grad_norm": 3.1901634314875045, "learning_rate": 9.419697589955645e-06, "loss": 0.0849, "step": 3887 }, { "epoch": 0.18066914498141265, "grad_norm": 2.953077741902349, "learning_rate": 9.419345663727805e-06, "loss": 0.1035, "step": 3888 }, { "epoch": 0.18071561338289963, "grad_norm": 2.737197193162187, "learning_rate": 9.418993637397138e-06, "loss": 0.0954, "step": 3889 }, { "epoch": 0.18076208178438663, "grad_norm": 0.8840035765677459, "learning_rate": 9.418641510971611e-06, "loss": 0.0824, "step": 3890 }, { "epoch": 0.1808085501858736, "grad_norm": 1.295653142265333, "learning_rate": 9.418289284459206e-06, "loss": 0.0858, "step": 3891 }, { "epoch": 0.18085501858736058, "grad_norm": 1.0011450420864805, "learning_rate": 9.417936957867896e-06, "loss": 0.069, "step": 3892 }, { "epoch": 0.1809014869888476, "grad_norm": 1.4500288818179807, "learning_rate": 9.417584531205667e-06, "loss": 0.0723, "step": 3893 }, { "epoch": 0.18094795539033456, "grad_norm": 0.6440579824922781, "learning_rate": 9.417232004480495e-06, "loss": 0.0513, "step": 3894 }, { "epoch": 0.18099442379182157, "grad_norm": 1.833548439716933, "learning_rate": 9.416879377700372e-06, "loss": 0.1415, "step": 3895 }, { "epoch": 0.18104089219330854, "grad_norm": 1.643242521466945, "learning_rate": 9.416526650873281e-06, "loss": 0.1101, "step": 3896 }, { "epoch": 0.18108736059479555, "grad_norm": 1.4961065814215584, "learning_rate": 9.416173824007213e-06, "loss": 0.1121, "step": 3897 }, { "epoch": 0.18113382899628253, "grad_norm": 2.0254206936809362, "learning_rate": 9.415820897110162e-06, "loss": 0.0825, "step": 3898 }, { "epoch": 0.18118029739776953, "grad_norm": 1.8430861287899118, "learning_rate": 9.415467870190118e-06, "loss": 0.0992, "step": 3899 }, { "epoch": 0.1812267657992565, "grad_norm": 1.142736722160695, "learning_rate": 9.415114743255081e-06, "loss": 0.0763, "step": 3900 }, { "epoch": 0.18127323420074348, "grad_norm": 2.001481152842702, "learning_rate": 9.414761516313046e-06, "loss": 0.0701, "step": 3901 }, { "epoch": 0.1813197026022305, "grad_norm": 1.907155401780785, "learning_rate": 9.414408189372018e-06, "loss": 0.1413, "step": 3902 }, { "epoch": 0.18136617100371746, "grad_norm": 1.726483508358083, "learning_rate": 9.414054762439998e-06, "loss": 0.0875, "step": 3903 }, { "epoch": 0.18141263940520447, "grad_norm": 1.7477047772405854, "learning_rate": 9.413701235524992e-06, "loss": 0.0836, "step": 3904 }, { "epoch": 0.18145910780669144, "grad_norm": 1.1483974237367787, "learning_rate": 9.413347608635008e-06, "loss": 0.0819, "step": 3905 }, { "epoch": 0.18150557620817845, "grad_norm": 2.066237404676065, "learning_rate": 9.412993881778054e-06, "loss": 0.1077, "step": 3906 }, { "epoch": 0.18155204460966543, "grad_norm": 1.3255334866783068, "learning_rate": 9.412640054962145e-06, "loss": 0.0779, "step": 3907 }, { "epoch": 0.18159851301115243, "grad_norm": 1.74248361163186, "learning_rate": 9.412286128195295e-06, "loss": 0.1009, "step": 3908 }, { "epoch": 0.1816449814126394, "grad_norm": 1.783623882679184, "learning_rate": 9.411932101485521e-06, "loss": 0.0796, "step": 3909 }, { "epoch": 0.18169144981412638, "grad_norm": 2.336537619827512, "learning_rate": 9.411577974840841e-06, "loss": 0.0834, "step": 3910 }, { "epoch": 0.1817379182156134, "grad_norm": 1.8710240889380914, "learning_rate": 9.411223748269275e-06, "loss": 0.0949, "step": 3911 }, { "epoch": 0.18178438661710036, "grad_norm": 4.50065494744586, "learning_rate": 9.410869421778852e-06, "loss": 0.1097, "step": 3912 }, { "epoch": 0.18183085501858737, "grad_norm": 3.2074659075159078, "learning_rate": 9.41051499537759e-06, "loss": 0.0858, "step": 3913 }, { "epoch": 0.18187732342007434, "grad_norm": 0.903876598602897, "learning_rate": 9.410160469073523e-06, "loss": 0.0577, "step": 3914 }, { "epoch": 0.18192379182156135, "grad_norm": 2.204633440286822, "learning_rate": 9.409805842874683e-06, "loss": 0.1106, "step": 3915 }, { "epoch": 0.18197026022304832, "grad_norm": 1.9467001597982267, "learning_rate": 9.409451116789094e-06, "loss": 0.1223, "step": 3916 }, { "epoch": 0.18201672862453533, "grad_norm": 1.4237886986039991, "learning_rate": 9.4090962908248e-06, "loss": 0.0758, "step": 3917 }, { "epoch": 0.1820631970260223, "grad_norm": 1.9846143631003756, "learning_rate": 9.408741364989833e-06, "loss": 0.0636, "step": 3918 }, { "epoch": 0.18210966542750928, "grad_norm": 1.5108700255156506, "learning_rate": 9.408386339292235e-06, "loss": 0.0942, "step": 3919 }, { "epoch": 0.1821561338289963, "grad_norm": 1.2194056460139149, "learning_rate": 9.408031213740045e-06, "loss": 0.0864, "step": 3920 }, { "epoch": 0.18220260223048326, "grad_norm": 1.4659921381165917, "learning_rate": 9.407675988341312e-06, "loss": 0.0725, "step": 3921 }, { "epoch": 0.18224907063197027, "grad_norm": 2.968995831334652, "learning_rate": 9.407320663104076e-06, "loss": 0.066, "step": 3922 }, { "epoch": 0.18229553903345724, "grad_norm": 1.2956326154087883, "learning_rate": 9.406965238036388e-06, "loss": 0.0481, "step": 3923 }, { "epoch": 0.18234200743494425, "grad_norm": 1.4215434424588544, "learning_rate": 9.4066097131463e-06, "loss": 0.1408, "step": 3924 }, { "epoch": 0.18238847583643122, "grad_norm": 1.9158912880065402, "learning_rate": 9.406254088441864e-06, "loss": 0.1155, "step": 3925 }, { "epoch": 0.18243494423791823, "grad_norm": 1.7571480253547411, "learning_rate": 9.405898363931137e-06, "loss": 0.0984, "step": 3926 }, { "epoch": 0.1824814126394052, "grad_norm": 1.5486197890908926, "learning_rate": 9.405542539622173e-06, "loss": 0.0952, "step": 3927 }, { "epoch": 0.18252788104089218, "grad_norm": 1.5694291406038776, "learning_rate": 9.405186615523035e-06, "loss": 0.1158, "step": 3928 }, { "epoch": 0.18257434944237919, "grad_norm": 1.1843150608700947, "learning_rate": 9.404830591641782e-06, "loss": 0.0648, "step": 3929 }, { "epoch": 0.18262081784386616, "grad_norm": 1.9727571206373558, "learning_rate": 9.404474467986483e-06, "loss": 0.1212, "step": 3930 }, { "epoch": 0.18266728624535317, "grad_norm": 0.9352866643523373, "learning_rate": 9.404118244565198e-06, "loss": 0.0573, "step": 3931 }, { "epoch": 0.18271375464684014, "grad_norm": 1.4838690048211514, "learning_rate": 9.403761921386e-06, "loss": 0.0912, "step": 3932 }, { "epoch": 0.18276022304832715, "grad_norm": 1.8676140815896158, "learning_rate": 9.403405498456962e-06, "loss": 0.0712, "step": 3933 }, { "epoch": 0.18280669144981412, "grad_norm": 1.121522326641278, "learning_rate": 9.403048975786154e-06, "loss": 0.0748, "step": 3934 }, { "epoch": 0.18285315985130113, "grad_norm": 1.042438970989464, "learning_rate": 9.402692353381653e-06, "loss": 0.0687, "step": 3935 }, { "epoch": 0.1828996282527881, "grad_norm": 0.9015383385795132, "learning_rate": 9.402335631251537e-06, "loss": 0.0694, "step": 3936 }, { "epoch": 0.18294609665427508, "grad_norm": 2.4001530671557427, "learning_rate": 9.401978809403882e-06, "loss": 0.1111, "step": 3937 }, { "epoch": 0.18299256505576209, "grad_norm": 2.122702728994406, "learning_rate": 9.401621887846777e-06, "loss": 0.0985, "step": 3938 }, { "epoch": 0.18303903345724906, "grad_norm": 2.755542416995443, "learning_rate": 9.401264866588303e-06, "loss": 0.1189, "step": 3939 }, { "epoch": 0.18308550185873607, "grad_norm": 0.9800809958748475, "learning_rate": 9.400907745636548e-06, "loss": 0.0641, "step": 3940 }, { "epoch": 0.18313197026022304, "grad_norm": 1.2986772511545894, "learning_rate": 9.4005505249996e-06, "loss": 0.073, "step": 3941 }, { "epoch": 0.18317843866171005, "grad_norm": 1.7758444781074967, "learning_rate": 9.400193204685552e-06, "loss": 0.0963, "step": 3942 }, { "epoch": 0.18322490706319702, "grad_norm": 1.6032435088033574, "learning_rate": 9.399835784702496e-06, "loss": 0.0778, "step": 3943 }, { "epoch": 0.18327137546468403, "grad_norm": 1.163316307731088, "learning_rate": 9.399478265058529e-06, "loss": 0.0397, "step": 3944 }, { "epoch": 0.183317843866171, "grad_norm": 1.3800525703047175, "learning_rate": 9.399120645761748e-06, "loss": 0.0914, "step": 3945 }, { "epoch": 0.18336431226765798, "grad_norm": 1.1958716234117792, "learning_rate": 9.398762926820256e-06, "loss": 0.0957, "step": 3946 }, { "epoch": 0.18341078066914498, "grad_norm": 0.9909681113416666, "learning_rate": 9.398405108242154e-06, "loss": 0.067, "step": 3947 }, { "epoch": 0.18345724907063196, "grad_norm": 2.3211114238625856, "learning_rate": 9.398047190035546e-06, "loss": 0.0577, "step": 3948 }, { "epoch": 0.18350371747211897, "grad_norm": 1.095673940926227, "learning_rate": 9.39768917220854e-06, "loss": 0.0715, "step": 3949 }, { "epoch": 0.18355018587360594, "grad_norm": 2.1955424338144534, "learning_rate": 9.397331054769247e-06, "loss": 0.0677, "step": 3950 }, { "epoch": 0.18359665427509295, "grad_norm": 1.8079769549032965, "learning_rate": 9.396972837725778e-06, "loss": 0.0829, "step": 3951 }, { "epoch": 0.18364312267657992, "grad_norm": 2.5045467937389394, "learning_rate": 9.396614521086245e-06, "loss": 0.1236, "step": 3952 }, { "epoch": 0.18368959107806693, "grad_norm": 1.6395845098687754, "learning_rate": 9.396256104858765e-06, "loss": 0.1242, "step": 3953 }, { "epoch": 0.1837360594795539, "grad_norm": 1.2363664907652816, "learning_rate": 9.39589758905146e-06, "loss": 0.0687, "step": 3954 }, { "epoch": 0.18378252788104088, "grad_norm": 1.0808234155616225, "learning_rate": 9.395538973672444e-06, "loss": 0.0538, "step": 3955 }, { "epoch": 0.18382899628252788, "grad_norm": 2.005379882110554, "learning_rate": 9.395180258729847e-06, "loss": 0.1041, "step": 3956 }, { "epoch": 0.18387546468401486, "grad_norm": 2.0048817524119533, "learning_rate": 9.394821444231792e-06, "loss": 0.0865, "step": 3957 }, { "epoch": 0.18392193308550187, "grad_norm": 1.5621823779846669, "learning_rate": 9.394462530186403e-06, "loss": 0.0943, "step": 3958 }, { "epoch": 0.18396840148698884, "grad_norm": 1.5196829076154494, "learning_rate": 9.394103516601815e-06, "loss": 0.0764, "step": 3959 }, { "epoch": 0.18401486988847585, "grad_norm": 0.916953754842361, "learning_rate": 9.393744403486157e-06, "loss": 0.0585, "step": 3960 }, { "epoch": 0.18406133828996282, "grad_norm": 1.5536502982812188, "learning_rate": 9.393385190847564e-06, "loss": 0.059, "step": 3961 }, { "epoch": 0.18410780669144983, "grad_norm": 2.4034017364458706, "learning_rate": 9.393025878694173e-06, "loss": 0.143, "step": 3962 }, { "epoch": 0.1841542750929368, "grad_norm": 1.338871521210979, "learning_rate": 9.39266646703412e-06, "loss": 0.0966, "step": 3963 }, { "epoch": 0.18420074349442378, "grad_norm": 2.2458902037264825, "learning_rate": 9.392306955875549e-06, "loss": 0.0699, "step": 3964 }, { "epoch": 0.18424721189591078, "grad_norm": 1.2636084524882496, "learning_rate": 9.391947345226605e-06, "loss": 0.0887, "step": 3965 }, { "epoch": 0.18429368029739776, "grad_norm": 2.088861195239055, "learning_rate": 9.391587635095429e-06, "loss": 0.0718, "step": 3966 }, { "epoch": 0.18434014869888476, "grad_norm": 1.7335638253484287, "learning_rate": 9.391227825490172e-06, "loss": 0.1185, "step": 3967 }, { "epoch": 0.18438661710037174, "grad_norm": 0.845787950382627, "learning_rate": 9.390867916418983e-06, "loss": 0.0869, "step": 3968 }, { "epoch": 0.18443308550185875, "grad_norm": 1.2774406195432255, "learning_rate": 9.390507907890015e-06, "loss": 0.0884, "step": 3969 }, { "epoch": 0.18447955390334572, "grad_norm": 0.9300790734547952, "learning_rate": 9.39014779991142e-06, "loss": 0.063, "step": 3970 }, { "epoch": 0.18452602230483273, "grad_norm": 2.2618388836034797, "learning_rate": 9.389787592491358e-06, "loss": 0.1576, "step": 3971 }, { "epoch": 0.1845724907063197, "grad_norm": 1.5829125131051898, "learning_rate": 9.389427285637986e-06, "loss": 0.1155, "step": 3972 }, { "epoch": 0.18461895910780668, "grad_norm": 1.5811817037244877, "learning_rate": 9.389066879359467e-06, "loss": 0.0682, "step": 3973 }, { "epoch": 0.18466542750929368, "grad_norm": 0.8877415077124957, "learning_rate": 9.388706373663965e-06, "loss": 0.065, "step": 3974 }, { "epoch": 0.18471189591078066, "grad_norm": 1.3171074614565255, "learning_rate": 9.388345768559644e-06, "loss": 0.0872, "step": 3975 }, { "epoch": 0.18475836431226766, "grad_norm": 1.0220072302531522, "learning_rate": 9.38798506405467e-06, "loss": 0.0713, "step": 3976 }, { "epoch": 0.18480483271375464, "grad_norm": 1.5291218515183362, "learning_rate": 9.387624260157219e-06, "loss": 0.0955, "step": 3977 }, { "epoch": 0.18485130111524165, "grad_norm": 1.4897326887538336, "learning_rate": 9.38726335687546e-06, "loss": 0.0768, "step": 3978 }, { "epoch": 0.18489776951672862, "grad_norm": 2.4531983561447395, "learning_rate": 9.386902354217566e-06, "loss": 0.1017, "step": 3979 }, { "epoch": 0.18494423791821563, "grad_norm": 2.0927801098500134, "learning_rate": 9.38654125219172e-06, "loss": 0.0833, "step": 3980 }, { "epoch": 0.1849907063197026, "grad_norm": 1.1813301352360976, "learning_rate": 9.386180050806097e-06, "loss": 0.0529, "step": 3981 }, { "epoch": 0.18503717472118958, "grad_norm": 1.1022764668382479, "learning_rate": 9.385818750068878e-06, "loss": 0.0659, "step": 3982 }, { "epoch": 0.18508364312267658, "grad_norm": 3.846497063360379, "learning_rate": 9.38545734998825e-06, "loss": 0.1679, "step": 3983 }, { "epoch": 0.18513011152416356, "grad_norm": 1.9616521329308176, "learning_rate": 9.385095850572397e-06, "loss": 0.0667, "step": 3984 }, { "epoch": 0.18517657992565056, "grad_norm": 1.0977725401146303, "learning_rate": 9.384734251829507e-06, "loss": 0.0711, "step": 3985 }, { "epoch": 0.18522304832713754, "grad_norm": 1.7633617751078174, "learning_rate": 9.38437255376777e-06, "loss": 0.1106, "step": 3986 }, { "epoch": 0.18526951672862454, "grad_norm": 1.133562229349542, "learning_rate": 9.384010756395385e-06, "loss": 0.0702, "step": 3987 }, { "epoch": 0.18531598513011152, "grad_norm": 1.1110956809422288, "learning_rate": 9.383648859720538e-06, "loss": 0.0646, "step": 3988 }, { "epoch": 0.18536245353159853, "grad_norm": 1.9831504148392558, "learning_rate": 9.383286863751433e-06, "loss": 0.1185, "step": 3989 }, { "epoch": 0.1854089219330855, "grad_norm": 2.0508481604249584, "learning_rate": 9.382924768496264e-06, "loss": 0.15, "step": 3990 }, { "epoch": 0.18545539033457248, "grad_norm": 2.0448035044562696, "learning_rate": 9.382562573963238e-06, "loss": 0.075, "step": 3991 }, { "epoch": 0.18550185873605948, "grad_norm": 1.6927147523086397, "learning_rate": 9.382200280160558e-06, "loss": 0.1554, "step": 3992 }, { "epoch": 0.18554832713754646, "grad_norm": 0.6481668915709726, "learning_rate": 9.38183788709643e-06, "loss": 0.0527, "step": 3993 }, { "epoch": 0.18559479553903346, "grad_norm": 1.341397596402195, "learning_rate": 9.38147539477906e-06, "loss": 0.0774, "step": 3994 }, { "epoch": 0.18564126394052044, "grad_norm": 1.7647065256992762, "learning_rate": 9.381112803216662e-06, "loss": 0.0785, "step": 3995 }, { "epoch": 0.18568773234200744, "grad_norm": 1.7443070888480532, "learning_rate": 9.380750112417447e-06, "loss": 0.0556, "step": 3996 }, { "epoch": 0.18573420074349442, "grad_norm": 0.8397967575142249, "learning_rate": 9.380387322389632e-06, "loss": 0.0781, "step": 3997 }, { "epoch": 0.18578066914498143, "grad_norm": 0.8953811922438328, "learning_rate": 9.380024433141435e-06, "loss": 0.0563, "step": 3998 }, { "epoch": 0.1858271375464684, "grad_norm": 1.1499429803855419, "learning_rate": 9.379661444681073e-06, "loss": 0.0591, "step": 3999 }, { "epoch": 0.18587360594795538, "grad_norm": 0.7944966149840775, "learning_rate": 9.37929835701677e-06, "loss": 0.0795, "step": 4000 }, { "epoch": 0.18592007434944238, "grad_norm": 1.633958449900572, "learning_rate": 9.37893517015675e-06, "loss": 0.0791, "step": 4001 }, { "epoch": 0.18596654275092936, "grad_norm": 1.0589761855498374, "learning_rate": 9.37857188410924e-06, "loss": 0.0808, "step": 4002 }, { "epoch": 0.18601301115241636, "grad_norm": 1.5458054028634354, "learning_rate": 9.37820849888247e-06, "loss": 0.114, "step": 4003 }, { "epoch": 0.18605947955390334, "grad_norm": 1.103423504251821, "learning_rate": 9.377845014484668e-06, "loss": 0.111, "step": 4004 }, { "epoch": 0.18610594795539034, "grad_norm": 0.9891681203420443, "learning_rate": 9.377481430924069e-06, "loss": 0.0545, "step": 4005 }, { "epoch": 0.18615241635687732, "grad_norm": 1.9654960864505286, "learning_rate": 9.377117748208909e-06, "loss": 0.1142, "step": 4006 }, { "epoch": 0.18619888475836432, "grad_norm": 1.3333770488444157, "learning_rate": 9.376753966347422e-06, "loss": 0.0803, "step": 4007 }, { "epoch": 0.1862453531598513, "grad_norm": 2.2989989518264196, "learning_rate": 9.376390085347854e-06, "loss": 0.0642, "step": 4008 }, { "epoch": 0.18629182156133828, "grad_norm": 1.3505404139779076, "learning_rate": 9.376026105218445e-06, "loss": 0.0731, "step": 4009 }, { "epoch": 0.18633828996282528, "grad_norm": 1.0085120112781172, "learning_rate": 9.375662025967438e-06, "loss": 0.1029, "step": 4010 }, { "epoch": 0.18638475836431226, "grad_norm": 2.495740319059485, "learning_rate": 9.375297847603082e-06, "loss": 0.1317, "step": 4011 }, { "epoch": 0.18643122676579926, "grad_norm": 1.5919548690793313, "learning_rate": 9.374933570133622e-06, "loss": 0.069, "step": 4012 }, { "epoch": 0.18647769516728624, "grad_norm": 1.378398304205718, "learning_rate": 9.374569193567315e-06, "loss": 0.1094, "step": 4013 }, { "epoch": 0.18652416356877324, "grad_norm": 1.7299634649986824, "learning_rate": 9.37420471791241e-06, "loss": 0.0476, "step": 4014 }, { "epoch": 0.18657063197026022, "grad_norm": 0.9632790757368043, "learning_rate": 9.373840143177165e-06, "loss": 0.055, "step": 4015 }, { "epoch": 0.18661710037174722, "grad_norm": 0.8789749587775683, "learning_rate": 9.373475469369837e-06, "loss": 0.0688, "step": 4016 }, { "epoch": 0.1866635687732342, "grad_norm": 0.7530377918361557, "learning_rate": 9.373110696498688e-06, "loss": 0.04, "step": 4017 }, { "epoch": 0.18671003717472118, "grad_norm": 1.3798677887737316, "learning_rate": 9.372745824571979e-06, "loss": 0.0871, "step": 4018 }, { "epoch": 0.18675650557620818, "grad_norm": 1.496597415921779, "learning_rate": 9.372380853597973e-06, "loss": 0.0754, "step": 4019 }, { "epoch": 0.18680297397769516, "grad_norm": 1.5387449892363756, "learning_rate": 9.372015783584938e-06, "loss": 0.1037, "step": 4020 }, { "epoch": 0.18684944237918216, "grad_norm": 0.9912580167216997, "learning_rate": 9.371650614541147e-06, "loss": 0.0581, "step": 4021 }, { "epoch": 0.18689591078066914, "grad_norm": 1.6162014560087488, "learning_rate": 9.371285346474868e-06, "loss": 0.1145, "step": 4022 }, { "epoch": 0.18694237918215614, "grad_norm": 2.634078523334295, "learning_rate": 9.370919979394374e-06, "loss": 0.1538, "step": 4023 }, { "epoch": 0.18698884758364312, "grad_norm": 0.7001086496896958, "learning_rate": 9.370554513307942e-06, "loss": 0.0371, "step": 4024 }, { "epoch": 0.18703531598513012, "grad_norm": 1.7199440021887549, "learning_rate": 9.370188948223851e-06, "loss": 0.0844, "step": 4025 }, { "epoch": 0.1870817843866171, "grad_norm": 3.2376693609302674, "learning_rate": 9.369823284150381e-06, "loss": 0.0844, "step": 4026 }, { "epoch": 0.18712825278810408, "grad_norm": 1.5098084894516883, "learning_rate": 9.369457521095814e-06, "loss": 0.111, "step": 4027 }, { "epoch": 0.18717472118959108, "grad_norm": 0.8777423145295387, "learning_rate": 9.369091659068434e-06, "loss": 0.0662, "step": 4028 }, { "epoch": 0.18722118959107806, "grad_norm": 2.5538615636392015, "learning_rate": 9.368725698076532e-06, "loss": 0.0524, "step": 4029 }, { "epoch": 0.18726765799256506, "grad_norm": 1.0695489611296338, "learning_rate": 9.368359638128394e-06, "loss": 0.0712, "step": 4030 }, { "epoch": 0.18731412639405204, "grad_norm": 1.3497085305739613, "learning_rate": 9.367993479232314e-06, "loss": 0.0748, "step": 4031 }, { "epoch": 0.18736059479553904, "grad_norm": 1.0136236327111474, "learning_rate": 9.367627221396583e-06, "loss": 0.0627, "step": 4032 }, { "epoch": 0.18740706319702602, "grad_norm": 8.861707811025978, "learning_rate": 9.367260864629498e-06, "loss": 0.0881, "step": 4033 }, { "epoch": 0.18745353159851302, "grad_norm": 1.2227596636736127, "learning_rate": 9.366894408939357e-06, "loss": 0.1015, "step": 4034 }, { "epoch": 0.1875, "grad_norm": 1.1319927675974284, "learning_rate": 9.366527854334464e-06, "loss": 0.0736, "step": 4035 }, { "epoch": 0.18754646840148698, "grad_norm": 0.9834501730854874, "learning_rate": 9.366161200823118e-06, "loss": 0.0612, "step": 4036 }, { "epoch": 0.18759293680297398, "grad_norm": 1.7868990961776363, "learning_rate": 9.365794448413628e-06, "loss": 0.1101, "step": 4037 }, { "epoch": 0.18763940520446096, "grad_norm": 1.74717039415251, "learning_rate": 9.365427597114296e-06, "loss": 0.0846, "step": 4038 }, { "epoch": 0.18768587360594796, "grad_norm": 3.3628439813948203, "learning_rate": 9.365060646933435e-06, "loss": 0.1109, "step": 4039 }, { "epoch": 0.18773234200743494, "grad_norm": 1.2486731703465326, "learning_rate": 9.364693597879357e-06, "loss": 0.0719, "step": 4040 }, { "epoch": 0.18777881040892194, "grad_norm": 1.25222140198088, "learning_rate": 9.364326449960375e-06, "loss": 0.0652, "step": 4041 }, { "epoch": 0.18782527881040892, "grad_norm": 3.9426916582787674, "learning_rate": 9.363959203184805e-06, "loss": 0.0822, "step": 4042 }, { "epoch": 0.18787174721189592, "grad_norm": 0.8387765869106345, "learning_rate": 9.363591857560967e-06, "loss": 0.0637, "step": 4043 }, { "epoch": 0.1879182156133829, "grad_norm": 1.0931431941582246, "learning_rate": 9.36322441309718e-06, "loss": 0.0653, "step": 4044 }, { "epoch": 0.18796468401486988, "grad_norm": 2.5313199756486537, "learning_rate": 9.362856869801768e-06, "loss": 0.1354, "step": 4045 }, { "epoch": 0.18801115241635688, "grad_norm": 2.10127535875277, "learning_rate": 9.362489227683056e-06, "loss": 0.0621, "step": 4046 }, { "epoch": 0.18805762081784386, "grad_norm": 1.7072117708058707, "learning_rate": 9.36212148674937e-06, "loss": 0.0827, "step": 4047 }, { "epoch": 0.18810408921933086, "grad_norm": 2.298025153988118, "learning_rate": 9.361753647009045e-06, "loss": 0.12, "step": 4048 }, { "epoch": 0.18815055762081784, "grad_norm": 4.708721245669059, "learning_rate": 9.361385708470406e-06, "loss": 0.0861, "step": 4049 }, { "epoch": 0.18819702602230484, "grad_norm": 0.9119900192251887, "learning_rate": 9.361017671141793e-06, "loss": 0.0659, "step": 4050 }, { "epoch": 0.18824349442379182, "grad_norm": 1.2465489907533487, "learning_rate": 9.360649535031538e-06, "loss": 0.0748, "step": 4051 }, { "epoch": 0.18828996282527882, "grad_norm": 1.1120517290958916, "learning_rate": 9.360281300147982e-06, "loss": 0.0743, "step": 4052 }, { "epoch": 0.1883364312267658, "grad_norm": 1.5342893908800102, "learning_rate": 9.359912966499465e-06, "loss": 0.0934, "step": 4053 }, { "epoch": 0.18838289962825278, "grad_norm": 1.066981753089455, "learning_rate": 9.35954453409433e-06, "loss": 0.076, "step": 4054 }, { "epoch": 0.18842936802973978, "grad_norm": 1.9085545617181976, "learning_rate": 9.359176002940925e-06, "loss": 0.0958, "step": 4055 }, { "epoch": 0.18847583643122676, "grad_norm": 2.0282457475106033, "learning_rate": 9.358807373047594e-06, "loss": 0.0582, "step": 4056 }, { "epoch": 0.18852230483271376, "grad_norm": 2.078786672965653, "learning_rate": 9.358438644422689e-06, "loss": 0.1259, "step": 4057 }, { "epoch": 0.18856877323420074, "grad_norm": 1.1630877358456575, "learning_rate": 9.358069817074562e-06, "loss": 0.0433, "step": 4058 }, { "epoch": 0.18861524163568774, "grad_norm": 2.219348276557146, "learning_rate": 9.357700891011566e-06, "loss": 0.1039, "step": 4059 }, { "epoch": 0.18866171003717472, "grad_norm": 1.5740635035894426, "learning_rate": 9.35733186624206e-06, "loss": 0.0773, "step": 4060 }, { "epoch": 0.18870817843866172, "grad_norm": 0.8579319698179391, "learning_rate": 9.356962742774399e-06, "loss": 0.0631, "step": 4061 }, { "epoch": 0.1887546468401487, "grad_norm": 2.3927575788503983, "learning_rate": 9.356593520616948e-06, "loss": 0.0845, "step": 4062 }, { "epoch": 0.18880111524163568, "grad_norm": 1.1825493822837652, "learning_rate": 9.356224199778068e-06, "loss": 0.0841, "step": 4063 }, { "epoch": 0.18884758364312268, "grad_norm": 1.5862806919479169, "learning_rate": 9.355854780266125e-06, "loss": 0.0949, "step": 4064 }, { "epoch": 0.18889405204460966, "grad_norm": 1.671484378940521, "learning_rate": 9.355485262089486e-06, "loss": 0.111, "step": 4065 }, { "epoch": 0.18894052044609666, "grad_norm": 1.1117425263155165, "learning_rate": 9.355115645256523e-06, "loss": 0.0599, "step": 4066 }, { "epoch": 0.18898698884758364, "grad_norm": 16.147686824852787, "learning_rate": 9.354745929775607e-06, "loss": 0.1086, "step": 4067 }, { "epoch": 0.18903345724907064, "grad_norm": 1.3945938416370078, "learning_rate": 9.354376115655111e-06, "loss": 0.0699, "step": 4068 }, { "epoch": 0.18907992565055762, "grad_norm": 1.7508401658309438, "learning_rate": 9.354006202903412e-06, "loss": 0.0969, "step": 4069 }, { "epoch": 0.18912639405204462, "grad_norm": 1.3415461431159406, "learning_rate": 9.353636191528893e-06, "loss": 0.0932, "step": 4070 }, { "epoch": 0.1891728624535316, "grad_norm": 2.5420964191616697, "learning_rate": 9.353266081539931e-06, "loss": 0.0717, "step": 4071 }, { "epoch": 0.18921933085501857, "grad_norm": 0.7898896733821469, "learning_rate": 9.352895872944911e-06, "loss": 0.0626, "step": 4072 }, { "epoch": 0.18926579925650558, "grad_norm": 2.155648879178465, "learning_rate": 9.352525565752217e-06, "loss": 0.0712, "step": 4073 }, { "epoch": 0.18931226765799256, "grad_norm": 3.3231290853295277, "learning_rate": 9.35215515997024e-06, "loss": 0.1086, "step": 4074 }, { "epoch": 0.18935873605947956, "grad_norm": 0.9986078030449388, "learning_rate": 9.351784655607366e-06, "loss": 0.0589, "step": 4075 }, { "epoch": 0.18940520446096654, "grad_norm": 2.526789945751638, "learning_rate": 9.35141405267199e-06, "loss": 0.154, "step": 4076 }, { "epoch": 0.18945167286245354, "grad_norm": 1.56698464946822, "learning_rate": 9.351043351172504e-06, "loss": 0.057, "step": 4077 }, { "epoch": 0.18949814126394052, "grad_norm": 5.281440962754179, "learning_rate": 9.35067255111731e-06, "loss": 0.0697, "step": 4078 }, { "epoch": 0.18954460966542752, "grad_norm": 1.2752701301649387, "learning_rate": 9.350301652514801e-06, "loss": 0.0494, "step": 4079 }, { "epoch": 0.1895910780669145, "grad_norm": 1.4765664181682558, "learning_rate": 9.349930655373383e-06, "loss": 0.0857, "step": 4080 }, { "epoch": 0.18963754646840147, "grad_norm": 0.963052520731429, "learning_rate": 9.349559559701457e-06, "loss": 0.0584, "step": 4081 }, { "epoch": 0.18968401486988848, "grad_norm": 2.3358016792368734, "learning_rate": 9.34918836550743e-06, "loss": 0.1027, "step": 4082 }, { "epoch": 0.18973048327137546, "grad_norm": 1.246722345695111, "learning_rate": 9.348817072799708e-06, "loss": 0.0844, "step": 4083 }, { "epoch": 0.18977695167286246, "grad_norm": 3.162255595372546, "learning_rate": 9.348445681586703e-06, "loss": 0.0774, "step": 4084 }, { "epoch": 0.18982342007434944, "grad_norm": 1.4254906180392701, "learning_rate": 9.348074191876825e-06, "loss": 0.0735, "step": 4085 }, { "epoch": 0.18986988847583644, "grad_norm": 2.1915641036997138, "learning_rate": 9.34770260367849e-06, "loss": 0.0914, "step": 4086 }, { "epoch": 0.18991635687732342, "grad_norm": 5.474013943730849, "learning_rate": 9.347330917000117e-06, "loss": 0.1067, "step": 4087 }, { "epoch": 0.18996282527881042, "grad_norm": 0.7388010963986323, "learning_rate": 9.346959131850123e-06, "loss": 0.0588, "step": 4088 }, { "epoch": 0.1900092936802974, "grad_norm": 0.8958653880540687, "learning_rate": 9.34658724823693e-06, "loss": 0.064, "step": 4089 }, { "epoch": 0.19005576208178437, "grad_norm": 3.3949897820871375, "learning_rate": 9.346215266168958e-06, "loss": 0.1154, "step": 4090 }, { "epoch": 0.19010223048327138, "grad_norm": 4.31461018627652, "learning_rate": 9.34584318565464e-06, "loss": 0.189, "step": 4091 }, { "epoch": 0.19014869888475835, "grad_norm": 1.8817322903294278, "learning_rate": 9.345471006702398e-06, "loss": 0.1145, "step": 4092 }, { "epoch": 0.19019516728624536, "grad_norm": 2.3133784412309084, "learning_rate": 9.345098729320664e-06, "loss": 0.1226, "step": 4093 }, { "epoch": 0.19024163568773234, "grad_norm": 1.4469966727408536, "learning_rate": 9.34472635351787e-06, "loss": 0.0954, "step": 4094 }, { "epoch": 0.19028810408921934, "grad_norm": 1.6619949738856465, "learning_rate": 9.344353879302454e-06, "loss": 0.1069, "step": 4095 }, { "epoch": 0.19033457249070632, "grad_norm": 3.140621995799914, "learning_rate": 9.343981306682848e-06, "loss": 0.094, "step": 4096 }, { "epoch": 0.19038104089219332, "grad_norm": 0.7788239238767941, "learning_rate": 9.343608635667493e-06, "loss": 0.0568, "step": 4097 }, { "epoch": 0.1904275092936803, "grad_norm": 1.4810326146499635, "learning_rate": 9.343235866264833e-06, "loss": 0.119, "step": 4098 }, { "epoch": 0.19047397769516727, "grad_norm": 4.304022989965541, "learning_rate": 9.342862998483308e-06, "loss": 0.1741, "step": 4099 }, { "epoch": 0.19052044609665428, "grad_norm": 1.1030773721180365, "learning_rate": 9.342490032331366e-06, "loss": 0.0697, "step": 4100 }, { "epoch": 0.19056691449814125, "grad_norm": 4.0597225338930185, "learning_rate": 9.342116967817454e-06, "loss": 0.1536, "step": 4101 }, { "epoch": 0.19061338289962826, "grad_norm": 2.1501389379465046, "learning_rate": 9.341743804950023e-06, "loss": 0.1355, "step": 4102 }, { "epoch": 0.19065985130111524, "grad_norm": 1.1785605884879364, "learning_rate": 9.341370543737525e-06, "loss": 0.0801, "step": 4103 }, { "epoch": 0.19070631970260224, "grad_norm": 1.580674521510797, "learning_rate": 9.340997184188413e-06, "loss": 0.1217, "step": 4104 }, { "epoch": 0.19075278810408922, "grad_norm": 3.50317706107466, "learning_rate": 9.34062372631115e-06, "loss": 0.118, "step": 4105 }, { "epoch": 0.19079925650557622, "grad_norm": 1.1788486034713497, "learning_rate": 9.34025017011419e-06, "loss": 0.0695, "step": 4106 }, { "epoch": 0.1908457249070632, "grad_norm": 1.865978589589338, "learning_rate": 9.339876515605993e-06, "loss": 0.1186, "step": 4107 }, { "epoch": 0.19089219330855017, "grad_norm": 1.2469176292963176, "learning_rate": 9.339502762795027e-06, "loss": 0.1001, "step": 4108 }, { "epoch": 0.19093866171003718, "grad_norm": 2.1043306870175074, "learning_rate": 9.339128911689757e-06, "loss": 0.0889, "step": 4109 }, { "epoch": 0.19098513011152415, "grad_norm": 2.5764279383663378, "learning_rate": 9.338754962298648e-06, "loss": 0.1681, "step": 4110 }, { "epoch": 0.19103159851301116, "grad_norm": 1.2912261527681541, "learning_rate": 9.338380914630175e-06, "loss": 0.1028, "step": 4111 }, { "epoch": 0.19107806691449813, "grad_norm": 1.5242876138367167, "learning_rate": 9.338006768692807e-06, "loss": 0.1168, "step": 4112 }, { "epoch": 0.19112453531598514, "grad_norm": 2.20098501864203, "learning_rate": 9.337632524495021e-06, "loss": 0.0694, "step": 4113 }, { "epoch": 0.19117100371747212, "grad_norm": 2.0691965324143258, "learning_rate": 9.337258182045293e-06, "loss": 0.0833, "step": 4114 }, { "epoch": 0.19121747211895912, "grad_norm": 1.796814936952983, "learning_rate": 9.336883741352101e-06, "loss": 0.1535, "step": 4115 }, { "epoch": 0.1912639405204461, "grad_norm": 1.1584268569424299, "learning_rate": 9.336509202423928e-06, "loss": 0.0658, "step": 4116 }, { "epoch": 0.19131040892193307, "grad_norm": 0.9825076447868819, "learning_rate": 9.336134565269258e-06, "loss": 0.0675, "step": 4117 }, { "epoch": 0.19135687732342008, "grad_norm": 1.400283245675585, "learning_rate": 9.335759829896575e-06, "loss": 0.0532, "step": 4118 }, { "epoch": 0.19140334572490705, "grad_norm": 3.203236977477495, "learning_rate": 9.335384996314371e-06, "loss": 0.0551, "step": 4119 }, { "epoch": 0.19144981412639406, "grad_norm": 0.8410282331122696, "learning_rate": 9.335010064531133e-06, "loss": 0.0968, "step": 4120 }, { "epoch": 0.19149628252788103, "grad_norm": 1.32549326666176, "learning_rate": 9.334635034555356e-06, "loss": 0.0859, "step": 4121 }, { "epoch": 0.19154275092936804, "grad_norm": 2.5033315714561697, "learning_rate": 9.33425990639553e-06, "loss": 0.077, "step": 4122 }, { "epoch": 0.19158921933085502, "grad_norm": 1.5071319920078334, "learning_rate": 9.333884680060156e-06, "loss": 0.0735, "step": 4123 }, { "epoch": 0.19163568773234202, "grad_norm": 0.9129425614545054, "learning_rate": 9.333509355557735e-06, "loss": 0.0607, "step": 4124 }, { "epoch": 0.191682156133829, "grad_norm": 1.4925539693562009, "learning_rate": 9.333133932896764e-06, "loss": 0.0941, "step": 4125 }, { "epoch": 0.19172862453531597, "grad_norm": 4.648390172981674, "learning_rate": 9.332758412085749e-06, "loss": 0.0975, "step": 4126 }, { "epoch": 0.19177509293680298, "grad_norm": 1.4406353361147832, "learning_rate": 9.332382793133198e-06, "loss": 0.0678, "step": 4127 }, { "epoch": 0.19182156133828995, "grad_norm": 1.324015053082608, "learning_rate": 9.332007076047616e-06, "loss": 0.0742, "step": 4128 }, { "epoch": 0.19186802973977696, "grad_norm": 2.4051267236107106, "learning_rate": 9.331631260837513e-06, "loss": 0.1245, "step": 4129 }, { "epoch": 0.19191449814126393, "grad_norm": 1.0710433164994124, "learning_rate": 9.331255347511403e-06, "loss": 0.101, "step": 4130 }, { "epoch": 0.19196096654275094, "grad_norm": 2.2848580231589577, "learning_rate": 9.330879336077803e-06, "loss": 0.0902, "step": 4131 }, { "epoch": 0.19200743494423791, "grad_norm": 0.8284213064353648, "learning_rate": 9.330503226545226e-06, "loss": 0.0653, "step": 4132 }, { "epoch": 0.19205390334572492, "grad_norm": 1.333542260544083, "learning_rate": 9.330127018922195e-06, "loss": 0.0605, "step": 4133 }, { "epoch": 0.1921003717472119, "grad_norm": 1.5118292728615599, "learning_rate": 9.329750713217227e-06, "loss": 0.0387, "step": 4134 }, { "epoch": 0.19214684014869887, "grad_norm": 1.4636876847526152, "learning_rate": 9.32937430943885e-06, "loss": 0.0608, "step": 4135 }, { "epoch": 0.19219330855018588, "grad_norm": 1.8121007138221006, "learning_rate": 9.328997807595588e-06, "loss": 0.0931, "step": 4136 }, { "epoch": 0.19223977695167285, "grad_norm": 2.4543674496966417, "learning_rate": 9.32862120769597e-06, "loss": 0.0873, "step": 4137 }, { "epoch": 0.19228624535315986, "grad_norm": 1.2383845581058246, "learning_rate": 9.328244509748526e-06, "loss": 0.0811, "step": 4138 }, { "epoch": 0.19233271375464683, "grad_norm": 1.6987409678314613, "learning_rate": 9.327867713761787e-06, "loss": 0.0863, "step": 4139 }, { "epoch": 0.19237918215613384, "grad_norm": 1.3360376097653954, "learning_rate": 9.32749081974429e-06, "loss": 0.0999, "step": 4140 }, { "epoch": 0.19242565055762081, "grad_norm": 1.1665959342903978, "learning_rate": 9.327113827704572e-06, "loss": 0.0839, "step": 4141 }, { "epoch": 0.19247211895910782, "grad_norm": 2.3448240621943137, "learning_rate": 9.326736737651172e-06, "loss": 0.1135, "step": 4142 }, { "epoch": 0.1925185873605948, "grad_norm": 0.7961095051952936, "learning_rate": 9.326359549592629e-06, "loss": 0.0486, "step": 4143 }, { "epoch": 0.19256505576208177, "grad_norm": 1.7816986463512872, "learning_rate": 9.325982263537489e-06, "loss": 0.1071, "step": 4144 }, { "epoch": 0.19261152416356878, "grad_norm": 2.570774655589997, "learning_rate": 9.3256048794943e-06, "loss": 0.0866, "step": 4145 }, { "epoch": 0.19265799256505575, "grad_norm": 1.4422334269075343, "learning_rate": 9.325227397471605e-06, "loss": 0.0924, "step": 4146 }, { "epoch": 0.19270446096654276, "grad_norm": 0.7038198446290148, "learning_rate": 9.324849817477958e-06, "loss": 0.0417, "step": 4147 }, { "epoch": 0.19275092936802973, "grad_norm": 1.6630963676876303, "learning_rate": 9.324472139521912e-06, "loss": 0.0843, "step": 4148 }, { "epoch": 0.19279739776951674, "grad_norm": 1.4516343455297993, "learning_rate": 9.324094363612021e-06, "loss": 0.0668, "step": 4149 }, { "epoch": 0.1928438661710037, "grad_norm": 1.0812640893062087, "learning_rate": 9.323716489756841e-06, "loss": 0.0726, "step": 4150 }, { "epoch": 0.19289033457249072, "grad_norm": 4.269285476382698, "learning_rate": 9.323338517964931e-06, "loss": 0.113, "step": 4151 }, { "epoch": 0.1929368029739777, "grad_norm": 1.7848498980583343, "learning_rate": 9.322960448244856e-06, "loss": 0.0854, "step": 4152 }, { "epoch": 0.19298327137546467, "grad_norm": 2.2767310924245945, "learning_rate": 9.322582280605175e-06, "loss": 0.0878, "step": 4153 }, { "epoch": 0.19302973977695168, "grad_norm": 1.2204339054186857, "learning_rate": 9.322204015054455e-06, "loss": 0.0801, "step": 4154 }, { "epoch": 0.19307620817843865, "grad_norm": 1.5503997568853591, "learning_rate": 9.321825651601267e-06, "loss": 0.0643, "step": 4155 }, { "epoch": 0.19312267657992566, "grad_norm": 3.0248735332067063, "learning_rate": 9.321447190254176e-06, "loss": 0.1525, "step": 4156 }, { "epoch": 0.19316914498141263, "grad_norm": 0.9626240355494148, "learning_rate": 9.321068631021762e-06, "loss": 0.0394, "step": 4157 }, { "epoch": 0.19321561338289964, "grad_norm": 3.132180868312904, "learning_rate": 9.320689973912593e-06, "loss": 0.1106, "step": 4158 }, { "epoch": 0.1932620817843866, "grad_norm": 2.085402777588455, "learning_rate": 9.32031121893525e-06, "loss": 0.0778, "step": 4159 }, { "epoch": 0.19330855018587362, "grad_norm": 1.9943557683325595, "learning_rate": 9.31993236609831e-06, "loss": 0.0885, "step": 4160 }, { "epoch": 0.1933550185873606, "grad_norm": 0.886311332708066, "learning_rate": 9.319553415410356e-06, "loss": 0.0581, "step": 4161 }, { "epoch": 0.19340148698884757, "grad_norm": 0.7845695988788399, "learning_rate": 9.31917436687997e-06, "loss": 0.0537, "step": 4162 }, { "epoch": 0.19344795539033457, "grad_norm": 1.0513886997844004, "learning_rate": 9.31879522051574e-06, "loss": 0.0855, "step": 4163 }, { "epoch": 0.19349442379182155, "grad_norm": 1.8847784346789227, "learning_rate": 9.318415976326252e-06, "loss": 0.0854, "step": 4164 }, { "epoch": 0.19354089219330856, "grad_norm": 1.0846305609718014, "learning_rate": 9.318036634320097e-06, "loss": 0.0665, "step": 4165 }, { "epoch": 0.19358736059479553, "grad_norm": 0.9016472824750621, "learning_rate": 9.317657194505868e-06, "loss": 0.0621, "step": 4166 }, { "epoch": 0.19363382899628254, "grad_norm": 1.8013891008244383, "learning_rate": 9.317277656892159e-06, "loss": 0.06, "step": 4167 }, { "epoch": 0.1936802973977695, "grad_norm": 1.2239141464918573, "learning_rate": 9.316898021487566e-06, "loss": 0.0545, "step": 4168 }, { "epoch": 0.19372676579925652, "grad_norm": 2.3977602813886323, "learning_rate": 9.316518288300692e-06, "loss": 0.1297, "step": 4169 }, { "epoch": 0.1937732342007435, "grad_norm": 1.5650812780294638, "learning_rate": 9.316138457340133e-06, "loss": 0.0728, "step": 4170 }, { "epoch": 0.19381970260223047, "grad_norm": 1.6793012258765354, "learning_rate": 9.315758528614499e-06, "loss": 0.1057, "step": 4171 }, { "epoch": 0.19386617100371747, "grad_norm": 1.223992527449361, "learning_rate": 9.315378502132389e-06, "loss": 0.0738, "step": 4172 }, { "epoch": 0.19391263940520445, "grad_norm": 2.3342830678616657, "learning_rate": 9.314998377902414e-06, "loss": 0.1346, "step": 4173 }, { "epoch": 0.19395910780669146, "grad_norm": 2.7097349945957667, "learning_rate": 9.314618155933183e-06, "loss": 0.0984, "step": 4174 }, { "epoch": 0.19400557620817843, "grad_norm": 2.566797910132611, "learning_rate": 9.314237836233314e-06, "loss": 0.1028, "step": 4175 }, { "epoch": 0.19405204460966544, "grad_norm": 1.0670372293333572, "learning_rate": 9.313857418811415e-06, "loss": 0.045, "step": 4176 }, { "epoch": 0.1940985130111524, "grad_norm": 0.5989530941235558, "learning_rate": 9.313476903676105e-06, "loss": 0.0514, "step": 4177 }, { "epoch": 0.19414498141263942, "grad_norm": 1.6003466668038382, "learning_rate": 9.313096290836003e-06, "loss": 0.1018, "step": 4178 }, { "epoch": 0.1941914498141264, "grad_norm": 1.572486333525513, "learning_rate": 9.312715580299732e-06, "loss": 0.0751, "step": 4179 }, { "epoch": 0.19423791821561337, "grad_norm": 2.166185025967915, "learning_rate": 9.312334772075912e-06, "loss": 0.0944, "step": 4180 }, { "epoch": 0.19428438661710037, "grad_norm": 1.0658963163746906, "learning_rate": 9.311953866173173e-06, "loss": 0.0725, "step": 4181 }, { "epoch": 0.19433085501858735, "grad_norm": 1.7841408047715845, "learning_rate": 9.31157286260014e-06, "loss": 0.0729, "step": 4182 }, { "epoch": 0.19437732342007435, "grad_norm": 1.8681049483209333, "learning_rate": 9.311191761365443e-06, "loss": 0.062, "step": 4183 }, { "epoch": 0.19442379182156133, "grad_norm": 1.1329456305653522, "learning_rate": 9.310810562477715e-06, "loss": 0.0691, "step": 4184 }, { "epoch": 0.19447026022304834, "grad_norm": 1.3437055020982587, "learning_rate": 9.310429265945593e-06, "loss": 0.0999, "step": 4185 }, { "epoch": 0.1945167286245353, "grad_norm": 1.4924015270448632, "learning_rate": 9.31004787177771e-06, "loss": 0.1018, "step": 4186 }, { "epoch": 0.19456319702602232, "grad_norm": 1.225217650675498, "learning_rate": 9.309666379982704e-06, "loss": 0.0766, "step": 4187 }, { "epoch": 0.1946096654275093, "grad_norm": 1.1030994533404255, "learning_rate": 9.309284790569221e-06, "loss": 0.0711, "step": 4188 }, { "epoch": 0.19465613382899627, "grad_norm": 0.7920040282470213, "learning_rate": 9.308903103545903e-06, "loss": 0.0698, "step": 4189 }, { "epoch": 0.19470260223048327, "grad_norm": 1.9234088809544707, "learning_rate": 9.308521318921394e-06, "loss": 0.1066, "step": 4190 }, { "epoch": 0.19474907063197025, "grad_norm": 3.5292689036879947, "learning_rate": 9.308139436704343e-06, "loss": 0.0651, "step": 4191 }, { "epoch": 0.19479553903345725, "grad_norm": 1.9577163761207261, "learning_rate": 9.307757456903399e-06, "loss": 0.0779, "step": 4192 }, { "epoch": 0.19484200743494423, "grad_norm": 1.1872293193637535, "learning_rate": 9.307375379527213e-06, "loss": 0.1125, "step": 4193 }, { "epoch": 0.19488847583643124, "grad_norm": 2.9382178395158185, "learning_rate": 9.306993204584442e-06, "loss": 0.0859, "step": 4194 }, { "epoch": 0.1949349442379182, "grad_norm": 1.6583074101257593, "learning_rate": 9.306610932083745e-06, "loss": 0.0768, "step": 4195 }, { "epoch": 0.19498141263940522, "grad_norm": 1.8251714646817692, "learning_rate": 9.306228562033775e-06, "loss": 0.1357, "step": 4196 }, { "epoch": 0.1950278810408922, "grad_norm": 1.3278186443868423, "learning_rate": 9.305846094443195e-06, "loss": 0.0975, "step": 4197 }, { "epoch": 0.19507434944237917, "grad_norm": 2.587561788729838, "learning_rate": 9.30546352932067e-06, "loss": 0.086, "step": 4198 }, { "epoch": 0.19512081784386617, "grad_norm": 1.7549581941523056, "learning_rate": 9.305080866674864e-06, "loss": 0.0899, "step": 4199 }, { "epoch": 0.19516728624535315, "grad_norm": 2.1601524600652358, "learning_rate": 9.304698106514448e-06, "loss": 0.0646, "step": 4200 }, { "epoch": 0.19521375464684015, "grad_norm": 3.1603223241349, "learning_rate": 9.304315248848088e-06, "loss": 0.0781, "step": 4201 }, { "epoch": 0.19526022304832713, "grad_norm": 1.2985348786962412, "learning_rate": 9.303932293684456e-06, "loss": 0.0842, "step": 4202 }, { "epoch": 0.19530669144981413, "grad_norm": 0.8113548252164247, "learning_rate": 9.30354924103223e-06, "loss": 0.0445, "step": 4203 }, { "epoch": 0.1953531598513011, "grad_norm": 1.0872783406669788, "learning_rate": 9.303166090900082e-06, "loss": 0.0541, "step": 4204 }, { "epoch": 0.19539962825278812, "grad_norm": 2.3991940906977747, "learning_rate": 9.302782843296694e-06, "loss": 0.069, "step": 4205 }, { "epoch": 0.1954460966542751, "grad_norm": 0.9779624524986281, "learning_rate": 9.302399498230746e-06, "loss": 0.0841, "step": 4206 }, { "epoch": 0.19549256505576207, "grad_norm": 0.8753254109749918, "learning_rate": 9.302016055710921e-06, "loss": 0.0416, "step": 4207 }, { "epoch": 0.19553903345724907, "grad_norm": 1.9140760778743247, "learning_rate": 9.301632515745905e-06, "loss": 0.1033, "step": 4208 }, { "epoch": 0.19558550185873605, "grad_norm": 2.467639069343262, "learning_rate": 9.301248878344384e-06, "loss": 0.0671, "step": 4209 }, { "epoch": 0.19563197026022305, "grad_norm": 1.5930624705864034, "learning_rate": 9.30086514351505e-06, "loss": 0.0947, "step": 4210 }, { "epoch": 0.19567843866171003, "grad_norm": 1.25011596045454, "learning_rate": 9.300481311266595e-06, "loss": 0.0457, "step": 4211 }, { "epoch": 0.19572490706319703, "grad_norm": 1.5370584465170176, "learning_rate": 9.30009738160771e-06, "loss": 0.0843, "step": 4212 }, { "epoch": 0.195771375464684, "grad_norm": 1.6076474818221487, "learning_rate": 9.299713354547096e-06, "loss": 0.0871, "step": 4213 }, { "epoch": 0.19581784386617102, "grad_norm": 1.3281088288713199, "learning_rate": 9.299329230093448e-06, "loss": 0.0616, "step": 4214 }, { "epoch": 0.195864312267658, "grad_norm": 1.1067977108980054, "learning_rate": 9.298945008255466e-06, "loss": 0.0667, "step": 4215 }, { "epoch": 0.19591078066914497, "grad_norm": 1.215571191097484, "learning_rate": 9.298560689041856e-06, "loss": 0.1381, "step": 4216 }, { "epoch": 0.19595724907063197, "grad_norm": 2.256853579076975, "learning_rate": 9.298176272461323e-06, "loss": 0.0724, "step": 4217 }, { "epoch": 0.19600371747211895, "grad_norm": 1.3460808731544225, "learning_rate": 9.297791758522575e-06, "loss": 0.0842, "step": 4218 }, { "epoch": 0.19605018587360595, "grad_norm": 1.395856494910829, "learning_rate": 9.297407147234319e-06, "loss": 0.0664, "step": 4219 }, { "epoch": 0.19609665427509293, "grad_norm": 1.8895987986469611, "learning_rate": 9.297022438605267e-06, "loss": 0.0608, "step": 4220 }, { "epoch": 0.19614312267657993, "grad_norm": 1.5570855898498461, "learning_rate": 9.296637632644134e-06, "loss": 0.0733, "step": 4221 }, { "epoch": 0.1961895910780669, "grad_norm": 1.7332615900675117, "learning_rate": 9.296252729359639e-06, "loss": 0.1107, "step": 4222 }, { "epoch": 0.19623605947955391, "grad_norm": 1.5855227758134436, "learning_rate": 9.295867728760497e-06, "loss": 0.0979, "step": 4223 }, { "epoch": 0.1962825278810409, "grad_norm": 1.7153371618295574, "learning_rate": 9.295482630855428e-06, "loss": 0.0691, "step": 4224 }, { "epoch": 0.19632899628252787, "grad_norm": 1.519304784371144, "learning_rate": 9.295097435653158e-06, "loss": 0.0911, "step": 4225 }, { "epoch": 0.19637546468401487, "grad_norm": 1.08884665047614, "learning_rate": 9.29471214316241e-06, "loss": 0.0607, "step": 4226 }, { "epoch": 0.19642193308550185, "grad_norm": 2.8528808014876232, "learning_rate": 9.294326753391913e-06, "loss": 0.0825, "step": 4227 }, { "epoch": 0.19646840148698885, "grad_norm": 1.112605368793047, "learning_rate": 9.293941266350394e-06, "loss": 0.0846, "step": 4228 }, { "epoch": 0.19651486988847583, "grad_norm": 1.5297494323799607, "learning_rate": 9.293555682046586e-06, "loss": 0.0978, "step": 4229 }, { "epoch": 0.19656133828996283, "grad_norm": 1.1216726440575386, "learning_rate": 9.293170000489225e-06, "loss": 0.0727, "step": 4230 }, { "epoch": 0.1966078066914498, "grad_norm": 1.486344577702708, "learning_rate": 9.292784221687043e-06, "loss": 0.0803, "step": 4231 }, { "epoch": 0.19665427509293681, "grad_norm": 0.9472343719264222, "learning_rate": 9.292398345648781e-06, "loss": 0.098, "step": 4232 }, { "epoch": 0.1967007434944238, "grad_norm": 1.3589304246975258, "learning_rate": 9.29201237238318e-06, "loss": 0.077, "step": 4233 }, { "epoch": 0.19674721189591077, "grad_norm": 3.7946904250483255, "learning_rate": 9.291626301898979e-06, "loss": 0.1207, "step": 4234 }, { "epoch": 0.19679368029739777, "grad_norm": 1.3096774302437393, "learning_rate": 9.291240134204926e-06, "loss": 0.0638, "step": 4235 }, { "epoch": 0.19684014869888475, "grad_norm": 1.8175766625100185, "learning_rate": 9.290853869309769e-06, "loss": 0.0874, "step": 4236 }, { "epoch": 0.19688661710037175, "grad_norm": 1.172446671020836, "learning_rate": 9.290467507222253e-06, "loss": 0.0556, "step": 4237 }, { "epoch": 0.19693308550185873, "grad_norm": 1.237427223313386, "learning_rate": 9.290081047951136e-06, "loss": 0.0864, "step": 4238 }, { "epoch": 0.19697955390334573, "grad_norm": 1.5125766962927563, "learning_rate": 9.289694491505166e-06, "loss": 0.0973, "step": 4239 }, { "epoch": 0.1970260223048327, "grad_norm": 1.1815276102200645, "learning_rate": 9.2893078378931e-06, "loss": 0.1122, "step": 4240 }, { "epoch": 0.19707249070631971, "grad_norm": 1.203525696976311, "learning_rate": 9.288921087123699e-06, "loss": 0.0638, "step": 4241 }, { "epoch": 0.1971189591078067, "grad_norm": 1.5592327757231388, "learning_rate": 9.28853423920572e-06, "loss": 0.0876, "step": 4242 }, { "epoch": 0.19716542750929367, "grad_norm": 1.5036714599397194, "learning_rate": 9.288147294147927e-06, "loss": 0.111, "step": 4243 }, { "epoch": 0.19721189591078067, "grad_norm": 1.146823324322295, "learning_rate": 9.287760251959086e-06, "loss": 0.054, "step": 4244 }, { "epoch": 0.19725836431226765, "grad_norm": 1.3569298752303007, "learning_rate": 9.287373112647962e-06, "loss": 0.1527, "step": 4245 }, { "epoch": 0.19730483271375465, "grad_norm": 1.4469104908312582, "learning_rate": 9.286985876223324e-06, "loss": 0.1016, "step": 4246 }, { "epoch": 0.19735130111524163, "grad_norm": 1.3052766532946376, "learning_rate": 9.286598542693943e-06, "loss": 0.0926, "step": 4247 }, { "epoch": 0.19739776951672863, "grad_norm": 2.0741514615300756, "learning_rate": 9.286211112068594e-06, "loss": 0.0978, "step": 4248 }, { "epoch": 0.1974442379182156, "grad_norm": 1.778403902578982, "learning_rate": 9.285823584356055e-06, "loss": 0.1247, "step": 4249 }, { "epoch": 0.1974907063197026, "grad_norm": 1.9526470893980805, "learning_rate": 9.2854359595651e-06, "loss": 0.1524, "step": 4250 }, { "epoch": 0.1975371747211896, "grad_norm": 1.2144797035813715, "learning_rate": 9.285048237704507e-06, "loss": 0.08, "step": 4251 }, { "epoch": 0.19758364312267657, "grad_norm": 1.9049300664422684, "learning_rate": 9.284660418783064e-06, "loss": 0.0807, "step": 4252 }, { "epoch": 0.19763011152416357, "grad_norm": 0.8750512009421688, "learning_rate": 9.284272502809552e-06, "loss": 0.062, "step": 4253 }, { "epoch": 0.19767657992565055, "grad_norm": 2.032704099285163, "learning_rate": 9.283884489792759e-06, "loss": 0.082, "step": 4254 }, { "epoch": 0.19772304832713755, "grad_norm": 0.8072945556764374, "learning_rate": 9.283496379741475e-06, "loss": 0.061, "step": 4255 }, { "epoch": 0.19776951672862453, "grad_norm": 1.7076537465255024, "learning_rate": 9.283108172664488e-06, "loss": 0.079, "step": 4256 }, { "epoch": 0.19781598513011153, "grad_norm": 1.2297155230788959, "learning_rate": 9.282719868570593e-06, "loss": 0.0793, "step": 4257 }, { "epoch": 0.1978624535315985, "grad_norm": 2.017380924547, "learning_rate": 9.282331467468585e-06, "loss": 0.1302, "step": 4258 }, { "epoch": 0.1979089219330855, "grad_norm": 1.060881831418178, "learning_rate": 9.281942969367263e-06, "loss": 0.0956, "step": 4259 }, { "epoch": 0.1979553903345725, "grad_norm": 0.9980608832620168, "learning_rate": 9.281554374275428e-06, "loss": 0.0843, "step": 4260 }, { "epoch": 0.19800185873605947, "grad_norm": 0.4949616310329698, "learning_rate": 9.281165682201878e-06, "loss": 0.035, "step": 4261 }, { "epoch": 0.19804832713754647, "grad_norm": 0.6260523618406172, "learning_rate": 9.280776893155418e-06, "loss": 0.038, "step": 4262 }, { "epoch": 0.19809479553903345, "grad_norm": 0.8274696842065343, "learning_rate": 9.280388007144859e-06, "loss": 0.0627, "step": 4263 }, { "epoch": 0.19814126394052045, "grad_norm": 1.250998156333739, "learning_rate": 9.279999024179004e-06, "loss": 0.074, "step": 4264 }, { "epoch": 0.19818773234200743, "grad_norm": 0.856990213187616, "learning_rate": 9.27960994426667e-06, "loss": 0.058, "step": 4265 }, { "epoch": 0.19823420074349443, "grad_norm": 0.7468510761318126, "learning_rate": 9.279220767416664e-06, "loss": 0.0574, "step": 4266 }, { "epoch": 0.1982806691449814, "grad_norm": 0.7895833709331102, "learning_rate": 9.278831493637805e-06, "loss": 0.0706, "step": 4267 }, { "epoch": 0.1983271375464684, "grad_norm": 4.19546170488973, "learning_rate": 9.278442122938908e-06, "loss": 0.1057, "step": 4268 }, { "epoch": 0.1983736059479554, "grad_norm": 1.7080908784849642, "learning_rate": 9.278052655328794e-06, "loss": 0.0737, "step": 4269 }, { "epoch": 0.19842007434944237, "grad_norm": 0.8527289438769955, "learning_rate": 9.277663090816284e-06, "loss": 0.0623, "step": 4270 }, { "epoch": 0.19846654275092937, "grad_norm": 2.139906613474113, "learning_rate": 9.277273429410205e-06, "loss": 0.0869, "step": 4271 }, { "epoch": 0.19851301115241635, "grad_norm": 0.6625173584260042, "learning_rate": 9.27688367111938e-06, "loss": 0.0624, "step": 4272 }, { "epoch": 0.19855947955390335, "grad_norm": 2.4006125139181442, "learning_rate": 9.276493815952638e-06, "loss": 0.0837, "step": 4273 }, { "epoch": 0.19860594795539033, "grad_norm": 2.0175639538227914, "learning_rate": 9.276103863918812e-06, "loss": 0.1197, "step": 4274 }, { "epoch": 0.19865241635687733, "grad_norm": 2.7014234109117456, "learning_rate": 9.275713815026732e-06, "loss": 0.1218, "step": 4275 }, { "epoch": 0.1986988847583643, "grad_norm": 1.5293763906567828, "learning_rate": 9.275323669285233e-06, "loss": 0.1069, "step": 4276 }, { "epoch": 0.1987453531598513, "grad_norm": 0.8419426359969985, "learning_rate": 9.274933426703154e-06, "loss": 0.0692, "step": 4277 }, { "epoch": 0.1987918215613383, "grad_norm": 0.8282706530589308, "learning_rate": 9.274543087289334e-06, "loss": 0.0485, "step": 4278 }, { "epoch": 0.19883828996282527, "grad_norm": 1.4732201834447085, "learning_rate": 9.274152651052614e-06, "loss": 0.0753, "step": 4279 }, { "epoch": 0.19888475836431227, "grad_norm": 1.1627540268601328, "learning_rate": 9.273762118001837e-06, "loss": 0.0733, "step": 4280 }, { "epoch": 0.19893122676579925, "grad_norm": 2.6195230492323733, "learning_rate": 9.273371488145852e-06, "loss": 0.115, "step": 4281 }, { "epoch": 0.19897769516728625, "grad_norm": 3.2942618863346484, "learning_rate": 9.272980761493503e-06, "loss": 0.0829, "step": 4282 }, { "epoch": 0.19902416356877323, "grad_norm": 1.0982877653424676, "learning_rate": 9.272589938053644e-06, "loss": 0.0556, "step": 4283 }, { "epoch": 0.19907063197026023, "grad_norm": 1.4463989975809528, "learning_rate": 9.272199017835124e-06, "loss": 0.0562, "step": 4284 }, { "epoch": 0.1991171003717472, "grad_norm": 0.6047711780842319, "learning_rate": 9.271808000846803e-06, "loss": 0.0481, "step": 4285 }, { "epoch": 0.1991635687732342, "grad_norm": 1.2793960439290468, "learning_rate": 9.271416887097532e-06, "loss": 0.0831, "step": 4286 }, { "epoch": 0.1992100371747212, "grad_norm": 1.354488744787146, "learning_rate": 9.271025676596174e-06, "loss": 0.0739, "step": 4287 }, { "epoch": 0.19925650557620816, "grad_norm": 2.6400357730291266, "learning_rate": 9.27063436935159e-06, "loss": 0.1246, "step": 4288 }, { "epoch": 0.19930297397769517, "grad_norm": 1.088613023700288, "learning_rate": 9.270242965372642e-06, "loss": 0.0828, "step": 4289 }, { "epoch": 0.19934944237918215, "grad_norm": 4.617896660430337, "learning_rate": 9.269851464668196e-06, "loss": 0.1224, "step": 4290 }, { "epoch": 0.19939591078066915, "grad_norm": 4.240457973656312, "learning_rate": 9.269459867247122e-06, "loss": 0.0545, "step": 4291 }, { "epoch": 0.19944237918215613, "grad_norm": 1.6229081020071707, "learning_rate": 9.269068173118286e-06, "loss": 0.1266, "step": 4292 }, { "epoch": 0.19948884758364313, "grad_norm": 4.611698902875402, "learning_rate": 9.268676382290564e-06, "loss": 0.0633, "step": 4293 }, { "epoch": 0.1995353159851301, "grad_norm": 1.5445383332623892, "learning_rate": 9.268284494772831e-06, "loss": 0.0943, "step": 4294 }, { "epoch": 0.1995817843866171, "grad_norm": 1.4087495983839398, "learning_rate": 9.267892510573959e-06, "loss": 0.0883, "step": 4295 }, { "epoch": 0.1996282527881041, "grad_norm": 4.591220082885536, "learning_rate": 9.267500429702832e-06, "loss": 0.1159, "step": 4296 }, { "epoch": 0.19967472118959106, "grad_norm": 1.9871699332814667, "learning_rate": 9.267108252168327e-06, "loss": 0.0877, "step": 4297 }, { "epoch": 0.19972118959107807, "grad_norm": 1.403774662221145, "learning_rate": 9.26671597797933e-06, "loss": 0.1169, "step": 4298 }, { "epoch": 0.19976765799256505, "grad_norm": 0.9099445824416132, "learning_rate": 9.266323607144725e-06, "loss": 0.0561, "step": 4299 }, { "epoch": 0.19981412639405205, "grad_norm": 1.031590037179852, "learning_rate": 9.2659311396734e-06, "loss": 0.1011, "step": 4300 }, { "epoch": 0.19986059479553903, "grad_norm": 2.395132204761422, "learning_rate": 9.265538575574246e-06, "loss": 0.0668, "step": 4301 }, { "epoch": 0.19990706319702603, "grad_norm": 1.718173040005979, "learning_rate": 9.265145914856153e-06, "loss": 0.1083, "step": 4302 }, { "epoch": 0.199953531598513, "grad_norm": 0.7779476603473531, "learning_rate": 9.264753157528017e-06, "loss": 0.0641, "step": 4303 }, { "epoch": 0.2, "grad_norm": 2.6452488938361887, "learning_rate": 9.264360303598732e-06, "loss": 0.1007, "step": 4304 }, { "epoch": 0.200046468401487, "grad_norm": 2.370857613006589, "learning_rate": 9.2639673530772e-06, "loss": 0.1641, "step": 4305 }, { "epoch": 0.20009293680297396, "grad_norm": 1.4991746233973045, "learning_rate": 9.263574305972317e-06, "loss": 0.1273, "step": 4306 }, { "epoch": 0.20013940520446097, "grad_norm": 0.9657464282952948, "learning_rate": 9.263181162292991e-06, "loss": 0.0782, "step": 4307 }, { "epoch": 0.20018587360594794, "grad_norm": 2.0197134199057345, "learning_rate": 9.262787922048121e-06, "loss": 0.0924, "step": 4308 }, { "epoch": 0.20023234200743495, "grad_norm": 2.612773732113385, "learning_rate": 9.262394585246622e-06, "loss": 0.1171, "step": 4309 }, { "epoch": 0.20027881040892193, "grad_norm": 2.440529729758746, "learning_rate": 9.2620011518974e-06, "loss": 0.1114, "step": 4310 }, { "epoch": 0.20032527881040893, "grad_norm": 1.019468305153116, "learning_rate": 9.261607622009363e-06, "loss": 0.0551, "step": 4311 }, { "epoch": 0.2003717472118959, "grad_norm": 0.6439220929162579, "learning_rate": 9.26121399559143e-06, "loss": 0.0496, "step": 4312 }, { "epoch": 0.2004182156133829, "grad_norm": 1.761883307808501, "learning_rate": 9.260820272652515e-06, "loss": 0.0974, "step": 4313 }, { "epoch": 0.2004646840148699, "grad_norm": 1.7275868303021809, "learning_rate": 9.260426453201536e-06, "loss": 0.1179, "step": 4314 }, { "epoch": 0.20051115241635686, "grad_norm": 1.0737904128508247, "learning_rate": 9.260032537247415e-06, "loss": 0.0712, "step": 4315 }, { "epoch": 0.20055762081784387, "grad_norm": 1.2760587109386272, "learning_rate": 9.259638524799072e-06, "loss": 0.1132, "step": 4316 }, { "epoch": 0.20060408921933084, "grad_norm": 1.026519506832853, "learning_rate": 9.259244415865433e-06, "loss": 0.0689, "step": 4317 }, { "epoch": 0.20065055762081785, "grad_norm": 1.494217368932753, "learning_rate": 9.258850210455426e-06, "loss": 0.0964, "step": 4318 }, { "epoch": 0.20069702602230483, "grad_norm": 1.3994223317271863, "learning_rate": 9.25845590857798e-06, "loss": 0.0682, "step": 4319 }, { "epoch": 0.20074349442379183, "grad_norm": 1.536655884834668, "learning_rate": 9.258061510242024e-06, "loss": 0.093, "step": 4320 }, { "epoch": 0.2007899628252788, "grad_norm": 1.5645977833769475, "learning_rate": 9.257667015456495e-06, "loss": 0.1025, "step": 4321 }, { "epoch": 0.2008364312267658, "grad_norm": 1.7034525235477156, "learning_rate": 9.257272424230326e-06, "loss": 0.0799, "step": 4322 }, { "epoch": 0.2008828996282528, "grad_norm": 1.7278412718233502, "learning_rate": 9.256877736572456e-06, "loss": 0.0996, "step": 4323 }, { "epoch": 0.20092936802973976, "grad_norm": 0.6978506094713344, "learning_rate": 9.256482952491825e-06, "loss": 0.0645, "step": 4324 }, { "epoch": 0.20097583643122677, "grad_norm": 1.792358941485831, "learning_rate": 9.256088071997375e-06, "loss": 0.1123, "step": 4325 }, { "epoch": 0.20102230483271374, "grad_norm": 2.160559598255565, "learning_rate": 9.255693095098052e-06, "loss": 0.1308, "step": 4326 }, { "epoch": 0.20106877323420075, "grad_norm": 2.529981303595907, "learning_rate": 9.2552980218028e-06, "loss": 0.1575, "step": 4327 }, { "epoch": 0.20111524163568772, "grad_norm": 2.024660722672691, "learning_rate": 9.25490285212057e-06, "loss": 0.0451, "step": 4328 }, { "epoch": 0.20116171003717473, "grad_norm": 1.357364402345107, "learning_rate": 9.25450758606031e-06, "loss": 0.0767, "step": 4329 }, { "epoch": 0.2012081784386617, "grad_norm": 2.0318677289690763, "learning_rate": 9.254112223630977e-06, "loss": 0.1007, "step": 4330 }, { "epoch": 0.2012546468401487, "grad_norm": 1.9842900151797858, "learning_rate": 9.253716764841527e-06, "loss": 0.148, "step": 4331 }, { "epoch": 0.2013011152416357, "grad_norm": 1.0676190812982627, "learning_rate": 9.253321209700913e-06, "loss": 0.0734, "step": 4332 }, { "epoch": 0.20134758364312266, "grad_norm": 1.3775246569253836, "learning_rate": 9.252925558218098e-06, "loss": 0.0802, "step": 4333 }, { "epoch": 0.20139405204460967, "grad_norm": 0.7982755555283254, "learning_rate": 9.252529810402043e-06, "loss": 0.0745, "step": 4334 }, { "epoch": 0.20144052044609664, "grad_norm": 1.7019374478447138, "learning_rate": 9.252133966261709e-06, "loss": 0.0652, "step": 4335 }, { "epoch": 0.20148698884758365, "grad_norm": 1.7902247376451548, "learning_rate": 9.251738025806069e-06, "loss": 0.1116, "step": 4336 }, { "epoch": 0.20153345724907062, "grad_norm": 2.750791677535726, "learning_rate": 9.251341989044087e-06, "loss": 0.125, "step": 4337 }, { "epoch": 0.20157992565055763, "grad_norm": 1.3188057933760828, "learning_rate": 9.250945855984735e-06, "loss": 0.0969, "step": 4338 }, { "epoch": 0.2016263940520446, "grad_norm": 2.051858357530296, "learning_rate": 9.250549626636983e-06, "loss": 0.118, "step": 4339 }, { "epoch": 0.2016728624535316, "grad_norm": 0.5877471224096568, "learning_rate": 9.25015330100981e-06, "loss": 0.065, "step": 4340 }, { "epoch": 0.20171933085501859, "grad_norm": 2.083823384431958, "learning_rate": 9.249756879112191e-06, "loss": 0.107, "step": 4341 }, { "epoch": 0.20176579925650556, "grad_norm": 1.762223232589524, "learning_rate": 9.249360360953106e-06, "loss": 0.0724, "step": 4342 }, { "epoch": 0.20181226765799257, "grad_norm": 1.1071376965181599, "learning_rate": 9.248963746541537e-06, "loss": 0.0914, "step": 4343 }, { "epoch": 0.20185873605947954, "grad_norm": 1.3320555957282318, "learning_rate": 9.248567035886466e-06, "loss": 0.0889, "step": 4344 }, { "epoch": 0.20190520446096655, "grad_norm": 1.6062574206289515, "learning_rate": 9.248170228996882e-06, "loss": 0.1275, "step": 4345 }, { "epoch": 0.20195167286245352, "grad_norm": 1.435926266651323, "learning_rate": 9.24777332588177e-06, "loss": 0.1849, "step": 4346 }, { "epoch": 0.20199814126394053, "grad_norm": 1.0300353389037642, "learning_rate": 9.247376326550122e-06, "loss": 0.0646, "step": 4347 }, { "epoch": 0.2020446096654275, "grad_norm": 1.674647034625747, "learning_rate": 9.246979231010929e-06, "loss": 0.0899, "step": 4348 }, { "epoch": 0.2020910780669145, "grad_norm": 2.660442378441546, "learning_rate": 9.246582039273187e-06, "loss": 0.0889, "step": 4349 }, { "epoch": 0.20213754646840149, "grad_norm": 0.9284165416420043, "learning_rate": 9.246184751345893e-06, "loss": 0.0477, "step": 4350 }, { "epoch": 0.20218401486988846, "grad_norm": 1.5749714857424926, "learning_rate": 9.245787367238044e-06, "loss": 0.0534, "step": 4351 }, { "epoch": 0.20223048327137547, "grad_norm": 0.9900334432024572, "learning_rate": 9.245389886958644e-06, "loss": 0.0695, "step": 4352 }, { "epoch": 0.20227695167286244, "grad_norm": 1.6759773727352636, "learning_rate": 9.244992310516693e-06, "loss": 0.0936, "step": 4353 }, { "epoch": 0.20232342007434945, "grad_norm": 1.1583918461317204, "learning_rate": 9.2445946379212e-06, "loss": 0.0593, "step": 4354 }, { "epoch": 0.20236988847583642, "grad_norm": 1.3535025542553945, "learning_rate": 9.244196869181173e-06, "loss": 0.0915, "step": 4355 }, { "epoch": 0.20241635687732343, "grad_norm": 1.6872105886765398, "learning_rate": 9.243799004305616e-06, "loss": 0.0894, "step": 4356 }, { "epoch": 0.2024628252788104, "grad_norm": 0.9471339893522535, "learning_rate": 9.243401043303548e-06, "loss": 0.0574, "step": 4357 }, { "epoch": 0.2025092936802974, "grad_norm": 1.2710234071546092, "learning_rate": 9.243002986183977e-06, "loss": 0.0681, "step": 4358 }, { "epoch": 0.20255576208178439, "grad_norm": 0.9545863954262761, "learning_rate": 9.242604832955927e-06, "loss": 0.0507, "step": 4359 }, { "epoch": 0.20260223048327136, "grad_norm": 1.1195140820903564, "learning_rate": 9.242206583628409e-06, "loss": 0.0617, "step": 4360 }, { "epoch": 0.20264869888475837, "grad_norm": 1.1736999516914066, "learning_rate": 9.241808238210448e-06, "loss": 0.0755, "step": 4361 }, { "epoch": 0.20269516728624534, "grad_norm": 2.6991080882567196, "learning_rate": 9.241409796711064e-06, "loss": 0.0876, "step": 4362 }, { "epoch": 0.20274163568773235, "grad_norm": 1.547346270779734, "learning_rate": 9.241011259139288e-06, "loss": 0.1059, "step": 4363 }, { "epoch": 0.20278810408921932, "grad_norm": 1.2751003659093483, "learning_rate": 9.240612625504142e-06, "loss": 0.0915, "step": 4364 }, { "epoch": 0.20283457249070633, "grad_norm": 2.7052675214640445, "learning_rate": 9.240213895814656e-06, "loss": 0.0846, "step": 4365 }, { "epoch": 0.2028810408921933, "grad_norm": 1.2571443656797858, "learning_rate": 9.239815070079862e-06, "loss": 0.0817, "step": 4366 }, { "epoch": 0.2029275092936803, "grad_norm": 1.8967852848061764, "learning_rate": 9.239416148308795e-06, "loss": 0.0628, "step": 4367 }, { "epoch": 0.20297397769516728, "grad_norm": 0.9900245018496073, "learning_rate": 9.23901713051049e-06, "loss": 0.0579, "step": 4368 }, { "epoch": 0.20302044609665426, "grad_norm": 1.909393221274175, "learning_rate": 9.238618016693987e-06, "loss": 0.0774, "step": 4369 }, { "epoch": 0.20306691449814127, "grad_norm": 1.0778935923011534, "learning_rate": 9.238218806868324e-06, "loss": 0.077, "step": 4370 }, { "epoch": 0.20311338289962824, "grad_norm": 1.5619518051742152, "learning_rate": 9.237819501042545e-06, "loss": 0.0845, "step": 4371 }, { "epoch": 0.20315985130111525, "grad_norm": 1.6989003469503554, "learning_rate": 9.237420099225694e-06, "loss": 0.0725, "step": 4372 }, { "epoch": 0.20320631970260222, "grad_norm": 1.3618195100429475, "learning_rate": 9.237020601426816e-06, "loss": 0.0541, "step": 4373 }, { "epoch": 0.20325278810408923, "grad_norm": 3.3707213195876133, "learning_rate": 9.236621007654963e-06, "loss": 0.0938, "step": 4374 }, { "epoch": 0.2032992565055762, "grad_norm": 1.9252052166959075, "learning_rate": 9.236221317919186e-06, "loss": 0.0839, "step": 4375 }, { "epoch": 0.2033457249070632, "grad_norm": 1.5401671538780253, "learning_rate": 9.235821532228537e-06, "loss": 0.1021, "step": 4376 }, { "epoch": 0.20339219330855018, "grad_norm": 1.6092482711467908, "learning_rate": 9.235421650592072e-06, "loss": 0.0555, "step": 4377 }, { "epoch": 0.2034386617100372, "grad_norm": 1.4178865608693865, "learning_rate": 9.235021673018849e-06, "loss": 0.102, "step": 4378 }, { "epoch": 0.20348513011152417, "grad_norm": 0.9097613169206678, "learning_rate": 9.234621599517928e-06, "loss": 0.0549, "step": 4379 }, { "epoch": 0.20353159851301114, "grad_norm": 1.9843411110766087, "learning_rate": 9.234221430098371e-06, "loss": 0.097, "step": 4380 }, { "epoch": 0.20357806691449815, "grad_norm": 1.1913078625618143, "learning_rate": 9.233821164769242e-06, "loss": 0.036, "step": 4381 }, { "epoch": 0.20362453531598512, "grad_norm": 2.1677852581742205, "learning_rate": 9.233420803539608e-06, "loss": 0.1135, "step": 4382 }, { "epoch": 0.20367100371747213, "grad_norm": 1.5981662981790805, "learning_rate": 9.233020346418538e-06, "loss": 0.1128, "step": 4383 }, { "epoch": 0.2037174721189591, "grad_norm": 1.7410217954785556, "learning_rate": 9.2326197934151e-06, "loss": 0.0749, "step": 4384 }, { "epoch": 0.2037639405204461, "grad_norm": 2.071902153248706, "learning_rate": 9.232219144538371e-06, "loss": 0.1274, "step": 4385 }, { "epoch": 0.20381040892193308, "grad_norm": 1.2473121772313087, "learning_rate": 9.231818399797421e-06, "loss": 0.0807, "step": 4386 }, { "epoch": 0.2038568773234201, "grad_norm": 1.3231801337505142, "learning_rate": 9.231417559201332e-06, "loss": 0.1051, "step": 4387 }, { "epoch": 0.20390334572490706, "grad_norm": 1.26521746557011, "learning_rate": 9.231016622759184e-06, "loss": 0.0826, "step": 4388 }, { "epoch": 0.20394981412639404, "grad_norm": 2.1112945465900355, "learning_rate": 9.230615590480053e-06, "loss": 0.0725, "step": 4389 }, { "epoch": 0.20399628252788105, "grad_norm": 2.5212485739326262, "learning_rate": 9.230214462373028e-06, "loss": 0.1246, "step": 4390 }, { "epoch": 0.20404275092936802, "grad_norm": 1.8513425334398093, "learning_rate": 9.229813238447193e-06, "loss": 0.1002, "step": 4391 }, { "epoch": 0.20408921933085503, "grad_norm": 0.6899030470280925, "learning_rate": 9.229411918711637e-06, "loss": 0.0482, "step": 4392 }, { "epoch": 0.204135687732342, "grad_norm": 1.197302265544262, "learning_rate": 9.229010503175448e-06, "loss": 0.1088, "step": 4393 }, { "epoch": 0.204182156133829, "grad_norm": 1.789147582295659, "learning_rate": 9.228608991847721e-06, "loss": 0.0957, "step": 4394 }, { "epoch": 0.20422862453531598, "grad_norm": 2.6466501770091804, "learning_rate": 9.22820738473755e-06, "loss": 0.0701, "step": 4395 }, { "epoch": 0.204275092936803, "grad_norm": 3.901675710277458, "learning_rate": 9.227805681854032e-06, "loss": 0.0771, "step": 4396 }, { "epoch": 0.20432156133828996, "grad_norm": 2.0802305577807574, "learning_rate": 9.227403883206266e-06, "loss": 0.1045, "step": 4397 }, { "epoch": 0.20436802973977694, "grad_norm": 1.6868075885116693, "learning_rate": 9.22700198880335e-06, "loss": 0.116, "step": 4398 }, { "epoch": 0.20441449814126395, "grad_norm": 1.2778837830183933, "learning_rate": 9.226599998654392e-06, "loss": 0.123, "step": 4399 }, { "epoch": 0.20446096654275092, "grad_norm": 2.268972202089449, "learning_rate": 9.226197912768496e-06, "loss": 0.067, "step": 4400 }, { "epoch": 0.20450743494423793, "grad_norm": 1.4180454401032132, "learning_rate": 9.225795731154768e-06, "loss": 0.0682, "step": 4401 }, { "epoch": 0.2045539033457249, "grad_norm": 0.8214006029816928, "learning_rate": 9.22539345382232e-06, "loss": 0.0676, "step": 4402 }, { "epoch": 0.2046003717472119, "grad_norm": 1.2669875363441296, "learning_rate": 9.224991080780263e-06, "loss": 0.086, "step": 4403 }, { "epoch": 0.20464684014869888, "grad_norm": 1.267835217789707, "learning_rate": 9.224588612037712e-06, "loss": 0.1032, "step": 4404 }, { "epoch": 0.2046933085501859, "grad_norm": 1.4723631927655236, "learning_rate": 9.224186047603782e-06, "loss": 0.1246, "step": 4405 }, { "epoch": 0.20473977695167286, "grad_norm": 1.8775227749009136, "learning_rate": 9.223783387487593e-06, "loss": 0.0806, "step": 4406 }, { "epoch": 0.20478624535315984, "grad_norm": 3.2405082090851045, "learning_rate": 9.223380631698264e-06, "loss": 0.09, "step": 4407 }, { "epoch": 0.20483271375464684, "grad_norm": 0.7546614740219177, "learning_rate": 9.222977780244918e-06, "loss": 0.0612, "step": 4408 }, { "epoch": 0.20487918215613382, "grad_norm": 1.6813486807807991, "learning_rate": 9.222574833136682e-06, "loss": 0.0991, "step": 4409 }, { "epoch": 0.20492565055762083, "grad_norm": 0.9521009394935261, "learning_rate": 9.222171790382681e-06, "loss": 0.0651, "step": 4410 }, { "epoch": 0.2049721189591078, "grad_norm": 1.208029497938691, "learning_rate": 9.221768651992045e-06, "loss": 0.0726, "step": 4411 }, { "epoch": 0.2050185873605948, "grad_norm": 1.0736864721004282, "learning_rate": 9.221365417973907e-06, "loss": 0.0703, "step": 4412 }, { "epoch": 0.20506505576208178, "grad_norm": 4.442904893717565, "learning_rate": 9.220962088337398e-06, "loss": 0.1532, "step": 4413 }, { "epoch": 0.2051115241635688, "grad_norm": 1.6469152784927952, "learning_rate": 9.220558663091657e-06, "loss": 0.1409, "step": 4414 }, { "epoch": 0.20515799256505576, "grad_norm": 1.6182952207676518, "learning_rate": 9.220155142245818e-06, "loss": 0.0909, "step": 4415 }, { "epoch": 0.20520446096654274, "grad_norm": 2.1523626468652997, "learning_rate": 9.219751525809024e-06, "loss": 0.0978, "step": 4416 }, { "epoch": 0.20525092936802974, "grad_norm": 1.7611455084837337, "learning_rate": 9.219347813790416e-06, "loss": 0.0945, "step": 4417 }, { "epoch": 0.20529739776951672, "grad_norm": 2.1499717311341473, "learning_rate": 9.218944006199142e-06, "loss": 0.0843, "step": 4418 }, { "epoch": 0.20534386617100372, "grad_norm": 2.288065991400189, "learning_rate": 9.218540103044344e-06, "loss": 0.0967, "step": 4419 }, { "epoch": 0.2053903345724907, "grad_norm": 0.8735283419326476, "learning_rate": 9.218136104335173e-06, "loss": 0.0468, "step": 4420 }, { "epoch": 0.2054368029739777, "grad_norm": 1.0935775265036762, "learning_rate": 9.21773201008078e-06, "loss": 0.0722, "step": 4421 }, { "epoch": 0.20548327137546468, "grad_norm": 1.5905630681915148, "learning_rate": 9.217327820290318e-06, "loss": 0.0869, "step": 4422 }, { "epoch": 0.2055297397769517, "grad_norm": 1.401538636489793, "learning_rate": 9.21692353497294e-06, "loss": 0.0909, "step": 4423 }, { "epoch": 0.20557620817843866, "grad_norm": 1.0882769105742245, "learning_rate": 9.216519154137808e-06, "loss": 0.0607, "step": 4424 }, { "epoch": 0.20562267657992564, "grad_norm": 3.645382215223284, "learning_rate": 9.216114677794078e-06, "loss": 0.1379, "step": 4425 }, { "epoch": 0.20566914498141264, "grad_norm": 1.6654392584039188, "learning_rate": 9.215710105950914e-06, "loss": 0.0763, "step": 4426 }, { "epoch": 0.20571561338289962, "grad_norm": 0.805935388443852, "learning_rate": 9.21530543861748e-06, "loss": 0.0616, "step": 4427 }, { "epoch": 0.20576208178438662, "grad_norm": 1.2747926460104413, "learning_rate": 9.21490067580294e-06, "loss": 0.0853, "step": 4428 }, { "epoch": 0.2058085501858736, "grad_norm": 1.67265332108495, "learning_rate": 9.214495817516464e-06, "loss": 0.0647, "step": 4429 }, { "epoch": 0.2058550185873606, "grad_norm": 1.069746618694546, "learning_rate": 9.214090863767223e-06, "loss": 0.0643, "step": 4430 }, { "epoch": 0.20590148698884758, "grad_norm": 0.6197875609873112, "learning_rate": 9.213685814564387e-06, "loss": 0.0613, "step": 4431 }, { "epoch": 0.20594795539033459, "grad_norm": 1.8336501777401317, "learning_rate": 9.213280669917132e-06, "loss": 0.0966, "step": 4432 }, { "epoch": 0.20599442379182156, "grad_norm": 0.611487995334926, "learning_rate": 9.212875429834638e-06, "loss": 0.038, "step": 4433 }, { "epoch": 0.20604089219330854, "grad_norm": 1.0827759682448774, "learning_rate": 9.21247009432608e-06, "loss": 0.1044, "step": 4434 }, { "epoch": 0.20608736059479554, "grad_norm": 1.4059794336497178, "learning_rate": 9.212064663400642e-06, "loss": 0.0708, "step": 4435 }, { "epoch": 0.20613382899628252, "grad_norm": 3.1403799994149026, "learning_rate": 9.211659137067505e-06, "loss": 0.0608, "step": 4436 }, { "epoch": 0.20618029739776952, "grad_norm": 2.940690985236793, "learning_rate": 9.211253515335856e-06, "loss": 0.0989, "step": 4437 }, { "epoch": 0.2062267657992565, "grad_norm": 2.0256407831113257, "learning_rate": 9.210847798214883e-06, "loss": 0.0913, "step": 4438 }, { "epoch": 0.2062732342007435, "grad_norm": 1.007226983088716, "learning_rate": 9.210441985713775e-06, "loss": 0.0417, "step": 4439 }, { "epoch": 0.20631970260223048, "grad_norm": 1.4118450046832838, "learning_rate": 9.210036077841725e-06, "loss": 0.066, "step": 4440 }, { "epoch": 0.20636617100371749, "grad_norm": 1.3175128290972637, "learning_rate": 9.209630074607927e-06, "loss": 0.0594, "step": 4441 }, { "epoch": 0.20641263940520446, "grad_norm": 1.506128217895401, "learning_rate": 9.209223976021577e-06, "loss": 0.1107, "step": 4442 }, { "epoch": 0.20645910780669144, "grad_norm": 1.887655203922404, "learning_rate": 9.208817782091874e-06, "loss": 0.1104, "step": 4443 }, { "epoch": 0.20650557620817844, "grad_norm": 0.9485215078696791, "learning_rate": 9.208411492828017e-06, "loss": 0.0593, "step": 4444 }, { "epoch": 0.20655204460966542, "grad_norm": 2.2662826359146924, "learning_rate": 9.208005108239212e-06, "loss": 0.096, "step": 4445 }, { "epoch": 0.20659851301115242, "grad_norm": 1.2604212149258271, "learning_rate": 9.207598628334664e-06, "loss": 0.0661, "step": 4446 }, { "epoch": 0.2066449814126394, "grad_norm": 1.4384690770296544, "learning_rate": 9.207192053123576e-06, "loss": 0.0918, "step": 4447 }, { "epoch": 0.2066914498141264, "grad_norm": 0.8727026705795715, "learning_rate": 9.20678538261516e-06, "loss": 0.0571, "step": 4448 }, { "epoch": 0.20673791821561338, "grad_norm": 0.838126452433466, "learning_rate": 9.20637861681863e-06, "loss": 0.0796, "step": 4449 }, { "epoch": 0.20678438661710039, "grad_norm": 1.172646974894135, "learning_rate": 9.205971755743194e-06, "loss": 0.0732, "step": 4450 }, { "epoch": 0.20683085501858736, "grad_norm": 2.4323473961502944, "learning_rate": 9.205564799398073e-06, "loss": 0.094, "step": 4451 }, { "epoch": 0.20687732342007434, "grad_norm": 0.7928883624009165, "learning_rate": 9.205157747792484e-06, "loss": 0.058, "step": 4452 }, { "epoch": 0.20692379182156134, "grad_norm": 1.228500375986433, "learning_rate": 9.204750600935644e-06, "loss": 0.1021, "step": 4453 }, { "epoch": 0.20697026022304832, "grad_norm": 2.708506884471084, "learning_rate": 9.204343358836779e-06, "loss": 0.1641, "step": 4454 }, { "epoch": 0.20701672862453532, "grad_norm": 1.574021231083418, "learning_rate": 9.203936021505111e-06, "loss": 0.0894, "step": 4455 }, { "epoch": 0.2070631970260223, "grad_norm": 1.1889762492340925, "learning_rate": 9.203528588949868e-06, "loss": 0.0943, "step": 4456 }, { "epoch": 0.2071096654275093, "grad_norm": 1.2944932415353125, "learning_rate": 9.203121061180277e-06, "loss": 0.089, "step": 4457 }, { "epoch": 0.20715613382899628, "grad_norm": 2.743230273855261, "learning_rate": 9.202713438205571e-06, "loss": 0.0958, "step": 4458 }, { "epoch": 0.20720260223048328, "grad_norm": 3.5926377038531276, "learning_rate": 9.202305720034984e-06, "loss": 0.0964, "step": 4459 }, { "epoch": 0.20724907063197026, "grad_norm": 2.1768554042068695, "learning_rate": 9.201897906677747e-06, "loss": 0.0772, "step": 4460 }, { "epoch": 0.20729553903345724, "grad_norm": 2.8629022602884566, "learning_rate": 9.201489998143101e-06, "loss": 0.1206, "step": 4461 }, { "epoch": 0.20734200743494424, "grad_norm": 1.2412965961816445, "learning_rate": 9.201081994440286e-06, "loss": 0.0727, "step": 4462 }, { "epoch": 0.20738847583643122, "grad_norm": 1.5166421284205793, "learning_rate": 9.200673895578541e-06, "loss": 0.0946, "step": 4463 }, { "epoch": 0.20743494423791822, "grad_norm": 0.6009472210140018, "learning_rate": 9.20026570156711e-06, "loss": 0.0397, "step": 4464 }, { "epoch": 0.2074814126394052, "grad_norm": 1.0874972419046325, "learning_rate": 9.199857412415243e-06, "loss": 0.084, "step": 4465 }, { "epoch": 0.2075278810408922, "grad_norm": 1.4702784295196167, "learning_rate": 9.199449028132183e-06, "loss": 0.0751, "step": 4466 }, { "epoch": 0.20757434944237918, "grad_norm": 0.8739553153799746, "learning_rate": 9.199040548727183e-06, "loss": 0.0671, "step": 4467 }, { "epoch": 0.20762081784386618, "grad_norm": 1.595463579614821, "learning_rate": 9.198631974209495e-06, "loss": 0.0499, "step": 4468 }, { "epoch": 0.20766728624535316, "grad_norm": 1.4887179104903812, "learning_rate": 9.198223304588375e-06, "loss": 0.0934, "step": 4469 }, { "epoch": 0.20771375464684014, "grad_norm": 0.8369300298808852, "learning_rate": 9.197814539873078e-06, "loss": 0.0633, "step": 4470 }, { "epoch": 0.20776022304832714, "grad_norm": 0.5675962891160401, "learning_rate": 9.197405680072863e-06, "loss": 0.0574, "step": 4471 }, { "epoch": 0.20780669144981412, "grad_norm": 2.9367988349347574, "learning_rate": 9.196996725196995e-06, "loss": 0.1482, "step": 4472 }, { "epoch": 0.20785315985130112, "grad_norm": 1.3918546305671704, "learning_rate": 9.196587675254729e-06, "loss": 0.1166, "step": 4473 }, { "epoch": 0.2078996282527881, "grad_norm": 1.1588833502612184, "learning_rate": 9.196178530255337e-06, "loss": 0.0689, "step": 4474 }, { "epoch": 0.2079460966542751, "grad_norm": 1.7166653672299779, "learning_rate": 9.195769290208087e-06, "loss": 0.1187, "step": 4475 }, { "epoch": 0.20799256505576208, "grad_norm": 2.430217748213502, "learning_rate": 9.195359955122244e-06, "loss": 0.1333, "step": 4476 }, { "epoch": 0.20803903345724908, "grad_norm": 1.1528473042580134, "learning_rate": 9.194950525007083e-06, "loss": 0.0757, "step": 4477 }, { "epoch": 0.20808550185873606, "grad_norm": 2.315711677988075, "learning_rate": 9.19454099987188e-06, "loss": 0.1258, "step": 4478 }, { "epoch": 0.20813197026022304, "grad_norm": 0.9258149852366092, "learning_rate": 9.194131379725906e-06, "loss": 0.0514, "step": 4479 }, { "epoch": 0.20817843866171004, "grad_norm": 1.211005870364728, "learning_rate": 9.193721664578441e-06, "loss": 0.1131, "step": 4480 }, { "epoch": 0.20822490706319702, "grad_norm": 1.7662548980608432, "learning_rate": 9.193311854438769e-06, "loss": 0.0567, "step": 4481 }, { "epoch": 0.20827137546468402, "grad_norm": 2.3475895323498612, "learning_rate": 9.19290194931617e-06, "loss": 0.1102, "step": 4482 }, { "epoch": 0.208317843866171, "grad_norm": 1.3125776605371984, "learning_rate": 9.192491949219928e-06, "loss": 0.0999, "step": 4483 }, { "epoch": 0.208364312267658, "grad_norm": 1.1961003483156625, "learning_rate": 9.192081854159331e-06, "loss": 0.0951, "step": 4484 }, { "epoch": 0.20841078066914498, "grad_norm": 2.666913798497642, "learning_rate": 9.191671664143666e-06, "loss": 0.1231, "step": 4485 }, { "epoch": 0.20845724907063198, "grad_norm": 1.356545860077052, "learning_rate": 9.191261379182228e-06, "loss": 0.1304, "step": 4486 }, { "epoch": 0.20850371747211896, "grad_norm": 1.1413050570066694, "learning_rate": 9.19085099928431e-06, "loss": 0.0565, "step": 4487 }, { "epoch": 0.20855018587360594, "grad_norm": 1.2026513287088134, "learning_rate": 9.190440524459203e-06, "loss": 0.0799, "step": 4488 }, { "epoch": 0.20859665427509294, "grad_norm": 1.2682905392773745, "learning_rate": 9.190029954716211e-06, "loss": 0.0917, "step": 4489 }, { "epoch": 0.20864312267657992, "grad_norm": 3.1588994484842003, "learning_rate": 9.189619290064628e-06, "loss": 0.0736, "step": 4490 }, { "epoch": 0.20868959107806692, "grad_norm": 1.9332221783624932, "learning_rate": 9.18920853051376e-06, "loss": 0.0768, "step": 4491 }, { "epoch": 0.2087360594795539, "grad_norm": 1.0437122514460082, "learning_rate": 9.188797676072909e-06, "loss": 0.0661, "step": 4492 }, { "epoch": 0.2087825278810409, "grad_norm": 1.1597160105231035, "learning_rate": 9.188386726751382e-06, "loss": 0.0803, "step": 4493 }, { "epoch": 0.20882899628252788, "grad_norm": 1.668265391100589, "learning_rate": 9.187975682558487e-06, "loss": 0.1071, "step": 4494 }, { "epoch": 0.20887546468401488, "grad_norm": 2.0975025341337243, "learning_rate": 9.187564543503537e-06, "loss": 0.0769, "step": 4495 }, { "epoch": 0.20892193308550186, "grad_norm": 1.6960294837155874, "learning_rate": 9.187153309595842e-06, "loss": 0.0809, "step": 4496 }, { "epoch": 0.20896840148698884, "grad_norm": 2.5385609349616476, "learning_rate": 9.186741980844717e-06, "loss": 0.0597, "step": 4497 }, { "epoch": 0.20901486988847584, "grad_norm": 1.5030666075790837, "learning_rate": 9.18633055725948e-06, "loss": 0.0749, "step": 4498 }, { "epoch": 0.20906133828996282, "grad_norm": 1.4093098628935985, "learning_rate": 9.185919038849449e-06, "loss": 0.0888, "step": 4499 }, { "epoch": 0.20910780669144982, "grad_norm": 2.563270103626069, "learning_rate": 9.185507425623947e-06, "loss": 0.0871, "step": 4500 }, { "epoch": 0.2091542750929368, "grad_norm": 2.339230726149532, "learning_rate": 9.185095717592295e-06, "loss": 0.0866, "step": 4501 }, { "epoch": 0.2092007434944238, "grad_norm": 1.254297248658388, "learning_rate": 9.184683914763823e-06, "loss": 0.0849, "step": 4502 }, { "epoch": 0.20924721189591078, "grad_norm": 2.3604775908498543, "learning_rate": 9.184272017147855e-06, "loss": 0.1486, "step": 4503 }, { "epoch": 0.20929368029739778, "grad_norm": 1.3783641163398017, "learning_rate": 9.18386002475372e-06, "loss": 0.0845, "step": 4504 }, { "epoch": 0.20934014869888476, "grad_norm": 2.5020241646270898, "learning_rate": 9.183447937590755e-06, "loss": 0.0599, "step": 4505 }, { "epoch": 0.20938661710037174, "grad_norm": 0.7320706680270708, "learning_rate": 9.18303575566829e-06, "loss": 0.042, "step": 4506 }, { "epoch": 0.20943308550185874, "grad_norm": 1.457890934287783, "learning_rate": 9.182623478995663e-06, "loss": 0.0872, "step": 4507 }, { "epoch": 0.20947955390334572, "grad_norm": 3.3366915240044683, "learning_rate": 9.182211107582212e-06, "loss": 0.0899, "step": 4508 }, { "epoch": 0.20952602230483272, "grad_norm": 2.5918001884746698, "learning_rate": 9.181798641437277e-06, "loss": 0.064, "step": 4509 }, { "epoch": 0.2095724907063197, "grad_norm": 1.122996614151323, "learning_rate": 9.1813860805702e-06, "loss": 0.083, "step": 4510 }, { "epoch": 0.2096189591078067, "grad_norm": 2.486862870779458, "learning_rate": 9.180973424990331e-06, "loss": 0.0388, "step": 4511 }, { "epoch": 0.20966542750929368, "grad_norm": 1.4756353739369894, "learning_rate": 9.180560674707012e-06, "loss": 0.0643, "step": 4512 }, { "epoch": 0.20971189591078068, "grad_norm": 1.936928484391384, "learning_rate": 9.180147829729593e-06, "loss": 0.0806, "step": 4513 }, { "epoch": 0.20975836431226766, "grad_norm": 0.9576545089002132, "learning_rate": 9.179734890067426e-06, "loss": 0.0396, "step": 4514 }, { "epoch": 0.20980483271375464, "grad_norm": 1.1648916494355404, "learning_rate": 9.179321855729866e-06, "loss": 0.0868, "step": 4515 }, { "epoch": 0.20985130111524164, "grad_norm": 1.8065907633745262, "learning_rate": 9.178908726726267e-06, "loss": 0.0952, "step": 4516 }, { "epoch": 0.20989776951672862, "grad_norm": 1.0392878540272548, "learning_rate": 9.178495503065988e-06, "loss": 0.0827, "step": 4517 }, { "epoch": 0.20994423791821562, "grad_norm": 1.7480602021354505, "learning_rate": 9.178082184758386e-06, "loss": 0.0796, "step": 4518 }, { "epoch": 0.2099907063197026, "grad_norm": 1.6953065619163925, "learning_rate": 9.177668771812826e-06, "loss": 0.0836, "step": 4519 }, { "epoch": 0.2100371747211896, "grad_norm": 0.7637907607505026, "learning_rate": 9.177255264238672e-06, "loss": 0.0545, "step": 4520 }, { "epoch": 0.21008364312267658, "grad_norm": 1.1771583445078795, "learning_rate": 9.176841662045289e-06, "loss": 0.1214, "step": 4521 }, { "epoch": 0.21013011152416358, "grad_norm": 0.9471110611478777, "learning_rate": 9.176427965242044e-06, "loss": 0.0786, "step": 4522 }, { "epoch": 0.21017657992565056, "grad_norm": 0.8730318571010854, "learning_rate": 9.176014173838315e-06, "loss": 0.0652, "step": 4523 }, { "epoch": 0.21022304832713753, "grad_norm": 2.5993737175191516, "learning_rate": 9.175600287843467e-06, "loss": 0.0782, "step": 4524 }, { "epoch": 0.21026951672862454, "grad_norm": 1.3323743106569075, "learning_rate": 9.175186307266877e-06, "loss": 0.0682, "step": 4525 }, { "epoch": 0.21031598513011152, "grad_norm": 1.2897665069668398, "learning_rate": 9.174772232117925e-06, "loss": 0.0997, "step": 4526 }, { "epoch": 0.21036245353159852, "grad_norm": 2.628735218631835, "learning_rate": 9.174358062405986e-06, "loss": 0.1867, "step": 4527 }, { "epoch": 0.2104089219330855, "grad_norm": 1.072914913047066, "learning_rate": 9.173943798140445e-06, "loss": 0.0651, "step": 4528 }, { "epoch": 0.2104553903345725, "grad_norm": 1.6676255741905481, "learning_rate": 9.173529439330683e-06, "loss": 0.0822, "step": 4529 }, { "epoch": 0.21050185873605948, "grad_norm": 1.1271437006557632, "learning_rate": 9.173114985986087e-06, "loss": 0.055, "step": 4530 }, { "epoch": 0.21054832713754648, "grad_norm": 2.467412005445707, "learning_rate": 9.172700438116045e-06, "loss": 0.1003, "step": 4531 }, { "epoch": 0.21059479553903346, "grad_norm": 1.3511714612717414, "learning_rate": 9.172285795729945e-06, "loss": 0.0762, "step": 4532 }, { "epoch": 0.21064126394052043, "grad_norm": 1.066041814874849, "learning_rate": 9.17187105883718e-06, "loss": 0.0687, "step": 4533 }, { "epoch": 0.21068773234200744, "grad_norm": 0.8786886637934758, "learning_rate": 9.171456227447147e-06, "loss": 0.0698, "step": 4534 }, { "epoch": 0.21073420074349442, "grad_norm": 1.7121513855101829, "learning_rate": 9.17104130156924e-06, "loss": 0.0875, "step": 4535 }, { "epoch": 0.21078066914498142, "grad_norm": 2.2829044325421637, "learning_rate": 9.170626281212856e-06, "loss": 0.046, "step": 4536 }, { "epoch": 0.2108271375464684, "grad_norm": 1.2249023163513328, "learning_rate": 9.170211166387399e-06, "loss": 0.064, "step": 4537 }, { "epoch": 0.2108736059479554, "grad_norm": 2.1778879281236647, "learning_rate": 9.169795957102268e-06, "loss": 0.1129, "step": 4538 }, { "epoch": 0.21092007434944238, "grad_norm": 1.6547453928817313, "learning_rate": 9.169380653366871e-06, "loss": 0.07, "step": 4539 }, { "epoch": 0.21096654275092938, "grad_norm": 1.6905016114934746, "learning_rate": 9.168965255190614e-06, "loss": 0.0445, "step": 4540 }, { "epoch": 0.21101301115241636, "grad_norm": 1.004515206520913, "learning_rate": 9.168549762582906e-06, "loss": 0.0627, "step": 4541 }, { "epoch": 0.21105947955390333, "grad_norm": 1.0260107330883117, "learning_rate": 9.168134175553158e-06, "loss": 0.0735, "step": 4542 }, { "epoch": 0.21110594795539034, "grad_norm": 4.62025818815602, "learning_rate": 9.167718494110784e-06, "loss": 0.0964, "step": 4543 }, { "epoch": 0.21115241635687731, "grad_norm": 1.2872935662799077, "learning_rate": 9.167302718265201e-06, "loss": 0.0935, "step": 4544 }, { "epoch": 0.21119888475836432, "grad_norm": 1.5162747608999223, "learning_rate": 9.166886848025824e-06, "loss": 0.0739, "step": 4545 }, { "epoch": 0.2112453531598513, "grad_norm": 1.6002010382605205, "learning_rate": 9.166470883402076e-06, "loss": 0.1148, "step": 4546 }, { "epoch": 0.2112918215613383, "grad_norm": 0.9531146686274238, "learning_rate": 9.166054824403377e-06, "loss": 0.0482, "step": 4547 }, { "epoch": 0.21133828996282528, "grad_norm": 1.2391018452863627, "learning_rate": 9.16563867103915e-06, "loss": 0.0992, "step": 4548 }, { "epoch": 0.21138475836431228, "grad_norm": 1.0605063356285314, "learning_rate": 9.165222423318825e-06, "loss": 0.0463, "step": 4549 }, { "epoch": 0.21143122676579926, "grad_norm": 1.153688730820876, "learning_rate": 9.164806081251826e-06, "loss": 0.0835, "step": 4550 }, { "epoch": 0.21147769516728623, "grad_norm": 1.6742977549293057, "learning_rate": 9.164389644847587e-06, "loss": 0.0952, "step": 4551 }, { "epoch": 0.21152416356877324, "grad_norm": 1.3916152394141095, "learning_rate": 9.16397311411554e-06, "loss": 0.088, "step": 4552 }, { "epoch": 0.21157063197026021, "grad_norm": 1.3188646287795098, "learning_rate": 9.163556489065119e-06, "loss": 0.0763, "step": 4553 }, { "epoch": 0.21161710037174722, "grad_norm": 1.200056457636396, "learning_rate": 9.163139769705762e-06, "loss": 0.0693, "step": 4554 }, { "epoch": 0.2116635687732342, "grad_norm": 2.0964840013129207, "learning_rate": 9.162722956046906e-06, "loss": 0.083, "step": 4555 }, { "epoch": 0.2117100371747212, "grad_norm": 2.107943064357693, "learning_rate": 9.162306048097995e-06, "loss": 0.0986, "step": 4556 }, { "epoch": 0.21175650557620818, "grad_norm": 0.9088684402429349, "learning_rate": 9.16188904586847e-06, "loss": 0.0569, "step": 4557 }, { "epoch": 0.21180297397769518, "grad_norm": 1.8331300706354856, "learning_rate": 9.16147194936778e-06, "loss": 0.1132, "step": 4558 }, { "epoch": 0.21184944237918216, "grad_norm": 1.322237622136069, "learning_rate": 9.16105475860537e-06, "loss": 0.0804, "step": 4559 }, { "epoch": 0.21189591078066913, "grad_norm": 1.4103640822078485, "learning_rate": 9.16063747359069e-06, "loss": 0.0807, "step": 4560 }, { "epoch": 0.21194237918215614, "grad_norm": 1.6701551720125287, "learning_rate": 9.160220094333193e-06, "loss": 0.1029, "step": 4561 }, { "epoch": 0.21198884758364311, "grad_norm": 1.0416766538126616, "learning_rate": 9.15980262084233e-06, "loss": 0.0513, "step": 4562 }, { "epoch": 0.21203531598513012, "grad_norm": 1.6966381662919614, "learning_rate": 9.159385053127561e-06, "loss": 0.1213, "step": 4563 }, { "epoch": 0.2120817843866171, "grad_norm": 3.384736612572424, "learning_rate": 9.158967391198343e-06, "loss": 0.075, "step": 4564 }, { "epoch": 0.2121282527881041, "grad_norm": 1.3468289831277864, "learning_rate": 9.158549635064136e-06, "loss": 0.068, "step": 4565 }, { "epoch": 0.21217472118959108, "grad_norm": 1.381700994932723, "learning_rate": 9.158131784734404e-06, "loss": 0.0979, "step": 4566 }, { "epoch": 0.21222118959107808, "grad_norm": 0.8622238145865575, "learning_rate": 9.157713840218609e-06, "loss": 0.0864, "step": 4567 }, { "epoch": 0.21226765799256506, "grad_norm": 1.17729381300957, "learning_rate": 9.15729580152622e-06, "loss": 0.0579, "step": 4568 }, { "epoch": 0.21231412639405203, "grad_norm": 0.7011785786346372, "learning_rate": 9.156877668666707e-06, "loss": 0.0606, "step": 4569 }, { "epoch": 0.21236059479553904, "grad_norm": 2.15219166474439, "learning_rate": 9.156459441649537e-06, "loss": 0.084, "step": 4570 }, { "epoch": 0.212407063197026, "grad_norm": 2.5485251130912356, "learning_rate": 9.156041120484188e-06, "loss": 0.0433, "step": 4571 }, { "epoch": 0.21245353159851302, "grad_norm": 1.3901250969479833, "learning_rate": 9.155622705180134e-06, "loss": 0.068, "step": 4572 }, { "epoch": 0.2125, "grad_norm": 3.262773069283591, "learning_rate": 9.15520419574685e-06, "loss": 0.1021, "step": 4573 }, { "epoch": 0.212546468401487, "grad_norm": 1.269914041080433, "learning_rate": 9.15478559219382e-06, "loss": 0.0628, "step": 4574 }, { "epoch": 0.21259293680297398, "grad_norm": 0.7996495279702015, "learning_rate": 9.154366894530522e-06, "loss": 0.0893, "step": 4575 }, { "epoch": 0.21263940520446098, "grad_norm": 2.134208104519811, "learning_rate": 9.153948102766442e-06, "loss": 0.1604, "step": 4576 }, { "epoch": 0.21268587360594796, "grad_norm": 1.2390837572909879, "learning_rate": 9.153529216911066e-06, "loss": 0.1334, "step": 4577 }, { "epoch": 0.21273234200743493, "grad_norm": 1.673711000976409, "learning_rate": 9.15311023697388e-06, "loss": 0.101, "step": 4578 }, { "epoch": 0.21277881040892194, "grad_norm": 0.8618484191442483, "learning_rate": 9.152691162964377e-06, "loss": 0.0608, "step": 4579 }, { "epoch": 0.2128252788104089, "grad_norm": 2.715174523088327, "learning_rate": 9.15227199489205e-06, "loss": 0.0764, "step": 4580 }, { "epoch": 0.21287174721189592, "grad_norm": 1.8133958886526826, "learning_rate": 9.151852732766392e-06, "loss": 0.1261, "step": 4581 }, { "epoch": 0.2129182156133829, "grad_norm": 1.1815544811650442, "learning_rate": 9.151433376596897e-06, "loss": 0.0602, "step": 4582 }, { "epoch": 0.2129646840148699, "grad_norm": 0.8474100568261312, "learning_rate": 9.151013926393068e-06, "loss": 0.0604, "step": 4583 }, { "epoch": 0.21301115241635687, "grad_norm": 1.3584481843761573, "learning_rate": 9.150594382164407e-06, "loss": 0.0605, "step": 4584 }, { "epoch": 0.21305762081784388, "grad_norm": 1.8217138510917406, "learning_rate": 9.150174743920413e-06, "loss": 0.1099, "step": 4585 }, { "epoch": 0.21310408921933086, "grad_norm": 0.9624962507088138, "learning_rate": 9.149755011670591e-06, "loss": 0.0639, "step": 4586 }, { "epoch": 0.21315055762081783, "grad_norm": 1.2311641658649641, "learning_rate": 9.149335185424453e-06, "loss": 0.0768, "step": 4587 }, { "epoch": 0.21319702602230484, "grad_norm": 0.9745048872056906, "learning_rate": 9.148915265191505e-06, "loss": 0.0751, "step": 4588 }, { "epoch": 0.2132434944237918, "grad_norm": 2.0320258732821923, "learning_rate": 9.148495250981261e-06, "loss": 0.114, "step": 4589 }, { "epoch": 0.21328996282527882, "grad_norm": 1.5455801696198357, "learning_rate": 9.148075142803232e-06, "loss": 0.1296, "step": 4590 }, { "epoch": 0.2133364312267658, "grad_norm": 1.8390957709900573, "learning_rate": 9.147654940666937e-06, "loss": 0.1042, "step": 4591 }, { "epoch": 0.2133828996282528, "grad_norm": 0.9343828724528761, "learning_rate": 9.14723464458189e-06, "loss": 0.0561, "step": 4592 }, { "epoch": 0.21342936802973977, "grad_norm": 3.5091961589524328, "learning_rate": 9.146814254557614e-06, "loss": 0.083, "step": 4593 }, { "epoch": 0.21347583643122678, "grad_norm": 2.6699632441363903, "learning_rate": 9.146393770603632e-06, "loss": 0.0928, "step": 4594 }, { "epoch": 0.21352230483271376, "grad_norm": 1.9437696345015019, "learning_rate": 9.145973192729465e-06, "loss": 0.1118, "step": 4595 }, { "epoch": 0.21356877323420073, "grad_norm": 1.6437884628752617, "learning_rate": 9.145552520944642e-06, "loss": 0.1311, "step": 4596 }, { "epoch": 0.21361524163568774, "grad_norm": 2.3916318093641733, "learning_rate": 9.145131755258694e-06, "loss": 0.1428, "step": 4597 }, { "epoch": 0.2136617100371747, "grad_norm": 0.9582261007165437, "learning_rate": 9.144710895681146e-06, "loss": 0.0485, "step": 4598 }, { "epoch": 0.21370817843866172, "grad_norm": 1.2740773222658754, "learning_rate": 9.144289942221536e-06, "loss": 0.0721, "step": 4599 }, { "epoch": 0.2137546468401487, "grad_norm": 0.5135226568357117, "learning_rate": 9.143868894889395e-06, "loss": 0.0434, "step": 4600 }, { "epoch": 0.2138011152416357, "grad_norm": 2.993201699979911, "learning_rate": 9.143447753694263e-06, "loss": 0.1804, "step": 4601 }, { "epoch": 0.21384758364312267, "grad_norm": 2.299215636743759, "learning_rate": 9.143026518645678e-06, "loss": 0.1396, "step": 4602 }, { "epoch": 0.21389405204460968, "grad_norm": 0.9115222064198345, "learning_rate": 9.142605189753185e-06, "loss": 0.0555, "step": 4603 }, { "epoch": 0.21394052044609665, "grad_norm": 2.317997071558005, "learning_rate": 9.142183767026322e-06, "loss": 0.0999, "step": 4604 }, { "epoch": 0.21398698884758363, "grad_norm": 2.317997071558005, "learning_rate": 9.142183767026322e-06, "loss": 0.0877, "step": 4605 }, { "epoch": 0.21403345724907064, "grad_norm": 1.3338813912966816, "learning_rate": 9.141762250474638e-06, "loss": 0.0883, "step": 4606 }, { "epoch": 0.2140799256505576, "grad_norm": 1.0468553916692047, "learning_rate": 9.141340640107678e-06, "loss": 0.0765, "step": 4607 }, { "epoch": 0.21412639405204462, "grad_norm": 1.0520860562307253, "learning_rate": 9.140918935934995e-06, "loss": 0.1059, "step": 4608 }, { "epoch": 0.2141728624535316, "grad_norm": 0.7513338295782197, "learning_rate": 9.14049713796614e-06, "loss": 0.0585, "step": 4609 }, { "epoch": 0.2142193308550186, "grad_norm": 1.050099540157534, "learning_rate": 9.140075246210667e-06, "loss": 0.0754, "step": 4610 }, { "epoch": 0.21426579925650557, "grad_norm": 1.3581736680504641, "learning_rate": 9.139653260678131e-06, "loss": 0.1146, "step": 4611 }, { "epoch": 0.21431226765799258, "grad_norm": 1.2945179930907633, "learning_rate": 9.139231181378094e-06, "loss": 0.0982, "step": 4612 }, { "epoch": 0.21435873605947955, "grad_norm": 1.6797013846486433, "learning_rate": 9.138809008320113e-06, "loss": 0.091, "step": 4613 }, { "epoch": 0.21440520446096653, "grad_norm": 2.1246855682039936, "learning_rate": 9.138386741513752e-06, "loss": 0.0664, "step": 4614 }, { "epoch": 0.21445167286245354, "grad_norm": 1.0990715444089505, "learning_rate": 9.137964380968577e-06, "loss": 0.1024, "step": 4615 }, { "epoch": 0.2144981412639405, "grad_norm": 1.4360327426385806, "learning_rate": 9.137541926694152e-06, "loss": 0.0837, "step": 4616 }, { "epoch": 0.21454460966542752, "grad_norm": 1.1100240236064256, "learning_rate": 9.137119378700048e-06, "loss": 0.0675, "step": 4617 }, { "epoch": 0.2145910780669145, "grad_norm": 1.197345788041036, "learning_rate": 9.136696736995836e-06, "loss": 0.1111, "step": 4618 }, { "epoch": 0.2146375464684015, "grad_norm": 2.055231608791289, "learning_rate": 9.136274001591089e-06, "loss": 0.0985, "step": 4619 }, { "epoch": 0.21468401486988847, "grad_norm": 2.403807753277557, "learning_rate": 9.135851172495383e-06, "loss": 0.1016, "step": 4620 }, { "epoch": 0.21473048327137548, "grad_norm": 2.0588225031943375, "learning_rate": 9.135428249718295e-06, "loss": 0.0802, "step": 4621 }, { "epoch": 0.21477695167286245, "grad_norm": 1.2163945325699572, "learning_rate": 9.135005233269403e-06, "loss": 0.0965, "step": 4622 }, { "epoch": 0.21482342007434943, "grad_norm": 0.6356624637902736, "learning_rate": 9.134582123158292e-06, "loss": 0.0587, "step": 4623 }, { "epoch": 0.21486988847583643, "grad_norm": 0.9010456573434472, "learning_rate": 9.134158919394545e-06, "loss": 0.0806, "step": 4624 }, { "epoch": 0.2149163568773234, "grad_norm": 0.9855990236938459, "learning_rate": 9.133735621987746e-06, "loss": 0.0502, "step": 4625 }, { "epoch": 0.21496282527881042, "grad_norm": 0.876497198620677, "learning_rate": 9.133312230947484e-06, "loss": 0.0684, "step": 4626 }, { "epoch": 0.2150092936802974, "grad_norm": 1.9158212826566237, "learning_rate": 9.132888746283351e-06, "loss": 0.0558, "step": 4627 }, { "epoch": 0.2150557620817844, "grad_norm": 1.0076137985745828, "learning_rate": 9.132465168004938e-06, "loss": 0.0496, "step": 4628 }, { "epoch": 0.21510223048327137, "grad_norm": 1.1916965567696152, "learning_rate": 9.13204149612184e-06, "loss": 0.0814, "step": 4629 }, { "epoch": 0.21514869888475838, "grad_norm": 1.205276274532629, "learning_rate": 9.131617730643653e-06, "loss": 0.0935, "step": 4630 }, { "epoch": 0.21519516728624535, "grad_norm": 1.8169005448890232, "learning_rate": 9.131193871579975e-06, "loss": 0.0646, "step": 4631 }, { "epoch": 0.21524163568773233, "grad_norm": 4.010911215244219, "learning_rate": 9.13076991894041e-06, "loss": 0.0962, "step": 4632 }, { "epoch": 0.21528810408921933, "grad_norm": 1.5982566381179084, "learning_rate": 9.130345872734555e-06, "loss": 0.0932, "step": 4633 }, { "epoch": 0.2153345724907063, "grad_norm": 1.1023650368924476, "learning_rate": 9.129921732972022e-06, "loss": 0.0534, "step": 4634 }, { "epoch": 0.21538104089219332, "grad_norm": 4.1772044227140315, "learning_rate": 9.129497499662414e-06, "loss": 0.1616, "step": 4635 }, { "epoch": 0.2154275092936803, "grad_norm": 0.9905196717007505, "learning_rate": 9.129073172815341e-06, "loss": 0.0737, "step": 4636 }, { "epoch": 0.2154739776951673, "grad_norm": 2.1639887369808135, "learning_rate": 9.128648752440416e-06, "loss": 0.1038, "step": 4637 }, { "epoch": 0.21552044609665427, "grad_norm": 1.444345963497009, "learning_rate": 9.12822423854725e-06, "loss": 0.0785, "step": 4638 }, { "epoch": 0.21556691449814128, "grad_norm": 1.5292634428556173, "learning_rate": 9.127799631145462e-06, "loss": 0.1139, "step": 4639 }, { "epoch": 0.21561338289962825, "grad_norm": 1.309381724909254, "learning_rate": 9.127374930244667e-06, "loss": 0.1101, "step": 4640 }, { "epoch": 0.21565985130111523, "grad_norm": 3.054614895155753, "learning_rate": 9.126950135854485e-06, "loss": 0.1352, "step": 4641 }, { "epoch": 0.21570631970260223, "grad_norm": 1.518360990907208, "learning_rate": 9.12652524798454e-06, "loss": 0.1101, "step": 4642 }, { "epoch": 0.2157527881040892, "grad_norm": 2.1192548313199957, "learning_rate": 9.126100266644456e-06, "loss": 0.068, "step": 4643 }, { "epoch": 0.21579925650557621, "grad_norm": 0.9079251167025852, "learning_rate": 9.125675191843855e-06, "loss": 0.068, "step": 4644 }, { "epoch": 0.2158457249070632, "grad_norm": 1.0336236467752438, "learning_rate": 9.125250023592372e-06, "loss": 0.0579, "step": 4645 }, { "epoch": 0.2158921933085502, "grad_norm": 1.7130521411688409, "learning_rate": 9.124824761899633e-06, "loss": 0.121, "step": 4646 }, { "epoch": 0.21593866171003717, "grad_norm": 2.3911977419571473, "learning_rate": 9.12439940677527e-06, "loss": 0.0752, "step": 4647 }, { "epoch": 0.21598513011152418, "grad_norm": 1.1128395698674542, "learning_rate": 9.123973958228923e-06, "loss": 0.0615, "step": 4648 }, { "epoch": 0.21603159851301115, "grad_norm": 0.3707064686342556, "learning_rate": 9.123548416270223e-06, "loss": 0.0413, "step": 4649 }, { "epoch": 0.21607806691449813, "grad_norm": 1.9023156511508135, "learning_rate": 9.123122780908813e-06, "loss": 0.1302, "step": 4650 }, { "epoch": 0.21612453531598513, "grad_norm": 1.6003729543725749, "learning_rate": 9.122697052154331e-06, "loss": 0.0779, "step": 4651 }, { "epoch": 0.2161710037174721, "grad_norm": 0.7312651621338924, "learning_rate": 9.122271230016423e-06, "loss": 0.0661, "step": 4652 }, { "epoch": 0.21621747211895911, "grad_norm": 0.5894271088811414, "learning_rate": 9.121845314504732e-06, "loss": 0.0568, "step": 4653 }, { "epoch": 0.2162639405204461, "grad_norm": 2.1451083034375897, "learning_rate": 9.121419305628907e-06, "loss": 0.0608, "step": 4654 }, { "epoch": 0.2163104089219331, "grad_norm": 1.3346238454101682, "learning_rate": 9.120993203398598e-06, "loss": 0.074, "step": 4655 }, { "epoch": 0.21635687732342007, "grad_norm": 1.1845610347756277, "learning_rate": 9.120567007823454e-06, "loss": 0.0615, "step": 4656 }, { "epoch": 0.21640334572490708, "grad_norm": 1.7265757405707198, "learning_rate": 9.120140718913131e-06, "loss": 0.1115, "step": 4657 }, { "epoch": 0.21644981412639405, "grad_norm": 2.080853935249514, "learning_rate": 9.119714336677284e-06, "loss": 0.0714, "step": 4658 }, { "epoch": 0.21649628252788103, "grad_norm": 1.9500103002969171, "learning_rate": 9.119287861125572e-06, "loss": 0.1022, "step": 4659 }, { "epoch": 0.21654275092936803, "grad_norm": 1.3684417214764082, "learning_rate": 9.118861292267655e-06, "loss": 0.0555, "step": 4660 }, { "epoch": 0.216589219330855, "grad_norm": 2.159962666421047, "learning_rate": 9.118434630113194e-06, "loss": 0.1154, "step": 4661 }, { "epoch": 0.216635687732342, "grad_norm": 1.735996086083297, "learning_rate": 9.118007874671853e-06, "loss": 0.0919, "step": 4662 }, { "epoch": 0.216682156133829, "grad_norm": 0.9263526075554891, "learning_rate": 9.117581025953302e-06, "loss": 0.0657, "step": 4663 }, { "epoch": 0.216728624535316, "grad_norm": 0.7853127032034275, "learning_rate": 9.117154083967205e-06, "loss": 0.0644, "step": 4664 }, { "epoch": 0.21677509293680297, "grad_norm": 3.92703850271177, "learning_rate": 9.116727048723235e-06, "loss": 0.1048, "step": 4665 }, { "epoch": 0.21682156133828998, "grad_norm": 15.701305025667928, "learning_rate": 9.116299920231065e-06, "loss": 0.1015, "step": 4666 }, { "epoch": 0.21686802973977695, "grad_norm": 2.2672267431240756, "learning_rate": 9.11587269850037e-06, "loss": 0.1175, "step": 4667 }, { "epoch": 0.21691449814126393, "grad_norm": 0.7800501078743199, "learning_rate": 9.115445383540825e-06, "loss": 0.0679, "step": 4668 }, { "epoch": 0.21696096654275093, "grad_norm": 1.3190425875367424, "learning_rate": 9.115017975362111e-06, "loss": 0.1087, "step": 4669 }, { "epoch": 0.2170074349442379, "grad_norm": 2.010646134154716, "learning_rate": 9.11459047397391e-06, "loss": 0.0815, "step": 4670 }, { "epoch": 0.2170539033457249, "grad_norm": 0.6007860895710422, "learning_rate": 9.114162879385904e-06, "loss": 0.042, "step": 4671 }, { "epoch": 0.2171003717472119, "grad_norm": 1.8288452845464014, "learning_rate": 9.113735191607778e-06, "loss": 0.1009, "step": 4672 }, { "epoch": 0.2171468401486989, "grad_norm": 2.463148402422358, "learning_rate": 9.113307410649222e-06, "loss": 0.097, "step": 4673 }, { "epoch": 0.21719330855018587, "grad_norm": 2.03388318665946, "learning_rate": 9.112879536519921e-06, "loss": 0.1209, "step": 4674 }, { "epoch": 0.21723977695167287, "grad_norm": 1.3794800494693642, "learning_rate": 9.112451569229572e-06, "loss": 0.0883, "step": 4675 }, { "epoch": 0.21728624535315985, "grad_norm": 1.2377321226265035, "learning_rate": 9.112023508787866e-06, "loss": 0.0711, "step": 4676 }, { "epoch": 0.21733271375464683, "grad_norm": 2.062912655063383, "learning_rate": 9.1115953552045e-06, "loss": 0.081, "step": 4677 }, { "epoch": 0.21737918215613383, "grad_norm": 2.160493248375229, "learning_rate": 9.111167108489172e-06, "loss": 0.1467, "step": 4678 }, { "epoch": 0.2174256505576208, "grad_norm": 2.1934660168983577, "learning_rate": 9.11073876865158e-06, "loss": 0.1138, "step": 4679 }, { "epoch": 0.2174721189591078, "grad_norm": 2.678563988055312, "learning_rate": 9.110310335701431e-06, "loss": 0.1366, "step": 4680 }, { "epoch": 0.2175185873605948, "grad_norm": 1.6277801437397936, "learning_rate": 9.109881809648427e-06, "loss": 0.068, "step": 4681 }, { "epoch": 0.2175650557620818, "grad_norm": 0.9714647483100882, "learning_rate": 9.109453190502273e-06, "loss": 0.0756, "step": 4682 }, { "epoch": 0.21761152416356877, "grad_norm": 2.2900183941065286, "learning_rate": 9.109024478272679e-06, "loss": 0.1012, "step": 4683 }, { "epoch": 0.21765799256505577, "grad_norm": 2.279639158002349, "learning_rate": 9.108595672969356e-06, "loss": 0.066, "step": 4684 }, { "epoch": 0.21770446096654275, "grad_norm": 0.9734545252501937, "learning_rate": 9.108166774602019e-06, "loss": 0.068, "step": 4685 }, { "epoch": 0.21775092936802973, "grad_norm": 2.0606072974106056, "learning_rate": 9.107737783180379e-06, "loss": 0.0989, "step": 4686 }, { "epoch": 0.21779739776951673, "grad_norm": 1.7083188620099863, "learning_rate": 9.107308698714155e-06, "loss": 0.1353, "step": 4687 }, { "epoch": 0.2178438661710037, "grad_norm": 1.624388385889405, "learning_rate": 9.106879521213066e-06, "loss": 0.0818, "step": 4688 }, { "epoch": 0.2178903345724907, "grad_norm": 3.0559641102754544, "learning_rate": 9.106450250686832e-06, "loss": 0.0693, "step": 4689 }, { "epoch": 0.2179368029739777, "grad_norm": 3.2713987329516625, "learning_rate": 9.106020887145179e-06, "loss": 0.1157, "step": 4690 }, { "epoch": 0.2179832713754647, "grad_norm": 1.8399899597761882, "learning_rate": 9.105591430597832e-06, "loss": 0.0508, "step": 4691 }, { "epoch": 0.21802973977695167, "grad_norm": 2.1314844116485436, "learning_rate": 9.105161881054519e-06, "loss": 0.1119, "step": 4692 }, { "epoch": 0.21807620817843867, "grad_norm": 1.5332798032664543, "learning_rate": 9.104732238524968e-06, "loss": 0.0884, "step": 4693 }, { "epoch": 0.21812267657992565, "grad_norm": 2.241498349222871, "learning_rate": 9.104302503018909e-06, "loss": 0.0722, "step": 4694 }, { "epoch": 0.21816914498141263, "grad_norm": 3.0973031386697802, "learning_rate": 9.103872674546081e-06, "loss": 0.1538, "step": 4695 }, { "epoch": 0.21821561338289963, "grad_norm": 1.4099362060602993, "learning_rate": 9.10344275311622e-06, "loss": 0.1256, "step": 4696 }, { "epoch": 0.2182620817843866, "grad_norm": 2.0720448860196985, "learning_rate": 9.103012738739057e-06, "loss": 0.1185, "step": 4697 }, { "epoch": 0.2183085501858736, "grad_norm": 1.5339289128244895, "learning_rate": 9.10258263142434e-06, "loss": 0.0697, "step": 4698 }, { "epoch": 0.2183550185873606, "grad_norm": 1.7548234624160985, "learning_rate": 9.102152431181808e-06, "loss": 0.0754, "step": 4699 }, { "epoch": 0.2184014869888476, "grad_norm": 1.4685390928575726, "learning_rate": 9.101722138021205e-06, "loss": 0.0477, "step": 4700 }, { "epoch": 0.21844795539033457, "grad_norm": 3.5511211220976207, "learning_rate": 9.10129175195228e-06, "loss": 0.0901, "step": 4701 }, { "epoch": 0.21849442379182157, "grad_norm": 1.698059646032786, "learning_rate": 9.10086127298478e-06, "loss": 0.1313, "step": 4702 }, { "epoch": 0.21854089219330855, "grad_norm": 1.22204444516817, "learning_rate": 9.100430701128455e-06, "loss": 0.101, "step": 4703 }, { "epoch": 0.21858736059479553, "grad_norm": 0.8942195727544903, "learning_rate": 9.100000036393061e-06, "loss": 0.0709, "step": 4704 }, { "epoch": 0.21863382899628253, "grad_norm": 2.6225915356861615, "learning_rate": 9.099569278788349e-06, "loss": 0.1042, "step": 4705 }, { "epoch": 0.2186802973977695, "grad_norm": 1.1802700652504665, "learning_rate": 9.09913842832408e-06, "loss": 0.0736, "step": 4706 }, { "epoch": 0.2187267657992565, "grad_norm": 2.46790828809982, "learning_rate": 9.098707485010009e-06, "loss": 0.1012, "step": 4707 }, { "epoch": 0.2187732342007435, "grad_norm": 1.8294794676582171, "learning_rate": 9.098276448855903e-06, "loss": 0.0883, "step": 4708 }, { "epoch": 0.2188197026022305, "grad_norm": 2.5045810084842763, "learning_rate": 9.097845319871519e-06, "loss": 0.1314, "step": 4709 }, { "epoch": 0.21886617100371747, "grad_norm": 1.9793306633201795, "learning_rate": 9.097414098066627e-06, "loss": 0.1418, "step": 4710 }, { "epoch": 0.21891263940520447, "grad_norm": 4.03948832320634, "learning_rate": 9.096982783450994e-06, "loss": 0.0883, "step": 4711 }, { "epoch": 0.21895910780669145, "grad_norm": 1.7624195374165115, "learning_rate": 9.096551376034388e-06, "loss": 0.1187, "step": 4712 }, { "epoch": 0.21900557620817843, "grad_norm": 1.5524582221326049, "learning_rate": 9.096119875826582e-06, "loss": 0.0969, "step": 4713 }, { "epoch": 0.21905204460966543, "grad_norm": 1.3273582210202173, "learning_rate": 9.095688282837348e-06, "loss": 0.0957, "step": 4714 }, { "epoch": 0.2190985130111524, "grad_norm": 2.223504244395058, "learning_rate": 9.095256597076464e-06, "loss": 0.0626, "step": 4715 }, { "epoch": 0.2191449814126394, "grad_norm": 1.649365384458187, "learning_rate": 9.09482481855371e-06, "loss": 0.1035, "step": 4716 }, { "epoch": 0.2191914498141264, "grad_norm": 2.7016314604583704, "learning_rate": 9.094392947278863e-06, "loss": 0.1318, "step": 4717 }, { "epoch": 0.2192379182156134, "grad_norm": 1.0772090702275479, "learning_rate": 9.093960983261707e-06, "loss": 0.1278, "step": 4718 }, { "epoch": 0.21928438661710037, "grad_norm": 0.7753463993298279, "learning_rate": 9.093528926512026e-06, "loss": 0.0595, "step": 4719 }, { "epoch": 0.21933085501858737, "grad_norm": 1.2807351840024745, "learning_rate": 9.093096777039605e-06, "loss": 0.085, "step": 4720 }, { "epoch": 0.21937732342007435, "grad_norm": 1.0361424921112585, "learning_rate": 9.092664534854237e-06, "loss": 0.0746, "step": 4721 }, { "epoch": 0.21942379182156133, "grad_norm": 3.3614249994417555, "learning_rate": 9.092232199965707e-06, "loss": 0.098, "step": 4722 }, { "epoch": 0.21947026022304833, "grad_norm": 0.9263723054184105, "learning_rate": 9.091799772383812e-06, "loss": 0.0699, "step": 4723 }, { "epoch": 0.2195167286245353, "grad_norm": 1.3538943611536793, "learning_rate": 9.091367252118344e-06, "loss": 0.0739, "step": 4724 }, { "epoch": 0.2195631970260223, "grad_norm": 1.7299015621853417, "learning_rate": 9.090934639179103e-06, "loss": 0.0974, "step": 4725 }, { "epoch": 0.2196096654275093, "grad_norm": 1.1280256514597886, "learning_rate": 9.090501933575887e-06, "loss": 0.0573, "step": 4726 }, { "epoch": 0.2196561338289963, "grad_norm": 1.3746908762453394, "learning_rate": 9.090069135318497e-06, "loss": 0.0881, "step": 4727 }, { "epoch": 0.21970260223048327, "grad_norm": 3.9992962446961196, "learning_rate": 9.089636244416736e-06, "loss": 0.1383, "step": 4728 }, { "epoch": 0.21974907063197027, "grad_norm": 1.5826765045731148, "learning_rate": 9.08920326088041e-06, "loss": 0.0921, "step": 4729 }, { "epoch": 0.21979553903345725, "grad_norm": 1.5433233247810558, "learning_rate": 9.088770184719327e-06, "loss": 0.0581, "step": 4730 }, { "epoch": 0.21984200743494423, "grad_norm": 0.9192699358002104, "learning_rate": 9.088337015943293e-06, "loss": 0.0508, "step": 4731 }, { "epoch": 0.21988847583643123, "grad_norm": 2.643823902880857, "learning_rate": 9.087903754562126e-06, "loss": 0.0766, "step": 4732 }, { "epoch": 0.2199349442379182, "grad_norm": 2.041961915320329, "learning_rate": 9.087470400585635e-06, "loss": 0.1085, "step": 4733 }, { "epoch": 0.2199814126394052, "grad_norm": 2.8608347611648246, "learning_rate": 9.087036954023639e-06, "loss": 0.0943, "step": 4734 }, { "epoch": 0.2200278810408922, "grad_norm": 1.1621240194472096, "learning_rate": 9.086603414885954e-06, "loss": 0.0988, "step": 4735 }, { "epoch": 0.2200743494423792, "grad_norm": 1.5402875207064104, "learning_rate": 9.086169783182399e-06, "loss": 0.0997, "step": 4736 }, { "epoch": 0.22012081784386617, "grad_norm": 5.778141283507369, "learning_rate": 9.085736058922799e-06, "loss": 0.1733, "step": 4737 }, { "epoch": 0.22016728624535317, "grad_norm": 1.9547320043172705, "learning_rate": 9.085302242116976e-06, "loss": 0.0564, "step": 4738 }, { "epoch": 0.22021375464684015, "grad_norm": 2.2195990129538834, "learning_rate": 9.084868332774758e-06, "loss": 0.0654, "step": 4739 }, { "epoch": 0.22026022304832713, "grad_norm": 2.8215920704349307, "learning_rate": 9.084434330905974e-06, "loss": 0.0699, "step": 4740 }, { "epoch": 0.22030669144981413, "grad_norm": 4.949484519824648, "learning_rate": 9.084000236520452e-06, "loss": 0.1181, "step": 4741 }, { "epoch": 0.2203531598513011, "grad_norm": 1.3880751121613313, "learning_rate": 9.083566049628027e-06, "loss": 0.0844, "step": 4742 }, { "epoch": 0.2203996282527881, "grad_norm": 1.4442978515542801, "learning_rate": 9.083131770238534e-06, "loss": 0.0844, "step": 4743 }, { "epoch": 0.2204460966542751, "grad_norm": 2.360320585972228, "learning_rate": 9.082697398361808e-06, "loss": 0.1092, "step": 4744 }, { "epoch": 0.2204925650557621, "grad_norm": 2.885426799384821, "learning_rate": 9.08226293400769e-06, "loss": 0.0884, "step": 4745 }, { "epoch": 0.22053903345724907, "grad_norm": 1.8622534982505583, "learning_rate": 9.08182837718602e-06, "loss": 0.0933, "step": 4746 }, { "epoch": 0.22058550185873607, "grad_norm": 1.2663555996227858, "learning_rate": 9.08139372790664e-06, "loss": 0.0816, "step": 4747 }, { "epoch": 0.22063197026022305, "grad_norm": 1.7238061667810813, "learning_rate": 9.080958986179395e-06, "loss": 0.0867, "step": 4748 }, { "epoch": 0.22067843866171002, "grad_norm": 1.6647298836986202, "learning_rate": 9.080524152014136e-06, "loss": 0.0935, "step": 4749 }, { "epoch": 0.22072490706319703, "grad_norm": 1.7263220806981145, "learning_rate": 9.080089225420712e-06, "loss": 0.1107, "step": 4750 }, { "epoch": 0.220771375464684, "grad_norm": 3.3917460294837936, "learning_rate": 9.07965420640897e-06, "loss": 0.0888, "step": 4751 }, { "epoch": 0.220817843866171, "grad_norm": 1.7716684637246665, "learning_rate": 9.079219094988766e-06, "loss": 0.09, "step": 4752 }, { "epoch": 0.220864312267658, "grad_norm": 1.4445798200295936, "learning_rate": 9.07878389116996e-06, "loss": 0.1012, "step": 4753 }, { "epoch": 0.220910780669145, "grad_norm": 1.2837113299406926, "learning_rate": 9.078348594962402e-06, "loss": 0.1412, "step": 4754 }, { "epoch": 0.22095724907063197, "grad_norm": 1.2523629980525866, "learning_rate": 9.07791320637596e-06, "loss": 0.0851, "step": 4755 }, { "epoch": 0.22100371747211897, "grad_norm": 1.3604342330502532, "learning_rate": 9.077477725420488e-06, "loss": 0.1114, "step": 4756 }, { "epoch": 0.22105018587360595, "grad_norm": 1.9773798418673856, "learning_rate": 9.077042152105855e-06, "loss": 0.0724, "step": 4757 }, { "epoch": 0.22109665427509292, "grad_norm": 0.945326639947411, "learning_rate": 9.076606486441928e-06, "loss": 0.0707, "step": 4758 }, { "epoch": 0.22114312267657993, "grad_norm": 2.1390912305733507, "learning_rate": 9.076170728438571e-06, "loss": 0.1139, "step": 4759 }, { "epoch": 0.2211895910780669, "grad_norm": 1.5010868825580457, "learning_rate": 9.07573487810566e-06, "loss": 0.0635, "step": 4760 }, { "epoch": 0.2212360594795539, "grad_norm": 1.1523112676506582, "learning_rate": 9.07529893545306e-06, "loss": 0.0733, "step": 4761 }, { "epoch": 0.22128252788104089, "grad_norm": 1.0218584771413695, "learning_rate": 9.074862900490655e-06, "loss": 0.0953, "step": 4762 }, { "epoch": 0.2213289962825279, "grad_norm": 1.6663371819433273, "learning_rate": 9.074426773228314e-06, "loss": 0.0849, "step": 4763 }, { "epoch": 0.22137546468401487, "grad_norm": 0.8834493335767055, "learning_rate": 9.073990553675918e-06, "loss": 0.0802, "step": 4764 }, { "epoch": 0.22142193308550187, "grad_norm": 1.9971245340624961, "learning_rate": 9.07355424184335e-06, "loss": 0.0935, "step": 4765 }, { "epoch": 0.22146840148698885, "grad_norm": 3.0066701630862993, "learning_rate": 9.073117837740489e-06, "loss": 0.0788, "step": 4766 }, { "epoch": 0.22151486988847582, "grad_norm": 3.1317894743869563, "learning_rate": 9.072681341377222e-06, "loss": 0.1602, "step": 4767 }, { "epoch": 0.22156133828996283, "grad_norm": 1.1737909317860498, "learning_rate": 9.072244752763437e-06, "loss": 0.0687, "step": 4768 }, { "epoch": 0.2216078066914498, "grad_norm": 1.4137256773911928, "learning_rate": 9.071808071909023e-06, "loss": 0.0782, "step": 4769 }, { "epoch": 0.2216542750929368, "grad_norm": 0.7540634981969634, "learning_rate": 9.07137129882387e-06, "loss": 0.0627, "step": 4770 }, { "epoch": 0.22170074349442379, "grad_norm": 2.6724381069029826, "learning_rate": 9.070934433517872e-06, "loss": 0.1279, "step": 4771 }, { "epoch": 0.2217472118959108, "grad_norm": 1.4134343410417611, "learning_rate": 9.070497476000925e-06, "loss": 0.0745, "step": 4772 }, { "epoch": 0.22179368029739777, "grad_norm": 2.2182483655840697, "learning_rate": 9.070060426282924e-06, "loss": 0.0846, "step": 4773 }, { "epoch": 0.22184014869888477, "grad_norm": 1.2980028448705165, "learning_rate": 9.069623284373773e-06, "loss": 0.0786, "step": 4774 }, { "epoch": 0.22188661710037175, "grad_norm": 0.7724256241460296, "learning_rate": 9.069186050283372e-06, "loss": 0.044, "step": 4775 }, { "epoch": 0.22193308550185872, "grad_norm": 1.7431223005452707, "learning_rate": 9.068748724021622e-06, "loss": 0.0835, "step": 4776 }, { "epoch": 0.22197955390334573, "grad_norm": 1.6784656020156516, "learning_rate": 9.068311305598433e-06, "loss": 0.0881, "step": 4777 }, { "epoch": 0.2220260223048327, "grad_norm": 1.2449766766204007, "learning_rate": 9.06787379502371e-06, "loss": 0.0794, "step": 4778 }, { "epoch": 0.2220724907063197, "grad_norm": 1.4588690047735384, "learning_rate": 9.067436192307364e-06, "loss": 0.0856, "step": 4779 }, { "epoch": 0.22211895910780668, "grad_norm": 1.2574507878976757, "learning_rate": 9.066998497459307e-06, "loss": 0.0923, "step": 4780 }, { "epoch": 0.2221654275092937, "grad_norm": 1.6688253344266921, "learning_rate": 9.066560710489455e-06, "loss": 0.0974, "step": 4781 }, { "epoch": 0.22221189591078067, "grad_norm": 1.187006139275598, "learning_rate": 9.066122831407722e-06, "loss": 0.0764, "step": 4782 }, { "epoch": 0.22225836431226767, "grad_norm": 3.60289992584757, "learning_rate": 9.065684860224028e-06, "loss": 0.1157, "step": 4783 }, { "epoch": 0.22230483271375465, "grad_norm": 3.6632949096142893, "learning_rate": 9.065246796948292e-06, "loss": 0.0798, "step": 4784 }, { "epoch": 0.22235130111524162, "grad_norm": 0.9965440031760561, "learning_rate": 9.064808641590439e-06, "loss": 0.1234, "step": 4785 }, { "epoch": 0.22239776951672863, "grad_norm": 2.109019877074865, "learning_rate": 9.06437039416039e-06, "loss": 0.0706, "step": 4786 }, { "epoch": 0.2224442379182156, "grad_norm": 1.5142607567434172, "learning_rate": 9.063932054668075e-06, "loss": 0.1429, "step": 4787 }, { "epoch": 0.2224907063197026, "grad_norm": 4.257899275698249, "learning_rate": 9.063493623123423e-06, "loss": 0.1475, "step": 4788 }, { "epoch": 0.22253717472118958, "grad_norm": 1.2722540810256318, "learning_rate": 9.063055099536363e-06, "loss": 0.072, "step": 4789 }, { "epoch": 0.2225836431226766, "grad_norm": 1.3851625622556865, "learning_rate": 9.062616483916827e-06, "loss": 0.1027, "step": 4790 }, { "epoch": 0.22263011152416357, "grad_norm": 1.1790071037109393, "learning_rate": 9.062177776274752e-06, "loss": 0.0882, "step": 4791 }, { "epoch": 0.22267657992565057, "grad_norm": 0.6450580431376265, "learning_rate": 9.061738976620076e-06, "loss": 0.0655, "step": 4792 }, { "epoch": 0.22272304832713755, "grad_norm": 1.836410082016358, "learning_rate": 9.061300084962738e-06, "loss": 0.1352, "step": 4793 }, { "epoch": 0.22276951672862452, "grad_norm": 1.3328575637477447, "learning_rate": 9.060861101312677e-06, "loss": 0.0887, "step": 4794 }, { "epoch": 0.22281598513011153, "grad_norm": 1.538553939704958, "learning_rate": 9.06042202567984e-06, "loss": 0.0862, "step": 4795 }, { "epoch": 0.2228624535315985, "grad_norm": 0.8234903562380294, "learning_rate": 9.05998285807417e-06, "loss": 0.0553, "step": 4796 }, { "epoch": 0.2229089219330855, "grad_norm": 1.8732938577970297, "learning_rate": 9.059543598505613e-06, "loss": 0.0397, "step": 4797 }, { "epoch": 0.22295539033457248, "grad_norm": 0.9844348277525514, "learning_rate": 9.059104246984123e-06, "loss": 0.0694, "step": 4798 }, { "epoch": 0.2230018587360595, "grad_norm": 1.5540148873501838, "learning_rate": 9.05866480351965e-06, "loss": 0.1027, "step": 4799 }, { "epoch": 0.22304832713754646, "grad_norm": 1.6733788365275692, "learning_rate": 9.058225268122148e-06, "loss": 0.0692, "step": 4800 }, { "epoch": 0.22309479553903347, "grad_norm": 0.5946068729790842, "learning_rate": 9.05778564080157e-06, "loss": 0.0535, "step": 4801 }, { "epoch": 0.22314126394052045, "grad_norm": 1.6762698124796234, "learning_rate": 9.057345921567878e-06, "loss": 0.1035, "step": 4802 }, { "epoch": 0.22318773234200742, "grad_norm": 1.1668114851518483, "learning_rate": 9.056906110431029e-06, "loss": 0.0864, "step": 4803 }, { "epoch": 0.22323420074349443, "grad_norm": 2.261913072789924, "learning_rate": 9.056466207400988e-06, "loss": 0.0949, "step": 4804 }, { "epoch": 0.2232806691449814, "grad_norm": 2.031939094897289, "learning_rate": 9.05602621248772e-06, "loss": 0.0823, "step": 4805 }, { "epoch": 0.2233271375464684, "grad_norm": 3.3853105715139975, "learning_rate": 9.055586125701187e-06, "loss": 0.0705, "step": 4806 }, { "epoch": 0.22337360594795538, "grad_norm": 0.7845103449008494, "learning_rate": 9.055145947051362e-06, "loss": 0.0496, "step": 4807 }, { "epoch": 0.2234200743494424, "grad_norm": 1.1585402157332376, "learning_rate": 9.05470567654821e-06, "loss": 0.1139, "step": 4808 }, { "epoch": 0.22346654275092936, "grad_norm": 0.7131038012819213, "learning_rate": 9.054265314201708e-06, "loss": 0.0602, "step": 4809 }, { "epoch": 0.22351301115241637, "grad_norm": 1.0424116710742062, "learning_rate": 9.053824860021832e-06, "loss": 0.0698, "step": 4810 }, { "epoch": 0.22355947955390335, "grad_norm": 1.586363897776284, "learning_rate": 9.053384314018554e-06, "loss": 0.0942, "step": 4811 }, { "epoch": 0.22360594795539032, "grad_norm": 2.055171156957815, "learning_rate": 9.052943676201857e-06, "loss": 0.0952, "step": 4812 }, { "epoch": 0.22365241635687733, "grad_norm": 1.9210091094605024, "learning_rate": 9.052502946581718e-06, "loss": 0.0903, "step": 4813 }, { "epoch": 0.2236988847583643, "grad_norm": 1.882163984525635, "learning_rate": 9.052062125168124e-06, "loss": 0.1662, "step": 4814 }, { "epoch": 0.2237453531598513, "grad_norm": 1.6011774007715123, "learning_rate": 9.051621211971056e-06, "loss": 0.0821, "step": 4815 }, { "epoch": 0.22379182156133828, "grad_norm": 1.7114058136168573, "learning_rate": 9.051180207000505e-06, "loss": 0.0828, "step": 4816 }, { "epoch": 0.2238382899628253, "grad_norm": 1.7396641198997966, "learning_rate": 9.050739110266457e-06, "loss": 0.0978, "step": 4817 }, { "epoch": 0.22388475836431226, "grad_norm": 1.0686577970150213, "learning_rate": 9.050297921778906e-06, "loss": 0.093, "step": 4818 }, { "epoch": 0.22393122676579927, "grad_norm": 1.1167274928993973, "learning_rate": 9.049856641547846e-06, "loss": 0.0812, "step": 4819 }, { "epoch": 0.22397769516728624, "grad_norm": 1.2178259818813277, "learning_rate": 9.049415269583268e-06, "loss": 0.0571, "step": 4820 }, { "epoch": 0.22402416356877322, "grad_norm": 3.037269448716474, "learning_rate": 9.048973805895171e-06, "loss": 0.1171, "step": 4821 }, { "epoch": 0.22407063197026023, "grad_norm": 1.1353747744395783, "learning_rate": 9.048532250493559e-06, "loss": 0.0795, "step": 4822 }, { "epoch": 0.2241171003717472, "grad_norm": 0.766077089950672, "learning_rate": 9.04809060338843e-06, "loss": 0.0581, "step": 4823 }, { "epoch": 0.2241635687732342, "grad_norm": 4.164092328560033, "learning_rate": 9.047648864589786e-06, "loss": 0.0821, "step": 4824 }, { "epoch": 0.22421003717472118, "grad_norm": 2.4204159076465004, "learning_rate": 9.047207034107636e-06, "loss": 0.0917, "step": 4825 }, { "epoch": 0.2242565055762082, "grad_norm": 1.2488995255675612, "learning_rate": 9.04676511195199e-06, "loss": 0.0544, "step": 4826 }, { "epoch": 0.22430297397769516, "grad_norm": 3.291727284159786, "learning_rate": 9.046323098132853e-06, "loss": 0.1258, "step": 4827 }, { "epoch": 0.22434944237918217, "grad_norm": 2.275582176494632, "learning_rate": 9.045880992660238e-06, "loss": 0.1056, "step": 4828 }, { "epoch": 0.22439591078066914, "grad_norm": 0.8516569523728125, "learning_rate": 9.045438795544161e-06, "loss": 0.0644, "step": 4829 }, { "epoch": 0.22444237918215612, "grad_norm": 1.586660295161562, "learning_rate": 9.044996506794638e-06, "loss": 0.1155, "step": 4830 }, { "epoch": 0.22448884758364313, "grad_norm": 0.8408335219074642, "learning_rate": 9.044554126421687e-06, "loss": 0.082, "step": 4831 }, { "epoch": 0.2245353159851301, "grad_norm": 1.1521345075049945, "learning_rate": 9.044111654435328e-06, "loss": 0.0724, "step": 4832 }, { "epoch": 0.2245817843866171, "grad_norm": 0.9012945700705869, "learning_rate": 9.043669090845584e-06, "loss": 0.0495, "step": 4833 }, { "epoch": 0.22462825278810408, "grad_norm": 1.5286120696341037, "learning_rate": 9.043226435662478e-06, "loss": 0.0828, "step": 4834 }, { "epoch": 0.2246747211895911, "grad_norm": 1.3666041210950044, "learning_rate": 9.042783688896039e-06, "loss": 0.0823, "step": 4835 }, { "epoch": 0.22472118959107806, "grad_norm": 2.344738819732348, "learning_rate": 9.042340850556294e-06, "loss": 0.0599, "step": 4836 }, { "epoch": 0.22476765799256507, "grad_norm": 1.161337567191643, "learning_rate": 9.041897920653274e-06, "loss": 0.0951, "step": 4837 }, { "epoch": 0.22481412639405204, "grad_norm": 1.281137892925639, "learning_rate": 9.041454899197013e-06, "loss": 0.0656, "step": 4838 }, { "epoch": 0.22486059479553902, "grad_norm": 1.4252882415982355, "learning_rate": 9.041011786197544e-06, "loss": 0.0936, "step": 4839 }, { "epoch": 0.22490706319702602, "grad_norm": 5.172873428284395, "learning_rate": 9.040568581664906e-06, "loss": 0.1153, "step": 4840 }, { "epoch": 0.224953531598513, "grad_norm": 1.0635777736298182, "learning_rate": 9.040125285609135e-06, "loss": 0.0655, "step": 4841 }, { "epoch": 0.225, "grad_norm": 2.262453240114287, "learning_rate": 9.039681898040276e-06, "loss": 0.0905, "step": 4842 }, { "epoch": 0.22504646840148698, "grad_norm": 0.9352818359966972, "learning_rate": 9.03923841896837e-06, "loss": 0.0693, "step": 4843 }, { "epoch": 0.225092936802974, "grad_norm": 1.577490461976426, "learning_rate": 9.038794848403463e-06, "loss": 0.1096, "step": 4844 }, { "epoch": 0.22513940520446096, "grad_norm": 2.272890878056086, "learning_rate": 9.0383511863556e-06, "loss": 0.1289, "step": 4845 }, { "epoch": 0.22518587360594797, "grad_norm": 0.9433902708439443, "learning_rate": 9.037907432834833e-06, "loss": 0.0889, "step": 4846 }, { "epoch": 0.22523234200743494, "grad_norm": 0.8297756067392708, "learning_rate": 9.037463587851214e-06, "loss": 0.0476, "step": 4847 }, { "epoch": 0.22527881040892192, "grad_norm": 1.670803190362835, "learning_rate": 9.037019651414795e-06, "loss": 0.0577, "step": 4848 }, { "epoch": 0.22532527881040892, "grad_norm": 2.6083079405513856, "learning_rate": 9.036575623535633e-06, "loss": 0.0683, "step": 4849 }, { "epoch": 0.2253717472118959, "grad_norm": 2.195088787726247, "learning_rate": 9.036131504223783e-06, "loss": 0.0981, "step": 4850 }, { "epoch": 0.2254182156133829, "grad_norm": 1.6540853648063083, "learning_rate": 9.035687293489308e-06, "loss": 0.0855, "step": 4851 }, { "epoch": 0.22546468401486988, "grad_norm": 1.4346625719819885, "learning_rate": 9.035242991342265e-06, "loss": 0.0468, "step": 4852 }, { "epoch": 0.22551115241635689, "grad_norm": 1.3863098834602516, "learning_rate": 9.034798597792725e-06, "loss": 0.1176, "step": 4853 }, { "epoch": 0.22555762081784386, "grad_norm": 2.6190054862828758, "learning_rate": 9.034354112850749e-06, "loss": 0.0955, "step": 4854 }, { "epoch": 0.22560408921933087, "grad_norm": 1.6414890292651132, "learning_rate": 9.033909536526406e-06, "loss": 0.0727, "step": 4855 }, { "epoch": 0.22565055762081784, "grad_norm": 3.333112877959099, "learning_rate": 9.033464868829767e-06, "loss": 0.1263, "step": 4856 }, { "epoch": 0.22569702602230482, "grad_norm": 1.0145157994345195, "learning_rate": 9.033020109770902e-06, "loss": 0.0625, "step": 4857 }, { "epoch": 0.22574349442379182, "grad_norm": 0.9327206876500592, "learning_rate": 9.032575259359888e-06, "loss": 0.0629, "step": 4858 }, { "epoch": 0.2257899628252788, "grad_norm": 2.2500520330197875, "learning_rate": 9.0321303176068e-06, "loss": 0.0762, "step": 4859 }, { "epoch": 0.2258364312267658, "grad_norm": 0.942283704103695, "learning_rate": 9.031685284521715e-06, "loss": 0.0673, "step": 4860 }, { "epoch": 0.22588289962825278, "grad_norm": 1.0243001733436423, "learning_rate": 9.031240160114718e-06, "loss": 0.0457, "step": 4861 }, { "epoch": 0.22592936802973979, "grad_norm": 1.4120497504743639, "learning_rate": 9.030794944395888e-06, "loss": 0.1324, "step": 4862 }, { "epoch": 0.22597583643122676, "grad_norm": 0.857041603386624, "learning_rate": 9.03034963737531e-06, "loss": 0.0655, "step": 4863 }, { "epoch": 0.22602230483271377, "grad_norm": 1.4534551447558868, "learning_rate": 9.02990423906307e-06, "loss": 0.0503, "step": 4864 }, { "epoch": 0.22606877323420074, "grad_norm": 1.6539357656486988, "learning_rate": 9.029458749469257e-06, "loss": 0.1545, "step": 4865 }, { "epoch": 0.22611524163568772, "grad_norm": 0.8166814380250994, "learning_rate": 9.029013168603965e-06, "loss": 0.0675, "step": 4866 }, { "epoch": 0.22616171003717472, "grad_norm": 1.0504451533905612, "learning_rate": 9.028567496477281e-06, "loss": 0.0776, "step": 4867 }, { "epoch": 0.2262081784386617, "grad_norm": 0.9662573521654697, "learning_rate": 9.028121733099304e-06, "loss": 0.066, "step": 4868 }, { "epoch": 0.2262546468401487, "grad_norm": 1.7504382786850101, "learning_rate": 9.027675878480131e-06, "loss": 0.1037, "step": 4869 }, { "epoch": 0.22630111524163568, "grad_norm": 1.5912380533786301, "learning_rate": 9.027229932629862e-06, "loss": 0.1142, "step": 4870 }, { "epoch": 0.22634758364312269, "grad_norm": 1.8172207350366183, "learning_rate": 9.026783895558592e-06, "loss": 0.1242, "step": 4871 }, { "epoch": 0.22639405204460966, "grad_norm": 2.4991963603782676, "learning_rate": 9.026337767276433e-06, "loss": 0.0734, "step": 4872 }, { "epoch": 0.22644052044609667, "grad_norm": 1.0531698143612678, "learning_rate": 9.025891547793485e-06, "loss": 0.089, "step": 4873 }, { "epoch": 0.22648698884758364, "grad_norm": 1.5709804525183573, "learning_rate": 9.025445237119855e-06, "loss": 0.1444, "step": 4874 }, { "epoch": 0.22653345724907062, "grad_norm": 2.31728030574012, "learning_rate": 9.024998835265654e-06, "loss": 0.1036, "step": 4875 }, { "epoch": 0.22657992565055762, "grad_norm": 2.2001482565895754, "learning_rate": 9.024552342240993e-06, "loss": 0.0562, "step": 4876 }, { "epoch": 0.2266263940520446, "grad_norm": 3.0160373159282883, "learning_rate": 9.024105758055986e-06, "loss": 0.0662, "step": 4877 }, { "epoch": 0.2266728624535316, "grad_norm": 1.3176317922876304, "learning_rate": 9.023659082720749e-06, "loss": 0.1053, "step": 4878 }, { "epoch": 0.22671933085501858, "grad_norm": 1.036767781414682, "learning_rate": 9.023212316245397e-06, "loss": 0.0815, "step": 4879 }, { "epoch": 0.22676579925650558, "grad_norm": 1.160203275045531, "learning_rate": 9.022765458640054e-06, "loss": 0.0968, "step": 4880 }, { "epoch": 0.22681226765799256, "grad_norm": 2.6740869489164227, "learning_rate": 9.022318509914838e-06, "loss": 0.1054, "step": 4881 }, { "epoch": 0.22685873605947957, "grad_norm": 1.086867481176666, "learning_rate": 9.021871470079875e-06, "loss": 0.0813, "step": 4882 }, { "epoch": 0.22690520446096654, "grad_norm": 1.4343736046269704, "learning_rate": 9.02142433914529e-06, "loss": 0.0793, "step": 4883 }, { "epoch": 0.22695167286245352, "grad_norm": 1.1356663854646032, "learning_rate": 9.020977117121212e-06, "loss": 0.0796, "step": 4884 }, { "epoch": 0.22699814126394052, "grad_norm": 0.6039947663841188, "learning_rate": 9.020529804017769e-06, "loss": 0.0174, "step": 4885 }, { "epoch": 0.2270446096654275, "grad_norm": 1.1517701514676542, "learning_rate": 9.020082399845095e-06, "loss": 0.0595, "step": 4886 }, { "epoch": 0.2270910780669145, "grad_norm": 1.9640480929959852, "learning_rate": 9.019634904613324e-06, "loss": 0.0761, "step": 4887 }, { "epoch": 0.22713754646840148, "grad_norm": 0.7597519939643143, "learning_rate": 9.019187318332592e-06, "loss": 0.0781, "step": 4888 }, { "epoch": 0.22718401486988848, "grad_norm": 1.264699746130359, "learning_rate": 9.018739641013034e-06, "loss": 0.0598, "step": 4889 }, { "epoch": 0.22723048327137546, "grad_norm": 1.1228509524097758, "learning_rate": 9.018291872664798e-06, "loss": 0.0713, "step": 4890 }, { "epoch": 0.22727695167286247, "grad_norm": 2.288080047732332, "learning_rate": 9.01784401329802e-06, "loss": 0.114, "step": 4891 }, { "epoch": 0.22732342007434944, "grad_norm": 0.9998725444075681, "learning_rate": 9.017396062922846e-06, "loss": 0.0912, "step": 4892 }, { "epoch": 0.22736988847583642, "grad_norm": 1.9959936991874352, "learning_rate": 9.016948021549422e-06, "loss": 0.0785, "step": 4893 }, { "epoch": 0.22741635687732342, "grad_norm": 1.214833417987921, "learning_rate": 9.0164998891879e-06, "loss": 0.1168, "step": 4894 }, { "epoch": 0.2274628252788104, "grad_norm": 1.4570178554773638, "learning_rate": 9.016051665848427e-06, "loss": 0.0817, "step": 4895 }, { "epoch": 0.2275092936802974, "grad_norm": 1.1159412465040721, "learning_rate": 9.015603351541155e-06, "loss": 0.0557, "step": 4896 }, { "epoch": 0.22755576208178438, "grad_norm": 1.0930295189949186, "learning_rate": 9.015154946276244e-06, "loss": 0.1044, "step": 4897 }, { "epoch": 0.22760223048327138, "grad_norm": 1.147679981196155, "learning_rate": 9.014706450063847e-06, "loss": 0.0742, "step": 4898 }, { "epoch": 0.22764869888475836, "grad_norm": 3.9347370204197167, "learning_rate": 9.014257862914122e-06, "loss": 0.1192, "step": 4899 }, { "epoch": 0.22769516728624536, "grad_norm": 1.2171701532969803, "learning_rate": 9.013809184837232e-06, "loss": 0.0782, "step": 4900 }, { "epoch": 0.22774163568773234, "grad_norm": 1.0130606734679766, "learning_rate": 9.013360415843339e-06, "loss": 0.063, "step": 4901 }, { "epoch": 0.22778810408921932, "grad_norm": 0.9609527063152848, "learning_rate": 9.01291155594261e-06, "loss": 0.0502, "step": 4902 }, { "epoch": 0.22783457249070632, "grad_norm": 1.276046210966069, "learning_rate": 9.012462605145211e-06, "loss": 0.095, "step": 4903 }, { "epoch": 0.2278810408921933, "grad_norm": 1.1837815256734043, "learning_rate": 9.01201356346131e-06, "loss": 0.0997, "step": 4904 }, { "epoch": 0.2279275092936803, "grad_norm": 1.2415180821230856, "learning_rate": 9.011564430901078e-06, "loss": 0.0686, "step": 4905 }, { "epoch": 0.22797397769516728, "grad_norm": 1.489111745343224, "learning_rate": 9.011115207474691e-06, "loss": 0.0851, "step": 4906 }, { "epoch": 0.22802044609665428, "grad_norm": 1.7252721806082087, "learning_rate": 9.010665893192324e-06, "loss": 0.0876, "step": 4907 }, { "epoch": 0.22806691449814126, "grad_norm": 1.4806800804591849, "learning_rate": 9.010216488064151e-06, "loss": 0.1029, "step": 4908 }, { "epoch": 0.22811338289962826, "grad_norm": 1.7704832479886108, "learning_rate": 9.009766992100355e-06, "loss": 0.0806, "step": 4909 }, { "epoch": 0.22815985130111524, "grad_norm": 0.9513183863981768, "learning_rate": 9.009317405311118e-06, "loss": 0.0801, "step": 4910 }, { "epoch": 0.22820631970260222, "grad_norm": 1.0492766389461634, "learning_rate": 9.00886772770662e-06, "loss": 0.0759, "step": 4911 }, { "epoch": 0.22825278810408922, "grad_norm": 1.4948463684683737, "learning_rate": 9.008417959297049e-06, "loss": 0.0857, "step": 4912 }, { "epoch": 0.2282992565055762, "grad_norm": 1.661562383963794, "learning_rate": 9.007968100092592e-06, "loss": 0.0742, "step": 4913 }, { "epoch": 0.2283457249070632, "grad_norm": 1.5102044262109526, "learning_rate": 9.007518150103441e-06, "loss": 0.0694, "step": 4914 }, { "epoch": 0.22839219330855018, "grad_norm": 1.8098865495801395, "learning_rate": 9.007068109339783e-06, "loss": 0.08, "step": 4915 }, { "epoch": 0.22843866171003718, "grad_norm": 1.7974754461345088, "learning_rate": 9.006617977811818e-06, "loss": 0.1237, "step": 4916 }, { "epoch": 0.22848513011152416, "grad_norm": 0.8563697192706231, "learning_rate": 9.006167755529738e-06, "loss": 0.0795, "step": 4917 }, { "epoch": 0.22853159851301116, "grad_norm": 1.4426556793088028, "learning_rate": 9.005717442503741e-06, "loss": 0.0974, "step": 4918 }, { "epoch": 0.22857806691449814, "grad_norm": 1.8543061312152955, "learning_rate": 9.005267038744029e-06, "loss": 0.115, "step": 4919 }, { "epoch": 0.22862453531598512, "grad_norm": 4.733426339154878, "learning_rate": 9.004816544260804e-06, "loss": 0.0651, "step": 4920 }, { "epoch": 0.22867100371747212, "grad_norm": 0.9625095240779431, "learning_rate": 9.004365959064268e-06, "loss": 0.0646, "step": 4921 }, { "epoch": 0.2287174721189591, "grad_norm": 1.0455831139706044, "learning_rate": 9.003915283164628e-06, "loss": 0.0696, "step": 4922 }, { "epoch": 0.2287639405204461, "grad_norm": 2.969099104332526, "learning_rate": 9.003464516572094e-06, "loss": 0.1411, "step": 4923 }, { "epoch": 0.22881040892193308, "grad_norm": 1.8279182090543824, "learning_rate": 9.003013659296874e-06, "loss": 0.1005, "step": 4924 }, { "epoch": 0.22885687732342008, "grad_norm": 0.9159361868109351, "learning_rate": 9.00256271134918e-06, "loss": 0.047, "step": 4925 }, { "epoch": 0.22890334572490706, "grad_norm": 1.193893815750032, "learning_rate": 9.002111672739232e-06, "loss": 0.072, "step": 4926 }, { "epoch": 0.22894981412639406, "grad_norm": 1.0284522580005657, "learning_rate": 9.00166054347724e-06, "loss": 0.0656, "step": 4927 }, { "epoch": 0.22899628252788104, "grad_norm": 1.5650577026081685, "learning_rate": 9.001209323573426e-06, "loss": 0.0475, "step": 4928 }, { "epoch": 0.22904275092936802, "grad_norm": 1.73578582385025, "learning_rate": 9.00075801303801e-06, "loss": 0.0983, "step": 4929 }, { "epoch": 0.22908921933085502, "grad_norm": 0.7331504693745089, "learning_rate": 9.000306611881211e-06, "loss": 0.0641, "step": 4930 }, { "epoch": 0.229135687732342, "grad_norm": 2.4649353812473676, "learning_rate": 8.99985512011326e-06, "loss": 0.146, "step": 4931 }, { "epoch": 0.229182156133829, "grad_norm": 1.7498588067248526, "learning_rate": 8.99940353774438e-06, "loss": 0.1047, "step": 4932 }, { "epoch": 0.22922862453531598, "grad_norm": 0.7630171601041552, "learning_rate": 8.9989518647848e-06, "loss": 0.0473, "step": 4933 }, { "epoch": 0.22927509293680298, "grad_norm": 2.88904017424593, "learning_rate": 8.998500101244754e-06, "loss": 0.0738, "step": 4934 }, { "epoch": 0.22932156133828996, "grad_norm": 1.2109041937264318, "learning_rate": 8.99804824713447e-06, "loss": 0.0975, "step": 4935 }, { "epoch": 0.22936802973977696, "grad_norm": 1.3136222657735384, "learning_rate": 8.997596302464185e-06, "loss": 0.1108, "step": 4936 }, { "epoch": 0.22941449814126394, "grad_norm": 1.2436220601157908, "learning_rate": 8.997144267244137e-06, "loss": 0.0667, "step": 4937 }, { "epoch": 0.22946096654275092, "grad_norm": 10.353557018340977, "learning_rate": 8.996692141484565e-06, "loss": 0.1555, "step": 4938 }, { "epoch": 0.22950743494423792, "grad_norm": 1.9255140553592147, "learning_rate": 8.996239925195708e-06, "loss": 0.0561, "step": 4939 }, { "epoch": 0.2295539033457249, "grad_norm": 2.4155117228990024, "learning_rate": 8.995787618387813e-06, "loss": 0.1213, "step": 4940 }, { "epoch": 0.2296003717472119, "grad_norm": 1.6803787958671228, "learning_rate": 8.995335221071123e-06, "loss": 0.1038, "step": 4941 }, { "epoch": 0.22964684014869888, "grad_norm": 1.6764678185177913, "learning_rate": 8.994882733255884e-06, "loss": 0.0995, "step": 4942 }, { "epoch": 0.22969330855018588, "grad_norm": 1.6619306855181415, "learning_rate": 8.994430154952347e-06, "loss": 0.0824, "step": 4943 }, { "epoch": 0.22973977695167286, "grad_norm": 1.416555359829016, "learning_rate": 8.993977486170761e-06, "loss": 0.0941, "step": 4944 }, { "epoch": 0.22978624535315986, "grad_norm": 1.1008554692439139, "learning_rate": 8.993524726921383e-06, "loss": 0.0772, "step": 4945 }, { "epoch": 0.22983271375464684, "grad_norm": 2.276305627330519, "learning_rate": 8.993071877214467e-06, "loss": 0.1367, "step": 4946 }, { "epoch": 0.22987918215613382, "grad_norm": 1.091840753089203, "learning_rate": 8.99261893706027e-06, "loss": 0.0891, "step": 4947 }, { "epoch": 0.22992565055762082, "grad_norm": 1.3323271602402997, "learning_rate": 8.992165906469052e-06, "loss": 0.0793, "step": 4948 }, { "epoch": 0.2299721189591078, "grad_norm": 4.602757415813125, "learning_rate": 8.991712785451074e-06, "loss": 0.1557, "step": 4949 }, { "epoch": 0.2300185873605948, "grad_norm": 8.305455310321342, "learning_rate": 8.991259574016603e-06, "loss": 0.1934, "step": 4950 }, { "epoch": 0.23006505576208178, "grad_norm": 1.6840098018121035, "learning_rate": 8.990806272175898e-06, "loss": 0.088, "step": 4951 }, { "epoch": 0.23011152416356878, "grad_norm": 2.071336003903814, "learning_rate": 8.990352879939232e-06, "loss": 0.075, "step": 4952 }, { "epoch": 0.23015799256505576, "grad_norm": 3.542818745201116, "learning_rate": 8.989899397316875e-06, "loss": 0.0747, "step": 4953 }, { "epoch": 0.23020446096654276, "grad_norm": 4.137966063636282, "learning_rate": 8.989445824319097e-06, "loss": 0.1149, "step": 4954 }, { "epoch": 0.23025092936802974, "grad_norm": 1.3653582178071488, "learning_rate": 8.98899216095617e-06, "loss": 0.0759, "step": 4955 }, { "epoch": 0.23029739776951672, "grad_norm": 1.1626272619563673, "learning_rate": 8.988538407238373e-06, "loss": 0.0869, "step": 4956 }, { "epoch": 0.23034386617100372, "grad_norm": 1.572980411635137, "learning_rate": 8.988084563175985e-06, "loss": 0.0483, "step": 4957 }, { "epoch": 0.2303903345724907, "grad_norm": 2.27415758530062, "learning_rate": 8.987630628779283e-06, "loss": 0.0897, "step": 4958 }, { "epoch": 0.2304368029739777, "grad_norm": 2.35273559463773, "learning_rate": 8.98717660405855e-06, "loss": 0.0854, "step": 4959 }, { "epoch": 0.23048327137546468, "grad_norm": 7.96641360605731, "learning_rate": 8.98672248902407e-06, "loss": 0.074, "step": 4960 }, { "epoch": 0.23052973977695168, "grad_norm": 1.6161963942894397, "learning_rate": 8.986268283686133e-06, "loss": 0.0378, "step": 4961 }, { "epoch": 0.23057620817843866, "grad_norm": 1.0775308971194122, "learning_rate": 8.98581398805502e-06, "loss": 0.0556, "step": 4962 }, { "epoch": 0.23062267657992566, "grad_norm": 1.3573512145986162, "learning_rate": 8.985359602141028e-06, "loss": 0.0754, "step": 4963 }, { "epoch": 0.23066914498141264, "grad_norm": 2.5703039685904963, "learning_rate": 8.984905125954444e-06, "loss": 0.1052, "step": 4964 }, { "epoch": 0.23071561338289961, "grad_norm": 1.4856977260197959, "learning_rate": 8.984450559505567e-06, "loss": 0.1109, "step": 4965 }, { "epoch": 0.23076208178438662, "grad_norm": 2.721801319587142, "learning_rate": 8.98399590280469e-06, "loss": 0.1275, "step": 4966 }, { "epoch": 0.2308085501858736, "grad_norm": 2.448741638698923, "learning_rate": 8.983541155862114e-06, "loss": 0.0669, "step": 4967 }, { "epoch": 0.2308550185873606, "grad_norm": 3.634237091126485, "learning_rate": 8.983086318688137e-06, "loss": 0.1244, "step": 4968 }, { "epoch": 0.23090148698884758, "grad_norm": 1.6549888097035272, "learning_rate": 8.982631391293064e-06, "loss": 0.0861, "step": 4969 }, { "epoch": 0.23094795539033458, "grad_norm": 3.06218973524868, "learning_rate": 8.982176373687197e-06, "loss": 0.0721, "step": 4970 }, { "epoch": 0.23099442379182156, "grad_norm": 1.8510587451136176, "learning_rate": 8.981721265880843e-06, "loss": 0.1251, "step": 4971 }, { "epoch": 0.23104089219330856, "grad_norm": 2.3904169811119202, "learning_rate": 8.981266067884314e-06, "loss": 0.0891, "step": 4972 }, { "epoch": 0.23108736059479554, "grad_norm": 1.1480601527157104, "learning_rate": 8.980810779707919e-06, "loss": 0.0849, "step": 4973 }, { "epoch": 0.23113382899628251, "grad_norm": 1.7664182955276577, "learning_rate": 8.980355401361968e-06, "loss": 0.1026, "step": 4974 }, { "epoch": 0.23118029739776952, "grad_norm": 1.695585906839542, "learning_rate": 8.979899932856778e-06, "loss": 0.1533, "step": 4975 }, { "epoch": 0.2312267657992565, "grad_norm": 2.2481659678730255, "learning_rate": 8.979444374202666e-06, "loss": 0.0916, "step": 4976 }, { "epoch": 0.2312732342007435, "grad_norm": 2.0106438234162476, "learning_rate": 8.97898872540995e-06, "loss": 0.1157, "step": 4977 }, { "epoch": 0.23131970260223048, "grad_norm": 1.0546180860607655, "learning_rate": 8.978532986488953e-06, "loss": 0.0293, "step": 4978 }, { "epoch": 0.23136617100371748, "grad_norm": 1.4775936344183724, "learning_rate": 8.978077157449996e-06, "loss": 0.1106, "step": 4979 }, { "epoch": 0.23141263940520446, "grad_norm": 1.289967949656818, "learning_rate": 8.977621238303405e-06, "loss": 0.0787, "step": 4980 }, { "epoch": 0.23145910780669146, "grad_norm": 2.3075165642517304, "learning_rate": 8.977165229059505e-06, "loss": 0.1045, "step": 4981 }, { "epoch": 0.23150557620817844, "grad_norm": 1.89190778173152, "learning_rate": 8.976709129728628e-06, "loss": 0.1292, "step": 4982 }, { "epoch": 0.23155204460966541, "grad_norm": 0.6691628289319862, "learning_rate": 8.976252940321103e-06, "loss": 0.0583, "step": 4983 }, { "epoch": 0.23159851301115242, "grad_norm": 1.2133184994108408, "learning_rate": 8.975796660847266e-06, "loss": 0.0611, "step": 4984 }, { "epoch": 0.2316449814126394, "grad_norm": 1.7203654772368546, "learning_rate": 8.975340291317448e-06, "loss": 0.1042, "step": 4985 }, { "epoch": 0.2316914498141264, "grad_norm": 1.845433963908992, "learning_rate": 8.97488383174199e-06, "loss": 0.069, "step": 4986 }, { "epoch": 0.23173791821561338, "grad_norm": 1.4167504314597028, "learning_rate": 8.974427282131229e-06, "loss": 0.0661, "step": 4987 }, { "epoch": 0.23178438661710038, "grad_norm": 1.8597701478457453, "learning_rate": 8.973970642495506e-06, "loss": 0.0685, "step": 4988 }, { "epoch": 0.23183085501858736, "grad_norm": 1.035009441277181, "learning_rate": 8.973513912845166e-06, "loss": 0.0599, "step": 4989 }, { "epoch": 0.23187732342007436, "grad_norm": 1.496685404183829, "learning_rate": 8.973057093190554e-06, "loss": 0.0516, "step": 4990 }, { "epoch": 0.23192379182156134, "grad_norm": 1.365208999077902, "learning_rate": 8.972600183542017e-06, "loss": 0.0673, "step": 4991 }, { "epoch": 0.2319702602230483, "grad_norm": 1.1333211775911254, "learning_rate": 8.972143183909905e-06, "loss": 0.072, "step": 4992 }, { "epoch": 0.23201672862453532, "grad_norm": 1.45739848250365, "learning_rate": 8.97168609430457e-06, "loss": 0.0989, "step": 4993 }, { "epoch": 0.2320631970260223, "grad_norm": 2.01770522005644, "learning_rate": 8.971228914736365e-06, "loss": 0.0539, "step": 4994 }, { "epoch": 0.2321096654275093, "grad_norm": 1.1873624907176714, "learning_rate": 8.970771645215643e-06, "loss": 0.0802, "step": 4995 }, { "epoch": 0.23215613382899628, "grad_norm": 1.3581218157445798, "learning_rate": 8.970314285752766e-06, "loss": 0.0848, "step": 4996 }, { "epoch": 0.23220260223048328, "grad_norm": 2.662848888577399, "learning_rate": 8.969856836358092e-06, "loss": 0.0823, "step": 4997 }, { "epoch": 0.23224907063197026, "grad_norm": 1.6111613083060328, "learning_rate": 8.969399297041983e-06, "loss": 0.0674, "step": 4998 }, { "epoch": 0.23229553903345726, "grad_norm": 3.2053773909105687, "learning_rate": 8.968941667814801e-06, "loss": 0.1322, "step": 4999 }, { "epoch": 0.23234200743494424, "grad_norm": 1.1964539023858005, "learning_rate": 8.968483948686912e-06, "loss": 0.0817, "step": 5000 }, { "epoch": 0.2323884758364312, "grad_norm": 1.0275353533378466, "learning_rate": 8.968026139668686e-06, "loss": 0.0843, "step": 5001 }, { "epoch": 0.23243494423791822, "grad_norm": 1.4741540044112649, "learning_rate": 8.967568240770492e-06, "loss": 0.0616, "step": 5002 }, { "epoch": 0.2324814126394052, "grad_norm": 1.4137112306042972, "learning_rate": 8.967110252002702e-06, "loss": 0.0856, "step": 5003 }, { "epoch": 0.2325278810408922, "grad_norm": 0.9310275712640277, "learning_rate": 8.96665217337569e-06, "loss": 0.079, "step": 5004 }, { "epoch": 0.23257434944237917, "grad_norm": 1.6821475240044936, "learning_rate": 8.966194004899829e-06, "loss": 0.0834, "step": 5005 }, { "epoch": 0.23262081784386618, "grad_norm": 1.261162111880674, "learning_rate": 8.9657357465855e-06, "loss": 0.0773, "step": 5006 }, { "epoch": 0.23266728624535316, "grad_norm": 2.015536392875122, "learning_rate": 8.965277398443085e-06, "loss": 0.1063, "step": 5007 }, { "epoch": 0.23271375464684016, "grad_norm": 2.2869999467012065, "learning_rate": 8.96481896048296e-06, "loss": 0.0846, "step": 5008 }, { "epoch": 0.23276022304832714, "grad_norm": 2.064906348136871, "learning_rate": 8.964360432715515e-06, "loss": 0.0844, "step": 5009 }, { "epoch": 0.2328066914498141, "grad_norm": 0.7772752621667226, "learning_rate": 8.963901815151132e-06, "loss": 0.0455, "step": 5010 }, { "epoch": 0.23285315985130112, "grad_norm": 1.3028184257766633, "learning_rate": 8.963443107800203e-06, "loss": 0.0638, "step": 5011 }, { "epoch": 0.2328996282527881, "grad_norm": 2.192424455030389, "learning_rate": 8.962984310673115e-06, "loss": 0.1375, "step": 5012 }, { "epoch": 0.2329460966542751, "grad_norm": 0.9532388534946412, "learning_rate": 8.962525423780261e-06, "loss": 0.0409, "step": 5013 }, { "epoch": 0.23299256505576207, "grad_norm": 1.5718406643488234, "learning_rate": 8.962066447132036e-06, "loss": 0.0588, "step": 5014 }, { "epoch": 0.23303903345724908, "grad_norm": 1.2483792927110726, "learning_rate": 8.961607380738836e-06, "loss": 0.089, "step": 5015 }, { "epoch": 0.23308550185873605, "grad_norm": 1.872518692766613, "learning_rate": 8.96114822461106e-06, "loss": 0.0759, "step": 5016 }, { "epoch": 0.23313197026022306, "grad_norm": 3.552868842773568, "learning_rate": 8.960688978759106e-06, "loss": 0.1099, "step": 5017 }, { "epoch": 0.23317843866171004, "grad_norm": 3.3982631688580796, "learning_rate": 8.960229643193381e-06, "loss": 0.1086, "step": 5018 }, { "epoch": 0.233224907063197, "grad_norm": 1.8354645563480427, "learning_rate": 8.959770217924284e-06, "loss": 0.0978, "step": 5019 }, { "epoch": 0.23327137546468402, "grad_norm": 2.9489380050640532, "learning_rate": 8.959310702962224e-06, "loss": 0.1304, "step": 5020 }, { "epoch": 0.233317843866171, "grad_norm": 2.3029501341787157, "learning_rate": 8.958851098317612e-06, "loss": 0.1171, "step": 5021 }, { "epoch": 0.233364312267658, "grad_norm": 0.49563765809856547, "learning_rate": 8.958391404000854e-06, "loss": 0.0467, "step": 5022 }, { "epoch": 0.23341078066914497, "grad_norm": 1.7520764599103178, "learning_rate": 8.957931620022364e-06, "loss": 0.0969, "step": 5023 }, { "epoch": 0.23345724907063198, "grad_norm": 1.8790935254502306, "learning_rate": 8.957471746392559e-06, "loss": 0.0988, "step": 5024 }, { "epoch": 0.23350371747211895, "grad_norm": 1.843112858914883, "learning_rate": 8.957011783121853e-06, "loss": 0.1072, "step": 5025 }, { "epoch": 0.23355018587360596, "grad_norm": 1.2660836435868466, "learning_rate": 8.956551730220664e-06, "loss": 0.0818, "step": 5026 }, { "epoch": 0.23359665427509294, "grad_norm": 1.248556818199349, "learning_rate": 8.956091587699415e-06, "loss": 0.0723, "step": 5027 }, { "epoch": 0.2336431226765799, "grad_norm": 1.0008084036113198, "learning_rate": 8.95563135556853e-06, "loss": 0.0683, "step": 5028 }, { "epoch": 0.23368959107806692, "grad_norm": 1.2029270550341664, "learning_rate": 8.955171033838428e-06, "loss": 0.0798, "step": 5029 }, { "epoch": 0.2337360594795539, "grad_norm": 1.1025308913883674, "learning_rate": 8.954710622519541e-06, "loss": 0.0799, "step": 5030 }, { "epoch": 0.2337825278810409, "grad_norm": 1.5052887717254955, "learning_rate": 8.954250121622296e-06, "loss": 0.0906, "step": 5031 }, { "epoch": 0.23382899628252787, "grad_norm": 0.7333740137820062, "learning_rate": 8.953789531157125e-06, "loss": 0.0474, "step": 5032 }, { "epoch": 0.23387546468401488, "grad_norm": 1.1058542434576577, "learning_rate": 8.95332885113446e-06, "loss": 0.0996, "step": 5033 }, { "epoch": 0.23392193308550185, "grad_norm": 1.4718759026014807, "learning_rate": 8.952868081564734e-06, "loss": 0.074, "step": 5034 }, { "epoch": 0.23396840148698886, "grad_norm": 0.9257024070650901, "learning_rate": 8.952407222458386e-06, "loss": 0.0604, "step": 5035 }, { "epoch": 0.23401486988847583, "grad_norm": 1.5388593343141526, "learning_rate": 8.951946273825856e-06, "loss": 0.0792, "step": 5036 }, { "epoch": 0.2340613382899628, "grad_norm": 1.1660764676109323, "learning_rate": 8.951485235677581e-06, "loss": 0.0634, "step": 5037 }, { "epoch": 0.23410780669144982, "grad_norm": 1.1379498008730444, "learning_rate": 8.95102410802401e-06, "loss": 0.1251, "step": 5038 }, { "epoch": 0.2341542750929368, "grad_norm": 4.327199072951553, "learning_rate": 8.95056289087558e-06, "loss": 0.1103, "step": 5039 }, { "epoch": 0.2342007434944238, "grad_norm": 2.5679728400729096, "learning_rate": 8.950101584242746e-06, "loss": 0.1234, "step": 5040 }, { "epoch": 0.23424721189591077, "grad_norm": 0.7308938774713394, "learning_rate": 8.949640188135953e-06, "loss": 0.0742, "step": 5041 }, { "epoch": 0.23429368029739778, "grad_norm": 0.48474978060103685, "learning_rate": 8.949178702565653e-06, "loss": 0.0311, "step": 5042 }, { "epoch": 0.23434014869888475, "grad_norm": 1.2902913259921596, "learning_rate": 8.948717127542299e-06, "loss": 0.0958, "step": 5043 }, { "epoch": 0.23438661710037176, "grad_norm": 0.7698043766280801, "learning_rate": 8.948255463076345e-06, "loss": 0.045, "step": 5044 }, { "epoch": 0.23443308550185873, "grad_norm": 2.7668046607246835, "learning_rate": 8.94779370917825e-06, "loss": 0.0584, "step": 5045 }, { "epoch": 0.23447955390334574, "grad_norm": 1.0518710633306902, "learning_rate": 8.947331865858473e-06, "loss": 0.0424, "step": 5046 }, { "epoch": 0.23452602230483272, "grad_norm": 2.028776688657799, "learning_rate": 8.946869933127476e-06, "loss": 0.1161, "step": 5047 }, { "epoch": 0.2345724907063197, "grad_norm": 1.3202682663732244, "learning_rate": 8.946407910995717e-06, "loss": 0.1091, "step": 5048 }, { "epoch": 0.2346189591078067, "grad_norm": 0.9486009511305179, "learning_rate": 8.945945799473668e-06, "loss": 0.08, "step": 5049 }, { "epoch": 0.23466542750929367, "grad_norm": 2.2239053348253455, "learning_rate": 8.945483598571793e-06, "loss": 0.069, "step": 5050 }, { "epoch": 0.23471189591078068, "grad_norm": 1.6044521068077817, "learning_rate": 8.945021308300563e-06, "loss": 0.1198, "step": 5051 }, { "epoch": 0.23475836431226765, "grad_norm": 2.1863558250412347, "learning_rate": 8.944558928670448e-06, "loss": 0.0862, "step": 5052 }, { "epoch": 0.23480483271375466, "grad_norm": 1.464587320049583, "learning_rate": 8.944096459691921e-06, "loss": 0.0743, "step": 5053 }, { "epoch": 0.23485130111524163, "grad_norm": 1.1626465940614938, "learning_rate": 8.94363390137546e-06, "loss": 0.0989, "step": 5054 }, { "epoch": 0.23489776951672864, "grad_norm": 1.8271294734903052, "learning_rate": 8.943171253731537e-06, "loss": 0.1408, "step": 5055 }, { "epoch": 0.23494423791821561, "grad_norm": 1.835118649707287, "learning_rate": 8.942708516770637e-06, "loss": 0.0756, "step": 5056 }, { "epoch": 0.2349907063197026, "grad_norm": 1.8905119041251075, "learning_rate": 8.94224569050324e-06, "loss": 0.0915, "step": 5057 }, { "epoch": 0.2350371747211896, "grad_norm": 2.362013489589403, "learning_rate": 8.941782774939827e-06, "loss": 0.0703, "step": 5058 }, { "epoch": 0.23508364312267657, "grad_norm": 1.4590052015024468, "learning_rate": 8.941319770090889e-06, "loss": 0.0771, "step": 5059 }, { "epoch": 0.23513011152416358, "grad_norm": 1.2142848734237253, "learning_rate": 8.940856675966907e-06, "loss": 0.0853, "step": 5060 }, { "epoch": 0.23517657992565055, "grad_norm": 0.7644108447226654, "learning_rate": 8.940393492578374e-06, "loss": 0.0547, "step": 5061 }, { "epoch": 0.23522304832713756, "grad_norm": 1.895317780466095, "learning_rate": 8.939930219935782e-06, "loss": 0.0935, "step": 5062 }, { "epoch": 0.23526951672862453, "grad_norm": 1.9523799559288268, "learning_rate": 8.939466858049624e-06, "loss": 0.1416, "step": 5063 }, { "epoch": 0.23531598513011154, "grad_norm": 1.1574841517720074, "learning_rate": 8.939003406930393e-06, "loss": 0.0772, "step": 5064 }, { "epoch": 0.23536245353159851, "grad_norm": 1.5734230948726158, "learning_rate": 8.938539866588593e-06, "loss": 0.072, "step": 5065 }, { "epoch": 0.2354089219330855, "grad_norm": 7.221625869199386, "learning_rate": 8.938076237034717e-06, "loss": 0.0822, "step": 5066 }, { "epoch": 0.2354553903345725, "grad_norm": 2.22741038665978, "learning_rate": 8.93761251827927e-06, "loss": 0.0889, "step": 5067 }, { "epoch": 0.23550185873605947, "grad_norm": 1.4847235985650484, "learning_rate": 8.937148710332755e-06, "loss": 0.0812, "step": 5068 }, { "epoch": 0.23554832713754648, "grad_norm": 1.3511398738128961, "learning_rate": 8.93668481320568e-06, "loss": 0.06, "step": 5069 }, { "epoch": 0.23559479553903345, "grad_norm": 2.259565427292304, "learning_rate": 8.936220826908548e-06, "loss": 0.1392, "step": 5070 }, { "epoch": 0.23564126394052046, "grad_norm": 1.8583827852238417, "learning_rate": 8.935756751451874e-06, "loss": 0.1178, "step": 5071 }, { "epoch": 0.23568773234200743, "grad_norm": 1.138224696316091, "learning_rate": 8.935292586846165e-06, "loss": 0.0821, "step": 5072 }, { "epoch": 0.23573420074349444, "grad_norm": 0.7683214816197412, "learning_rate": 8.934828333101938e-06, "loss": 0.0684, "step": 5073 }, { "epoch": 0.23578066914498141, "grad_norm": 2.130522506704115, "learning_rate": 8.934363990229709e-06, "loss": 0.0972, "step": 5074 }, { "epoch": 0.2358271375464684, "grad_norm": 1.6848967983105645, "learning_rate": 8.933899558239993e-06, "loss": 0.0823, "step": 5075 }, { "epoch": 0.2358736059479554, "grad_norm": 1.4692961558216215, "learning_rate": 8.933435037143312e-06, "loss": 0.0925, "step": 5076 }, { "epoch": 0.23592007434944237, "grad_norm": 1.8846335201915076, "learning_rate": 8.932970426950188e-06, "loss": 0.1024, "step": 5077 }, { "epoch": 0.23596654275092938, "grad_norm": 2.5444012360952413, "learning_rate": 8.932505727671143e-06, "loss": 0.158, "step": 5078 }, { "epoch": 0.23601301115241635, "grad_norm": 1.2635838513580733, "learning_rate": 8.932040939316707e-06, "loss": 0.0451, "step": 5079 }, { "epoch": 0.23605947955390336, "grad_norm": 2.0831793534133864, "learning_rate": 8.931576061897404e-06, "loss": 0.0866, "step": 5080 }, { "epoch": 0.23610594795539033, "grad_norm": 2.6533818924857298, "learning_rate": 8.931111095423766e-06, "loss": 0.1185, "step": 5081 }, { "epoch": 0.23615241635687734, "grad_norm": 1.6800056253317968, "learning_rate": 8.930646039906324e-06, "loss": 0.1193, "step": 5082 }, { "epoch": 0.2361988847583643, "grad_norm": 1.9485374352574685, "learning_rate": 8.930180895355611e-06, "loss": 0.1307, "step": 5083 }, { "epoch": 0.2362453531598513, "grad_norm": 1.525292984379711, "learning_rate": 8.929715661782166e-06, "loss": 0.0974, "step": 5084 }, { "epoch": 0.2362918215613383, "grad_norm": 1.336476335416742, "learning_rate": 8.929250339196523e-06, "loss": 0.1271, "step": 5085 }, { "epoch": 0.23633828996282527, "grad_norm": 1.8449988624820406, "learning_rate": 8.928784927609226e-06, "loss": 0.0893, "step": 5086 }, { "epoch": 0.23638475836431228, "grad_norm": 1.458919753051283, "learning_rate": 8.928319427030816e-06, "loss": 0.1044, "step": 5087 }, { "epoch": 0.23643122676579925, "grad_norm": 1.356779320394581, "learning_rate": 8.927853837471835e-06, "loss": 0.0914, "step": 5088 }, { "epoch": 0.23647769516728626, "grad_norm": 1.32051747798862, "learning_rate": 8.927388158942831e-06, "loss": 0.0819, "step": 5089 }, { "epoch": 0.23652416356877323, "grad_norm": 1.0228497206510523, "learning_rate": 8.926922391454353e-06, "loss": 0.0581, "step": 5090 }, { "epoch": 0.23657063197026024, "grad_norm": 1.7778823830382955, "learning_rate": 8.926456535016948e-06, "loss": 0.0845, "step": 5091 }, { "epoch": 0.2366171003717472, "grad_norm": 1.4137325554828861, "learning_rate": 8.925990589641171e-06, "loss": 0.0929, "step": 5092 }, { "epoch": 0.2366635687732342, "grad_norm": 1.5488201916967126, "learning_rate": 8.925524555337575e-06, "loss": 0.0896, "step": 5093 }, { "epoch": 0.2367100371747212, "grad_norm": 1.4927617427866238, "learning_rate": 8.925058432116716e-06, "loss": 0.1, "step": 5094 }, { "epoch": 0.23675650557620817, "grad_norm": 1.6893413161163229, "learning_rate": 8.924592219989151e-06, "loss": 0.1346, "step": 5095 }, { "epoch": 0.23680297397769517, "grad_norm": 2.230691067445527, "learning_rate": 8.924125918965445e-06, "loss": 0.1136, "step": 5096 }, { "epoch": 0.23684944237918215, "grad_norm": 1.428639344151798, "learning_rate": 8.923659529056156e-06, "loss": 0.082, "step": 5097 }, { "epoch": 0.23689591078066916, "grad_norm": 1.5240650056270395, "learning_rate": 8.923193050271848e-06, "loss": 0.1048, "step": 5098 }, { "epoch": 0.23694237918215613, "grad_norm": 1.1599321976670887, "learning_rate": 8.922726482623088e-06, "loss": 0.0973, "step": 5099 }, { "epoch": 0.23698884758364314, "grad_norm": 1.3494583091214967, "learning_rate": 8.922259826120445e-06, "loss": 0.0803, "step": 5100 }, { "epoch": 0.2370353159851301, "grad_norm": 1.1718592468468465, "learning_rate": 8.921793080774489e-06, "loss": 0.0758, "step": 5101 }, { "epoch": 0.2370817843866171, "grad_norm": 0.9700206153673299, "learning_rate": 8.921326246595792e-06, "loss": 0.0648, "step": 5102 }, { "epoch": 0.2371282527881041, "grad_norm": 1.3312057521629166, "learning_rate": 8.92085932359493e-06, "loss": 0.0447, "step": 5103 }, { "epoch": 0.23717472118959107, "grad_norm": 4.995195200971145, "learning_rate": 8.920392311782475e-06, "loss": 0.0912, "step": 5104 }, { "epoch": 0.23722118959107807, "grad_norm": 1.2108736961543782, "learning_rate": 8.91992521116901e-06, "loss": 0.0705, "step": 5105 }, { "epoch": 0.23726765799256505, "grad_norm": 2.123205038042184, "learning_rate": 8.919458021765111e-06, "loss": 0.0791, "step": 5106 }, { "epoch": 0.23731412639405206, "grad_norm": 1.1798164583891333, "learning_rate": 8.918990743581366e-06, "loss": 0.0928, "step": 5107 }, { "epoch": 0.23736059479553903, "grad_norm": 0.9194333233895015, "learning_rate": 8.918523376628352e-06, "loss": 0.0406, "step": 5108 }, { "epoch": 0.23740706319702604, "grad_norm": 1.0886012287095892, "learning_rate": 8.918055920916662e-06, "loss": 0.0761, "step": 5109 }, { "epoch": 0.237453531598513, "grad_norm": 1.8871847483671442, "learning_rate": 8.917588376456883e-06, "loss": 0.0726, "step": 5110 }, { "epoch": 0.2375, "grad_norm": 1.3391422644252964, "learning_rate": 8.917120743259601e-06, "loss": 0.091, "step": 5111 }, { "epoch": 0.237546468401487, "grad_norm": 1.7020248596619394, "learning_rate": 8.916653021335413e-06, "loss": 0.0649, "step": 5112 }, { "epoch": 0.23759293680297397, "grad_norm": 1.3877938806472323, "learning_rate": 8.916185210694912e-06, "loss": 0.0753, "step": 5113 }, { "epoch": 0.23763940520446097, "grad_norm": 2.5075100226905676, "learning_rate": 8.915717311348693e-06, "loss": 0.1046, "step": 5114 }, { "epoch": 0.23768587360594795, "grad_norm": 0.7466060423760701, "learning_rate": 8.915249323307358e-06, "loss": 0.0609, "step": 5115 }, { "epoch": 0.23773234200743495, "grad_norm": 1.556083041085988, "learning_rate": 8.914781246581504e-06, "loss": 0.0337, "step": 5116 }, { "epoch": 0.23777881040892193, "grad_norm": 2.3812157406938006, "learning_rate": 8.914313081181734e-06, "loss": 0.0993, "step": 5117 }, { "epoch": 0.23782527881040894, "grad_norm": 1.0682591382547058, "learning_rate": 8.913844827118652e-06, "loss": 0.061, "step": 5118 }, { "epoch": 0.2378717472118959, "grad_norm": 1.76679622080519, "learning_rate": 8.913376484402867e-06, "loss": 0.0963, "step": 5119 }, { "epoch": 0.2379182156133829, "grad_norm": 2.079073310209199, "learning_rate": 8.912908053044985e-06, "loss": 0.1379, "step": 5120 }, { "epoch": 0.2379646840148699, "grad_norm": 1.3090400824613273, "learning_rate": 8.91243953305562e-06, "loss": 0.0932, "step": 5121 }, { "epoch": 0.23801115241635687, "grad_norm": 1.6646312892813582, "learning_rate": 8.911970924445377e-06, "loss": 0.0977, "step": 5122 }, { "epoch": 0.23805762081784387, "grad_norm": 1.3032726315885557, "learning_rate": 8.911502227224879e-06, "loss": 0.066, "step": 5123 }, { "epoch": 0.23810408921933085, "grad_norm": 1.0942911354298124, "learning_rate": 8.911033441404737e-06, "loss": 0.0621, "step": 5124 }, { "epoch": 0.23815055762081785, "grad_norm": 4.015117631262661, "learning_rate": 8.910564566995571e-06, "loss": 0.1165, "step": 5125 }, { "epoch": 0.23819702602230483, "grad_norm": 1.3033006276730912, "learning_rate": 8.910095604008005e-06, "loss": 0.0794, "step": 5126 }, { "epoch": 0.23824349442379184, "grad_norm": 1.161406180827882, "learning_rate": 8.909626552452655e-06, "loss": 0.079, "step": 5127 }, { "epoch": 0.2382899628252788, "grad_norm": 0.6092368732165894, "learning_rate": 8.90915741234015e-06, "loss": 0.0575, "step": 5128 }, { "epoch": 0.2383364312267658, "grad_norm": 1.0666070013611926, "learning_rate": 8.908688183681114e-06, "loss": 0.0464, "step": 5129 }, { "epoch": 0.2383828996282528, "grad_norm": 0.6800836198387795, "learning_rate": 8.908218866486178e-06, "loss": 0.058, "step": 5130 }, { "epoch": 0.23842936802973977, "grad_norm": 1.0584343412726396, "learning_rate": 8.907749460765972e-06, "loss": 0.0642, "step": 5131 }, { "epoch": 0.23847583643122677, "grad_norm": 1.5038125333286396, "learning_rate": 8.907279966531127e-06, "loss": 0.0918, "step": 5132 }, { "epoch": 0.23852230483271375, "grad_norm": 2.2834692242458683, "learning_rate": 8.90681038379228e-06, "loss": 0.0826, "step": 5133 }, { "epoch": 0.23856877323420075, "grad_norm": 0.821987178509534, "learning_rate": 8.906340712560062e-06, "loss": 0.0617, "step": 5134 }, { "epoch": 0.23861524163568773, "grad_norm": 1.2315685834133154, "learning_rate": 8.90587095284512e-06, "loss": 0.1281, "step": 5135 }, { "epoch": 0.23866171003717473, "grad_norm": 1.6201027671729324, "learning_rate": 8.905401104658087e-06, "loss": 0.0867, "step": 5136 }, { "epoch": 0.2387081784386617, "grad_norm": 2.059100338705917, "learning_rate": 8.90493116800961e-06, "loss": 0.0532, "step": 5137 }, { "epoch": 0.2387546468401487, "grad_norm": 2.5740778195855376, "learning_rate": 8.904461142910332e-06, "loss": 0.0478, "step": 5138 }, { "epoch": 0.2388011152416357, "grad_norm": 2.6282142462469036, "learning_rate": 8.9039910293709e-06, "loss": 0.0922, "step": 5139 }, { "epoch": 0.23884758364312267, "grad_norm": 1.4277956022267155, "learning_rate": 8.903520827401961e-06, "loss": 0.086, "step": 5140 }, { "epoch": 0.23889405204460967, "grad_norm": 1.4161681929368515, "learning_rate": 8.903050537014168e-06, "loss": 0.0639, "step": 5141 }, { "epoch": 0.23894052044609665, "grad_norm": 1.4266255041957718, "learning_rate": 8.902580158218172e-06, "loss": 0.1, "step": 5142 }, { "epoch": 0.23898698884758365, "grad_norm": 1.2508013389798838, "learning_rate": 8.90210969102463e-06, "loss": 0.0967, "step": 5143 }, { "epoch": 0.23903345724907063, "grad_norm": 1.2836751754685338, "learning_rate": 8.901639135444193e-06, "loss": 0.061, "step": 5144 }, { "epoch": 0.23907992565055763, "grad_norm": 1.8407371774216368, "learning_rate": 8.901168491487526e-06, "loss": 0.1782, "step": 5145 }, { "epoch": 0.2391263940520446, "grad_norm": 1.4040342205106369, "learning_rate": 8.900697759165286e-06, "loss": 0.0944, "step": 5146 }, { "epoch": 0.2391728624535316, "grad_norm": 0.8871622265143009, "learning_rate": 8.900226938488136e-06, "loss": 0.0552, "step": 5147 }, { "epoch": 0.2392193308550186, "grad_norm": 1.2851586149118963, "learning_rate": 8.89975602946674e-06, "loss": 0.0528, "step": 5148 }, { "epoch": 0.23926579925650557, "grad_norm": 1.8426650175977357, "learning_rate": 8.899285032111768e-06, "loss": 0.0795, "step": 5149 }, { "epoch": 0.23931226765799257, "grad_norm": 0.8039956680802812, "learning_rate": 8.898813946433884e-06, "loss": 0.068, "step": 5150 }, { "epoch": 0.23935873605947955, "grad_norm": 3.3803830701448296, "learning_rate": 8.898342772443762e-06, "loss": 0.078, "step": 5151 }, { "epoch": 0.23940520446096655, "grad_norm": 1.2896100505397965, "learning_rate": 8.897871510152071e-06, "loss": 0.0844, "step": 5152 }, { "epoch": 0.23945167286245353, "grad_norm": 1.3848497718441444, "learning_rate": 8.897400159569492e-06, "loss": 0.0788, "step": 5153 }, { "epoch": 0.23949814126394053, "grad_norm": 1.2877945740135572, "learning_rate": 8.896928720706693e-06, "loss": 0.0946, "step": 5154 }, { "epoch": 0.2395446096654275, "grad_norm": 1.204216232896277, "learning_rate": 8.89645719357436e-06, "loss": 0.0722, "step": 5155 }, { "epoch": 0.2395910780669145, "grad_norm": 1.4705253766333213, "learning_rate": 8.89598557818317e-06, "loss": 0.0974, "step": 5156 }, { "epoch": 0.2396375464684015, "grad_norm": 1.494740765721779, "learning_rate": 8.895513874543805e-06, "loss": 0.093, "step": 5157 }, { "epoch": 0.23968401486988847, "grad_norm": 6.361427782570121, "learning_rate": 8.895042082666952e-06, "loss": 0.0996, "step": 5158 }, { "epoch": 0.23973048327137547, "grad_norm": 1.5656030747365568, "learning_rate": 8.894570202563295e-06, "loss": 0.16, "step": 5159 }, { "epoch": 0.23977695167286245, "grad_norm": 1.3691138046289901, "learning_rate": 8.894098234243526e-06, "loss": 0.064, "step": 5160 }, { "epoch": 0.23982342007434945, "grad_norm": 2.0012044409165552, "learning_rate": 8.893626177718332e-06, "loss": 0.0598, "step": 5161 }, { "epoch": 0.23986988847583643, "grad_norm": 1.2798750651620705, "learning_rate": 8.893154032998409e-06, "loss": 0.1473, "step": 5162 }, { "epoch": 0.23991635687732343, "grad_norm": 1.4996573927209829, "learning_rate": 8.892681800094447e-06, "loss": 0.0722, "step": 5163 }, { "epoch": 0.2399628252788104, "grad_norm": 0.5288490843135978, "learning_rate": 8.892209479017146e-06, "loss": 0.0758, "step": 5164 }, { "epoch": 0.2400092936802974, "grad_norm": 1.5676083970638968, "learning_rate": 8.891737069777205e-06, "loss": 0.126, "step": 5165 }, { "epoch": 0.2400557620817844, "grad_norm": 1.400099297574647, "learning_rate": 8.891264572385323e-06, "loss": 0.1027, "step": 5166 }, { "epoch": 0.24010223048327137, "grad_norm": 0.7014220702912475, "learning_rate": 8.890791986852203e-06, "loss": 0.0602, "step": 5167 }, { "epoch": 0.24014869888475837, "grad_norm": 1.363067789138595, "learning_rate": 8.890319313188549e-06, "loss": 0.0888, "step": 5168 }, { "epoch": 0.24019516728624535, "grad_norm": 4.250955607189921, "learning_rate": 8.889846551405067e-06, "loss": 0.0819, "step": 5169 }, { "epoch": 0.24024163568773235, "grad_norm": 1.2655005471310499, "learning_rate": 8.88937370151247e-06, "loss": 0.1065, "step": 5170 }, { "epoch": 0.24028810408921933, "grad_norm": 2.2790651637189585, "learning_rate": 8.888900763521462e-06, "loss": 0.0635, "step": 5171 }, { "epoch": 0.24033457249070633, "grad_norm": 1.8032233676552922, "learning_rate": 8.888427737442759e-06, "loss": 0.0813, "step": 5172 }, { "epoch": 0.2403810408921933, "grad_norm": 1.7688473433583438, "learning_rate": 8.887954623287076e-06, "loss": 0.1126, "step": 5173 }, { "epoch": 0.24042750929368029, "grad_norm": 1.1505499599875375, "learning_rate": 8.887481421065129e-06, "loss": 0.0712, "step": 5174 }, { "epoch": 0.2404739776951673, "grad_norm": 1.1756298977518376, "learning_rate": 8.887008130787637e-06, "loss": 0.076, "step": 5175 }, { "epoch": 0.24052044609665427, "grad_norm": 1.7852666847253584, "learning_rate": 8.886534752465319e-06, "loss": 0.0977, "step": 5176 }, { "epoch": 0.24056691449814127, "grad_norm": 1.0627768064839644, "learning_rate": 8.886061286108898e-06, "loss": 0.0873, "step": 5177 }, { "epoch": 0.24061338289962825, "grad_norm": 2.584959090440067, "learning_rate": 8.8855877317291e-06, "loss": 0.095, "step": 5178 }, { "epoch": 0.24065985130111525, "grad_norm": 1.1184386292700037, "learning_rate": 8.885114089336649e-06, "loss": 0.1315, "step": 5179 }, { "epoch": 0.24070631970260223, "grad_norm": 1.017074589587853, "learning_rate": 8.884640358942276e-06, "loss": 0.0827, "step": 5180 }, { "epoch": 0.24075278810408923, "grad_norm": 1.2032258832277825, "learning_rate": 8.884166540556709e-06, "loss": 0.0914, "step": 5181 }, { "epoch": 0.2407992565055762, "grad_norm": 1.3195120163058411, "learning_rate": 8.883692634190682e-06, "loss": 0.0683, "step": 5182 }, { "epoch": 0.24084572490706319, "grad_norm": 1.097989473532077, "learning_rate": 8.88321863985493e-06, "loss": 0.0567, "step": 5183 }, { "epoch": 0.2408921933085502, "grad_norm": 2.7540730940538474, "learning_rate": 8.88274455756019e-06, "loss": 0.1156, "step": 5184 }, { "epoch": 0.24093866171003717, "grad_norm": 3.4216546512191077, "learning_rate": 8.882270387317198e-06, "loss": 0.1605, "step": 5185 }, { "epoch": 0.24098513011152417, "grad_norm": 0.6995239910659624, "learning_rate": 8.881796129136694e-06, "loss": 0.0636, "step": 5186 }, { "epoch": 0.24103159851301115, "grad_norm": 1.1275125225934826, "learning_rate": 8.881321783029423e-06, "loss": 0.0588, "step": 5187 }, { "epoch": 0.24107806691449815, "grad_norm": 1.039260154201148, "learning_rate": 8.880847349006131e-06, "loss": 0.0527, "step": 5188 }, { "epoch": 0.24112453531598513, "grad_norm": 1.3874133476180086, "learning_rate": 8.880372827077561e-06, "loss": 0.1161, "step": 5189 }, { "epoch": 0.24117100371747213, "grad_norm": 1.8029007810401252, "learning_rate": 8.879898217254461e-06, "loss": 0.1231, "step": 5190 }, { "epoch": 0.2412174721189591, "grad_norm": 1.020610064734688, "learning_rate": 8.879423519547583e-06, "loss": 0.0652, "step": 5191 }, { "epoch": 0.24126394052044609, "grad_norm": 1.0988969094223033, "learning_rate": 8.878948733967681e-06, "loss": 0.0705, "step": 5192 }, { "epoch": 0.2413104089219331, "grad_norm": 1.3319706304908174, "learning_rate": 8.878473860525507e-06, "loss": 0.0832, "step": 5193 }, { "epoch": 0.24135687732342007, "grad_norm": 1.5917580402344815, "learning_rate": 8.877998899231818e-06, "loss": 0.0735, "step": 5194 }, { "epoch": 0.24140334572490707, "grad_norm": 0.8488669902430095, "learning_rate": 8.87752385009737e-06, "loss": 0.0463, "step": 5195 }, { "epoch": 0.24144981412639405, "grad_norm": 1.038581461713351, "learning_rate": 8.87704871313293e-06, "loss": 0.0698, "step": 5196 }, { "epoch": 0.24149628252788105, "grad_norm": 1.438833789956859, "learning_rate": 8.876573488349253e-06, "loss": 0.0974, "step": 5197 }, { "epoch": 0.24154275092936803, "grad_norm": 0.8778916502847878, "learning_rate": 8.876098175757108e-06, "loss": 0.0751, "step": 5198 }, { "epoch": 0.24158921933085503, "grad_norm": 1.6025259777725607, "learning_rate": 8.87562277536726e-06, "loss": 0.0854, "step": 5199 }, { "epoch": 0.241635687732342, "grad_norm": 1.4347334387655013, "learning_rate": 8.875147287190476e-06, "loss": 0.1503, "step": 5200 }, { "epoch": 0.24168215613382898, "grad_norm": 1.1676813984312595, "learning_rate": 8.874671711237529e-06, "loss": 0.0894, "step": 5201 }, { "epoch": 0.241728624535316, "grad_norm": 1.1266850673982425, "learning_rate": 8.874196047519188e-06, "loss": 0.057, "step": 5202 }, { "epoch": 0.24177509293680297, "grad_norm": 1.7514515064262792, "learning_rate": 8.873720296046229e-06, "loss": 0.0847, "step": 5203 }, { "epoch": 0.24182156133828997, "grad_norm": 1.6861253408506893, "learning_rate": 8.873244456829428e-06, "loss": 0.0747, "step": 5204 }, { "epoch": 0.24186802973977695, "grad_norm": 0.9542533479219126, "learning_rate": 8.872768529879565e-06, "loss": 0.0782, "step": 5205 }, { "epoch": 0.24191449814126395, "grad_norm": 2.4030651151098326, "learning_rate": 8.872292515207419e-06, "loss": 0.1193, "step": 5206 }, { "epoch": 0.24196096654275093, "grad_norm": 1.5677836063615431, "learning_rate": 8.871816412823771e-06, "loss": 0.0962, "step": 5207 }, { "epoch": 0.24200743494423793, "grad_norm": 1.3472050264714266, "learning_rate": 8.871340222739406e-06, "loss": 0.0401, "step": 5208 }, { "epoch": 0.2420539033457249, "grad_norm": 1.3376206563423179, "learning_rate": 8.870863944965112e-06, "loss": 0.0772, "step": 5209 }, { "epoch": 0.24210037174721188, "grad_norm": 0.8004561866723476, "learning_rate": 8.870387579511674e-06, "loss": 0.0464, "step": 5210 }, { "epoch": 0.2421468401486989, "grad_norm": 0.6589099427307819, "learning_rate": 8.869911126389885e-06, "loss": 0.0434, "step": 5211 }, { "epoch": 0.24219330855018587, "grad_norm": 1.453965471512029, "learning_rate": 8.869434585610534e-06, "loss": 0.0924, "step": 5212 }, { "epoch": 0.24223977695167287, "grad_norm": 1.6026576159724857, "learning_rate": 8.868957957184419e-06, "loss": 0.114, "step": 5213 }, { "epoch": 0.24228624535315985, "grad_norm": 0.7920403710410577, "learning_rate": 8.868481241122334e-06, "loss": 0.0495, "step": 5214 }, { "epoch": 0.24233271375464685, "grad_norm": 1.1587140996743643, "learning_rate": 8.868004437435076e-06, "loss": 0.0761, "step": 5215 }, { "epoch": 0.24237918215613383, "grad_norm": 1.099047316342221, "learning_rate": 8.86752754613345e-06, "loss": 0.0528, "step": 5216 }, { "epoch": 0.24242565055762083, "grad_norm": 0.5730500843170762, "learning_rate": 8.867050567228251e-06, "loss": 0.0548, "step": 5217 }, { "epoch": 0.2424721189591078, "grad_norm": 1.865867300235781, "learning_rate": 8.866573500730288e-06, "loss": 0.0886, "step": 5218 }, { "epoch": 0.24251858736059478, "grad_norm": 1.274527713080361, "learning_rate": 8.866096346650365e-06, "loss": 0.0614, "step": 5219 }, { "epoch": 0.2425650557620818, "grad_norm": 1.3871704994883691, "learning_rate": 8.865619104999292e-06, "loss": 0.0772, "step": 5220 }, { "epoch": 0.24261152416356876, "grad_norm": 1.5491645619007335, "learning_rate": 8.865141775787876e-06, "loss": 0.0602, "step": 5221 }, { "epoch": 0.24265799256505577, "grad_norm": 0.9524893532294244, "learning_rate": 8.864664359026934e-06, "loss": 0.0452, "step": 5222 }, { "epoch": 0.24270446096654275, "grad_norm": 1.3517714553055338, "learning_rate": 8.864186854727273e-06, "loss": 0.0616, "step": 5223 }, { "epoch": 0.24275092936802975, "grad_norm": 1.1136422902532606, "learning_rate": 8.863709262899715e-06, "loss": 0.0408, "step": 5224 }, { "epoch": 0.24279739776951673, "grad_norm": 1.514677873552291, "learning_rate": 8.863231583555073e-06, "loss": 0.1126, "step": 5225 }, { "epoch": 0.24284386617100373, "grad_norm": 1.5344530755477364, "learning_rate": 8.862753816704174e-06, "loss": 0.0732, "step": 5226 }, { "epoch": 0.2428903345724907, "grad_norm": 0.8146711383101726, "learning_rate": 8.862275962357833e-06, "loss": 0.0531, "step": 5227 }, { "epoch": 0.24293680297397768, "grad_norm": 1.0824126664136806, "learning_rate": 8.861798020526878e-06, "loss": 0.0627, "step": 5228 }, { "epoch": 0.2429832713754647, "grad_norm": 0.7593669264751327, "learning_rate": 8.861319991222132e-06, "loss": 0.0614, "step": 5229 }, { "epoch": 0.24302973977695166, "grad_norm": 0.5847833563143531, "learning_rate": 8.860841874454425e-06, "loss": 0.0478, "step": 5230 }, { "epoch": 0.24307620817843867, "grad_norm": 1.7436813593582832, "learning_rate": 8.860363670234584e-06, "loss": 0.0846, "step": 5231 }, { "epoch": 0.24312267657992565, "grad_norm": 0.9638826429552544, "learning_rate": 8.859885378573444e-06, "loss": 0.0654, "step": 5232 }, { "epoch": 0.24316914498141265, "grad_norm": 0.864175984356204, "learning_rate": 8.859406999481839e-06, "loss": 0.0437, "step": 5233 }, { "epoch": 0.24321561338289963, "grad_norm": 1.5206060699365864, "learning_rate": 8.858928532970602e-06, "loss": 0.0876, "step": 5234 }, { "epoch": 0.24326208178438663, "grad_norm": 0.6635570025981, "learning_rate": 8.858449979050575e-06, "loss": 0.0465, "step": 5235 }, { "epoch": 0.2433085501858736, "grad_norm": 0.9479886091835996, "learning_rate": 8.857971337732591e-06, "loss": 0.0961, "step": 5236 }, { "epoch": 0.24335501858736058, "grad_norm": 1.5327880307627946, "learning_rate": 8.857492609027498e-06, "loss": 0.0865, "step": 5237 }, { "epoch": 0.2434014869888476, "grad_norm": 1.790192036294245, "learning_rate": 8.857013792946137e-06, "loss": 0.0806, "step": 5238 }, { "epoch": 0.24344795539033456, "grad_norm": 0.9101833902365972, "learning_rate": 8.856534889499356e-06, "loss": 0.0704, "step": 5239 }, { "epoch": 0.24349442379182157, "grad_norm": 1.5068147207216152, "learning_rate": 8.856055898697998e-06, "loss": 0.0591, "step": 5240 }, { "epoch": 0.24354089219330854, "grad_norm": 0.5106337901079199, "learning_rate": 8.855576820552917e-06, "loss": 0.0545, "step": 5241 }, { "epoch": 0.24358736059479555, "grad_norm": 1.072631280579208, "learning_rate": 8.855097655074962e-06, "loss": 0.0493, "step": 5242 }, { "epoch": 0.24363382899628253, "grad_norm": 0.9080413016036509, "learning_rate": 8.854618402274989e-06, "loss": 0.0578, "step": 5243 }, { "epoch": 0.24368029739776953, "grad_norm": 0.9812517625709152, "learning_rate": 8.854139062163852e-06, "loss": 0.0654, "step": 5244 }, { "epoch": 0.2437267657992565, "grad_norm": 2.0186944608996273, "learning_rate": 8.853659634752409e-06, "loss": 0.0519, "step": 5245 }, { "epoch": 0.24377323420074348, "grad_norm": 1.219724627063, "learning_rate": 8.853180120051518e-06, "loss": 0.077, "step": 5246 }, { "epoch": 0.2438197026022305, "grad_norm": 2.462063273509443, "learning_rate": 8.852700518072043e-06, "loss": 0.1093, "step": 5247 }, { "epoch": 0.24386617100371746, "grad_norm": 2.2685686719916247, "learning_rate": 8.852220828824847e-06, "loss": 0.1072, "step": 5248 }, { "epoch": 0.24391263940520447, "grad_norm": 1.0547537861783918, "learning_rate": 8.851741052320794e-06, "loss": 0.0774, "step": 5249 }, { "epoch": 0.24395910780669144, "grad_norm": 9.198114741027586, "learning_rate": 8.851261188570752e-06, "loss": 0.1331, "step": 5250 }, { "epoch": 0.24400557620817845, "grad_norm": 1.4697162765001726, "learning_rate": 8.85078123758559e-06, "loss": 0.0852, "step": 5251 }, { "epoch": 0.24405204460966543, "grad_norm": 0.6555554081410188, "learning_rate": 8.85030119937618e-06, "loss": 0.0498, "step": 5252 }, { "epoch": 0.24409851301115243, "grad_norm": 0.63823654542423, "learning_rate": 8.849821073953397e-06, "loss": 0.0734, "step": 5253 }, { "epoch": 0.2441449814126394, "grad_norm": 2.3375351248275917, "learning_rate": 8.849340861328115e-06, "loss": 0.0524, "step": 5254 }, { "epoch": 0.24419144981412638, "grad_norm": 2.7516465490317277, "learning_rate": 8.84886056151121e-06, "loss": 0.0848, "step": 5255 }, { "epoch": 0.2442379182156134, "grad_norm": 0.7511383405872083, "learning_rate": 8.848380174513565e-06, "loss": 0.0452, "step": 5256 }, { "epoch": 0.24428438661710036, "grad_norm": 1.52784027841771, "learning_rate": 8.847899700346056e-06, "loss": 0.0637, "step": 5257 }, { "epoch": 0.24433085501858737, "grad_norm": 1.614814353595599, "learning_rate": 8.84741913901957e-06, "loss": 0.0694, "step": 5258 }, { "epoch": 0.24437732342007434, "grad_norm": 1.2385915970061105, "learning_rate": 8.84693849054499e-06, "loss": 0.0794, "step": 5259 }, { "epoch": 0.24442379182156135, "grad_norm": 0.8816577344706431, "learning_rate": 8.846457754933205e-06, "loss": 0.0647, "step": 5260 }, { "epoch": 0.24447026022304832, "grad_norm": 1.5079689475798332, "learning_rate": 8.845976932195104e-06, "loss": 0.0869, "step": 5261 }, { "epoch": 0.24451672862453533, "grad_norm": 1.255614729644642, "learning_rate": 8.845496022341577e-06, "loss": 0.0693, "step": 5262 }, { "epoch": 0.2445631970260223, "grad_norm": 1.7078256754923593, "learning_rate": 8.84501502538352e-06, "loss": 0.0839, "step": 5263 }, { "epoch": 0.24460966542750928, "grad_norm": 1.7536196443677006, "learning_rate": 8.844533941331823e-06, "loss": 0.1095, "step": 5264 }, { "epoch": 0.2446561338289963, "grad_norm": 0.7523992327112934, "learning_rate": 8.844052770197387e-06, "loss": 0.0464, "step": 5265 }, { "epoch": 0.24470260223048326, "grad_norm": 1.868531770464004, "learning_rate": 8.84357151199111e-06, "loss": 0.0949, "step": 5266 }, { "epoch": 0.24474907063197027, "grad_norm": 0.706251017475993, "learning_rate": 8.843090166723891e-06, "loss": 0.0669, "step": 5267 }, { "epoch": 0.24479553903345724, "grad_norm": 0.8827551380292944, "learning_rate": 8.842608734406636e-06, "loss": 0.0698, "step": 5268 }, { "epoch": 0.24484200743494425, "grad_norm": 3.139532694113292, "learning_rate": 8.84212721505025e-06, "loss": 0.051, "step": 5269 }, { "epoch": 0.24488847583643122, "grad_norm": 3.1900072830642223, "learning_rate": 8.84164560866564e-06, "loss": 0.0776, "step": 5270 }, { "epoch": 0.24493494423791823, "grad_norm": 1.2439452018420016, "learning_rate": 8.841163915263713e-06, "loss": 0.0494, "step": 5271 }, { "epoch": 0.2449814126394052, "grad_norm": 1.4797938559721129, "learning_rate": 8.840682134855378e-06, "loss": 0.1176, "step": 5272 }, { "epoch": 0.24502788104089218, "grad_norm": 0.8948070408180143, "learning_rate": 8.840200267451553e-06, "loss": 0.0697, "step": 5273 }, { "epoch": 0.24507434944237919, "grad_norm": 2.3621275186564223, "learning_rate": 8.839718313063149e-06, "loss": 0.1158, "step": 5274 }, { "epoch": 0.24512081784386616, "grad_norm": 1.4429132763212986, "learning_rate": 8.839236271701083e-06, "loss": 0.1182, "step": 5275 }, { "epoch": 0.24516728624535317, "grad_norm": 1.5920931887534282, "learning_rate": 8.838754143376278e-06, "loss": 0.1066, "step": 5276 }, { "epoch": 0.24521375464684014, "grad_norm": 0.966264766373483, "learning_rate": 8.838271928099649e-06, "loss": 0.0817, "step": 5277 }, { "epoch": 0.24526022304832715, "grad_norm": 1.0742380385705925, "learning_rate": 8.837789625882122e-06, "loss": 0.0777, "step": 5278 }, { "epoch": 0.24530669144981412, "grad_norm": 1.1688256564074113, "learning_rate": 8.83730723673462e-06, "loss": 0.0863, "step": 5279 }, { "epoch": 0.24535315985130113, "grad_norm": 0.7221549853150039, "learning_rate": 8.83682476066807e-06, "loss": 0.0643, "step": 5280 }, { "epoch": 0.2453996282527881, "grad_norm": 1.2129795986269862, "learning_rate": 8.836342197693402e-06, "loss": 0.0617, "step": 5281 }, { "epoch": 0.24544609665427508, "grad_norm": 2.5363301583950215, "learning_rate": 8.835859547821546e-06, "loss": 0.1004, "step": 5282 }, { "epoch": 0.24549256505576209, "grad_norm": 2.964091074956711, "learning_rate": 8.835376811063433e-06, "loss": 0.0658, "step": 5283 }, { "epoch": 0.24553903345724906, "grad_norm": 2.633414681926421, "learning_rate": 8.83489398743e-06, "loss": 0.0984, "step": 5284 }, { "epoch": 0.24558550185873607, "grad_norm": 1.0745527870029787, "learning_rate": 8.83441107693218e-06, "loss": 0.0599, "step": 5285 }, { "epoch": 0.24563197026022304, "grad_norm": 2.337068675228871, "learning_rate": 8.833928079580914e-06, "loss": 0.0745, "step": 5286 }, { "epoch": 0.24567843866171005, "grad_norm": 0.6491448703006287, "learning_rate": 8.833444995387142e-06, "loss": 0.0708, "step": 5287 }, { "epoch": 0.24572490706319702, "grad_norm": 1.136816144809182, "learning_rate": 8.832961824361807e-06, "loss": 0.0727, "step": 5288 }, { "epoch": 0.24577137546468403, "grad_norm": 0.9393555524901304, "learning_rate": 8.832478566515852e-06, "loss": 0.0423, "step": 5289 }, { "epoch": 0.245817843866171, "grad_norm": 0.986159049769606, "learning_rate": 8.831995221860225e-06, "loss": 0.0616, "step": 5290 }, { "epoch": 0.24586431226765798, "grad_norm": 4.825146439977571, "learning_rate": 8.831511790405871e-06, "loss": 0.0493, "step": 5291 }, { "epoch": 0.24591078066914498, "grad_norm": 1.2157762731103026, "learning_rate": 8.831028272163745e-06, "loss": 0.0646, "step": 5292 }, { "epoch": 0.24595724907063196, "grad_norm": 1.431160950496307, "learning_rate": 8.830544667144794e-06, "loss": 0.081, "step": 5293 }, { "epoch": 0.24600371747211897, "grad_norm": 1.448875034944396, "learning_rate": 8.830060975359975e-06, "loss": 0.0772, "step": 5294 }, { "epoch": 0.24605018587360594, "grad_norm": 1.9367439649887772, "learning_rate": 8.829577196820245e-06, "loss": 0.1032, "step": 5295 }, { "epoch": 0.24609665427509295, "grad_norm": 1.3767310116684208, "learning_rate": 8.82909333153656e-06, "loss": 0.0589, "step": 5296 }, { "epoch": 0.24614312267657992, "grad_norm": 1.4579081996989485, "learning_rate": 8.828609379519882e-06, "loss": 0.0798, "step": 5297 }, { "epoch": 0.24618959107806693, "grad_norm": 2.7676539139688976, "learning_rate": 8.828125340781171e-06, "loss": 0.0809, "step": 5298 }, { "epoch": 0.2462360594795539, "grad_norm": 0.9451487785319864, "learning_rate": 8.827641215331393e-06, "loss": 0.07, "step": 5299 }, { "epoch": 0.24628252788104088, "grad_norm": 1.8929314449705688, "learning_rate": 8.827157003181513e-06, "loss": 0.0983, "step": 5300 }, { "epoch": 0.24632899628252788, "grad_norm": 0.7710479234632945, "learning_rate": 8.826672704342498e-06, "loss": 0.048, "step": 5301 }, { "epoch": 0.24637546468401486, "grad_norm": 1.5945610060370243, "learning_rate": 8.82618831882532e-06, "loss": 0.1007, "step": 5302 }, { "epoch": 0.24642193308550187, "grad_norm": 2.6355097659378743, "learning_rate": 8.825703846640948e-06, "loss": 0.1088, "step": 5303 }, { "epoch": 0.24646840148698884, "grad_norm": 0.99088558422696, "learning_rate": 8.825219287800357e-06, "loss": 0.0658, "step": 5304 }, { "epoch": 0.24651486988847585, "grad_norm": 2.050457098504385, "learning_rate": 8.824734642314525e-06, "loss": 0.0634, "step": 5305 }, { "epoch": 0.24656133828996282, "grad_norm": 1.989304668878881, "learning_rate": 8.824249910194427e-06, "loss": 0.0854, "step": 5306 }, { "epoch": 0.24660780669144983, "grad_norm": 0.8146954188800696, "learning_rate": 8.823765091451045e-06, "loss": 0.0613, "step": 5307 }, { "epoch": 0.2466542750929368, "grad_norm": 2.0833227424333955, "learning_rate": 8.823280186095357e-06, "loss": 0.1088, "step": 5308 }, { "epoch": 0.24670074349442378, "grad_norm": 2.0833227424333955, "learning_rate": 8.823280186095357e-06, "loss": 0.1068, "step": 5309 }, { "epoch": 0.24674721189591078, "grad_norm": 2.3876635565355686, "learning_rate": 8.82279519413835e-06, "loss": 0.0885, "step": 5310 }, { "epoch": 0.24679368029739776, "grad_norm": 3.7410748365631425, "learning_rate": 8.822310115591008e-06, "loss": 0.094, "step": 5311 }, { "epoch": 0.24684014869888476, "grad_norm": 2.587529363821231, "learning_rate": 8.82182495046432e-06, "loss": 0.0585, "step": 5312 }, { "epoch": 0.24688661710037174, "grad_norm": 1.5189876311697783, "learning_rate": 8.821339698769272e-06, "loss": 0.0886, "step": 5313 }, { "epoch": 0.24693308550185875, "grad_norm": 2.2964692567083294, "learning_rate": 8.82085436051686e-06, "loss": 0.066, "step": 5314 }, { "epoch": 0.24697955390334572, "grad_norm": 2.1820585054429547, "learning_rate": 8.820368935718076e-06, "loss": 0.0879, "step": 5315 }, { "epoch": 0.24702602230483273, "grad_norm": 1.2296979696714834, "learning_rate": 8.819883424383913e-06, "loss": 0.0565, "step": 5316 }, { "epoch": 0.2470724907063197, "grad_norm": 1.6202171997168238, "learning_rate": 8.81939782652537e-06, "loss": 0.0529, "step": 5317 }, { "epoch": 0.24711895910780668, "grad_norm": 2.0047699329340247, "learning_rate": 8.818912142153449e-06, "loss": 0.0675, "step": 5318 }, { "epoch": 0.24716542750929368, "grad_norm": 1.0295001425616896, "learning_rate": 8.818426371279145e-06, "loss": 0.0547, "step": 5319 }, { "epoch": 0.24721189591078066, "grad_norm": 1.1664645536333467, "learning_rate": 8.817940513913468e-06, "loss": 0.0966, "step": 5320 }, { "epoch": 0.24725836431226766, "grad_norm": 2.2011926402510325, "learning_rate": 8.817454570067418e-06, "loss": 0.1309, "step": 5321 }, { "epoch": 0.24730483271375464, "grad_norm": 2.2026271267622493, "learning_rate": 8.816968539752004e-06, "loss": 0.0973, "step": 5322 }, { "epoch": 0.24735130111524165, "grad_norm": 1.2128443287879982, "learning_rate": 8.816482422978237e-06, "loss": 0.0695, "step": 5323 }, { "epoch": 0.24739776951672862, "grad_norm": 1.269355156306986, "learning_rate": 8.815996219757125e-06, "loss": 0.0778, "step": 5324 }, { "epoch": 0.24744423791821563, "grad_norm": 1.2440714214751194, "learning_rate": 8.815509930099683e-06, "loss": 0.0724, "step": 5325 }, { "epoch": 0.2474907063197026, "grad_norm": 0.7409260573060463, "learning_rate": 8.815023554016926e-06, "loss": 0.0697, "step": 5326 }, { "epoch": 0.24753717472118958, "grad_norm": 0.996508386105754, "learning_rate": 8.81453709151987e-06, "loss": 0.0865, "step": 5327 }, { "epoch": 0.24758364312267658, "grad_norm": 1.0950128366539171, "learning_rate": 8.814050542619532e-06, "loss": 0.0595, "step": 5328 }, { "epoch": 0.24763011152416356, "grad_norm": 1.0338842846012968, "learning_rate": 8.813563907326938e-06, "loss": 0.0751, "step": 5329 }, { "epoch": 0.24767657992565056, "grad_norm": 1.5501565880042532, "learning_rate": 8.813077185653106e-06, "loss": 0.1106, "step": 5330 }, { "epoch": 0.24772304832713754, "grad_norm": 0.975783571187496, "learning_rate": 8.812590377609062e-06, "loss": 0.0719, "step": 5331 }, { "epoch": 0.24776951672862454, "grad_norm": 3.301874741010887, "learning_rate": 8.812103483205835e-06, "loss": 0.1142, "step": 5332 }, { "epoch": 0.24781598513011152, "grad_norm": 1.0158665057393605, "learning_rate": 8.811616502454451e-06, "loss": 0.067, "step": 5333 }, { "epoch": 0.24786245353159853, "grad_norm": 0.7075053714993954, "learning_rate": 8.811129435365941e-06, "loss": 0.0756, "step": 5334 }, { "epoch": 0.2479089219330855, "grad_norm": 1.5400173483256285, "learning_rate": 8.81064228195134e-06, "loss": 0.0865, "step": 5335 }, { "epoch": 0.24795539033457248, "grad_norm": 2.522959356393249, "learning_rate": 8.81015504222168e-06, "loss": 0.1191, "step": 5336 }, { "epoch": 0.24800185873605948, "grad_norm": 0.579932189270826, "learning_rate": 8.809667716187998e-06, "loss": 0.0468, "step": 5337 }, { "epoch": 0.24804832713754646, "grad_norm": 0.7496727743537818, "learning_rate": 8.80918030386133e-06, "loss": 0.0474, "step": 5338 }, { "epoch": 0.24809479553903346, "grad_norm": 1.1595102996843107, "learning_rate": 8.808692805252723e-06, "loss": 0.0823, "step": 5339 }, { "epoch": 0.24814126394052044, "grad_norm": 1.8451083617528716, "learning_rate": 8.808205220373214e-06, "loss": 0.0851, "step": 5340 }, { "epoch": 0.24818773234200744, "grad_norm": 0.6574780048627307, "learning_rate": 8.80771754923385e-06, "loss": 0.0558, "step": 5341 }, { "epoch": 0.24823420074349442, "grad_norm": 2.2292012444317417, "learning_rate": 8.807229791845673e-06, "loss": 0.1118, "step": 5342 }, { "epoch": 0.24828066914498143, "grad_norm": 1.3200023559849712, "learning_rate": 8.806741948219736e-06, "loss": 0.0819, "step": 5343 }, { "epoch": 0.2483271375464684, "grad_norm": 1.3745142689898715, "learning_rate": 8.806254018367086e-06, "loss": 0.0724, "step": 5344 }, { "epoch": 0.24837360594795538, "grad_norm": 1.3425471824629092, "learning_rate": 8.805766002298778e-06, "loss": 0.0495, "step": 5345 }, { "epoch": 0.24842007434944238, "grad_norm": 1.853538853071791, "learning_rate": 8.805277900025863e-06, "loss": 0.0617, "step": 5346 }, { "epoch": 0.24846654275092936, "grad_norm": 1.4565990070437305, "learning_rate": 8.804789711559401e-06, "loss": 0.066, "step": 5347 }, { "epoch": 0.24851301115241636, "grad_norm": 1.3560197835217032, "learning_rate": 8.804301436910445e-06, "loss": 0.0999, "step": 5348 }, { "epoch": 0.24855947955390334, "grad_norm": 1.6275381750585114, "learning_rate": 8.80381307609006e-06, "loss": 0.0999, "step": 5349 }, { "epoch": 0.24860594795539034, "grad_norm": 1.3756910041609234, "learning_rate": 8.803324629109305e-06, "loss": 0.054, "step": 5350 }, { "epoch": 0.24865241635687732, "grad_norm": 0.9703197729371467, "learning_rate": 8.802836095979241e-06, "loss": 0.0792, "step": 5351 }, { "epoch": 0.24869888475836432, "grad_norm": 1.2795509599298251, "learning_rate": 8.80234747671094e-06, "loss": 0.0871, "step": 5352 }, { "epoch": 0.2487453531598513, "grad_norm": 0.790827015757868, "learning_rate": 8.801858771315466e-06, "loss": 0.0557, "step": 5353 }, { "epoch": 0.24879182156133828, "grad_norm": 1.5860563874973992, "learning_rate": 8.80136997980389e-06, "loss": 0.0625, "step": 5354 }, { "epoch": 0.24883828996282528, "grad_norm": 1.0975061705121376, "learning_rate": 8.800881102187282e-06, "loss": 0.0708, "step": 5355 }, { "epoch": 0.24888475836431226, "grad_norm": 1.5420236688950546, "learning_rate": 8.800392138476718e-06, "loss": 0.092, "step": 5356 }, { "epoch": 0.24893122676579926, "grad_norm": 1.8148964687669449, "learning_rate": 8.79990308868327e-06, "loss": 0.1022, "step": 5357 }, { "epoch": 0.24897769516728624, "grad_norm": 1.1323815313869876, "learning_rate": 8.799413952818021e-06, "loss": 0.0734, "step": 5358 }, { "epoch": 0.24902416356877324, "grad_norm": 1.1277567025547557, "learning_rate": 8.798924730892044e-06, "loss": 0.0567, "step": 5359 }, { "epoch": 0.24907063197026022, "grad_norm": 0.43152235973249037, "learning_rate": 8.798435422916425e-06, "loss": 0.0491, "step": 5360 }, { "epoch": 0.24911710037174722, "grad_norm": 0.7678996144653694, "learning_rate": 8.797946028902247e-06, "loss": 0.0891, "step": 5361 }, { "epoch": 0.2491635687732342, "grad_norm": 0.9252054604355162, "learning_rate": 8.797456548860592e-06, "loss": 0.062, "step": 5362 }, { "epoch": 0.24921003717472118, "grad_norm": 0.7131331830478292, "learning_rate": 8.79696698280255e-06, "loss": 0.0607, "step": 5363 }, { "epoch": 0.24925650557620818, "grad_norm": 1.2329519991060918, "learning_rate": 8.79647733073921e-06, "loss": 0.0831, "step": 5364 }, { "epoch": 0.24930297397769516, "grad_norm": 1.8890773601580761, "learning_rate": 8.795987592681663e-06, "loss": 0.0644, "step": 5365 }, { "epoch": 0.24934944237918216, "grad_norm": 1.7474101631589345, "learning_rate": 8.795497768641001e-06, "loss": 0.1152, "step": 5366 }, { "epoch": 0.24939591078066914, "grad_norm": 1.9356266499941401, "learning_rate": 8.795007858628322e-06, "loss": 0.0681, "step": 5367 }, { "epoch": 0.24944237918215614, "grad_norm": 5.316959590108893, "learning_rate": 8.794517862654718e-06, "loss": 0.0837, "step": 5368 }, { "epoch": 0.24948884758364312, "grad_norm": 1.2358790964782167, "learning_rate": 8.794027780731292e-06, "loss": 0.0775, "step": 5369 }, { "epoch": 0.24953531598513012, "grad_norm": 0.9139196439689635, "learning_rate": 8.793537612869143e-06, "loss": 0.0553, "step": 5370 }, { "epoch": 0.2495817843866171, "grad_norm": 0.9185646759526294, "learning_rate": 8.793047359079376e-06, "loss": 0.0585, "step": 5371 }, { "epoch": 0.24962825278810408, "grad_norm": 1.7969583991948783, "learning_rate": 8.792557019373094e-06, "loss": 0.0575, "step": 5372 }, { "epoch": 0.24967472118959108, "grad_norm": 1.1440504959894615, "learning_rate": 8.792066593761404e-06, "loss": 0.055, "step": 5373 }, { "epoch": 0.24972118959107806, "grad_norm": 1.0093634221742902, "learning_rate": 8.791576082255414e-06, "loss": 0.0596, "step": 5374 }, { "epoch": 0.24976765799256506, "grad_norm": 4.677557742241432, "learning_rate": 8.791085484866233e-06, "loss": 0.1398, "step": 5375 }, { "epoch": 0.24981412639405204, "grad_norm": 0.9169208124699322, "learning_rate": 8.790594801604978e-06, "loss": 0.0757, "step": 5376 }, { "epoch": 0.24986059479553904, "grad_norm": 1.1803553692503777, "learning_rate": 8.790104032482761e-06, "loss": 0.0731, "step": 5377 }, { "epoch": 0.24990706319702602, "grad_norm": 0.6692068367098044, "learning_rate": 8.789613177510698e-06, "loss": 0.0542, "step": 5378 }, { "epoch": 0.24995353159851302, "grad_norm": 1.383443741523499, "learning_rate": 8.78912223669991e-06, "loss": 0.075, "step": 5379 }, { "epoch": 0.25, "grad_norm": 1.2949723882897937, "learning_rate": 8.788631210061514e-06, "loss": 0.1269, "step": 5380 }, { "epoch": 0.250046468401487, "grad_norm": 1.1206301754363464, "learning_rate": 8.788140097606633e-06, "loss": 0.0923, "step": 5381 }, { "epoch": 0.25009293680297395, "grad_norm": 1.0033559389564215, "learning_rate": 8.787648899346393e-06, "loss": 0.0619, "step": 5382 }, { "epoch": 0.250139405204461, "grad_norm": 1.1168706287136534, "learning_rate": 8.787157615291918e-06, "loss": 0.0464, "step": 5383 }, { "epoch": 0.25018587360594796, "grad_norm": 1.9848666452612105, "learning_rate": 8.786666245454338e-06, "loss": 0.0532, "step": 5384 }, { "epoch": 0.25023234200743494, "grad_norm": 1.309685303852879, "learning_rate": 8.786174789844782e-06, "loss": 0.0736, "step": 5385 }, { "epoch": 0.2502788104089219, "grad_norm": 1.0924094125814354, "learning_rate": 8.78568324847438e-06, "loss": 0.083, "step": 5386 }, { "epoch": 0.25032527881040895, "grad_norm": 1.3537013441988612, "learning_rate": 8.78519162135427e-06, "loss": 0.035, "step": 5387 }, { "epoch": 0.2503717472118959, "grad_norm": 0.8387299815962832, "learning_rate": 8.784699908495587e-06, "loss": 0.0591, "step": 5388 }, { "epoch": 0.2504182156133829, "grad_norm": 1.7175140160446989, "learning_rate": 8.784208109909465e-06, "loss": 0.0957, "step": 5389 }, { "epoch": 0.2504646840148699, "grad_norm": 1.1414220713179137, "learning_rate": 8.783716225607048e-06, "loss": 0.0762, "step": 5390 }, { "epoch": 0.25051115241635685, "grad_norm": 1.3256639017048282, "learning_rate": 8.783224255599476e-06, "loss": 0.0649, "step": 5391 }, { "epoch": 0.2505576208178439, "grad_norm": 1.4502141551957588, "learning_rate": 8.782732199897894e-06, "loss": 0.0631, "step": 5392 }, { "epoch": 0.25060408921933086, "grad_norm": 9.347254469647265, "learning_rate": 8.782240058513444e-06, "loss": 0.0549, "step": 5393 }, { "epoch": 0.25065055762081784, "grad_norm": 2.082583862970823, "learning_rate": 8.78174783145728e-06, "loss": 0.1328, "step": 5394 }, { "epoch": 0.2506970260223048, "grad_norm": 0.838305174014438, "learning_rate": 8.781255518740543e-06, "loss": 0.0563, "step": 5395 }, { "epoch": 0.25074349442379185, "grad_norm": 1.861979658269817, "learning_rate": 8.780763120374391e-06, "loss": 0.0903, "step": 5396 }, { "epoch": 0.2507899628252788, "grad_norm": 0.9724927515054748, "learning_rate": 8.780270636369974e-06, "loss": 0.0879, "step": 5397 }, { "epoch": 0.2508364312267658, "grad_norm": 1.2478789202789393, "learning_rate": 8.779778066738451e-06, "loss": 0.064, "step": 5398 }, { "epoch": 0.2508828996282528, "grad_norm": 1.4894976615813986, "learning_rate": 8.779285411490974e-06, "loss": 0.077, "step": 5399 }, { "epoch": 0.25092936802973975, "grad_norm": 1.0290879118771459, "learning_rate": 8.778792670638705e-06, "loss": 0.0731, "step": 5400 }, { "epoch": 0.2509758364312268, "grad_norm": 1.881316285876013, "learning_rate": 8.778299844192806e-06, "loss": 0.0722, "step": 5401 }, { "epoch": 0.25102230483271376, "grad_norm": 2.2212916396771236, "learning_rate": 8.777806932164438e-06, "loss": 0.1072, "step": 5402 }, { "epoch": 0.25106877323420074, "grad_norm": 2.8380109421650257, "learning_rate": 8.777313934564768e-06, "loss": 0.0655, "step": 5403 }, { "epoch": 0.2511152416356877, "grad_norm": 1.8275814929541316, "learning_rate": 8.776820851404962e-06, "loss": 0.0805, "step": 5404 }, { "epoch": 0.25116171003717475, "grad_norm": 1.2425294315601085, "learning_rate": 8.776327682696187e-06, "loss": 0.0831, "step": 5405 }, { "epoch": 0.2512081784386617, "grad_norm": 1.3358106332235797, "learning_rate": 8.775834428449617e-06, "loss": 0.067, "step": 5406 }, { "epoch": 0.2512546468401487, "grad_norm": 0.6568893193371584, "learning_rate": 8.775341088676422e-06, "loss": 0.0604, "step": 5407 }, { "epoch": 0.2513011152416357, "grad_norm": 1.9392915129996664, "learning_rate": 8.77484766338778e-06, "loss": 0.0755, "step": 5408 }, { "epoch": 0.25134758364312265, "grad_norm": 1.1967339426124732, "learning_rate": 8.774354152594862e-06, "loss": 0.0805, "step": 5409 }, { "epoch": 0.2513940520446097, "grad_norm": 1.7033512101879422, "learning_rate": 8.773860556308852e-06, "loss": 0.0896, "step": 5410 }, { "epoch": 0.25144052044609666, "grad_norm": 1.1594900696439476, "learning_rate": 8.77336687454093e-06, "loss": 0.0762, "step": 5411 }, { "epoch": 0.25148698884758364, "grad_norm": 1.102695539357091, "learning_rate": 8.772873107302274e-06, "loss": 0.0686, "step": 5412 }, { "epoch": 0.2515334572490706, "grad_norm": 0.9209829887623318, "learning_rate": 8.772379254604074e-06, "loss": 0.0412, "step": 5413 }, { "epoch": 0.25157992565055765, "grad_norm": 1.385407395349413, "learning_rate": 8.771885316457512e-06, "loss": 0.0984, "step": 5414 }, { "epoch": 0.2516263940520446, "grad_norm": 0.8510724855157134, "learning_rate": 8.771391292873777e-06, "loss": 0.0652, "step": 5415 }, { "epoch": 0.2516728624535316, "grad_norm": 1.5644439633283513, "learning_rate": 8.77089718386406e-06, "loss": 0.1457, "step": 5416 }, { "epoch": 0.2517193308550186, "grad_norm": 2.2621866769714147, "learning_rate": 8.770402989439556e-06, "loss": 0.0806, "step": 5417 }, { "epoch": 0.25176579925650555, "grad_norm": 1.3065166441136038, "learning_rate": 8.769908709611453e-06, "loss": 0.0842, "step": 5418 }, { "epoch": 0.2518122676579926, "grad_norm": 0.9636704896550705, "learning_rate": 8.76941434439095e-06, "loss": 0.0747, "step": 5419 }, { "epoch": 0.25185873605947956, "grad_norm": 1.5707689272448073, "learning_rate": 8.768919893789244e-06, "loss": 0.1085, "step": 5420 }, { "epoch": 0.25190520446096654, "grad_norm": 1.390435381223245, "learning_rate": 8.768425357817538e-06, "loss": 0.0777, "step": 5421 }, { "epoch": 0.2519516728624535, "grad_norm": 1.0574878732037256, "learning_rate": 8.767930736487031e-06, "loss": 0.0654, "step": 5422 }, { "epoch": 0.25199814126394054, "grad_norm": 1.8111086649703116, "learning_rate": 8.767436029808926e-06, "loss": 0.1262, "step": 5423 }, { "epoch": 0.2520446096654275, "grad_norm": 1.340626783013559, "learning_rate": 8.766941237794432e-06, "loss": 0.0906, "step": 5424 }, { "epoch": 0.2520910780669145, "grad_norm": 0.9477914541190341, "learning_rate": 8.766446360454753e-06, "loss": 0.0696, "step": 5425 }, { "epoch": 0.2521375464684015, "grad_norm": 0.5431415485270258, "learning_rate": 8.765951397801099e-06, "loss": 0.0385, "step": 5426 }, { "epoch": 0.25218401486988845, "grad_norm": 0.8910455775390588, "learning_rate": 8.765456349844683e-06, "loss": 0.0869, "step": 5427 }, { "epoch": 0.2522304832713755, "grad_norm": 1.3684977683895176, "learning_rate": 8.764961216596718e-06, "loss": 0.0616, "step": 5428 }, { "epoch": 0.25227695167286246, "grad_norm": 0.9731672302434503, "learning_rate": 8.764465998068418e-06, "loss": 0.0546, "step": 5429 }, { "epoch": 0.25232342007434944, "grad_norm": 1.387832872256046, "learning_rate": 8.763970694271003e-06, "loss": 0.1147, "step": 5430 }, { "epoch": 0.2523698884758364, "grad_norm": 1.014683245387092, "learning_rate": 8.763475305215688e-06, "loss": 0.1261, "step": 5431 }, { "epoch": 0.25241635687732344, "grad_norm": 0.8022474207425178, "learning_rate": 8.762979830913697e-06, "loss": 0.0618, "step": 5432 }, { "epoch": 0.2524628252788104, "grad_norm": 0.5433092253901854, "learning_rate": 8.762484271376252e-06, "loss": 0.0422, "step": 5433 }, { "epoch": 0.2525092936802974, "grad_norm": 1.4636739170270103, "learning_rate": 8.761988626614578e-06, "loss": 0.0806, "step": 5434 }, { "epoch": 0.2525557620817844, "grad_norm": 1.6277312301418343, "learning_rate": 8.761492896639905e-06, "loss": 0.0558, "step": 5435 }, { "epoch": 0.25260223048327135, "grad_norm": 1.8623983759331169, "learning_rate": 8.760997081463458e-06, "loss": 0.1006, "step": 5436 }, { "epoch": 0.2526486988847584, "grad_norm": 1.1261417700570255, "learning_rate": 8.760501181096467e-06, "loss": 0.1008, "step": 5437 }, { "epoch": 0.25269516728624536, "grad_norm": 1.79379504703238, "learning_rate": 8.760005195550168e-06, "loss": 0.0737, "step": 5438 }, { "epoch": 0.25274163568773234, "grad_norm": 1.8479073641195989, "learning_rate": 8.759509124835791e-06, "loss": 0.1569, "step": 5439 }, { "epoch": 0.2527881040892193, "grad_norm": 0.6722001660000874, "learning_rate": 8.75901296896458e-06, "loss": 0.0613, "step": 5440 }, { "epoch": 0.25283457249070634, "grad_norm": 1.8897150388206532, "learning_rate": 8.758516727947765e-06, "loss": 0.1149, "step": 5441 }, { "epoch": 0.2528810408921933, "grad_norm": 1.4643661929190683, "learning_rate": 8.758020401796593e-06, "loss": 0.0839, "step": 5442 }, { "epoch": 0.2529275092936803, "grad_norm": 0.5752651293572775, "learning_rate": 8.757523990522304e-06, "loss": 0.0432, "step": 5443 }, { "epoch": 0.2529739776951673, "grad_norm": 0.6776868275387035, "learning_rate": 8.75702749413614e-06, "loss": 0.0388, "step": 5444 }, { "epoch": 0.25302044609665425, "grad_norm": 1.4147211725320956, "learning_rate": 8.756530912649348e-06, "loss": 0.0752, "step": 5445 }, { "epoch": 0.2530669144981413, "grad_norm": 1.816025355688588, "learning_rate": 8.75603424607318e-06, "loss": 0.0806, "step": 5446 }, { "epoch": 0.25311338289962826, "grad_norm": 3.50113043947025, "learning_rate": 8.755537494418882e-06, "loss": 0.1057, "step": 5447 }, { "epoch": 0.25315985130111524, "grad_norm": 1.352782754337955, "learning_rate": 8.755040657697707e-06, "loss": 0.056, "step": 5448 }, { "epoch": 0.2532063197026022, "grad_norm": 0.9827716903406742, "learning_rate": 8.754543735920908e-06, "loss": 0.0902, "step": 5449 }, { "epoch": 0.25325278810408924, "grad_norm": 0.9229417690792321, "learning_rate": 8.754046729099744e-06, "loss": 0.0585, "step": 5450 }, { "epoch": 0.2532992565055762, "grad_norm": 2.054562690218911, "learning_rate": 8.753549637245469e-06, "loss": 0.0938, "step": 5451 }, { "epoch": 0.2533457249070632, "grad_norm": 1.303970864074269, "learning_rate": 8.753052460369345e-06, "loss": 0.0811, "step": 5452 }, { "epoch": 0.2533921933085502, "grad_norm": 1.6644696237001562, "learning_rate": 8.752555198482632e-06, "loss": 0.1651, "step": 5453 }, { "epoch": 0.25343866171003715, "grad_norm": 0.8038804499713873, "learning_rate": 8.752057851596594e-06, "loss": 0.0394, "step": 5454 }, { "epoch": 0.2534851301115242, "grad_norm": 1.3254904313585913, "learning_rate": 8.751560419722498e-06, "loss": 0.0747, "step": 5455 }, { "epoch": 0.25353159851301116, "grad_norm": 1.8485678615321126, "learning_rate": 8.751062902871608e-06, "loss": 0.1044, "step": 5456 }, { "epoch": 0.25357806691449813, "grad_norm": 1.6676365829913211, "learning_rate": 8.750565301055198e-06, "loss": 0.096, "step": 5457 }, { "epoch": 0.2536245353159851, "grad_norm": 1.0288133467324239, "learning_rate": 8.750067614284534e-06, "loss": 0.056, "step": 5458 }, { "epoch": 0.25367100371747214, "grad_norm": 2.061895148682722, "learning_rate": 8.749569842570893e-06, "loss": 0.0739, "step": 5459 }, { "epoch": 0.2537174721189591, "grad_norm": 0.7683749750961444, "learning_rate": 8.74907198592555e-06, "loss": 0.0811, "step": 5460 }, { "epoch": 0.2537639405204461, "grad_norm": 4.38750252513919, "learning_rate": 8.74857404435978e-06, "loss": 0.0785, "step": 5461 }, { "epoch": 0.2538104089219331, "grad_norm": 0.8468585974379308, "learning_rate": 8.74807601788486e-06, "loss": 0.0633, "step": 5462 }, { "epoch": 0.25385687732342005, "grad_norm": 1.783863666416928, "learning_rate": 8.747577906512076e-06, "loss": 0.0869, "step": 5463 }, { "epoch": 0.2539033457249071, "grad_norm": 1.142491980529738, "learning_rate": 8.747079710252707e-06, "loss": 0.0675, "step": 5464 }, { "epoch": 0.25394981412639406, "grad_norm": 1.728347233071497, "learning_rate": 8.74658142911804e-06, "loss": 0.0799, "step": 5465 }, { "epoch": 0.25399628252788103, "grad_norm": 1.178189893795703, "learning_rate": 8.74608306311936e-06, "loss": 0.088, "step": 5466 }, { "epoch": 0.254042750929368, "grad_norm": 0.8359196375592831, "learning_rate": 8.745584612267955e-06, "loss": 0.0638, "step": 5467 }, { "epoch": 0.25408921933085504, "grad_norm": 1.7976865062213456, "learning_rate": 8.745086076575116e-06, "loss": 0.0843, "step": 5468 }, { "epoch": 0.254135687732342, "grad_norm": 1.5838868481190331, "learning_rate": 8.744587456052138e-06, "loss": 0.1062, "step": 5469 }, { "epoch": 0.254182156133829, "grad_norm": 1.4570970520796158, "learning_rate": 8.74408875071031e-06, "loss": 0.0611, "step": 5470 }, { "epoch": 0.254228624535316, "grad_norm": 1.389058407363047, "learning_rate": 8.743589960560934e-06, "loss": 0.099, "step": 5471 }, { "epoch": 0.25427509293680295, "grad_norm": 3.0792453493712437, "learning_rate": 8.743091085615305e-06, "loss": 0.1145, "step": 5472 }, { "epoch": 0.25432156133829, "grad_norm": 0.7355607462350341, "learning_rate": 8.742592125884722e-06, "loss": 0.0429, "step": 5473 }, { "epoch": 0.25436802973977696, "grad_norm": 2.1548152209649305, "learning_rate": 8.74209308138049e-06, "loss": 0.0835, "step": 5474 }, { "epoch": 0.25441449814126393, "grad_norm": 1.190953585680269, "learning_rate": 8.741593952113909e-06, "loss": 0.0766, "step": 5475 }, { "epoch": 0.2544609665427509, "grad_norm": 1.436414214351826, "learning_rate": 8.74109473809629e-06, "loss": 0.0782, "step": 5476 }, { "epoch": 0.25450743494423794, "grad_norm": 2.8903902692594445, "learning_rate": 8.740595439338934e-06, "loss": 0.0757, "step": 5477 }, { "epoch": 0.2545539033457249, "grad_norm": 0.8515300974134834, "learning_rate": 8.740096055853158e-06, "loss": 0.0707, "step": 5478 }, { "epoch": 0.2546003717472119, "grad_norm": 1.1572242549636698, "learning_rate": 8.739596587650267e-06, "loss": 0.0785, "step": 5479 }, { "epoch": 0.25464684014869887, "grad_norm": 0.6122744964064876, "learning_rate": 8.739097034741579e-06, "loss": 0.0586, "step": 5480 }, { "epoch": 0.25469330855018585, "grad_norm": 0.9332456490755097, "learning_rate": 8.738597397138409e-06, "loss": 0.0393, "step": 5481 }, { "epoch": 0.2547397769516729, "grad_norm": 1.6410788585098952, "learning_rate": 8.738097674852071e-06, "loss": 0.1133, "step": 5482 }, { "epoch": 0.25478624535315986, "grad_norm": 1.974079770657937, "learning_rate": 8.737597867893887e-06, "loss": 0.0789, "step": 5483 }, { "epoch": 0.25483271375464683, "grad_norm": 1.8267073176499575, "learning_rate": 8.737097976275177e-06, "loss": 0.1657, "step": 5484 }, { "epoch": 0.2548791821561338, "grad_norm": 1.9365166417211725, "learning_rate": 8.736598000007266e-06, "loss": 0.0898, "step": 5485 }, { "epoch": 0.25492565055762084, "grad_norm": 0.7540722887248604, "learning_rate": 8.736097939101477e-06, "loss": 0.0737, "step": 5486 }, { "epoch": 0.2549721189591078, "grad_norm": 0.9528331089471161, "learning_rate": 8.735597793569137e-06, "loss": 0.0618, "step": 5487 }, { "epoch": 0.2550185873605948, "grad_norm": 1.155384027891273, "learning_rate": 8.735097563421577e-06, "loss": 0.088, "step": 5488 }, { "epoch": 0.25506505576208177, "grad_norm": 1.1389982649419046, "learning_rate": 8.734597248670125e-06, "loss": 0.0632, "step": 5489 }, { "epoch": 0.25511152416356875, "grad_norm": 1.9399879375391784, "learning_rate": 8.734096849326113e-06, "loss": 0.1092, "step": 5490 }, { "epoch": 0.2551579925650558, "grad_norm": 2.304630255127737, "learning_rate": 8.733596365400879e-06, "loss": 0.1008, "step": 5491 }, { "epoch": 0.25520446096654276, "grad_norm": 0.943999003109665, "learning_rate": 8.733095796905758e-06, "loss": 0.0688, "step": 5492 }, { "epoch": 0.25525092936802973, "grad_norm": 2.06310765236626, "learning_rate": 8.732595143852088e-06, "loss": 0.0868, "step": 5493 }, { "epoch": 0.2552973977695167, "grad_norm": 0.8906116204796467, "learning_rate": 8.732094406251209e-06, "loss": 0.0588, "step": 5494 }, { "epoch": 0.25534386617100374, "grad_norm": 1.4994430230666733, "learning_rate": 8.731593584114465e-06, "loss": 0.1009, "step": 5495 }, { "epoch": 0.2553903345724907, "grad_norm": 0.8492948696598038, "learning_rate": 8.731092677453197e-06, "loss": 0.0651, "step": 5496 }, { "epoch": 0.2554368029739777, "grad_norm": 1.278080540900851, "learning_rate": 8.730591686278755e-06, "loss": 0.0734, "step": 5497 }, { "epoch": 0.25548327137546467, "grad_norm": 1.2147885352223295, "learning_rate": 8.730090610602484e-06, "loss": 0.1017, "step": 5498 }, { "epoch": 0.25552973977695165, "grad_norm": 1.0667987529080045, "learning_rate": 8.729589450435733e-06, "loss": 0.0634, "step": 5499 }, { "epoch": 0.2555762081784387, "grad_norm": 0.9571998895317938, "learning_rate": 8.729088205789858e-06, "loss": 0.0578, "step": 5500 }, { "epoch": 0.25562267657992566, "grad_norm": 1.0463976011448746, "learning_rate": 8.72858687667621e-06, "loss": 0.0576, "step": 5501 }, { "epoch": 0.25566914498141263, "grad_norm": 1.2068482716096953, "learning_rate": 8.728085463106144e-06, "loss": 0.0722, "step": 5502 }, { "epoch": 0.2557156133828996, "grad_norm": 3.0844602764368796, "learning_rate": 8.72758396509102e-06, "loss": 0.0901, "step": 5503 }, { "epoch": 0.25576208178438664, "grad_norm": 1.064630195850018, "learning_rate": 8.727082382642195e-06, "loss": 0.0577, "step": 5504 }, { "epoch": 0.2558085501858736, "grad_norm": 0.6639184422495639, "learning_rate": 8.726580715771032e-06, "loss": 0.0685, "step": 5505 }, { "epoch": 0.2558550185873606, "grad_norm": 0.9879677352210645, "learning_rate": 8.726078964488893e-06, "loss": 0.0573, "step": 5506 }, { "epoch": 0.25590148698884757, "grad_norm": 0.8540373125645443, "learning_rate": 8.725577128807144e-06, "loss": 0.0691, "step": 5507 }, { "epoch": 0.25594795539033455, "grad_norm": 1.751738055451408, "learning_rate": 8.72507520873715e-06, "loss": 0.0598, "step": 5508 }, { "epoch": 0.2559944237918216, "grad_norm": 1.7972861974552237, "learning_rate": 8.724573204290285e-06, "loss": 0.0682, "step": 5509 }, { "epoch": 0.25604089219330856, "grad_norm": 2.2899320855177456, "learning_rate": 8.724071115477915e-06, "loss": 0.1198, "step": 5510 }, { "epoch": 0.25608736059479553, "grad_norm": 1.330708621565504, "learning_rate": 8.723568942311416e-06, "loss": 0.0652, "step": 5511 }, { "epoch": 0.2561338289962825, "grad_norm": 1.6840350306705352, "learning_rate": 8.723066684802162e-06, "loss": 0.1174, "step": 5512 }, { "epoch": 0.25618029739776954, "grad_norm": 0.7790142788933946, "learning_rate": 8.722564342961528e-06, "loss": 0.082, "step": 5513 }, { "epoch": 0.2562267657992565, "grad_norm": 1.1816474461825948, "learning_rate": 8.722061916800892e-06, "loss": 0.0704, "step": 5514 }, { "epoch": 0.2562732342007435, "grad_norm": 1.1765642618901258, "learning_rate": 8.72155940633164e-06, "loss": 0.0784, "step": 5515 }, { "epoch": 0.25631970260223047, "grad_norm": 1.2783087652985006, "learning_rate": 8.721056811565147e-06, "loss": 0.0483, "step": 5516 }, { "epoch": 0.25636617100371745, "grad_norm": 1.0267219167173, "learning_rate": 8.720554132512802e-06, "loss": 0.0498, "step": 5517 }, { "epoch": 0.2564126394052045, "grad_norm": 2.6494280300573276, "learning_rate": 8.72005136918599e-06, "loss": 0.0373, "step": 5518 }, { "epoch": 0.25645910780669146, "grad_norm": 1.248296149242599, "learning_rate": 8.719548521596102e-06, "loss": 0.0588, "step": 5519 }, { "epoch": 0.25650557620817843, "grad_norm": 1.8527888397262262, "learning_rate": 8.719045589754522e-06, "loss": 0.1205, "step": 5520 }, { "epoch": 0.2565520446096654, "grad_norm": 1.6316119708128634, "learning_rate": 8.718542573672646e-06, "loss": 0.0996, "step": 5521 }, { "epoch": 0.25659851301115244, "grad_norm": 1.948948421353827, "learning_rate": 8.718039473361867e-06, "loss": 0.1092, "step": 5522 }, { "epoch": 0.2566449814126394, "grad_norm": 0.7329860371762073, "learning_rate": 8.71753628883358e-06, "loss": 0.0828, "step": 5523 }, { "epoch": 0.2566914498141264, "grad_norm": 1.742460323253493, "learning_rate": 8.717033020099186e-06, "loss": 0.0847, "step": 5524 }, { "epoch": 0.25673791821561337, "grad_norm": 2.1608906436121442, "learning_rate": 8.71652966717008e-06, "loss": 0.1095, "step": 5525 }, { "epoch": 0.25678438661710035, "grad_norm": 1.1715074001171621, "learning_rate": 8.716026230057664e-06, "loss": 0.0607, "step": 5526 }, { "epoch": 0.2568308550185874, "grad_norm": 0.8764513013294063, "learning_rate": 8.715522708773345e-06, "loss": 0.0389, "step": 5527 }, { "epoch": 0.25687732342007435, "grad_norm": 0.9079397025724278, "learning_rate": 8.715019103328525e-06, "loss": 0.0842, "step": 5528 }, { "epoch": 0.25692379182156133, "grad_norm": 0.7840586088126091, "learning_rate": 8.714515413734614e-06, "loss": 0.0676, "step": 5529 }, { "epoch": 0.2569702602230483, "grad_norm": 1.0116952022640042, "learning_rate": 8.714011640003018e-06, "loss": 0.065, "step": 5530 }, { "epoch": 0.25701672862453534, "grad_norm": 1.7566209557676735, "learning_rate": 8.71350778214515e-06, "loss": 0.0969, "step": 5531 }, { "epoch": 0.2570631970260223, "grad_norm": 0.8063463438322477, "learning_rate": 8.713003840172422e-06, "loss": 0.0804, "step": 5532 }, { "epoch": 0.2571096654275093, "grad_norm": 0.9320320882929731, "learning_rate": 8.71249981409625e-06, "loss": 0.0786, "step": 5533 }, { "epoch": 0.25715613382899627, "grad_norm": 0.6186163153031045, "learning_rate": 8.711995703928049e-06, "loss": 0.0514, "step": 5534 }, { "epoch": 0.25720260223048325, "grad_norm": 0.7024639124591052, "learning_rate": 8.71149150967924e-06, "loss": 0.0471, "step": 5535 }, { "epoch": 0.2572490706319703, "grad_norm": 0.781404271957857, "learning_rate": 8.71098723136124e-06, "loss": 0.0431, "step": 5536 }, { "epoch": 0.25729553903345725, "grad_norm": 1.8588410377706721, "learning_rate": 8.710482868985475e-06, "loss": 0.0822, "step": 5537 }, { "epoch": 0.25734200743494423, "grad_norm": 1.1144231316205435, "learning_rate": 8.709978422563366e-06, "loss": 0.0595, "step": 5538 }, { "epoch": 0.2573884758364312, "grad_norm": 1.0727076434198723, "learning_rate": 8.709473892106342e-06, "loss": 0.063, "step": 5539 }, { "epoch": 0.25743494423791824, "grad_norm": 1.7163003549900286, "learning_rate": 8.70896927762583e-06, "loss": 0.0775, "step": 5540 }, { "epoch": 0.2574814126394052, "grad_norm": 1.7855694244353437, "learning_rate": 8.708464579133261e-06, "loss": 0.1153, "step": 5541 }, { "epoch": 0.2575278810408922, "grad_norm": 1.659774492261617, "learning_rate": 8.707959796640065e-06, "loss": 0.1008, "step": 5542 }, { "epoch": 0.25757434944237917, "grad_norm": 2.3084593954860275, "learning_rate": 8.707454930157678e-06, "loss": 0.1212, "step": 5543 }, { "epoch": 0.25762081784386615, "grad_norm": 0.8076597973425224, "learning_rate": 8.706949979697535e-06, "loss": 0.0645, "step": 5544 }, { "epoch": 0.2576672862453532, "grad_norm": 3.0267463675511155, "learning_rate": 8.706444945271073e-06, "loss": 0.0516, "step": 5545 }, { "epoch": 0.25771375464684015, "grad_norm": 2.8747048953425898, "learning_rate": 8.705939826889732e-06, "loss": 0.0598, "step": 5546 }, { "epoch": 0.25776022304832713, "grad_norm": 1.5792619058431558, "learning_rate": 8.705434624564954e-06, "loss": 0.1601, "step": 5547 }, { "epoch": 0.2578066914498141, "grad_norm": 1.6447977198788002, "learning_rate": 8.704929338308181e-06, "loss": 0.0642, "step": 5548 }, { "epoch": 0.25785315985130114, "grad_norm": 0.82081540130352, "learning_rate": 8.70442396813086e-06, "loss": 0.0728, "step": 5549 }, { "epoch": 0.2578996282527881, "grad_norm": 1.6721879810395262, "learning_rate": 8.703918514044435e-06, "loss": 0.0912, "step": 5550 }, { "epoch": 0.2579460966542751, "grad_norm": 1.2317220816650276, "learning_rate": 8.703412976060361e-06, "loss": 0.0687, "step": 5551 }, { "epoch": 0.25799256505576207, "grad_norm": 1.6937531258742242, "learning_rate": 8.702907354190081e-06, "loss": 0.0564, "step": 5552 }, { "epoch": 0.25803903345724905, "grad_norm": 0.9829518868460922, "learning_rate": 8.702401648445057e-06, "loss": 0.0705, "step": 5553 }, { "epoch": 0.2580855018587361, "grad_norm": 1.5246993247609857, "learning_rate": 8.701895858836736e-06, "loss": 0.0764, "step": 5554 }, { "epoch": 0.25813197026022305, "grad_norm": 0.9853542830157485, "learning_rate": 8.701389985376578e-06, "loss": 0.0813, "step": 5555 }, { "epoch": 0.25817843866171003, "grad_norm": 1.295006121543516, "learning_rate": 8.700884028076042e-06, "loss": 0.0812, "step": 5556 }, { "epoch": 0.258224907063197, "grad_norm": 0.972265237244358, "learning_rate": 8.700377986946586e-06, "loss": 0.0449, "step": 5557 }, { "epoch": 0.25827137546468404, "grad_norm": 0.9516523932565712, "learning_rate": 8.699871861999676e-06, "loss": 0.069, "step": 5558 }, { "epoch": 0.258317843866171, "grad_norm": 1.4616000731870313, "learning_rate": 8.699365653246774e-06, "loss": 0.0793, "step": 5559 }, { "epoch": 0.258364312267658, "grad_norm": 0.9861514102901004, "learning_rate": 8.698859360699345e-06, "loss": 0.0736, "step": 5560 }, { "epoch": 0.25841078066914497, "grad_norm": 1.1110374172507358, "learning_rate": 8.69835298436886e-06, "loss": 0.0672, "step": 5561 }, { "epoch": 0.25845724907063194, "grad_norm": 1.1370415194460841, "learning_rate": 8.697846524266786e-06, "loss": 0.0686, "step": 5562 }, { "epoch": 0.258503717472119, "grad_norm": 0.7737514010180403, "learning_rate": 8.697339980404597e-06, "loss": 0.0294, "step": 5563 }, { "epoch": 0.25855018587360595, "grad_norm": 1.1923770350313079, "learning_rate": 8.696833352793766e-06, "loss": 0.0579, "step": 5564 }, { "epoch": 0.25859665427509293, "grad_norm": 1.250829507948748, "learning_rate": 8.696326641445768e-06, "loss": 0.0839, "step": 5565 }, { "epoch": 0.2586431226765799, "grad_norm": 0.8983963104465331, "learning_rate": 8.695819846372085e-06, "loss": 0.0665, "step": 5566 }, { "epoch": 0.25868959107806694, "grad_norm": 1.548194768945726, "learning_rate": 8.695312967584188e-06, "loss": 0.0824, "step": 5567 }, { "epoch": 0.2587360594795539, "grad_norm": 1.1251230134072276, "learning_rate": 8.694806005093566e-06, "loss": 0.0746, "step": 5568 }, { "epoch": 0.2587825278810409, "grad_norm": 1.6067795073222986, "learning_rate": 8.694298958911699e-06, "loss": 0.1177, "step": 5569 }, { "epoch": 0.25882899628252787, "grad_norm": 0.8263360838341024, "learning_rate": 8.693791829050072e-06, "loss": 0.0483, "step": 5570 }, { "epoch": 0.25887546468401484, "grad_norm": 1.6682927999140034, "learning_rate": 8.69328461552017e-06, "loss": 0.0729, "step": 5571 }, { "epoch": 0.2589219330855019, "grad_norm": 2.5067310262235547, "learning_rate": 8.692777318333485e-06, "loss": 0.0651, "step": 5572 }, { "epoch": 0.25896840148698885, "grad_norm": 1.5569864075145625, "learning_rate": 8.692269937501509e-06, "loss": 0.064, "step": 5573 }, { "epoch": 0.25901486988847583, "grad_norm": 2.3793516323875465, "learning_rate": 8.691762473035734e-06, "loss": 0.0898, "step": 5574 }, { "epoch": 0.2590613382899628, "grad_norm": 5.5990339682853145, "learning_rate": 8.691254924947652e-06, "loss": 0.135, "step": 5575 }, { "epoch": 0.25910780669144984, "grad_norm": 0.7888009170275784, "learning_rate": 8.690747293248761e-06, "loss": 0.051, "step": 5576 }, { "epoch": 0.2591542750929368, "grad_norm": 0.9206639403607023, "learning_rate": 8.690239577950558e-06, "loss": 0.07, "step": 5577 }, { "epoch": 0.2592007434944238, "grad_norm": 2.3663838903446304, "learning_rate": 8.689731779064546e-06, "loss": 0.1015, "step": 5578 }, { "epoch": 0.25924721189591077, "grad_norm": 0.8344333044158462, "learning_rate": 8.689223896602226e-06, "loss": 0.06, "step": 5579 }, { "epoch": 0.25929368029739774, "grad_norm": 1.2616791742181819, "learning_rate": 8.688715930575102e-06, "loss": 0.0809, "step": 5580 }, { "epoch": 0.2593401486988848, "grad_norm": 2.0523935621445446, "learning_rate": 8.688207880994679e-06, "loss": 0.0974, "step": 5581 }, { "epoch": 0.25938661710037175, "grad_norm": 1.3171719789890228, "learning_rate": 8.687699747872468e-06, "loss": 0.0829, "step": 5582 }, { "epoch": 0.25943308550185873, "grad_norm": 1.5654763985190516, "learning_rate": 8.687191531219973e-06, "loss": 0.1012, "step": 5583 }, { "epoch": 0.2594795539033457, "grad_norm": 1.4825034687230951, "learning_rate": 8.68668323104871e-06, "loss": 0.0943, "step": 5584 }, { "epoch": 0.25952602230483274, "grad_norm": 0.9564821407476102, "learning_rate": 8.686174847370192e-06, "loss": 0.0808, "step": 5585 }, { "epoch": 0.2595724907063197, "grad_norm": 1.4003369059847997, "learning_rate": 8.685666380195935e-06, "loss": 0.0806, "step": 5586 }, { "epoch": 0.2596189591078067, "grad_norm": 0.9460659795658336, "learning_rate": 8.685157829537455e-06, "loss": 0.063, "step": 5587 }, { "epoch": 0.25966542750929367, "grad_norm": 1.0015859416305815, "learning_rate": 8.684649195406271e-06, "loss": 0.0741, "step": 5588 }, { "epoch": 0.25971189591078064, "grad_norm": 0.8748730142622864, "learning_rate": 8.684140477813906e-06, "loss": 0.0562, "step": 5589 }, { "epoch": 0.2597583643122677, "grad_norm": 1.1768842951789895, "learning_rate": 8.68363167677188e-06, "loss": 0.1302, "step": 5590 }, { "epoch": 0.25980483271375465, "grad_norm": 1.059968679271064, "learning_rate": 8.68312279229172e-06, "loss": 0.1038, "step": 5591 }, { "epoch": 0.25985130111524163, "grad_norm": 1.3810592791075098, "learning_rate": 8.682613824384953e-06, "loss": 0.0686, "step": 5592 }, { "epoch": 0.2598977695167286, "grad_norm": 0.9077231157762361, "learning_rate": 8.682104773063107e-06, "loss": 0.051, "step": 5593 }, { "epoch": 0.25994423791821564, "grad_norm": 1.0477421480048097, "learning_rate": 8.681595638337713e-06, "loss": 0.0754, "step": 5594 }, { "epoch": 0.2599907063197026, "grad_norm": 1.759452442141568, "learning_rate": 8.681086420220301e-06, "loss": 0.0747, "step": 5595 }, { "epoch": 0.2600371747211896, "grad_norm": 2.279180147264709, "learning_rate": 8.68057711872241e-06, "loss": 0.1148, "step": 5596 }, { "epoch": 0.26008364312267657, "grad_norm": 1.3332260363277044, "learning_rate": 8.680067733855573e-06, "loss": 0.0657, "step": 5597 }, { "epoch": 0.26013011152416354, "grad_norm": 2.814918084386146, "learning_rate": 8.679558265631327e-06, "loss": 0.103, "step": 5598 }, { "epoch": 0.2601765799256506, "grad_norm": 2.0606887305552473, "learning_rate": 8.679048714061215e-06, "loss": 0.1068, "step": 5599 }, { "epoch": 0.26022304832713755, "grad_norm": 1.8424201141129752, "learning_rate": 8.678539079156776e-06, "loss": 0.0947, "step": 5600 }, { "epoch": 0.26026951672862453, "grad_norm": 1.6601367935928435, "learning_rate": 8.678029360929556e-06, "loss": 0.1261, "step": 5601 }, { "epoch": 0.2603159851301115, "grad_norm": 1.279710320724446, "learning_rate": 8.6775195593911e-06, "loss": 0.0546, "step": 5602 }, { "epoch": 0.26036245353159854, "grad_norm": 1.5511591555796682, "learning_rate": 8.677009674552956e-06, "loss": 0.0846, "step": 5603 }, { "epoch": 0.2604089219330855, "grad_norm": 1.4996213946759782, "learning_rate": 8.676499706426674e-06, "loss": 0.0799, "step": 5604 }, { "epoch": 0.2604553903345725, "grad_norm": 1.2076005433816697, "learning_rate": 8.675989655023802e-06, "loss": 0.0755, "step": 5605 }, { "epoch": 0.26050185873605947, "grad_norm": 1.9208524573896297, "learning_rate": 8.675479520355895e-06, "loss": 0.1075, "step": 5606 }, { "epoch": 0.26054832713754644, "grad_norm": 1.294846940499804, "learning_rate": 8.674969302434509e-06, "loss": 0.0645, "step": 5607 }, { "epoch": 0.2605947955390335, "grad_norm": 1.7121644165647176, "learning_rate": 8.6744590012712e-06, "loss": 0.0728, "step": 5608 }, { "epoch": 0.26064126394052045, "grad_norm": 1.091110264766153, "learning_rate": 8.67394861687753e-06, "loss": 0.0733, "step": 5609 }, { "epoch": 0.2606877323420074, "grad_norm": 2.1972563697507623, "learning_rate": 8.673438149265055e-06, "loss": 0.0732, "step": 5610 }, { "epoch": 0.2607342007434944, "grad_norm": 1.9445498990725614, "learning_rate": 8.67292759844534e-06, "loss": 0.0615, "step": 5611 }, { "epoch": 0.26078066914498144, "grad_norm": 0.9883277432475595, "learning_rate": 8.67241696442995e-06, "loss": 0.0686, "step": 5612 }, { "epoch": 0.2608271375464684, "grad_norm": 1.409231513889417, "learning_rate": 8.671906247230449e-06, "loss": 0.0615, "step": 5613 }, { "epoch": 0.2608736059479554, "grad_norm": 2.6640355965802733, "learning_rate": 8.671395446858409e-06, "loss": 0.0922, "step": 5614 }, { "epoch": 0.26092007434944237, "grad_norm": 1.8821227289446212, "learning_rate": 8.670884563325397e-06, "loss": 0.0473, "step": 5615 }, { "epoch": 0.26096654275092934, "grad_norm": 2.5470265059625534, "learning_rate": 8.670373596642985e-06, "loss": 0.0844, "step": 5616 }, { "epoch": 0.2610130111524164, "grad_norm": 1.5948786486906805, "learning_rate": 8.669862546822749e-06, "loss": 0.0864, "step": 5617 }, { "epoch": 0.26105947955390335, "grad_norm": 0.8869628827653844, "learning_rate": 8.669351413876264e-06, "loss": 0.0581, "step": 5618 }, { "epoch": 0.2611059479553903, "grad_norm": 1.4261802792996472, "learning_rate": 8.668840197815107e-06, "loss": 0.077, "step": 5619 }, { "epoch": 0.2611524163568773, "grad_norm": 0.9177986527952865, "learning_rate": 8.668328898650859e-06, "loss": 0.0719, "step": 5620 }, { "epoch": 0.26119888475836434, "grad_norm": 0.7130573039038856, "learning_rate": 8.6678175163951e-06, "loss": 0.0711, "step": 5621 }, { "epoch": 0.2612453531598513, "grad_norm": 0.9444158032403597, "learning_rate": 8.667306051059414e-06, "loss": 0.1066, "step": 5622 }, { "epoch": 0.2612918215613383, "grad_norm": 1.5260842867614997, "learning_rate": 8.666794502655387e-06, "loss": 0.0662, "step": 5623 }, { "epoch": 0.26133828996282527, "grad_norm": 1.2815040507664288, "learning_rate": 8.666282871194604e-06, "loss": 0.0642, "step": 5624 }, { "epoch": 0.26138475836431224, "grad_norm": 1.6616072705290643, "learning_rate": 8.665771156688656e-06, "loss": 0.0624, "step": 5625 }, { "epoch": 0.2614312267657993, "grad_norm": 1.7887600130669004, "learning_rate": 8.665259359149132e-06, "loss": 0.0724, "step": 5626 }, { "epoch": 0.26147769516728625, "grad_norm": 2.2385218441914883, "learning_rate": 8.664747478587627e-06, "loss": 0.1213, "step": 5627 }, { "epoch": 0.2615241635687732, "grad_norm": 2.452804996162286, "learning_rate": 8.664235515015736e-06, "loss": 0.0799, "step": 5628 }, { "epoch": 0.2615706319702602, "grad_norm": 0.9860762419373474, "learning_rate": 8.66372346844505e-06, "loss": 0.0737, "step": 5629 }, { "epoch": 0.26161710037174724, "grad_norm": 1.7217819723726016, "learning_rate": 8.663211338887174e-06, "loss": 0.0658, "step": 5630 }, { "epoch": 0.2616635687732342, "grad_norm": 0.8291654215407063, "learning_rate": 8.662699126353707e-06, "loss": 0.0625, "step": 5631 }, { "epoch": 0.2617100371747212, "grad_norm": 1.9114964481933032, "learning_rate": 8.662186830856248e-06, "loss": 0.135, "step": 5632 }, { "epoch": 0.26175650557620816, "grad_norm": 1.6513346914191536, "learning_rate": 8.661674452406404e-06, "loss": 0.1194, "step": 5633 }, { "epoch": 0.26180297397769514, "grad_norm": 2.567335547652312, "learning_rate": 8.661161991015778e-06, "loss": 0.0859, "step": 5634 }, { "epoch": 0.2618494423791822, "grad_norm": 0.9526069713299515, "learning_rate": 8.660649446695981e-06, "loss": 0.0602, "step": 5635 }, { "epoch": 0.26189591078066915, "grad_norm": 0.9536212284383668, "learning_rate": 8.660136819458622e-06, "loss": 0.0642, "step": 5636 }, { "epoch": 0.2619423791821561, "grad_norm": 1.1932820126927846, "learning_rate": 8.65962410931531e-06, "loss": 0.0783, "step": 5637 }, { "epoch": 0.2619888475836431, "grad_norm": 2.6715310265953005, "learning_rate": 8.659111316277662e-06, "loss": 0.0878, "step": 5638 }, { "epoch": 0.26203531598513014, "grad_norm": 1.4905120542748227, "learning_rate": 8.65859844035729e-06, "loss": 0.0598, "step": 5639 }, { "epoch": 0.2620817843866171, "grad_norm": 1.0137749225014563, "learning_rate": 8.658085481565813e-06, "loss": 0.0268, "step": 5640 }, { "epoch": 0.2621282527881041, "grad_norm": 1.3846602667145849, "learning_rate": 8.65757243991485e-06, "loss": 0.0596, "step": 5641 }, { "epoch": 0.26217472118959106, "grad_norm": 1.2447110508599812, "learning_rate": 8.657059315416021e-06, "loss": 0.0769, "step": 5642 }, { "epoch": 0.26222118959107804, "grad_norm": 0.5646889513492013, "learning_rate": 8.656546108080952e-06, "loss": 0.058, "step": 5643 }, { "epoch": 0.2622676579925651, "grad_norm": 3.2182256259611752, "learning_rate": 8.656032817921262e-06, "loss": 0.0416, "step": 5644 }, { "epoch": 0.26231412639405205, "grad_norm": 0.7862370426197549, "learning_rate": 8.655519444948583e-06, "loss": 0.051, "step": 5645 }, { "epoch": 0.262360594795539, "grad_norm": 0.7436609810644961, "learning_rate": 8.65500598917454e-06, "loss": 0.0327, "step": 5646 }, { "epoch": 0.262407063197026, "grad_norm": 1.511713089853115, "learning_rate": 8.654492450610765e-06, "loss": 0.0828, "step": 5647 }, { "epoch": 0.26245353159851303, "grad_norm": 1.9562578142967466, "learning_rate": 8.65397882926889e-06, "loss": 0.0841, "step": 5648 }, { "epoch": 0.2625, "grad_norm": 0.7916704335907614, "learning_rate": 8.65346512516055e-06, "loss": 0.0512, "step": 5649 }, { "epoch": 0.262546468401487, "grad_norm": 1.331437844389385, "learning_rate": 8.652951338297377e-06, "loss": 0.077, "step": 5650 }, { "epoch": 0.26259293680297396, "grad_norm": 1.0232445125662544, "learning_rate": 8.652437468691012e-06, "loss": 0.0618, "step": 5651 }, { "epoch": 0.26263940520446094, "grad_norm": 2.2315338377325764, "learning_rate": 8.651923516353096e-06, "loss": 0.11, "step": 5652 }, { "epoch": 0.262685873605948, "grad_norm": 2.6931579992589176, "learning_rate": 8.651409481295267e-06, "loss": 0.0968, "step": 5653 }, { "epoch": 0.26273234200743495, "grad_norm": 1.853649015488594, "learning_rate": 8.650895363529172e-06, "loss": 0.0829, "step": 5654 }, { "epoch": 0.2627788104089219, "grad_norm": 0.9549478554939408, "learning_rate": 8.650381163066454e-06, "loss": 0.0616, "step": 5655 }, { "epoch": 0.2628252788104089, "grad_norm": 0.6593350786358133, "learning_rate": 8.64986687991876e-06, "loss": 0.0684, "step": 5656 }, { "epoch": 0.26287174721189593, "grad_norm": 1.234041832370311, "learning_rate": 8.64935251409774e-06, "loss": 0.1032, "step": 5657 }, { "epoch": 0.2629182156133829, "grad_norm": 1.0468531148156992, "learning_rate": 8.648838065615045e-06, "loss": 0.0688, "step": 5658 }, { "epoch": 0.2629646840148699, "grad_norm": 1.0163135395754803, "learning_rate": 8.648323534482326e-06, "loss": 0.0759, "step": 5659 }, { "epoch": 0.26301115241635686, "grad_norm": 1.5554344706255032, "learning_rate": 8.64780892071124e-06, "loss": 0.0994, "step": 5660 }, { "epoch": 0.26305762081784384, "grad_norm": 0.7996773789537457, "learning_rate": 8.647294224313443e-06, "loss": 0.0453, "step": 5661 }, { "epoch": 0.26310408921933087, "grad_norm": 0.9220544275500326, "learning_rate": 8.646779445300593e-06, "loss": 0.0961, "step": 5662 }, { "epoch": 0.26315055762081785, "grad_norm": 0.9240925085371832, "learning_rate": 8.646264583684351e-06, "loss": 0.0651, "step": 5663 }, { "epoch": 0.2631970260223048, "grad_norm": 3.0041478117629072, "learning_rate": 8.645749639476378e-06, "loss": 0.08, "step": 5664 }, { "epoch": 0.2632434944237918, "grad_norm": 1.4750452357790644, "learning_rate": 8.645234612688338e-06, "loss": 0.097, "step": 5665 }, { "epoch": 0.26328996282527883, "grad_norm": 1.1765949464921923, "learning_rate": 8.6447195033319e-06, "loss": 0.0652, "step": 5666 }, { "epoch": 0.2633364312267658, "grad_norm": 0.6281023274799762, "learning_rate": 8.644204311418728e-06, "loss": 0.0589, "step": 5667 }, { "epoch": 0.2633828996282528, "grad_norm": 0.9069972334853404, "learning_rate": 8.64368903696049e-06, "loss": 0.0695, "step": 5668 }, { "epoch": 0.26342936802973976, "grad_norm": 1.3099788797062162, "learning_rate": 8.643173679968863e-06, "loss": 0.051, "step": 5669 }, { "epoch": 0.26347583643122674, "grad_norm": 1.5624535605460885, "learning_rate": 8.642658240455518e-06, "loss": 0.0845, "step": 5670 }, { "epoch": 0.26352230483271377, "grad_norm": 1.7658891114691084, "learning_rate": 8.642142718432129e-06, "loss": 0.0616, "step": 5671 }, { "epoch": 0.26356877323420075, "grad_norm": 2.3387868668107563, "learning_rate": 8.641627113910374e-06, "loss": 0.0907, "step": 5672 }, { "epoch": 0.2636152416356877, "grad_norm": 1.2957034558062803, "learning_rate": 8.641111426901933e-06, "loss": 0.0973, "step": 5673 }, { "epoch": 0.2636617100371747, "grad_norm": 0.5843383123889628, "learning_rate": 8.640595657418487e-06, "loss": 0.0549, "step": 5674 }, { "epoch": 0.26370817843866173, "grad_norm": 2.0978699675183305, "learning_rate": 8.640079805471715e-06, "loss": 0.1018, "step": 5675 }, { "epoch": 0.2637546468401487, "grad_norm": 0.7062691098889826, "learning_rate": 8.639563871073306e-06, "loss": 0.0586, "step": 5676 }, { "epoch": 0.2638011152416357, "grad_norm": 0.7125669623462905, "learning_rate": 8.639047854234945e-06, "loss": 0.0704, "step": 5677 }, { "epoch": 0.26384758364312266, "grad_norm": 2.0672081201774843, "learning_rate": 8.63853175496832e-06, "loss": 0.0958, "step": 5678 }, { "epoch": 0.26389405204460964, "grad_norm": 1.3434338081929558, "learning_rate": 8.63801557328512e-06, "loss": 0.0794, "step": 5679 }, { "epoch": 0.26394052044609667, "grad_norm": 1.445540311611333, "learning_rate": 8.63749930919704e-06, "loss": 0.0854, "step": 5680 }, { "epoch": 0.26398698884758365, "grad_norm": 0.6735828859228814, "learning_rate": 8.63698296271577e-06, "loss": 0.0592, "step": 5681 }, { "epoch": 0.2640334572490706, "grad_norm": 1.6223997440989109, "learning_rate": 8.63646653385301e-06, "loss": 0.0583, "step": 5682 }, { "epoch": 0.2640799256505576, "grad_norm": 0.9607936524675361, "learning_rate": 8.635950022620453e-06, "loss": 0.0834, "step": 5683 }, { "epoch": 0.26412639405204463, "grad_norm": 1.869315738127304, "learning_rate": 8.635433429029805e-06, "loss": 0.1003, "step": 5684 }, { "epoch": 0.2641728624535316, "grad_norm": 1.3109734996370854, "learning_rate": 8.634916753092761e-06, "loss": 0.084, "step": 5685 }, { "epoch": 0.2642193308550186, "grad_norm": 1.273151316877236, "learning_rate": 8.634399994821029e-06, "loss": 0.0875, "step": 5686 }, { "epoch": 0.26426579925650556, "grad_norm": 1.0679005317058203, "learning_rate": 8.63388315422631e-06, "loss": 0.0792, "step": 5687 }, { "epoch": 0.26431226765799254, "grad_norm": 1.1655124321789097, "learning_rate": 8.633366231320315e-06, "loss": 0.0706, "step": 5688 }, { "epoch": 0.26435873605947957, "grad_norm": 1.022067841330479, "learning_rate": 8.632849226114748e-06, "loss": 0.1123, "step": 5689 }, { "epoch": 0.26440520446096655, "grad_norm": 1.0550314669760124, "learning_rate": 8.632332138621324e-06, "loss": 0.0794, "step": 5690 }, { "epoch": 0.2644516728624535, "grad_norm": 1.2302122647212193, "learning_rate": 8.631814968851754e-06, "loss": 0.0684, "step": 5691 }, { "epoch": 0.2644981412639405, "grad_norm": 2.564756934697069, "learning_rate": 8.631297716817754e-06, "loss": 0.045, "step": 5692 }, { "epoch": 0.26454460966542753, "grad_norm": 1.1626399044189917, "learning_rate": 8.630780382531037e-06, "loss": 0.0651, "step": 5693 }, { "epoch": 0.2645910780669145, "grad_norm": 1.0214044413907606, "learning_rate": 8.630262966003324e-06, "loss": 0.0725, "step": 5694 }, { "epoch": 0.2646375464684015, "grad_norm": 0.90304896676457, "learning_rate": 8.629745467246333e-06, "loss": 0.0589, "step": 5695 }, { "epoch": 0.26468401486988846, "grad_norm": 1.3975414460468794, "learning_rate": 8.629227886271787e-06, "loss": 0.0862, "step": 5696 }, { "epoch": 0.26473048327137544, "grad_norm": 1.0064710751387755, "learning_rate": 8.62871022309141e-06, "loss": 0.0743, "step": 5697 }, { "epoch": 0.26477695167286247, "grad_norm": 0.8866358476229612, "learning_rate": 8.628192477716927e-06, "loss": 0.0577, "step": 5698 }, { "epoch": 0.26482342007434945, "grad_norm": 0.6566509521609569, "learning_rate": 8.627674650160067e-06, "loss": 0.0417, "step": 5699 }, { "epoch": 0.2648698884758364, "grad_norm": 1.182281177176067, "learning_rate": 8.627156740432559e-06, "loss": 0.0954, "step": 5700 }, { "epoch": 0.2649163568773234, "grad_norm": 0.720838915930597, "learning_rate": 8.62663874854613e-06, "loss": 0.0637, "step": 5701 }, { "epoch": 0.26496282527881043, "grad_norm": 3.772460332610454, "learning_rate": 8.626120674512519e-06, "loss": 0.0982, "step": 5702 }, { "epoch": 0.2650092936802974, "grad_norm": 1.5619312953682198, "learning_rate": 8.625602518343457e-06, "loss": 0.1057, "step": 5703 }, { "epoch": 0.2650557620817844, "grad_norm": 1.192575662950234, "learning_rate": 8.625084280050683e-06, "loss": 0.0924, "step": 5704 }, { "epoch": 0.26510223048327136, "grad_norm": 0.6705670520842403, "learning_rate": 8.624565959645935e-06, "loss": 0.0584, "step": 5705 }, { "epoch": 0.26514869888475834, "grad_norm": 2.489159539898225, "learning_rate": 8.624047557140952e-06, "loss": 0.0508, "step": 5706 }, { "epoch": 0.26519516728624537, "grad_norm": 1.7020181723720211, "learning_rate": 8.623529072547476e-06, "loss": 0.1122, "step": 5707 }, { "epoch": 0.26524163568773235, "grad_norm": 1.774007700830784, "learning_rate": 8.623010505877256e-06, "loss": 0.0705, "step": 5708 }, { "epoch": 0.2652881040892193, "grad_norm": 0.9349441105028844, "learning_rate": 8.622491857142034e-06, "loss": 0.0922, "step": 5709 }, { "epoch": 0.2653345724907063, "grad_norm": 0.42330719844767634, "learning_rate": 8.621973126353557e-06, "loss": 0.0465, "step": 5710 }, { "epoch": 0.26538104089219333, "grad_norm": 1.2699334371780981, "learning_rate": 8.62145431352358e-06, "loss": 0.1413, "step": 5711 }, { "epoch": 0.2654275092936803, "grad_norm": 0.8091115068108309, "learning_rate": 8.620935418663849e-06, "loss": 0.0401, "step": 5712 }, { "epoch": 0.2654739776951673, "grad_norm": 0.7555867360207366, "learning_rate": 8.62041644178612e-06, "loss": 0.0755, "step": 5713 }, { "epoch": 0.26552044609665426, "grad_norm": 1.3446367860523503, "learning_rate": 8.619897382902149e-06, "loss": 0.1024, "step": 5714 }, { "epoch": 0.26556691449814124, "grad_norm": 1.1283114898060593, "learning_rate": 8.619378242023692e-06, "loss": 0.1154, "step": 5715 }, { "epoch": 0.26561338289962827, "grad_norm": 1.126281916517488, "learning_rate": 8.618859019162509e-06, "loss": 0.092, "step": 5716 }, { "epoch": 0.26565985130111525, "grad_norm": 0.5804939538869495, "learning_rate": 8.618339714330359e-06, "loss": 0.0454, "step": 5717 }, { "epoch": 0.2657063197026022, "grad_norm": 1.5120759073753667, "learning_rate": 8.617820327539006e-06, "loss": 0.0888, "step": 5718 }, { "epoch": 0.2657527881040892, "grad_norm": 0.7803443493061177, "learning_rate": 8.617300858800217e-06, "loss": 0.0636, "step": 5719 }, { "epoch": 0.26579925650557623, "grad_norm": 1.16660155957991, "learning_rate": 8.616781308125755e-06, "loss": 0.061, "step": 5720 }, { "epoch": 0.2658457249070632, "grad_norm": 1.452321039046305, "learning_rate": 8.616261675527389e-06, "loss": 0.0555, "step": 5721 }, { "epoch": 0.2658921933085502, "grad_norm": 1.5056060200192551, "learning_rate": 8.615741961016892e-06, "loss": 0.1059, "step": 5722 }, { "epoch": 0.26593866171003716, "grad_norm": 0.5476169065274774, "learning_rate": 8.615222164606032e-06, "loss": 0.056, "step": 5723 }, { "epoch": 0.26598513011152414, "grad_norm": 0.9098666586975727, "learning_rate": 8.614702286306587e-06, "loss": 0.1015, "step": 5724 }, { "epoch": 0.26603159851301117, "grad_norm": 1.3386456861911098, "learning_rate": 8.61418232613033e-06, "loss": 0.0673, "step": 5725 }, { "epoch": 0.26607806691449815, "grad_norm": 0.5917358445222249, "learning_rate": 8.61366228408904e-06, "loss": 0.0519, "step": 5726 }, { "epoch": 0.2661245353159851, "grad_norm": 1.205256155941515, "learning_rate": 8.613142160194495e-06, "loss": 0.057, "step": 5727 }, { "epoch": 0.2661710037174721, "grad_norm": 0.9000571667203465, "learning_rate": 8.612621954458477e-06, "loss": 0.054, "step": 5728 }, { "epoch": 0.26621747211895913, "grad_norm": 0.8566115893726675, "learning_rate": 8.612101666892771e-06, "loss": 0.061, "step": 5729 }, { "epoch": 0.2662639405204461, "grad_norm": 0.7970369318991025, "learning_rate": 8.61158129750916e-06, "loss": 0.05, "step": 5730 }, { "epoch": 0.2663104089219331, "grad_norm": 2.3757822887284084, "learning_rate": 8.61106084631943e-06, "loss": 0.0966, "step": 5731 }, { "epoch": 0.26635687732342006, "grad_norm": 1.0163727571245484, "learning_rate": 8.610540313335374e-06, "loss": 0.0808, "step": 5732 }, { "epoch": 0.26640334572490704, "grad_norm": 1.1061171924620046, "learning_rate": 8.610019698568779e-06, "loss": 0.073, "step": 5733 }, { "epoch": 0.26644981412639407, "grad_norm": 1.2155200660738439, "learning_rate": 8.609499002031438e-06, "loss": 0.0722, "step": 5734 }, { "epoch": 0.26649628252788105, "grad_norm": 1.1373352235791745, "learning_rate": 8.608978223735145e-06, "loss": 0.0382, "step": 5735 }, { "epoch": 0.266542750929368, "grad_norm": 0.71833816792794, "learning_rate": 8.608457363691697e-06, "loss": 0.0697, "step": 5736 }, { "epoch": 0.266589219330855, "grad_norm": 2.263901490782359, "learning_rate": 8.607936421912892e-06, "loss": 0.072, "step": 5737 }, { "epoch": 0.26663568773234203, "grad_norm": 0.8553205779239473, "learning_rate": 8.60741539841053e-06, "loss": 0.062, "step": 5738 }, { "epoch": 0.266682156133829, "grad_norm": 0.868522062750557, "learning_rate": 8.606894293196413e-06, "loss": 0.0749, "step": 5739 }, { "epoch": 0.266728624535316, "grad_norm": 1.4297891119940156, "learning_rate": 8.606373106282342e-06, "loss": 0.1121, "step": 5740 }, { "epoch": 0.26677509293680296, "grad_norm": 1.1733546156990293, "learning_rate": 8.605851837680127e-06, "loss": 0.123, "step": 5741 }, { "epoch": 0.26682156133828994, "grad_norm": 1.936947697080058, "learning_rate": 8.605330487401571e-06, "loss": 0.0774, "step": 5742 }, { "epoch": 0.26686802973977697, "grad_norm": 1.457350825887219, "learning_rate": 8.604809055458487e-06, "loss": 0.0971, "step": 5743 }, { "epoch": 0.26691449814126395, "grad_norm": 3.5502081827732397, "learning_rate": 8.604287541862681e-06, "loss": 0.1353, "step": 5744 }, { "epoch": 0.2669609665427509, "grad_norm": 1.6965872009813023, "learning_rate": 8.603765946625971e-06, "loss": 0.0728, "step": 5745 }, { "epoch": 0.2670074349442379, "grad_norm": 3.985142291302167, "learning_rate": 8.60324426976017e-06, "loss": 0.0856, "step": 5746 }, { "epoch": 0.26705390334572493, "grad_norm": 1.036212207839871, "learning_rate": 8.602722511277091e-06, "loss": 0.0512, "step": 5747 }, { "epoch": 0.2671003717472119, "grad_norm": 1.4989551611598988, "learning_rate": 8.602200671188557e-06, "loss": 0.1001, "step": 5748 }, { "epoch": 0.2671468401486989, "grad_norm": 0.7798519141120118, "learning_rate": 8.601678749506388e-06, "loss": 0.0679, "step": 5749 }, { "epoch": 0.26719330855018586, "grad_norm": 1.269047166234263, "learning_rate": 8.601156746242403e-06, "loss": 0.0887, "step": 5750 }, { "epoch": 0.26723977695167284, "grad_norm": 1.4852180686518301, "learning_rate": 8.600634661408426e-06, "loss": 0.1015, "step": 5751 }, { "epoch": 0.26728624535315987, "grad_norm": 1.1436391659518323, "learning_rate": 8.600112495016289e-06, "loss": 0.0712, "step": 5752 }, { "epoch": 0.26733271375464684, "grad_norm": 1.9185972520911152, "learning_rate": 8.59959024707781e-06, "loss": 0.1651, "step": 5753 }, { "epoch": 0.2673791821561338, "grad_norm": 1.9497916213617623, "learning_rate": 8.599067917604827e-06, "loss": 0.1334, "step": 5754 }, { "epoch": 0.2674256505576208, "grad_norm": 2.1242956321861652, "learning_rate": 8.598545506609165e-06, "loss": 0.1014, "step": 5755 }, { "epoch": 0.26747211895910783, "grad_norm": 0.7741728085571689, "learning_rate": 8.59802301410266e-06, "loss": 0.059, "step": 5756 }, { "epoch": 0.2675185873605948, "grad_norm": 0.703596579752211, "learning_rate": 8.597500440097149e-06, "loss": 0.0656, "step": 5757 }, { "epoch": 0.2675650557620818, "grad_norm": 2.789673329906087, "learning_rate": 8.596977784604465e-06, "loss": 0.0935, "step": 5758 }, { "epoch": 0.26761152416356876, "grad_norm": 0.7369484041977631, "learning_rate": 8.59645504763645e-06, "loss": 0.0519, "step": 5759 }, { "epoch": 0.26765799256505574, "grad_norm": 2.4211745050633824, "learning_rate": 8.595932229204941e-06, "loss": 0.1218, "step": 5760 }, { "epoch": 0.26770446096654277, "grad_norm": 1.7169716411416789, "learning_rate": 8.595409329321784e-06, "loss": 0.0862, "step": 5761 }, { "epoch": 0.26775092936802974, "grad_norm": 0.8427715610278386, "learning_rate": 8.59488634799882e-06, "loss": 0.0835, "step": 5762 }, { "epoch": 0.2677973977695167, "grad_norm": 1.0300333734447285, "learning_rate": 8.594363285247899e-06, "loss": 0.0576, "step": 5763 }, { "epoch": 0.2678438661710037, "grad_norm": 1.0812918118254942, "learning_rate": 8.593840141080864e-06, "loss": 0.1254, "step": 5764 }, { "epoch": 0.26789033457249073, "grad_norm": 1.120160786265383, "learning_rate": 8.593316915509569e-06, "loss": 0.0801, "step": 5765 }, { "epoch": 0.2679368029739777, "grad_norm": 1.360392602161295, "learning_rate": 8.592793608545863e-06, "loss": 0.0806, "step": 5766 }, { "epoch": 0.2679832713754647, "grad_norm": 1.403203675808037, "learning_rate": 8.592270220201602e-06, "loss": 0.0722, "step": 5767 }, { "epoch": 0.26802973977695166, "grad_norm": 2.4381876800145084, "learning_rate": 8.591746750488639e-06, "loss": 0.0726, "step": 5768 }, { "epoch": 0.26807620817843864, "grad_norm": 2.0559273838109062, "learning_rate": 8.591223199418832e-06, "loss": 0.0752, "step": 5769 }, { "epoch": 0.26812267657992567, "grad_norm": 1.992476092524669, "learning_rate": 8.590699567004039e-06, "loss": 0.098, "step": 5770 }, { "epoch": 0.26816914498141264, "grad_norm": 1.3921836656196231, "learning_rate": 8.590175853256121e-06, "loss": 0.0595, "step": 5771 }, { "epoch": 0.2682156133828996, "grad_norm": 4.130521331044076, "learning_rate": 8.589652058186944e-06, "loss": 0.0907, "step": 5772 }, { "epoch": 0.2682620817843866, "grad_norm": 2.019918751703223, "learning_rate": 8.589128181808367e-06, "loss": 0.0844, "step": 5773 }, { "epoch": 0.26830855018587363, "grad_norm": 1.092593129901473, "learning_rate": 8.58860422413226e-06, "loss": 0.0669, "step": 5774 }, { "epoch": 0.2683550185873606, "grad_norm": 1.37204127319846, "learning_rate": 8.588080185170491e-06, "loss": 0.0729, "step": 5775 }, { "epoch": 0.2684014869888476, "grad_norm": 1.876497852271771, "learning_rate": 8.58755606493493e-06, "loss": 0.0902, "step": 5776 }, { "epoch": 0.26844795539033456, "grad_norm": 1.0100084332867367, "learning_rate": 8.587031863437447e-06, "loss": 0.056, "step": 5777 }, { "epoch": 0.26849442379182153, "grad_norm": 1.5768629513371204, "learning_rate": 8.586507580689917e-06, "loss": 0.0725, "step": 5778 }, { "epoch": 0.26854089219330857, "grad_norm": 3.589066334022472, "learning_rate": 8.585983216704217e-06, "loss": 0.0991, "step": 5779 }, { "epoch": 0.26858736059479554, "grad_norm": 1.9196808920605728, "learning_rate": 8.585458771492223e-06, "loss": 0.0641, "step": 5780 }, { "epoch": 0.2686338289962825, "grad_norm": 2.5286073440021406, "learning_rate": 8.584934245065813e-06, "loss": 0.1059, "step": 5781 }, { "epoch": 0.2686802973977695, "grad_norm": 1.01947276853546, "learning_rate": 8.584409637436869e-06, "loss": 0.0613, "step": 5782 }, { "epoch": 0.26872676579925653, "grad_norm": 1.0112192355296594, "learning_rate": 8.583884948617274e-06, "loss": 0.0559, "step": 5783 }, { "epoch": 0.2687732342007435, "grad_norm": 2.09448273283987, "learning_rate": 8.583360178618913e-06, "loss": 0.1302, "step": 5784 }, { "epoch": 0.2688197026022305, "grad_norm": 19.00412974560039, "learning_rate": 8.582835327453674e-06, "loss": 0.1004, "step": 5785 }, { "epoch": 0.26886617100371746, "grad_norm": 1.0636489334047043, "learning_rate": 8.582310395133445e-06, "loss": 0.0664, "step": 5786 }, { "epoch": 0.26891263940520443, "grad_norm": 3.7807430991401945, "learning_rate": 8.58178538167011e-06, "loss": 0.0863, "step": 5787 }, { "epoch": 0.26895910780669147, "grad_norm": 1.2741514162461829, "learning_rate": 8.58126028707557e-06, "loss": 0.0679, "step": 5788 }, { "epoch": 0.26900557620817844, "grad_norm": 1.736725086212963, "learning_rate": 8.580735111361717e-06, "loss": 0.0657, "step": 5789 }, { "epoch": 0.2690520446096654, "grad_norm": 5.061444847248129, "learning_rate": 8.580209854540442e-06, "loss": 0.1908, "step": 5790 }, { "epoch": 0.2690985130111524, "grad_norm": 1.7267852856602768, "learning_rate": 8.579684516623648e-06, "loss": 0.0922, "step": 5791 }, { "epoch": 0.26914498141263943, "grad_norm": 3.329650867396829, "learning_rate": 8.579159097623231e-06, "loss": 0.1023, "step": 5792 }, { "epoch": 0.2691914498141264, "grad_norm": 1.7005694238491456, "learning_rate": 8.578633597551093e-06, "loss": 0.0858, "step": 5793 }, { "epoch": 0.2692379182156134, "grad_norm": 2.452290582832128, "learning_rate": 8.578108016419138e-06, "loss": 0.0793, "step": 5794 }, { "epoch": 0.26928438661710036, "grad_norm": 0.9086468017844753, "learning_rate": 8.577582354239272e-06, "loss": 0.1122, "step": 5795 }, { "epoch": 0.26933085501858733, "grad_norm": 2.3325940289200973, "learning_rate": 8.577056611023397e-06, "loss": 0.0959, "step": 5796 }, { "epoch": 0.26937732342007437, "grad_norm": 1.8780441612584287, "learning_rate": 8.576530786783426e-06, "loss": 0.0747, "step": 5797 }, { "epoch": 0.26942379182156134, "grad_norm": 26.27296523802553, "learning_rate": 8.576004881531271e-06, "loss": 0.4507, "step": 5798 }, { "epoch": 0.2694702602230483, "grad_norm": 1.2954522331977605, "learning_rate": 8.575478895278839e-06, "loss": 0.0799, "step": 5799 }, { "epoch": 0.2695167286245353, "grad_norm": 1.6624949155563042, "learning_rate": 8.574952828038047e-06, "loss": 0.1053, "step": 5800 }, { "epoch": 0.2695631970260223, "grad_norm": 2.1094027637849546, "learning_rate": 8.574426679820813e-06, "loss": 0.0978, "step": 5801 }, { "epoch": 0.2696096654275093, "grad_norm": 2.55266758092862, "learning_rate": 8.573900450639052e-06, "loss": 0.1292, "step": 5802 }, { "epoch": 0.2696561338289963, "grad_norm": 2.7461074025694185, "learning_rate": 8.573374140504684e-06, "loss": 0.1679, "step": 5803 }, { "epoch": 0.26970260223048326, "grad_norm": 4.514838002036577, "learning_rate": 8.572847749429631e-06, "loss": 0.187, "step": 5804 }, { "epoch": 0.26974907063197023, "grad_norm": 1.7465021835384409, "learning_rate": 8.572321277425815e-06, "loss": 0.0978, "step": 5805 }, { "epoch": 0.26979553903345727, "grad_norm": 2.9702084876377883, "learning_rate": 8.571794724505163e-06, "loss": 0.1964, "step": 5806 }, { "epoch": 0.26984200743494424, "grad_norm": 1.5391780414980527, "learning_rate": 8.571268090679601e-06, "loss": 0.1005, "step": 5807 }, { "epoch": 0.2698884758364312, "grad_norm": 1.2369145366348269, "learning_rate": 8.570741375961059e-06, "loss": 0.1072, "step": 5808 }, { "epoch": 0.2699349442379182, "grad_norm": 1.8918779514642867, "learning_rate": 8.570214580361466e-06, "loss": 0.0682, "step": 5809 }, { "epoch": 0.2699814126394052, "grad_norm": 0.817685137619319, "learning_rate": 8.569687703892754e-06, "loss": 0.0529, "step": 5810 }, { "epoch": 0.2700278810408922, "grad_norm": 1.3159390442496866, "learning_rate": 8.56916074656686e-06, "loss": 0.1013, "step": 5811 }, { "epoch": 0.2700743494423792, "grad_norm": 4.0124113376471975, "learning_rate": 8.568633708395719e-06, "loss": 0.0746, "step": 5812 }, { "epoch": 0.27012081784386616, "grad_norm": 1.614314933679916, "learning_rate": 8.568106589391268e-06, "loss": 0.0888, "step": 5813 }, { "epoch": 0.27016728624535313, "grad_norm": 1.3927947182400422, "learning_rate": 8.567579389565447e-06, "loss": 0.0865, "step": 5814 }, { "epoch": 0.27021375464684017, "grad_norm": 2.00984055705105, "learning_rate": 8.567052108930199e-06, "loss": 0.1225, "step": 5815 }, { "epoch": 0.27026022304832714, "grad_norm": 1.3895254740561118, "learning_rate": 8.566524747497464e-06, "loss": 0.0564, "step": 5816 }, { "epoch": 0.2703066914498141, "grad_norm": 1.8677927745667275, "learning_rate": 8.565997305279193e-06, "loss": 0.1204, "step": 5817 }, { "epoch": 0.2703531598513011, "grad_norm": 3.0679743846883336, "learning_rate": 8.565469782287327e-06, "loss": 0.1679, "step": 5818 }, { "epoch": 0.2703996282527881, "grad_norm": 2.0367558131646057, "learning_rate": 8.56494217853382e-06, "loss": 0.1146, "step": 5819 }, { "epoch": 0.2704460966542751, "grad_norm": 0.9725575998255802, "learning_rate": 8.56441449403062e-06, "loss": 0.0716, "step": 5820 }, { "epoch": 0.2704925650557621, "grad_norm": 2.137673199234072, "learning_rate": 8.56388672878968e-06, "loss": 0.0622, "step": 5821 }, { "epoch": 0.27053903345724906, "grad_norm": 1.2038810548056555, "learning_rate": 8.563358882822954e-06, "loss": 0.0768, "step": 5822 }, { "epoch": 0.27058550185873603, "grad_norm": 3.140239556181532, "learning_rate": 8.562830956142401e-06, "loss": 0.1013, "step": 5823 }, { "epoch": 0.27063197026022306, "grad_norm": 0.9877040455594027, "learning_rate": 8.562302948759974e-06, "loss": 0.0677, "step": 5824 }, { "epoch": 0.27067843866171004, "grad_norm": 1.5070646193900554, "learning_rate": 8.561774860687639e-06, "loss": 0.0645, "step": 5825 }, { "epoch": 0.270724907063197, "grad_norm": 1.738150336385428, "learning_rate": 8.561246691937353e-06, "loss": 0.0806, "step": 5826 }, { "epoch": 0.270771375464684, "grad_norm": 2.7121775251584865, "learning_rate": 8.560718442521083e-06, "loss": 0.1188, "step": 5827 }, { "epoch": 0.270817843866171, "grad_norm": 2.565181819804814, "learning_rate": 8.56019011245079e-06, "loss": 0.1376, "step": 5828 }, { "epoch": 0.270864312267658, "grad_norm": 3.647997447706652, "learning_rate": 8.559661701738445e-06, "loss": 0.1195, "step": 5829 }, { "epoch": 0.270910780669145, "grad_norm": 1.497339710241409, "learning_rate": 8.559133210396017e-06, "loss": 0.038, "step": 5830 }, { "epoch": 0.27095724907063196, "grad_norm": 1.1564789396336235, "learning_rate": 8.558604638435476e-06, "loss": 0.0969, "step": 5831 }, { "epoch": 0.27100371747211893, "grad_norm": 2.2168120455904106, "learning_rate": 8.558075985868793e-06, "loss": 0.0867, "step": 5832 }, { "epoch": 0.27105018587360596, "grad_norm": 1.6382697704243017, "learning_rate": 8.557547252707946e-06, "loss": 0.0772, "step": 5833 }, { "epoch": 0.27109665427509294, "grad_norm": 1.5979645243093936, "learning_rate": 8.557018438964907e-06, "loss": 0.0653, "step": 5834 }, { "epoch": 0.2711431226765799, "grad_norm": 1.2481856637837658, "learning_rate": 8.556489544651658e-06, "loss": 0.0752, "step": 5835 }, { "epoch": 0.2711895910780669, "grad_norm": 2.869546597986672, "learning_rate": 8.555960569780178e-06, "loss": 0.0891, "step": 5836 }, { "epoch": 0.2712360594795539, "grad_norm": 4.04377808915112, "learning_rate": 8.555431514362447e-06, "loss": 0.1263, "step": 5837 }, { "epoch": 0.2712825278810409, "grad_norm": 1.427526243222184, "learning_rate": 8.554902378410452e-06, "loss": 0.0669, "step": 5838 }, { "epoch": 0.2713289962825279, "grad_norm": 1.1558948715345245, "learning_rate": 8.554373161936176e-06, "loss": 0.0615, "step": 5839 }, { "epoch": 0.27137546468401486, "grad_norm": 1.7974106279107986, "learning_rate": 8.553843864951608e-06, "loss": 0.1016, "step": 5840 }, { "epoch": 0.27142193308550183, "grad_norm": 2.3180596291929745, "learning_rate": 8.553314487468733e-06, "loss": 0.0676, "step": 5841 }, { "epoch": 0.27146840148698886, "grad_norm": 3.380136010929301, "learning_rate": 8.552785029499548e-06, "loss": 0.1127, "step": 5842 }, { "epoch": 0.27151486988847584, "grad_norm": 1.5952202133272704, "learning_rate": 8.552255491056043e-06, "loss": 0.1168, "step": 5843 }, { "epoch": 0.2715613382899628, "grad_norm": 1.8230904550812916, "learning_rate": 8.55172587215021e-06, "loss": 0.0979, "step": 5844 }, { "epoch": 0.2716078066914498, "grad_norm": 1.7719962465258585, "learning_rate": 8.55119617279405e-06, "loss": 0.0793, "step": 5845 }, { "epoch": 0.2716542750929368, "grad_norm": 2.2936948715422028, "learning_rate": 8.550666392999558e-06, "loss": 0.0776, "step": 5846 }, { "epoch": 0.2717007434944238, "grad_norm": 0.7988852751339487, "learning_rate": 8.550136532778736e-06, "loss": 0.0471, "step": 5847 }, { "epoch": 0.2717472118959108, "grad_norm": 1.8820576143613426, "learning_rate": 8.549606592143584e-06, "loss": 0.0893, "step": 5848 }, { "epoch": 0.27179368029739776, "grad_norm": 1.5348902644984361, "learning_rate": 8.549076571106107e-06, "loss": 0.0839, "step": 5849 }, { "epoch": 0.27184014869888473, "grad_norm": 2.9253357619458806, "learning_rate": 8.548546469678311e-06, "loss": 0.0831, "step": 5850 }, { "epoch": 0.27188661710037176, "grad_norm": 0.7030212598033081, "learning_rate": 8.548016287872203e-06, "loss": 0.0506, "step": 5851 }, { "epoch": 0.27193308550185874, "grad_norm": 6.9879288236779695, "learning_rate": 8.547486025699792e-06, "loss": 0.1589, "step": 5852 }, { "epoch": 0.2719795539033457, "grad_norm": 0.8051824795161773, "learning_rate": 8.546955683173089e-06, "loss": 0.1062, "step": 5853 }, { "epoch": 0.2720260223048327, "grad_norm": 9.339872335345634, "learning_rate": 8.546425260304106e-06, "loss": 0.1088, "step": 5854 }, { "epoch": 0.2720724907063197, "grad_norm": 3.0485456153106725, "learning_rate": 8.545894757104858e-06, "loss": 0.1663, "step": 5855 }, { "epoch": 0.2721189591078067, "grad_norm": 3.3684155038130768, "learning_rate": 8.54536417358736e-06, "loss": 0.0646, "step": 5856 }, { "epoch": 0.2721654275092937, "grad_norm": 1.783715058075964, "learning_rate": 8.544833509763635e-06, "loss": 0.0836, "step": 5857 }, { "epoch": 0.27221189591078065, "grad_norm": 0.9613195916810303, "learning_rate": 8.5443027656457e-06, "loss": 0.0668, "step": 5858 }, { "epoch": 0.27225836431226763, "grad_norm": 1.4828793044595228, "learning_rate": 8.543771941245576e-06, "loss": 0.0658, "step": 5859 }, { "epoch": 0.27230483271375466, "grad_norm": 2.2915977758211405, "learning_rate": 8.543241036575288e-06, "loss": 0.1141, "step": 5860 }, { "epoch": 0.27235130111524164, "grad_norm": 2.091052665948503, "learning_rate": 8.542710051646861e-06, "loss": 0.0956, "step": 5861 }, { "epoch": 0.2723977695167286, "grad_norm": 2.1154493843111775, "learning_rate": 8.542178986472323e-06, "loss": 0.0823, "step": 5862 }, { "epoch": 0.2724442379182156, "grad_norm": 3.970750647549495, "learning_rate": 8.541647841063703e-06, "loss": 0.0883, "step": 5863 }, { "epoch": 0.2724907063197026, "grad_norm": 1.8558230601457935, "learning_rate": 8.541116615433032e-06, "loss": 0.104, "step": 5864 }, { "epoch": 0.2725371747211896, "grad_norm": 2.557385488367736, "learning_rate": 8.540585309592341e-06, "loss": 0.1346, "step": 5865 }, { "epoch": 0.2725836431226766, "grad_norm": 1.9419327554244028, "learning_rate": 8.540053923553668e-06, "loss": 0.0832, "step": 5866 }, { "epoch": 0.27263011152416355, "grad_norm": 2.7014292332795793, "learning_rate": 8.539522457329048e-06, "loss": 0.1195, "step": 5867 }, { "epoch": 0.27267657992565053, "grad_norm": 0.9454311638018893, "learning_rate": 8.53899091093052e-06, "loss": 0.0841, "step": 5868 }, { "epoch": 0.27272304832713756, "grad_norm": 3.0302303431267297, "learning_rate": 8.53845928437012e-06, "loss": 0.1056, "step": 5869 }, { "epoch": 0.27276951672862454, "grad_norm": 1.6068575027869612, "learning_rate": 8.537927577659894e-06, "loss": 0.066, "step": 5870 }, { "epoch": 0.2728159851301115, "grad_norm": 3.7655474705573226, "learning_rate": 8.537395790811886e-06, "loss": 0.163, "step": 5871 }, { "epoch": 0.2728624535315985, "grad_norm": 2.206314523248017, "learning_rate": 8.536863923838141e-06, "loss": 0.1119, "step": 5872 }, { "epoch": 0.2729089219330855, "grad_norm": 0.8590534987111944, "learning_rate": 8.536331976750705e-06, "loss": 0.0766, "step": 5873 }, { "epoch": 0.2729553903345725, "grad_norm": 3.2105285941349533, "learning_rate": 8.535799949561628e-06, "loss": 0.0846, "step": 5874 }, { "epoch": 0.2730018587360595, "grad_norm": 1.4615603388382317, "learning_rate": 8.53526784228296e-06, "loss": 0.0792, "step": 5875 }, { "epoch": 0.27304832713754645, "grad_norm": 1.4203614045201773, "learning_rate": 8.534735654926753e-06, "loss": 0.098, "step": 5876 }, { "epoch": 0.27309479553903343, "grad_norm": 1.8811516421606136, "learning_rate": 8.534203387505066e-06, "loss": 0.0521, "step": 5877 }, { "epoch": 0.27314126394052046, "grad_norm": 1.5418419464514392, "learning_rate": 8.533671040029952e-06, "loss": 0.0985, "step": 5878 }, { "epoch": 0.27318773234200744, "grad_norm": 1.6615377690647308, "learning_rate": 8.533138612513468e-06, "loss": 0.0775, "step": 5879 }, { "epoch": 0.2732342007434944, "grad_norm": 0.899228058448961, "learning_rate": 8.532606104967677e-06, "loss": 0.0443, "step": 5880 }, { "epoch": 0.2732806691449814, "grad_norm": 0.9823035584194219, "learning_rate": 8.53207351740464e-06, "loss": 0.0807, "step": 5881 }, { "epoch": 0.2733271375464684, "grad_norm": 1.1203777597497608, "learning_rate": 8.531540849836422e-06, "loss": 0.0759, "step": 5882 }, { "epoch": 0.2733736059479554, "grad_norm": 1.4974875736748678, "learning_rate": 8.531008102275085e-06, "loss": 0.0848, "step": 5883 }, { "epoch": 0.2734200743494424, "grad_norm": 2.0016369631764204, "learning_rate": 8.530475274732698e-06, "loss": 0.0548, "step": 5884 }, { "epoch": 0.27346654275092935, "grad_norm": 1.3765791440490132, "learning_rate": 8.52994236722133e-06, "loss": 0.0842, "step": 5885 }, { "epoch": 0.27351301115241633, "grad_norm": 1.6035273530584309, "learning_rate": 8.529409379753052e-06, "loss": 0.0793, "step": 5886 }, { "epoch": 0.27355947955390336, "grad_norm": 2.6732273528921238, "learning_rate": 8.528876312339938e-06, "loss": 0.1173, "step": 5887 }, { "epoch": 0.27360594795539034, "grad_norm": 1.5474662458186927, "learning_rate": 8.528343164994061e-06, "loss": 0.0986, "step": 5888 }, { "epoch": 0.2736524163568773, "grad_norm": 1.348656833633751, "learning_rate": 8.527809937727499e-06, "loss": 0.0736, "step": 5889 }, { "epoch": 0.2736988847583643, "grad_norm": 1.9163672994479781, "learning_rate": 8.527276630552327e-06, "loss": 0.0821, "step": 5890 }, { "epoch": 0.2737453531598513, "grad_norm": 1.385297872986423, "learning_rate": 8.526743243480629e-06, "loss": 0.0769, "step": 5891 }, { "epoch": 0.2737918215613383, "grad_norm": 1.2126125322197079, "learning_rate": 8.526209776524482e-06, "loss": 0.0508, "step": 5892 }, { "epoch": 0.2738382899628253, "grad_norm": 1.4673440659280919, "learning_rate": 8.525676229695974e-06, "loss": 0.0998, "step": 5893 }, { "epoch": 0.27388475836431225, "grad_norm": 2.8551088281102657, "learning_rate": 8.525142603007187e-06, "loss": 0.1793, "step": 5894 }, { "epoch": 0.27393122676579923, "grad_norm": 1.2445310658425435, "learning_rate": 8.524608896470213e-06, "loss": 0.0971, "step": 5895 }, { "epoch": 0.27397769516728626, "grad_norm": 1.901263083754215, "learning_rate": 8.524075110097136e-06, "loss": 0.1173, "step": 5896 }, { "epoch": 0.27402416356877324, "grad_norm": 0.7410120517419246, "learning_rate": 8.523541243900048e-06, "loss": 0.0536, "step": 5897 }, { "epoch": 0.2740706319702602, "grad_norm": 1.4593160365175746, "learning_rate": 8.523007297891044e-06, "loss": 0.103, "step": 5898 }, { "epoch": 0.2741171003717472, "grad_norm": 2.5589925421218362, "learning_rate": 8.522473272082217e-06, "loss": 0.1196, "step": 5899 }, { "epoch": 0.2741635687732342, "grad_norm": 3.864477112277123, "learning_rate": 8.521939166485663e-06, "loss": 0.0832, "step": 5900 }, { "epoch": 0.2742100371747212, "grad_norm": 1.2696708364985556, "learning_rate": 8.521404981113479e-06, "loss": 0.0663, "step": 5901 }, { "epoch": 0.2742565055762082, "grad_norm": 1.605614375065866, "learning_rate": 8.520870715977766e-06, "loss": 0.0696, "step": 5902 }, { "epoch": 0.27430297397769515, "grad_norm": 1.0752488789251238, "learning_rate": 8.520336371090627e-06, "loss": 0.0767, "step": 5903 }, { "epoch": 0.27434944237918213, "grad_norm": 2.462051952996488, "learning_rate": 8.519801946464162e-06, "loss": 0.0857, "step": 5904 }, { "epoch": 0.27439591078066916, "grad_norm": 1.6425939621071248, "learning_rate": 8.51926744211048e-06, "loss": 0.0697, "step": 5905 }, { "epoch": 0.27444237918215614, "grad_norm": 1.9018856694864579, "learning_rate": 8.518732858041685e-06, "loss": 0.069, "step": 5906 }, { "epoch": 0.2744888475836431, "grad_norm": 0.9049949462545649, "learning_rate": 8.518198194269888e-06, "loss": 0.0554, "step": 5907 }, { "epoch": 0.2745353159851301, "grad_norm": 2.0046412560514364, "learning_rate": 8.5176634508072e-06, "loss": 0.1018, "step": 5908 }, { "epoch": 0.2745817843866171, "grad_norm": 1.2308809304985673, "learning_rate": 8.517128627665732e-06, "loss": 0.0618, "step": 5909 }, { "epoch": 0.2746282527881041, "grad_norm": 1.8932299958759409, "learning_rate": 8.516593724857598e-06, "loss": 0.0658, "step": 5910 }, { "epoch": 0.2746747211895911, "grad_norm": 0.9437232888965846, "learning_rate": 8.516058742394917e-06, "loss": 0.0519, "step": 5911 }, { "epoch": 0.27472118959107805, "grad_norm": 2.197958013998803, "learning_rate": 8.515523680289803e-06, "loss": 0.1051, "step": 5912 }, { "epoch": 0.27476765799256503, "grad_norm": 1.5289251379114164, "learning_rate": 8.51498853855438e-06, "loss": 0.0835, "step": 5913 }, { "epoch": 0.27481412639405206, "grad_norm": 3.6798279688981586, "learning_rate": 8.514453317200763e-06, "loss": 0.0844, "step": 5914 }, { "epoch": 0.27486059479553904, "grad_norm": 0.8766785032790227, "learning_rate": 8.513918016241083e-06, "loss": 0.0466, "step": 5915 }, { "epoch": 0.274907063197026, "grad_norm": 1.643093204683166, "learning_rate": 8.51338263568746e-06, "loss": 0.0917, "step": 5916 }, { "epoch": 0.274953531598513, "grad_norm": 1.5594175308776568, "learning_rate": 8.512847175552023e-06, "loss": 0.0623, "step": 5917 }, { "epoch": 0.275, "grad_norm": 4.44851296107441, "learning_rate": 8.512311635846901e-06, "loss": 0.1209, "step": 5918 }, { "epoch": 0.275046468401487, "grad_norm": 1.4517541502682967, "learning_rate": 8.511776016584224e-06, "loss": 0.0555, "step": 5919 }, { "epoch": 0.275092936802974, "grad_norm": 3.47128949650911, "learning_rate": 8.511240317776121e-06, "loss": 0.0813, "step": 5920 }, { "epoch": 0.27513940520446095, "grad_norm": 2.8064451337127414, "learning_rate": 8.510704539434734e-06, "loss": 0.0856, "step": 5921 }, { "epoch": 0.27518587360594793, "grad_norm": 2.1888383357785943, "learning_rate": 8.51016868157219e-06, "loss": 0.1192, "step": 5922 }, { "epoch": 0.27523234200743496, "grad_norm": 1.0075970620444972, "learning_rate": 8.509632744200633e-06, "loss": 0.0817, "step": 5923 }, { "epoch": 0.27527881040892194, "grad_norm": 3.4426409722834497, "learning_rate": 8.509096727332201e-06, "loss": 0.0848, "step": 5924 }, { "epoch": 0.2753252788104089, "grad_norm": 1.9969775807175505, "learning_rate": 8.508560630979034e-06, "loss": 0.083, "step": 5925 }, { "epoch": 0.2753717472118959, "grad_norm": 1.6829330055834228, "learning_rate": 8.508024455153276e-06, "loss": 0.1172, "step": 5926 }, { "epoch": 0.2754182156133829, "grad_norm": 3.153833852943343, "learning_rate": 8.507488199867072e-06, "loss": 0.1077, "step": 5927 }, { "epoch": 0.2754646840148699, "grad_norm": 1.7765255346356128, "learning_rate": 8.506951865132569e-06, "loss": 0.1217, "step": 5928 }, { "epoch": 0.2755111524163569, "grad_norm": 2.7144398369174763, "learning_rate": 8.506415450961916e-06, "loss": 0.1212, "step": 5929 }, { "epoch": 0.27555762081784385, "grad_norm": 0.7474264075391567, "learning_rate": 8.505878957367262e-06, "loss": 0.0648, "step": 5930 }, { "epoch": 0.27560408921933083, "grad_norm": 2.939604976748661, "learning_rate": 8.505342384360759e-06, "loss": 0.0964, "step": 5931 }, { "epoch": 0.27565055762081786, "grad_norm": 2.0803549151750915, "learning_rate": 8.504805731954562e-06, "loss": 0.0784, "step": 5932 }, { "epoch": 0.27569702602230484, "grad_norm": 1.6425123698308606, "learning_rate": 8.504269000160826e-06, "loss": 0.0795, "step": 5933 }, { "epoch": 0.2757434944237918, "grad_norm": 1.8095792307432061, "learning_rate": 8.50373218899171e-06, "loss": 0.1655, "step": 5934 }, { "epoch": 0.2757899628252788, "grad_norm": 1.1963592669734107, "learning_rate": 8.503195298459372e-06, "loss": 0.0977, "step": 5935 }, { "epoch": 0.2758364312267658, "grad_norm": 1.2708878800414147, "learning_rate": 8.502658328575972e-06, "loss": 0.078, "step": 5936 }, { "epoch": 0.2758828996282528, "grad_norm": 1.9179318923187925, "learning_rate": 8.502121279353677e-06, "loss": 0.0905, "step": 5937 }, { "epoch": 0.2759293680297398, "grad_norm": 1.932863286392397, "learning_rate": 8.501584150804648e-06, "loss": 0.0546, "step": 5938 }, { "epoch": 0.27597583643122675, "grad_norm": 1.0134091694142437, "learning_rate": 8.501046942941051e-06, "loss": 0.0585, "step": 5939 }, { "epoch": 0.2760223048327137, "grad_norm": 0.801081105877187, "learning_rate": 8.500509655775058e-06, "loss": 0.0687, "step": 5940 }, { "epoch": 0.27606877323420076, "grad_norm": 1.2771320478516412, "learning_rate": 8.499972289318836e-06, "loss": 0.0926, "step": 5941 }, { "epoch": 0.27611524163568774, "grad_norm": 1.4055488905430766, "learning_rate": 8.49943484358456e-06, "loss": 0.1469, "step": 5942 }, { "epoch": 0.2761617100371747, "grad_norm": 0.7594393720744811, "learning_rate": 8.4988973185844e-06, "loss": 0.0588, "step": 5943 }, { "epoch": 0.2762081784386617, "grad_norm": 1.192377989463255, "learning_rate": 8.498359714330533e-06, "loss": 0.1078, "step": 5944 }, { "epoch": 0.2762546468401487, "grad_norm": 1.4435098743318342, "learning_rate": 8.497822030835137e-06, "loss": 0.07, "step": 5945 }, { "epoch": 0.2763011152416357, "grad_norm": 0.5549184828792062, "learning_rate": 8.497284268110389e-06, "loss": 0.0356, "step": 5946 }, { "epoch": 0.2763475836431227, "grad_norm": 0.7541359200838617, "learning_rate": 8.496746426168475e-06, "loss": 0.04, "step": 5947 }, { "epoch": 0.27639405204460965, "grad_norm": 2.0701388861178067, "learning_rate": 8.496208505021572e-06, "loss": 0.0657, "step": 5948 }, { "epoch": 0.2764405204460966, "grad_norm": 1.2131358640734349, "learning_rate": 8.495670504681866e-06, "loss": 0.0687, "step": 5949 }, { "epoch": 0.27648698884758366, "grad_norm": 2.5606115320704976, "learning_rate": 8.495132425161546e-06, "loss": 0.1002, "step": 5950 }, { "epoch": 0.27653345724907064, "grad_norm": 0.8470666670440083, "learning_rate": 8.494594266472796e-06, "loss": 0.0681, "step": 5951 }, { "epoch": 0.2765799256505576, "grad_norm": 2.2032649798534387, "learning_rate": 8.494056028627808e-06, "loss": 0.1002, "step": 5952 }, { "epoch": 0.2766263940520446, "grad_norm": 1.3409153125027757, "learning_rate": 8.493517711638775e-06, "loss": 0.1259, "step": 5953 }, { "epoch": 0.2766728624535316, "grad_norm": 1.1835968244852806, "learning_rate": 8.492979315517888e-06, "loss": 0.1113, "step": 5954 }, { "epoch": 0.2767193308550186, "grad_norm": 0.8304070642309561, "learning_rate": 8.492440840277342e-06, "loss": 0.0482, "step": 5955 }, { "epoch": 0.2767657992565056, "grad_norm": 1.3908299511547288, "learning_rate": 8.491902285929337e-06, "loss": 0.0861, "step": 5956 }, { "epoch": 0.27681226765799255, "grad_norm": 1.854187406023268, "learning_rate": 8.49136365248607e-06, "loss": 0.0893, "step": 5957 }, { "epoch": 0.2768587360594795, "grad_norm": 1.5181601619944909, "learning_rate": 8.490824939959742e-06, "loss": 0.0896, "step": 5958 }, { "epoch": 0.27690520446096656, "grad_norm": 1.205828864361539, "learning_rate": 8.490286148362555e-06, "loss": 0.0722, "step": 5959 }, { "epoch": 0.27695167286245354, "grad_norm": 1.3776653810793456, "learning_rate": 8.489747277706712e-06, "loss": 0.1226, "step": 5960 }, { "epoch": 0.2769981412639405, "grad_norm": 1.7215739137547965, "learning_rate": 8.489208328004422e-06, "loss": 0.0904, "step": 5961 }, { "epoch": 0.2770446096654275, "grad_norm": 0.8964559632576762, "learning_rate": 8.488669299267888e-06, "loss": 0.0445, "step": 5962 }, { "epoch": 0.2770910780669145, "grad_norm": 1.0532522222529455, "learning_rate": 8.488130191509324e-06, "loss": 0.1112, "step": 5963 }, { "epoch": 0.2771375464684015, "grad_norm": 2.1150864249286867, "learning_rate": 8.487591004740942e-06, "loss": 0.0996, "step": 5964 }, { "epoch": 0.2771840148698885, "grad_norm": 1.3203268306243838, "learning_rate": 8.487051738974952e-06, "loss": 0.0686, "step": 5965 }, { "epoch": 0.27723048327137545, "grad_norm": 1.0241937723298022, "learning_rate": 8.486512394223569e-06, "loss": 0.0768, "step": 5966 }, { "epoch": 0.2772769516728624, "grad_norm": 2.6466764735518056, "learning_rate": 8.485972970499012e-06, "loss": 0.0503, "step": 5967 }, { "epoch": 0.27732342007434946, "grad_norm": 2.727674448540914, "learning_rate": 8.485433467813497e-06, "loss": 0.0526, "step": 5968 }, { "epoch": 0.27736988847583643, "grad_norm": 0.9470972678271641, "learning_rate": 8.484893886179247e-06, "loss": 0.0734, "step": 5969 }, { "epoch": 0.2774163568773234, "grad_norm": 1.3954084744336883, "learning_rate": 8.484354225608481e-06, "loss": 0.1031, "step": 5970 }, { "epoch": 0.2774628252788104, "grad_norm": 0.8933146952355906, "learning_rate": 8.483814486113428e-06, "loss": 0.0627, "step": 5971 }, { "epoch": 0.2775092936802974, "grad_norm": 0.8155616439632778, "learning_rate": 8.483274667706307e-06, "loss": 0.0516, "step": 5972 }, { "epoch": 0.2775557620817844, "grad_norm": 1.3631632427953977, "learning_rate": 8.48273477039935e-06, "loss": 0.1063, "step": 5973 }, { "epoch": 0.2776022304832714, "grad_norm": 2.365286622765197, "learning_rate": 8.482194794204785e-06, "loss": 0.1017, "step": 5974 }, { "epoch": 0.27764869888475835, "grad_norm": 6.037693222972294, "learning_rate": 8.48165473913484e-06, "loss": 0.0939, "step": 5975 }, { "epoch": 0.2776951672862453, "grad_norm": 1.3891619063632605, "learning_rate": 8.481114605201755e-06, "loss": 0.0831, "step": 5976 }, { "epoch": 0.27774163568773236, "grad_norm": 3.2505482048477425, "learning_rate": 8.48057439241776e-06, "loss": 0.1376, "step": 5977 }, { "epoch": 0.27778810408921933, "grad_norm": 1.8568810863435241, "learning_rate": 8.48003410079509e-06, "loss": 0.1125, "step": 5978 }, { "epoch": 0.2778345724907063, "grad_norm": 1.4205258671518994, "learning_rate": 8.479493730345985e-06, "loss": 0.0799, "step": 5979 }, { "epoch": 0.2778810408921933, "grad_norm": 1.6162145657754985, "learning_rate": 8.478953281082686e-06, "loss": 0.0821, "step": 5980 }, { "epoch": 0.2779275092936803, "grad_norm": 4.9234240648717265, "learning_rate": 8.478412753017433e-06, "loss": 0.137, "step": 5981 }, { "epoch": 0.2779739776951673, "grad_norm": 1.0996254923270334, "learning_rate": 8.47787214616247e-06, "loss": 0.0811, "step": 5982 }, { "epoch": 0.2780204460966543, "grad_norm": 1.1254735255146129, "learning_rate": 8.477331460530042e-06, "loss": 0.0836, "step": 5983 }, { "epoch": 0.27806691449814125, "grad_norm": 1.3081721575988492, "learning_rate": 8.476790696132397e-06, "loss": 0.0806, "step": 5984 }, { "epoch": 0.2781133828996282, "grad_norm": 1.2425962382196172, "learning_rate": 8.476249852981784e-06, "loss": 0.064, "step": 5985 }, { "epoch": 0.27815985130111526, "grad_norm": 4.208138275335618, "learning_rate": 8.475708931090453e-06, "loss": 0.102, "step": 5986 }, { "epoch": 0.27820631970260223, "grad_norm": 0.7666670711474741, "learning_rate": 8.475167930470657e-06, "loss": 0.0478, "step": 5987 }, { "epoch": 0.2782527881040892, "grad_norm": 0.8665013705716561, "learning_rate": 8.474626851134648e-06, "loss": 0.0745, "step": 5988 }, { "epoch": 0.2782992565055762, "grad_norm": 1.7752984327724426, "learning_rate": 8.474085693094687e-06, "loss": 0.051, "step": 5989 }, { "epoch": 0.2783457249070632, "grad_norm": 1.287216993619765, "learning_rate": 8.473544456363027e-06, "loss": 0.0354, "step": 5990 }, { "epoch": 0.2783921933085502, "grad_norm": 1.9452785024216128, "learning_rate": 8.473003140951929e-06, "loss": 0.0809, "step": 5991 }, { "epoch": 0.27843866171003717, "grad_norm": 2.2988246836915667, "learning_rate": 8.472461746873655e-06, "loss": 0.1123, "step": 5992 }, { "epoch": 0.27848513011152415, "grad_norm": 2.845159380011875, "learning_rate": 8.471920274140466e-06, "loss": 0.0546, "step": 5993 }, { "epoch": 0.2785315985130111, "grad_norm": 1.805492037365869, "learning_rate": 8.471378722764631e-06, "loss": 0.0795, "step": 5994 }, { "epoch": 0.27857806691449816, "grad_norm": 0.9150133188755165, "learning_rate": 8.470837092758414e-06, "loss": 0.0524, "step": 5995 }, { "epoch": 0.27862453531598513, "grad_norm": 3.3298132603364787, "learning_rate": 8.470295384134082e-06, "loss": 0.1112, "step": 5996 }, { "epoch": 0.2786710037174721, "grad_norm": 1.8520851689320528, "learning_rate": 8.469753596903909e-06, "loss": 0.1006, "step": 5997 }, { "epoch": 0.2787174721189591, "grad_norm": 2.6404186714817866, "learning_rate": 8.469211731080165e-06, "loss": 0.1214, "step": 5998 }, { "epoch": 0.2787639405204461, "grad_norm": 1.7885739683171857, "learning_rate": 8.468669786675124e-06, "loss": 0.0756, "step": 5999 }, { "epoch": 0.2788104089219331, "grad_norm": 1.2823912218965903, "learning_rate": 8.46812776370106e-06, "loss": 0.0549, "step": 6000 }, { "epoch": 0.27885687732342007, "grad_norm": 2.102504268815378, "learning_rate": 8.467585662170254e-06, "loss": 0.0653, "step": 6001 }, { "epoch": 0.27890334572490705, "grad_norm": 3.349911847244895, "learning_rate": 8.467043482094982e-06, "loss": 0.0827, "step": 6002 }, { "epoch": 0.278949814126394, "grad_norm": 1.4532382884940578, "learning_rate": 8.466501223487526e-06, "loss": 0.0768, "step": 6003 }, { "epoch": 0.27899628252788106, "grad_norm": 1.1016389478772488, "learning_rate": 8.46595888636017e-06, "loss": 0.1034, "step": 6004 }, { "epoch": 0.27904275092936803, "grad_norm": 1.7784742435592444, "learning_rate": 8.465416470725197e-06, "loss": 0.0995, "step": 6005 }, { "epoch": 0.279089219330855, "grad_norm": 1.1583914749615682, "learning_rate": 8.464873976594894e-06, "loss": 0.0765, "step": 6006 }, { "epoch": 0.279135687732342, "grad_norm": 1.780897108485317, "learning_rate": 8.464331403981547e-06, "loss": 0.0657, "step": 6007 }, { "epoch": 0.279182156133829, "grad_norm": 1.196652657520024, "learning_rate": 8.463788752897448e-06, "loss": 0.0947, "step": 6008 }, { "epoch": 0.279228624535316, "grad_norm": 0.5633885527861607, "learning_rate": 8.46324602335489e-06, "loss": 0.0654, "step": 6009 }, { "epoch": 0.27927509293680297, "grad_norm": 0.8336126823546519, "learning_rate": 8.462703215366165e-06, "loss": 0.0578, "step": 6010 }, { "epoch": 0.27932156133828995, "grad_norm": 1.6455700052468403, "learning_rate": 8.462160328943564e-06, "loss": 0.0679, "step": 6011 }, { "epoch": 0.2793680297397769, "grad_norm": 1.2942299857630901, "learning_rate": 8.461617364099391e-06, "loss": 0.0513, "step": 6012 }, { "epoch": 0.27941449814126396, "grad_norm": 1.4488510566913173, "learning_rate": 8.46107432084594e-06, "loss": 0.0616, "step": 6013 }, { "epoch": 0.27946096654275093, "grad_norm": 1.6066587777749488, "learning_rate": 8.460531199195511e-06, "loss": 0.0845, "step": 6014 }, { "epoch": 0.2795074349442379, "grad_norm": 0.95042000925921, "learning_rate": 8.459987999160413e-06, "loss": 0.0585, "step": 6015 }, { "epoch": 0.2795539033457249, "grad_norm": 1.0970922296675572, "learning_rate": 8.459444720752941e-06, "loss": 0.0998, "step": 6016 }, { "epoch": 0.2796003717472119, "grad_norm": 1.095363027740787, "learning_rate": 8.458901363985408e-06, "loss": 0.0549, "step": 6017 }, { "epoch": 0.2796468401486989, "grad_norm": 2.686025689738951, "learning_rate": 8.458357928870117e-06, "loss": 0.0806, "step": 6018 }, { "epoch": 0.27969330855018587, "grad_norm": 1.453005008407167, "learning_rate": 8.45781441541938e-06, "loss": 0.0463, "step": 6019 }, { "epoch": 0.27973977695167285, "grad_norm": 1.6869020943640518, "learning_rate": 8.457270823645506e-06, "loss": 0.1002, "step": 6020 }, { "epoch": 0.2797862453531598, "grad_norm": 1.5031536608265934, "learning_rate": 8.45672715356081e-06, "loss": 0.071, "step": 6021 }, { "epoch": 0.27983271375464686, "grad_norm": 1.1577815334165151, "learning_rate": 8.456183405177604e-06, "loss": 0.0698, "step": 6022 }, { "epoch": 0.27987918215613383, "grad_norm": 0.9564431227308804, "learning_rate": 8.45563957850821e-06, "loss": 0.0865, "step": 6023 }, { "epoch": 0.2799256505576208, "grad_norm": 1.2800024228196492, "learning_rate": 8.455095673564941e-06, "loss": 0.0782, "step": 6024 }, { "epoch": 0.2799721189591078, "grad_norm": 0.9921936312095145, "learning_rate": 8.454551690360117e-06, "loss": 0.0821, "step": 6025 }, { "epoch": 0.2800185873605948, "grad_norm": 1.2436761611062146, "learning_rate": 8.454007628906063e-06, "loss": 0.083, "step": 6026 }, { "epoch": 0.2800650557620818, "grad_norm": 1.5497873892736889, "learning_rate": 8.4534634892151e-06, "loss": 0.0861, "step": 6027 }, { "epoch": 0.28011152416356877, "grad_norm": 0.9622904715708692, "learning_rate": 8.452919271299554e-06, "loss": 0.0501, "step": 6028 }, { "epoch": 0.28015799256505575, "grad_norm": 0.566374202740756, "learning_rate": 8.452374975171754e-06, "loss": 0.0555, "step": 6029 }, { "epoch": 0.2802044609665427, "grad_norm": 0.7539192451879254, "learning_rate": 8.451830600844027e-06, "loss": 0.0578, "step": 6030 }, { "epoch": 0.28025092936802976, "grad_norm": 1.2476429963003606, "learning_rate": 8.451286148328701e-06, "loss": 0.078, "step": 6031 }, { "epoch": 0.28029739776951673, "grad_norm": 1.1675538599825455, "learning_rate": 8.450741617638115e-06, "loss": 0.0666, "step": 6032 }, { "epoch": 0.2803438661710037, "grad_norm": 0.991246076839696, "learning_rate": 8.450197008784598e-06, "loss": 0.0574, "step": 6033 }, { "epoch": 0.2803903345724907, "grad_norm": 2.590003426029871, "learning_rate": 8.449652321780487e-06, "loss": 0.0812, "step": 6034 }, { "epoch": 0.2804368029739777, "grad_norm": 2.183068916376716, "learning_rate": 8.44910755663812e-06, "loss": 0.0993, "step": 6035 }, { "epoch": 0.2804832713754647, "grad_norm": 0.9077552643739469, "learning_rate": 8.448562713369839e-06, "loss": 0.0704, "step": 6036 }, { "epoch": 0.28052973977695167, "grad_norm": 3.162784294174978, "learning_rate": 8.448017791987982e-06, "loss": 0.0875, "step": 6037 }, { "epoch": 0.28057620817843865, "grad_norm": 1.088086509889508, "learning_rate": 8.447472792504891e-06, "loss": 0.0606, "step": 6038 }, { "epoch": 0.2806226765799256, "grad_norm": 2.669494840147646, "learning_rate": 8.446927714932914e-06, "loss": 0.081, "step": 6039 }, { "epoch": 0.28066914498141265, "grad_norm": 1.120554543021906, "learning_rate": 8.446382559284396e-06, "loss": 0.066, "step": 6040 }, { "epoch": 0.28071561338289963, "grad_norm": 0.5988183429151956, "learning_rate": 8.445837325571688e-06, "loss": 0.0497, "step": 6041 }, { "epoch": 0.2807620817843866, "grad_norm": 0.6174970473976304, "learning_rate": 8.445292013807135e-06, "loss": 0.0313, "step": 6042 }, { "epoch": 0.2808085501858736, "grad_norm": 1.8353658878130492, "learning_rate": 8.444746624003095e-06, "loss": 0.0708, "step": 6043 }, { "epoch": 0.2808550185873606, "grad_norm": 1.8405720888362826, "learning_rate": 8.444201156171915e-06, "loss": 0.1107, "step": 6044 }, { "epoch": 0.2809014869888476, "grad_norm": 1.3347233159236396, "learning_rate": 8.443655610325957e-06, "loss": 0.0714, "step": 6045 }, { "epoch": 0.28094795539033457, "grad_norm": 1.2837214125614704, "learning_rate": 8.443109986477574e-06, "loss": 0.0521, "step": 6046 }, { "epoch": 0.28099442379182155, "grad_norm": 1.1878183168090857, "learning_rate": 8.442564284639124e-06, "loss": 0.0658, "step": 6047 }, { "epoch": 0.2810408921933085, "grad_norm": 1.4488357009514576, "learning_rate": 8.442018504822972e-06, "loss": 0.0778, "step": 6048 }, { "epoch": 0.28108736059479555, "grad_norm": 1.5607637562988033, "learning_rate": 8.441472647041477e-06, "loss": 0.1288, "step": 6049 }, { "epoch": 0.28113382899628253, "grad_norm": 1.6935137154565703, "learning_rate": 8.440926711307005e-06, "loss": 0.0736, "step": 6050 }, { "epoch": 0.2811802973977695, "grad_norm": 1.8110294293825007, "learning_rate": 8.44038069763192e-06, "loss": 0.0933, "step": 6051 }, { "epoch": 0.2812267657992565, "grad_norm": 1.7450324365485723, "learning_rate": 8.439834606028594e-06, "loss": 0.1088, "step": 6052 }, { "epoch": 0.2812732342007435, "grad_norm": 2.1761451725718923, "learning_rate": 8.439288436509393e-06, "loss": 0.0755, "step": 6053 }, { "epoch": 0.2813197026022305, "grad_norm": 2.036472021560639, "learning_rate": 8.438742189086688e-06, "loss": 0.0923, "step": 6054 }, { "epoch": 0.28136617100371747, "grad_norm": 1.0355997748528556, "learning_rate": 8.438195863772854e-06, "loss": 0.0818, "step": 6055 }, { "epoch": 0.28141263940520445, "grad_norm": 1.2115092275395267, "learning_rate": 8.437649460580265e-06, "loss": 0.0528, "step": 6056 }, { "epoch": 0.2814591078066915, "grad_norm": 0.6811784324993723, "learning_rate": 8.437102979521297e-06, "loss": 0.0553, "step": 6057 }, { "epoch": 0.28150557620817845, "grad_norm": 2.4636484014706177, "learning_rate": 8.436556420608331e-06, "loss": 0.0869, "step": 6058 }, { "epoch": 0.28155204460966543, "grad_norm": 1.4166579311147796, "learning_rate": 8.436009783853742e-06, "loss": 0.092, "step": 6059 }, { "epoch": 0.2815985130111524, "grad_norm": 0.7980370805961333, "learning_rate": 8.435463069269919e-06, "loss": 0.0708, "step": 6060 }, { "epoch": 0.2816449814126394, "grad_norm": 1.5929509120109986, "learning_rate": 8.434916276869238e-06, "loss": 0.0752, "step": 6061 }, { "epoch": 0.2816914498141264, "grad_norm": 1.1325044872879937, "learning_rate": 8.434369406664091e-06, "loss": 0.0611, "step": 6062 }, { "epoch": 0.2817379182156134, "grad_norm": 1.07321699084909, "learning_rate": 8.433822458666861e-06, "loss": 0.0618, "step": 6063 }, { "epoch": 0.28178438661710037, "grad_norm": 1.195063464653602, "learning_rate": 8.433275432889936e-06, "loss": 0.0777, "step": 6064 }, { "epoch": 0.28183085501858735, "grad_norm": 1.9111737508632891, "learning_rate": 8.432728329345713e-06, "loss": 0.0844, "step": 6065 }, { "epoch": 0.2818773234200744, "grad_norm": 1.4710211796917807, "learning_rate": 8.432181148046578e-06, "loss": 0.0701, "step": 6066 }, { "epoch": 0.28192379182156135, "grad_norm": 2.106925813990715, "learning_rate": 8.431633889004928e-06, "loss": 0.1093, "step": 6067 }, { "epoch": 0.28197026022304833, "grad_norm": 1.2432866424158744, "learning_rate": 8.43108655223316e-06, "loss": 0.0714, "step": 6068 }, { "epoch": 0.2820167286245353, "grad_norm": 0.986428204747173, "learning_rate": 8.43053913774367e-06, "loss": 0.0666, "step": 6069 }, { "epoch": 0.2820631970260223, "grad_norm": 1.301734844605666, "learning_rate": 8.429991645548857e-06, "loss": 0.0874, "step": 6070 }, { "epoch": 0.2821096654275093, "grad_norm": 0.9937773265546518, "learning_rate": 8.429444075661124e-06, "loss": 0.0905, "step": 6071 }, { "epoch": 0.2821561338289963, "grad_norm": 1.7348345358297563, "learning_rate": 8.428896428092871e-06, "loss": 0.0977, "step": 6072 }, { "epoch": 0.28220260223048327, "grad_norm": 1.9558707449436394, "learning_rate": 8.428348702856508e-06, "loss": 0.0768, "step": 6073 }, { "epoch": 0.28224907063197024, "grad_norm": 1.7345063268822385, "learning_rate": 8.427800899964438e-06, "loss": 0.0926, "step": 6074 }, { "epoch": 0.2822955390334573, "grad_norm": 0.7647841001030559, "learning_rate": 8.427253019429068e-06, "loss": 0.0375, "step": 6075 }, { "epoch": 0.28234200743494425, "grad_norm": 1.96168371980853, "learning_rate": 8.426705061262812e-06, "loss": 0.0605, "step": 6076 }, { "epoch": 0.28238847583643123, "grad_norm": 1.0082706979199638, "learning_rate": 8.426157025478082e-06, "loss": 0.0674, "step": 6077 }, { "epoch": 0.2824349442379182, "grad_norm": 1.0977123482408344, "learning_rate": 8.425608912087285e-06, "loss": 0.0607, "step": 6078 }, { "epoch": 0.2824814126394052, "grad_norm": 2.46178295862391, "learning_rate": 8.425060721102842e-06, "loss": 0.0898, "step": 6079 }, { "epoch": 0.2825278810408922, "grad_norm": 0.8182353115586128, "learning_rate": 8.42451245253717e-06, "loss": 0.0626, "step": 6080 }, { "epoch": 0.2825743494423792, "grad_norm": 2.4643268519022588, "learning_rate": 8.423964106402687e-06, "loss": 0.0849, "step": 6081 }, { "epoch": 0.28262081784386617, "grad_norm": 0.677878733677574, "learning_rate": 8.42341568271181e-06, "loss": 0.047, "step": 6082 }, { "epoch": 0.28266728624535314, "grad_norm": 1.0107063146125244, "learning_rate": 8.42286718147697e-06, "loss": 0.041, "step": 6083 }, { "epoch": 0.2827137546468402, "grad_norm": 0.8386667055420943, "learning_rate": 8.422318602710581e-06, "loss": 0.0642, "step": 6084 }, { "epoch": 0.28276022304832715, "grad_norm": 1.8940642439168764, "learning_rate": 8.421769946425075e-06, "loss": 0.0676, "step": 6085 }, { "epoch": 0.28280669144981413, "grad_norm": 1.6271771383354163, "learning_rate": 8.42122121263288e-06, "loss": 0.1103, "step": 6086 }, { "epoch": 0.2828531598513011, "grad_norm": 1.1243245631319392, "learning_rate": 8.420672401346424e-06, "loss": 0.0451, "step": 6087 }, { "epoch": 0.2828996282527881, "grad_norm": 0.9000848422155759, "learning_rate": 8.420123512578136e-06, "loss": 0.0489, "step": 6088 }, { "epoch": 0.2829460966542751, "grad_norm": 1.5059982152683875, "learning_rate": 8.419574546340453e-06, "loss": 0.1259, "step": 6089 }, { "epoch": 0.2829925650557621, "grad_norm": 1.7073674888306078, "learning_rate": 8.419025502645804e-06, "loss": 0.0861, "step": 6090 }, { "epoch": 0.28303903345724907, "grad_norm": 1.5712093450446487, "learning_rate": 8.418476381506631e-06, "loss": 0.0845, "step": 6091 }, { "epoch": 0.28308550185873604, "grad_norm": 1.5002370804743224, "learning_rate": 8.417927182935371e-06, "loss": 0.0655, "step": 6092 }, { "epoch": 0.2831319702602231, "grad_norm": 1.8389446069349287, "learning_rate": 8.417377906944462e-06, "loss": 0.0683, "step": 6093 }, { "epoch": 0.28317843866171005, "grad_norm": 1.3564833880003928, "learning_rate": 8.416828553546345e-06, "loss": 0.1057, "step": 6094 }, { "epoch": 0.28322490706319703, "grad_norm": 0.6545147473670343, "learning_rate": 8.416279122753468e-06, "loss": 0.0498, "step": 6095 }, { "epoch": 0.283271375464684, "grad_norm": 1.5189863562684487, "learning_rate": 8.415729614578273e-06, "loss": 0.0926, "step": 6096 }, { "epoch": 0.283317843866171, "grad_norm": 3.1343469112874165, "learning_rate": 8.415180029033205e-06, "loss": 0.08, "step": 6097 }, { "epoch": 0.283364312267658, "grad_norm": 0.9233178387964055, "learning_rate": 8.414630366130718e-06, "loss": 0.0716, "step": 6098 }, { "epoch": 0.283410780669145, "grad_norm": 1.4326780161996489, "learning_rate": 8.414080625883256e-06, "loss": 0.0839, "step": 6099 }, { "epoch": 0.28345724907063197, "grad_norm": 0.993914661459488, "learning_rate": 8.413530808303276e-06, "loss": 0.0548, "step": 6100 }, { "epoch": 0.28350371747211894, "grad_norm": 1.014825268529659, "learning_rate": 8.412980913403233e-06, "loss": 0.0873, "step": 6101 }, { "epoch": 0.283550185873606, "grad_norm": 2.081255218788415, "learning_rate": 8.412430941195577e-06, "loss": 0.0572, "step": 6102 }, { "epoch": 0.28359665427509295, "grad_norm": 1.0198958599977022, "learning_rate": 8.41188089169277e-06, "loss": 0.0643, "step": 6103 }, { "epoch": 0.28364312267657993, "grad_norm": 0.6282378281219797, "learning_rate": 8.411330764907268e-06, "loss": 0.0462, "step": 6104 }, { "epoch": 0.2836895910780669, "grad_norm": 1.4891393720839043, "learning_rate": 8.410780560851537e-06, "loss": 0.0818, "step": 6105 }, { "epoch": 0.2837360594795539, "grad_norm": 0.6359607269731814, "learning_rate": 8.410230279538033e-06, "loss": 0.042, "step": 6106 }, { "epoch": 0.2837825278810409, "grad_norm": 1.3769886589285563, "learning_rate": 8.409679920979228e-06, "loss": 0.1309, "step": 6107 }, { "epoch": 0.2838289962825279, "grad_norm": 2.8698707443113824, "learning_rate": 8.40912948518758e-06, "loss": 0.0433, "step": 6108 }, { "epoch": 0.28387546468401487, "grad_norm": 1.3347279821523277, "learning_rate": 8.408578972175563e-06, "loss": 0.0594, "step": 6109 }, { "epoch": 0.28392193308550184, "grad_norm": 1.1520945628020758, "learning_rate": 8.408028381955646e-06, "loss": 0.0572, "step": 6110 }, { "epoch": 0.2839684014869889, "grad_norm": 1.3195835732969636, "learning_rate": 8.407477714540297e-06, "loss": 0.1063, "step": 6111 }, { "epoch": 0.28401486988847585, "grad_norm": 1.0924948532597358, "learning_rate": 8.406926969941994e-06, "loss": 0.0615, "step": 6112 }, { "epoch": 0.28406133828996283, "grad_norm": 1.007354382959541, "learning_rate": 8.406376148173207e-06, "loss": 0.0602, "step": 6113 }, { "epoch": 0.2841078066914498, "grad_norm": 1.257228488058906, "learning_rate": 8.405825249246416e-06, "loss": 0.0969, "step": 6114 }, { "epoch": 0.2841542750929368, "grad_norm": 2.2705401429356105, "learning_rate": 8.4052742731741e-06, "loss": 0.0642, "step": 6115 }, { "epoch": 0.2842007434944238, "grad_norm": 1.6662140305165718, "learning_rate": 8.404723219968736e-06, "loss": 0.067, "step": 6116 }, { "epoch": 0.2842472118959108, "grad_norm": 2.586033599999208, "learning_rate": 8.404172089642808e-06, "loss": 0.0918, "step": 6117 }, { "epoch": 0.28429368029739777, "grad_norm": 0.895913942775619, "learning_rate": 8.403620882208801e-06, "loss": 0.0835, "step": 6118 }, { "epoch": 0.28434014869888474, "grad_norm": 1.4152349390013566, "learning_rate": 8.403069597679195e-06, "loss": 0.0672, "step": 6119 }, { "epoch": 0.2843866171003718, "grad_norm": 0.9884828787878127, "learning_rate": 8.402518236066484e-06, "loss": 0.0671, "step": 6120 }, { "epoch": 0.28443308550185875, "grad_norm": 1.313795319050625, "learning_rate": 8.401966797383153e-06, "loss": 0.1047, "step": 6121 }, { "epoch": 0.2844795539033457, "grad_norm": 1.5067873923680197, "learning_rate": 8.401415281641694e-06, "loss": 0.0826, "step": 6122 }, { "epoch": 0.2845260223048327, "grad_norm": 1.9366289960983254, "learning_rate": 8.400863688854598e-06, "loss": 0.1577, "step": 6123 }, { "epoch": 0.2845724907063197, "grad_norm": 1.596766212136352, "learning_rate": 8.40031201903436e-06, "loss": 0.0567, "step": 6124 }, { "epoch": 0.2846189591078067, "grad_norm": 3.3230022911496016, "learning_rate": 8.399760272193476e-06, "loss": 0.1008, "step": 6125 }, { "epoch": 0.2846654275092937, "grad_norm": 1.244849396661052, "learning_rate": 8.399208448344445e-06, "loss": 0.0704, "step": 6126 }, { "epoch": 0.28471189591078067, "grad_norm": 0.8793145384284325, "learning_rate": 8.398656547499764e-06, "loss": 0.0519, "step": 6127 }, { "epoch": 0.28475836431226764, "grad_norm": 0.6062722993424968, "learning_rate": 8.398104569671937e-06, "loss": 0.0645, "step": 6128 }, { "epoch": 0.2848048327137547, "grad_norm": 0.5153978261502191, "learning_rate": 8.397552514873462e-06, "loss": 0.0347, "step": 6129 }, { "epoch": 0.28485130111524165, "grad_norm": 1.3852636579015443, "learning_rate": 8.397000383116849e-06, "loss": 0.0724, "step": 6130 }, { "epoch": 0.2848977695167286, "grad_norm": 2.0400698395984387, "learning_rate": 8.396448174414602e-06, "loss": 0.0997, "step": 6131 }, { "epoch": 0.2849442379182156, "grad_norm": 0.8518534659700884, "learning_rate": 8.395895888779227e-06, "loss": 0.0719, "step": 6132 }, { "epoch": 0.2849907063197026, "grad_norm": 0.5968169432518446, "learning_rate": 8.395343526223239e-06, "loss": 0.0595, "step": 6133 }, { "epoch": 0.2850371747211896, "grad_norm": 0.9515816966228933, "learning_rate": 8.394791086759144e-06, "loss": 0.0504, "step": 6134 }, { "epoch": 0.2850836431226766, "grad_norm": 1.3198551736135402, "learning_rate": 8.394238570399458e-06, "loss": 0.0647, "step": 6135 }, { "epoch": 0.28513011152416357, "grad_norm": 1.6791753573271404, "learning_rate": 8.393685977156698e-06, "loss": 0.1124, "step": 6136 }, { "epoch": 0.28517657992565054, "grad_norm": 0.6812962256369506, "learning_rate": 8.393133307043377e-06, "loss": 0.0487, "step": 6137 }, { "epoch": 0.2852230483271376, "grad_norm": 2.8492410381486306, "learning_rate": 8.392580560072017e-06, "loss": 0.0862, "step": 6138 }, { "epoch": 0.28526951672862455, "grad_norm": 0.8163676704931183, "learning_rate": 8.392027736255135e-06, "loss": 0.0807, "step": 6139 }, { "epoch": 0.2853159851301115, "grad_norm": 1.1663626666655884, "learning_rate": 8.391474835605256e-06, "loss": 0.0529, "step": 6140 }, { "epoch": 0.2853624535315985, "grad_norm": 1.0055032675543958, "learning_rate": 8.390921858134902e-06, "loss": 0.0374, "step": 6141 }, { "epoch": 0.2854089219330855, "grad_norm": 0.9311078989551065, "learning_rate": 8.390368803856599e-06, "loss": 0.0624, "step": 6142 }, { "epoch": 0.2854553903345725, "grad_norm": 0.6668145133164707, "learning_rate": 8.389815672782874e-06, "loss": 0.0519, "step": 6143 }, { "epoch": 0.2855018587360595, "grad_norm": 1.3263072156304463, "learning_rate": 8.389262464926256e-06, "loss": 0.058, "step": 6144 }, { "epoch": 0.28554832713754646, "grad_norm": 1.341449671483184, "learning_rate": 8.388709180299278e-06, "loss": 0.0557, "step": 6145 }, { "epoch": 0.28559479553903344, "grad_norm": 0.5839237004851177, "learning_rate": 8.38815581891447e-06, "loss": 0.0612, "step": 6146 }, { "epoch": 0.2856412639405205, "grad_norm": 2.0306801808546844, "learning_rate": 8.387602380784366e-06, "loss": 0.0692, "step": 6147 }, { "epoch": 0.28568773234200745, "grad_norm": 0.8053645786184052, "learning_rate": 8.387048865921502e-06, "loss": 0.0512, "step": 6148 }, { "epoch": 0.2857342007434944, "grad_norm": 1.0457669775177707, "learning_rate": 8.386495274338416e-06, "loss": 0.0828, "step": 6149 }, { "epoch": 0.2857806691449814, "grad_norm": 1.0317462392977352, "learning_rate": 8.38594160604765e-06, "loss": 0.054, "step": 6150 }, { "epoch": 0.2858271375464684, "grad_norm": 1.3718774043356774, "learning_rate": 8.385387861061743e-06, "loss": 0.0738, "step": 6151 }, { "epoch": 0.2858736059479554, "grad_norm": 1.3699996857409287, "learning_rate": 8.384834039393235e-06, "loss": 0.0638, "step": 6152 }, { "epoch": 0.2859200743494424, "grad_norm": 0.8237988301741306, "learning_rate": 8.384280141054674e-06, "loss": 0.0553, "step": 6153 }, { "epoch": 0.28596654275092936, "grad_norm": 1.3011257819599693, "learning_rate": 8.383726166058609e-06, "loss": 0.0534, "step": 6154 }, { "epoch": 0.28601301115241634, "grad_norm": 1.563980229366189, "learning_rate": 8.383172114417583e-06, "loss": 0.075, "step": 6155 }, { "epoch": 0.2860594795539034, "grad_norm": 1.9672896454534514, "learning_rate": 8.382617986144148e-06, "loss": 0.0908, "step": 6156 }, { "epoch": 0.28610594795539035, "grad_norm": 0.826640631703899, "learning_rate": 8.382063781250855e-06, "loss": 0.0542, "step": 6157 }, { "epoch": 0.2861524163568773, "grad_norm": 1.620648231944591, "learning_rate": 8.38150949975026e-06, "loss": 0.0752, "step": 6158 }, { "epoch": 0.2861988847583643, "grad_norm": 1.0809371882481629, "learning_rate": 8.380955141654914e-06, "loss": 0.0589, "step": 6159 }, { "epoch": 0.2862453531598513, "grad_norm": 0.5113401489103832, "learning_rate": 8.380400706977374e-06, "loss": 0.0476, "step": 6160 }, { "epoch": 0.2862918215613383, "grad_norm": 3.80185822838264, "learning_rate": 8.379846195730203e-06, "loss": 0.0744, "step": 6161 }, { "epoch": 0.2863382899628253, "grad_norm": 0.6579943006875304, "learning_rate": 8.379291607925956e-06, "loss": 0.0559, "step": 6162 }, { "epoch": 0.28638475836431226, "grad_norm": 1.507285179370283, "learning_rate": 8.378736943577202e-06, "loss": 0.0816, "step": 6163 }, { "epoch": 0.28643122676579924, "grad_norm": 0.6739747770904817, "learning_rate": 8.378182202696495e-06, "loss": 0.0692, "step": 6164 }, { "epoch": 0.2864776951672863, "grad_norm": 2.519262475543871, "learning_rate": 8.377627385296407e-06, "loss": 0.1127, "step": 6165 }, { "epoch": 0.28652416356877325, "grad_norm": 1.2109609586234835, "learning_rate": 8.377072491389506e-06, "loss": 0.0566, "step": 6166 }, { "epoch": 0.2865706319702602, "grad_norm": 0.5831147981575739, "learning_rate": 8.376517520988356e-06, "loss": 0.0536, "step": 6167 }, { "epoch": 0.2866171003717472, "grad_norm": 0.9329565628912039, "learning_rate": 8.375962474105532e-06, "loss": 0.0768, "step": 6168 }, { "epoch": 0.2866635687732342, "grad_norm": 0.752935267979748, "learning_rate": 8.375407350753606e-06, "loss": 0.0627, "step": 6169 }, { "epoch": 0.2867100371747212, "grad_norm": 2.2624697204637556, "learning_rate": 8.37485215094515e-06, "loss": 0.0834, "step": 6170 }, { "epoch": 0.2867565055762082, "grad_norm": 0.9853658639332079, "learning_rate": 8.37429687469274e-06, "loss": 0.058, "step": 6171 }, { "epoch": 0.28680297397769516, "grad_norm": 1.1649086809819895, "learning_rate": 8.373741522008954e-06, "loss": 0.0694, "step": 6172 }, { "epoch": 0.28684944237918214, "grad_norm": 1.7286833066053557, "learning_rate": 8.373186092906373e-06, "loss": 0.1028, "step": 6173 }, { "epoch": 0.28689591078066917, "grad_norm": 1.6935558800049575, "learning_rate": 8.372630587397576e-06, "loss": 0.0581, "step": 6174 }, { "epoch": 0.28694237918215615, "grad_norm": 1.2795215552270949, "learning_rate": 8.372075005495147e-06, "loss": 0.0656, "step": 6175 }, { "epoch": 0.2869888475836431, "grad_norm": 1.4940458695941383, "learning_rate": 8.37151934721167e-06, "loss": 0.0734, "step": 6176 }, { "epoch": 0.2870353159851301, "grad_norm": 1.2146010654717936, "learning_rate": 8.370963612559732e-06, "loss": 0.085, "step": 6177 }, { "epoch": 0.2870817843866171, "grad_norm": 5.821364120778986, "learning_rate": 8.37040780155192e-06, "loss": 0.0854, "step": 6178 }, { "epoch": 0.2871282527881041, "grad_norm": 1.7538977676545104, "learning_rate": 8.369851914200822e-06, "loss": 0.0672, "step": 6179 }, { "epoch": 0.2871747211895911, "grad_norm": 1.3100910020614374, "learning_rate": 8.369295950519034e-06, "loss": 0.0852, "step": 6180 }, { "epoch": 0.28722118959107806, "grad_norm": 1.6738144409254998, "learning_rate": 8.368739910519146e-06, "loss": 0.0729, "step": 6181 }, { "epoch": 0.28726765799256504, "grad_norm": 1.0419972577618903, "learning_rate": 8.368183794213752e-06, "loss": 0.0982, "step": 6182 }, { "epoch": 0.28731412639405207, "grad_norm": 1.1522104861820368, "learning_rate": 8.36762760161545e-06, "loss": 0.0869, "step": 6183 }, { "epoch": 0.28736059479553905, "grad_norm": 3.3152918553556114, "learning_rate": 8.367071332736841e-06, "loss": 0.0883, "step": 6184 }, { "epoch": 0.287407063197026, "grad_norm": 1.9874370932608831, "learning_rate": 8.366514987590521e-06, "loss": 0.0789, "step": 6185 }, { "epoch": 0.287453531598513, "grad_norm": 4.633375849320501, "learning_rate": 8.365958566189095e-06, "loss": 0.0861, "step": 6186 }, { "epoch": 0.2875, "grad_norm": 1.1224591332640714, "learning_rate": 8.365402068545164e-06, "loss": 0.0919, "step": 6187 }, { "epoch": 0.287546468401487, "grad_norm": 0.9141357795153389, "learning_rate": 8.364845494671332e-06, "loss": 0.0593, "step": 6188 }, { "epoch": 0.287592936802974, "grad_norm": 1.4372664584216097, "learning_rate": 8.36428884458021e-06, "loss": 0.0864, "step": 6189 }, { "epoch": 0.28763940520446096, "grad_norm": 0.5254146038012145, "learning_rate": 8.363732118284404e-06, "loss": 0.0462, "step": 6190 }, { "epoch": 0.28768587360594794, "grad_norm": 1.4258979092246389, "learning_rate": 8.363175315796525e-06, "loss": 0.0676, "step": 6191 }, { "epoch": 0.28773234200743497, "grad_norm": 1.2914299066187624, "learning_rate": 8.362618437129187e-06, "loss": 0.0847, "step": 6192 }, { "epoch": 0.28777881040892195, "grad_norm": 1.511053617725459, "learning_rate": 8.362061482295001e-06, "loss": 0.0558, "step": 6193 }, { "epoch": 0.2878252788104089, "grad_norm": 1.965876129480507, "learning_rate": 8.361504451306585e-06, "loss": 0.0906, "step": 6194 }, { "epoch": 0.2878717472118959, "grad_norm": 2.5581861112265667, "learning_rate": 8.360947344176554e-06, "loss": 0.1197, "step": 6195 }, { "epoch": 0.2879182156133829, "grad_norm": 1.1648035713251144, "learning_rate": 8.360390160917528e-06, "loss": 0.0668, "step": 6196 }, { "epoch": 0.2879646840148699, "grad_norm": 1.167455182285652, "learning_rate": 8.359832901542131e-06, "loss": 0.0603, "step": 6197 }, { "epoch": 0.2880111524163569, "grad_norm": 0.7372880418233928, "learning_rate": 8.35927556606298e-06, "loss": 0.0611, "step": 6198 }, { "epoch": 0.28805762081784386, "grad_norm": 1.4755162211034425, "learning_rate": 8.358718154492702e-06, "loss": 0.0655, "step": 6199 }, { "epoch": 0.28810408921933084, "grad_norm": 1.0683682982390252, "learning_rate": 8.358160666843924e-06, "loss": 0.1007, "step": 6200 }, { "epoch": 0.28815055762081787, "grad_norm": 0.6683086941391855, "learning_rate": 8.357603103129273e-06, "loss": 0.0396, "step": 6201 }, { "epoch": 0.28819702602230485, "grad_norm": 0.7955822323419333, "learning_rate": 8.357045463361376e-06, "loss": 0.0757, "step": 6202 }, { "epoch": 0.2882434944237918, "grad_norm": 0.7103092044980999, "learning_rate": 8.356487747552866e-06, "loss": 0.0593, "step": 6203 }, { "epoch": 0.2882899628252788, "grad_norm": 2.14267400840097, "learning_rate": 8.355929955716378e-06, "loss": 0.095, "step": 6204 }, { "epoch": 0.2883364312267658, "grad_norm": 2.0782807369581358, "learning_rate": 8.355372087864542e-06, "loss": 0.0557, "step": 6205 }, { "epoch": 0.2883828996282528, "grad_norm": 0.985380873785128, "learning_rate": 8.354814144009997e-06, "loss": 0.0605, "step": 6206 }, { "epoch": 0.2884293680297398, "grad_norm": 1.3818664917572052, "learning_rate": 8.354256124165383e-06, "loss": 0.1012, "step": 6207 }, { "epoch": 0.28847583643122676, "grad_norm": 1.3196542479676923, "learning_rate": 8.353698028343336e-06, "loss": 0.0777, "step": 6208 }, { "epoch": 0.28852230483271374, "grad_norm": 1.3625081730329505, "learning_rate": 8.353139856556498e-06, "loss": 0.0739, "step": 6209 }, { "epoch": 0.28856877323420077, "grad_norm": 1.0662871753909744, "learning_rate": 8.352581608817516e-06, "loss": 0.0727, "step": 6210 }, { "epoch": 0.28861524163568775, "grad_norm": 0.5635504260380221, "learning_rate": 8.352023285139029e-06, "loss": 0.0486, "step": 6211 }, { "epoch": 0.2886617100371747, "grad_norm": 1.990626690752517, "learning_rate": 8.351464885533688e-06, "loss": 0.0793, "step": 6212 }, { "epoch": 0.2887081784386617, "grad_norm": 2.0573876095838117, "learning_rate": 8.35090641001414e-06, "loss": 0.0405, "step": 6213 }, { "epoch": 0.2887546468401487, "grad_norm": 1.0623120995919846, "learning_rate": 8.350347858593035e-06, "loss": 0.0719, "step": 6214 }, { "epoch": 0.2888011152416357, "grad_norm": 0.7205571727418514, "learning_rate": 8.349789231283025e-06, "loss": 0.0579, "step": 6215 }, { "epoch": 0.2888475836431227, "grad_norm": 1.2715313661322114, "learning_rate": 8.349230528096764e-06, "loss": 0.0917, "step": 6216 }, { "epoch": 0.28889405204460966, "grad_norm": 2.0099854298769686, "learning_rate": 8.348671749046905e-06, "loss": 0.0863, "step": 6217 }, { "epoch": 0.28894052044609664, "grad_norm": 1.0257816249132283, "learning_rate": 8.348112894146106e-06, "loss": 0.0553, "step": 6218 }, { "epoch": 0.28898698884758367, "grad_norm": 0.8682253861005813, "learning_rate": 8.347553963407028e-06, "loss": 0.0353, "step": 6219 }, { "epoch": 0.28903345724907065, "grad_norm": 1.7024001763110272, "learning_rate": 8.346994956842327e-06, "loss": 0.0643, "step": 6220 }, { "epoch": 0.2890799256505576, "grad_norm": 0.8425944764692667, "learning_rate": 8.34643587446467e-06, "loss": 0.054, "step": 6221 }, { "epoch": 0.2891263940520446, "grad_norm": 1.7982511027764523, "learning_rate": 8.345876716286719e-06, "loss": 0.0857, "step": 6222 }, { "epoch": 0.2891728624535316, "grad_norm": 1.4079326295871586, "learning_rate": 8.345317482321137e-06, "loss": 0.055, "step": 6223 }, { "epoch": 0.2892193308550186, "grad_norm": 1.9787326160151328, "learning_rate": 8.344758172580593e-06, "loss": 0.0818, "step": 6224 }, { "epoch": 0.2892657992565056, "grad_norm": 2.563382092875322, "learning_rate": 8.344198787077758e-06, "loss": 0.1076, "step": 6225 }, { "epoch": 0.28931226765799256, "grad_norm": 2.2314806460527423, "learning_rate": 8.343639325825298e-06, "loss": 0.0742, "step": 6226 }, { "epoch": 0.28935873605947954, "grad_norm": 0.6133306749542334, "learning_rate": 8.34307978883589e-06, "loss": 0.0535, "step": 6227 }, { "epoch": 0.28940520446096657, "grad_norm": 0.5962166299387399, "learning_rate": 8.342520176122207e-06, "loss": 0.0421, "step": 6228 }, { "epoch": 0.28945167286245355, "grad_norm": 0.9742860294772309, "learning_rate": 8.341960487696924e-06, "loss": 0.0743, "step": 6229 }, { "epoch": 0.2894981412639405, "grad_norm": 0.9279319426235001, "learning_rate": 8.341400723572718e-06, "loss": 0.111, "step": 6230 }, { "epoch": 0.2895446096654275, "grad_norm": 0.6566760710240473, "learning_rate": 8.340840883762267e-06, "loss": 0.057, "step": 6231 }, { "epoch": 0.2895910780669145, "grad_norm": 1.3300857844692477, "learning_rate": 8.340280968278255e-06, "loss": 0.0834, "step": 6232 }, { "epoch": 0.2896375464684015, "grad_norm": 1.069519261462884, "learning_rate": 8.339720977133365e-06, "loss": 0.0395, "step": 6233 }, { "epoch": 0.2896840148698885, "grad_norm": 1.7996964727335711, "learning_rate": 8.339160910340278e-06, "loss": 0.1142, "step": 6234 }, { "epoch": 0.28973048327137546, "grad_norm": 1.0774753153033578, "learning_rate": 8.338600767911683e-06, "loss": 0.0717, "step": 6235 }, { "epoch": 0.28977695167286244, "grad_norm": 1.1101855341466051, "learning_rate": 8.338040549860265e-06, "loss": 0.0578, "step": 6236 }, { "epoch": 0.28982342007434947, "grad_norm": 0.7763381026706846, "learning_rate": 8.337480256198718e-06, "loss": 0.0801, "step": 6237 }, { "epoch": 0.28986988847583645, "grad_norm": 2.9415778830105395, "learning_rate": 8.336919886939729e-06, "loss": 0.128, "step": 6238 }, { "epoch": 0.2899163568773234, "grad_norm": 1.4105463592816394, "learning_rate": 8.336359442095994e-06, "loss": 0.0922, "step": 6239 }, { "epoch": 0.2899628252788104, "grad_norm": 0.9116967319468582, "learning_rate": 8.335798921680204e-06, "loss": 0.0322, "step": 6240 }, { "epoch": 0.2900092936802974, "grad_norm": 0.6191794223827793, "learning_rate": 8.335238325705058e-06, "loss": 0.0409, "step": 6241 }, { "epoch": 0.2900557620817844, "grad_norm": 2.300248271925929, "learning_rate": 8.334677654183254e-06, "loss": 0.1038, "step": 6242 }, { "epoch": 0.2901022304832714, "grad_norm": 0.7090097946347336, "learning_rate": 8.334116907127491e-06, "loss": 0.0641, "step": 6243 }, { "epoch": 0.29014869888475836, "grad_norm": 1.4811532972107033, "learning_rate": 8.333556084550472e-06, "loss": 0.0745, "step": 6244 }, { "epoch": 0.29019516728624534, "grad_norm": 2.3609757513415084, "learning_rate": 8.3329951864649e-06, "loss": 0.0923, "step": 6245 }, { "epoch": 0.29024163568773237, "grad_norm": 0.7745000294051748, "learning_rate": 8.332434212883477e-06, "loss": 0.0752, "step": 6246 }, { "epoch": 0.29028810408921935, "grad_norm": 1.6789897043887891, "learning_rate": 8.331873163818913e-06, "loss": 0.0626, "step": 6247 }, { "epoch": 0.2903345724907063, "grad_norm": 1.0892714643350419, "learning_rate": 8.331312039283915e-06, "loss": 0.0608, "step": 6248 }, { "epoch": 0.2903810408921933, "grad_norm": 1.0513773393493366, "learning_rate": 8.330750839291193e-06, "loss": 0.0664, "step": 6249 }, { "epoch": 0.2904275092936803, "grad_norm": 0.8232976688684875, "learning_rate": 8.330189563853458e-06, "loss": 0.079, "step": 6250 }, { "epoch": 0.2904739776951673, "grad_norm": 1.7172055165385371, "learning_rate": 8.329628212983427e-06, "loss": 0.0594, "step": 6251 }, { "epoch": 0.2905204460966543, "grad_norm": 0.7453452389560468, "learning_rate": 8.32906678669381e-06, "loss": 0.0555, "step": 6252 }, { "epoch": 0.29056691449814126, "grad_norm": 0.9106079448349734, "learning_rate": 8.328505284997329e-06, "loss": 0.0604, "step": 6253 }, { "epoch": 0.29061338289962824, "grad_norm": 0.5963208223228307, "learning_rate": 8.327943707906698e-06, "loss": 0.0562, "step": 6254 }, { "epoch": 0.29065985130111527, "grad_norm": 0.9796213573456004, "learning_rate": 8.327382055434639e-06, "loss": 0.0692, "step": 6255 }, { "epoch": 0.29070631970260224, "grad_norm": 1.1209669950182606, "learning_rate": 8.326820327593874e-06, "loss": 0.042, "step": 6256 }, { "epoch": 0.2907527881040892, "grad_norm": 2.406052375024512, "learning_rate": 8.32625852439713e-06, "loss": 0.0745, "step": 6257 }, { "epoch": 0.2907992565055762, "grad_norm": 0.702598172572007, "learning_rate": 8.325696645857128e-06, "loss": 0.0476, "step": 6258 }, { "epoch": 0.2908457249070632, "grad_norm": 1.3873708004048597, "learning_rate": 8.325134691986598e-06, "loss": 0.0587, "step": 6259 }, { "epoch": 0.2908921933085502, "grad_norm": 1.265015931734346, "learning_rate": 8.324572662798267e-06, "loss": 0.0708, "step": 6260 }, { "epoch": 0.2909386617100372, "grad_norm": 1.1750210319096277, "learning_rate": 8.324010558304865e-06, "loss": 0.085, "step": 6261 }, { "epoch": 0.29098513011152416, "grad_norm": 0.9323386976580039, "learning_rate": 8.323448378519126e-06, "loss": 0.0681, "step": 6262 }, { "epoch": 0.29103159851301114, "grad_norm": 1.3556108108629654, "learning_rate": 8.322886123453784e-06, "loss": 0.0586, "step": 6263 }, { "epoch": 0.29107806691449817, "grad_norm": 1.000376711597098, "learning_rate": 8.322323793121571e-06, "loss": 0.0731, "step": 6264 }, { "epoch": 0.29112453531598514, "grad_norm": 1.8942453300262743, "learning_rate": 8.321761387535231e-06, "loss": 0.1144, "step": 6265 }, { "epoch": 0.2911710037174721, "grad_norm": 1.9457894520891108, "learning_rate": 8.321198906707498e-06, "loss": 0.0808, "step": 6266 }, { "epoch": 0.2912174721189591, "grad_norm": 1.456338333380479, "learning_rate": 8.320636350651115e-06, "loss": 0.0701, "step": 6267 }, { "epoch": 0.2912639405204461, "grad_norm": 1.2739469672762818, "learning_rate": 8.320073719378822e-06, "loss": 0.0851, "step": 6268 }, { "epoch": 0.2913104089219331, "grad_norm": 0.770106609662313, "learning_rate": 8.319511012903367e-06, "loss": 0.0668, "step": 6269 }, { "epoch": 0.2913568773234201, "grad_norm": 1.6606879062390505, "learning_rate": 8.318948231237493e-06, "loss": 0.1037, "step": 6270 }, { "epoch": 0.29140334572490706, "grad_norm": 1.0762300486011065, "learning_rate": 8.318385374393949e-06, "loss": 0.0551, "step": 6271 }, { "epoch": 0.29144981412639404, "grad_norm": 1.5051733991692948, "learning_rate": 8.317822442385483e-06, "loss": 0.1025, "step": 6272 }, { "epoch": 0.29149628252788107, "grad_norm": 2.240763140103894, "learning_rate": 8.317259435224848e-06, "loss": 0.0493, "step": 6273 }, { "epoch": 0.29154275092936804, "grad_norm": 0.8452926906870667, "learning_rate": 8.316696352924794e-06, "loss": 0.0938, "step": 6274 }, { "epoch": 0.291589219330855, "grad_norm": 2.1238460594020134, "learning_rate": 8.316133195498078e-06, "loss": 0.0595, "step": 6275 }, { "epoch": 0.291635687732342, "grad_norm": 0.8654212027992635, "learning_rate": 8.315569962957452e-06, "loss": 0.0482, "step": 6276 }, { "epoch": 0.291682156133829, "grad_norm": 0.6868738601033094, "learning_rate": 8.31500665531568e-06, "loss": 0.0567, "step": 6277 }, { "epoch": 0.291728624535316, "grad_norm": 0.7938755319314996, "learning_rate": 8.314443272585518e-06, "loss": 0.0518, "step": 6278 }, { "epoch": 0.291775092936803, "grad_norm": 1.0771160022322206, "learning_rate": 8.313879814779726e-06, "loss": 0.0539, "step": 6279 }, { "epoch": 0.29182156133828996, "grad_norm": 0.9361843644442904, "learning_rate": 8.313316281911068e-06, "loss": 0.069, "step": 6280 }, { "epoch": 0.29186802973977694, "grad_norm": 0.820960803561652, "learning_rate": 8.312752673992312e-06, "loss": 0.0539, "step": 6281 }, { "epoch": 0.29191449814126397, "grad_norm": 1.5165166863697812, "learning_rate": 8.31218899103622e-06, "loss": 0.1325, "step": 6282 }, { "epoch": 0.29196096654275094, "grad_norm": 1.6597868779482754, "learning_rate": 8.31162523305556e-06, "loss": 0.0748, "step": 6283 }, { "epoch": 0.2920074349442379, "grad_norm": 1.080447436755187, "learning_rate": 8.311061400063104e-06, "loss": 0.0412, "step": 6284 }, { "epoch": 0.2920539033457249, "grad_norm": 1.953331405211946, "learning_rate": 8.310497492071622e-06, "loss": 0.0795, "step": 6285 }, { "epoch": 0.2921003717472119, "grad_norm": 1.869650120252811, "learning_rate": 8.309933509093888e-06, "loss": 0.0775, "step": 6286 }, { "epoch": 0.2921468401486989, "grad_norm": 1.478113709313865, "learning_rate": 8.309369451142677e-06, "loss": 0.0632, "step": 6287 }, { "epoch": 0.2921933085501859, "grad_norm": 1.7498357719205888, "learning_rate": 8.308805318230764e-06, "loss": 0.1043, "step": 6288 }, { "epoch": 0.29223977695167286, "grad_norm": 0.911127146038045, "learning_rate": 8.308241110370927e-06, "loss": 0.0459, "step": 6289 }, { "epoch": 0.29228624535315983, "grad_norm": 5.000775900556824, "learning_rate": 8.30767682757595e-06, "loss": 0.1218, "step": 6290 }, { "epoch": 0.29233271375464687, "grad_norm": 1.040622630419069, "learning_rate": 8.307112469858608e-06, "loss": 0.0885, "step": 6291 }, { "epoch": 0.29237918215613384, "grad_norm": 1.5385364506116455, "learning_rate": 8.306548037231691e-06, "loss": 0.0986, "step": 6292 }, { "epoch": 0.2924256505576208, "grad_norm": 0.7745660809939839, "learning_rate": 8.30598352970798e-06, "loss": 0.073, "step": 6293 }, { "epoch": 0.2924721189591078, "grad_norm": 1.0068387732512305, "learning_rate": 8.305418947300262e-06, "loss": 0.0725, "step": 6294 }, { "epoch": 0.2925185873605948, "grad_norm": 0.9219181991459952, "learning_rate": 8.304854290021326e-06, "loss": 0.086, "step": 6295 }, { "epoch": 0.2925650557620818, "grad_norm": 0.6721401835728393, "learning_rate": 8.304289557883962e-06, "loss": 0.0424, "step": 6296 }, { "epoch": 0.2926115241635688, "grad_norm": 0.6287488571272783, "learning_rate": 8.303724750900961e-06, "loss": 0.0607, "step": 6297 }, { "epoch": 0.29265799256505576, "grad_norm": 0.9486244515169198, "learning_rate": 8.30315986908512e-06, "loss": 0.0615, "step": 6298 }, { "epoch": 0.29270446096654273, "grad_norm": 0.9396793398415443, "learning_rate": 8.302594912449228e-06, "loss": 0.0582, "step": 6299 }, { "epoch": 0.29275092936802977, "grad_norm": 1.013206262713589, "learning_rate": 8.302029881006089e-06, "loss": 0.057, "step": 6300 }, { "epoch": 0.29279739776951674, "grad_norm": 1.562168045565516, "learning_rate": 8.301464774768496e-06, "loss": 0.0712, "step": 6301 }, { "epoch": 0.2928438661710037, "grad_norm": 0.8952222547938159, "learning_rate": 8.30089959374925e-06, "loss": 0.0856, "step": 6302 }, { "epoch": 0.2928903345724907, "grad_norm": 0.873846682087961, "learning_rate": 8.300334337961154e-06, "loss": 0.0691, "step": 6303 }, { "epoch": 0.2929368029739777, "grad_norm": 1.3030313598898742, "learning_rate": 8.299769007417013e-06, "loss": 0.1397, "step": 6304 }, { "epoch": 0.2929832713754647, "grad_norm": 0.5330293531088134, "learning_rate": 8.29920360212963e-06, "loss": 0.0473, "step": 6305 }, { "epoch": 0.2930297397769517, "grad_norm": 1.45069040741519, "learning_rate": 8.298638122111815e-06, "loss": 0.0668, "step": 6306 }, { "epoch": 0.29307620817843866, "grad_norm": 1.4580988268495931, "learning_rate": 8.298072567376374e-06, "loss": 0.0846, "step": 6307 }, { "epoch": 0.29312267657992563, "grad_norm": 0.9209992116158313, "learning_rate": 8.297506937936118e-06, "loss": 0.0466, "step": 6308 }, { "epoch": 0.29316914498141267, "grad_norm": 0.3512052494326205, "learning_rate": 8.296941233803859e-06, "loss": 0.0331, "step": 6309 }, { "epoch": 0.29321561338289964, "grad_norm": 0.6004314002392723, "learning_rate": 8.296375454992412e-06, "loss": 0.0445, "step": 6310 }, { "epoch": 0.2932620817843866, "grad_norm": 1.110168701579223, "learning_rate": 8.29580960151459e-06, "loss": 0.0666, "step": 6311 }, { "epoch": 0.2933085501858736, "grad_norm": 1.777513099492694, "learning_rate": 8.295243673383213e-06, "loss": 0.1104, "step": 6312 }, { "epoch": 0.29335501858736057, "grad_norm": 4.463949191340277, "learning_rate": 8.294677670611098e-06, "loss": 0.0988, "step": 6313 }, { "epoch": 0.2934014869888476, "grad_norm": 1.5716915198368318, "learning_rate": 8.294111593211065e-06, "loss": 0.0908, "step": 6314 }, { "epoch": 0.2934479553903346, "grad_norm": 0.8893309953019488, "learning_rate": 8.293545441195941e-06, "loss": 0.0511, "step": 6315 }, { "epoch": 0.29349442379182156, "grad_norm": 1.611556780238864, "learning_rate": 8.292979214578542e-06, "loss": 0.0839, "step": 6316 }, { "epoch": 0.29354089219330853, "grad_norm": 1.148395124116483, "learning_rate": 8.292412913371702e-06, "loss": 0.0931, "step": 6317 }, { "epoch": 0.29358736059479557, "grad_norm": 2.057659499661605, "learning_rate": 8.291846537588242e-06, "loss": 0.0995, "step": 6318 }, { "epoch": 0.29363382899628254, "grad_norm": 1.5223947968389073, "learning_rate": 8.291280087240994e-06, "loss": 0.0907, "step": 6319 }, { "epoch": 0.2936802973977695, "grad_norm": 2.0885301026232517, "learning_rate": 8.290713562342787e-06, "loss": 0.0907, "step": 6320 }, { "epoch": 0.2937267657992565, "grad_norm": 1.3870093236401375, "learning_rate": 8.290146962906456e-06, "loss": 0.0809, "step": 6321 }, { "epoch": 0.29377323420074347, "grad_norm": 0.7840112391190254, "learning_rate": 8.289580288944832e-06, "loss": 0.0832, "step": 6322 }, { "epoch": 0.2938197026022305, "grad_norm": 1.2291261078249838, "learning_rate": 8.289013540470753e-06, "loss": 0.0797, "step": 6323 }, { "epoch": 0.2938661710037175, "grad_norm": 1.428753355451174, "learning_rate": 8.288446717497057e-06, "loss": 0.0648, "step": 6324 }, { "epoch": 0.29391263940520446, "grad_norm": 0.9047995510624938, "learning_rate": 8.287879820036579e-06, "loss": 0.0731, "step": 6325 }, { "epoch": 0.29395910780669143, "grad_norm": 1.9032505287552832, "learning_rate": 8.287312848102164e-06, "loss": 0.0931, "step": 6326 }, { "epoch": 0.29400557620817847, "grad_norm": 5.2813790143632335, "learning_rate": 8.286745801706652e-06, "loss": 0.0643, "step": 6327 }, { "epoch": 0.29405204460966544, "grad_norm": 0.8128239031608002, "learning_rate": 8.286178680862891e-06, "loss": 0.0628, "step": 6328 }, { "epoch": 0.2940985130111524, "grad_norm": 1.8442965731763015, "learning_rate": 8.285611485583722e-06, "loss": 0.083, "step": 6329 }, { "epoch": 0.2941449814126394, "grad_norm": 1.0107390429946594, "learning_rate": 8.285044215881996e-06, "loss": 0.0612, "step": 6330 }, { "epoch": 0.29419144981412637, "grad_norm": 1.111526556777749, "learning_rate": 8.28447687177056e-06, "loss": 0.0756, "step": 6331 }, { "epoch": 0.2942379182156134, "grad_norm": 1.2653964917414042, "learning_rate": 8.283909453262267e-06, "loss": 0.0764, "step": 6332 }, { "epoch": 0.2942843866171004, "grad_norm": 1.2784189738332465, "learning_rate": 8.283341960369968e-06, "loss": 0.0781, "step": 6333 }, { "epoch": 0.29433085501858736, "grad_norm": 1.0891336414473969, "learning_rate": 8.282774393106519e-06, "loss": 0.0689, "step": 6334 }, { "epoch": 0.29437732342007433, "grad_norm": 0.9495028885831533, "learning_rate": 8.282206751484774e-06, "loss": 0.0672, "step": 6335 }, { "epoch": 0.29442379182156136, "grad_norm": 1.9033918035639161, "learning_rate": 8.281639035517591e-06, "loss": 0.0925, "step": 6336 }, { "epoch": 0.29447026022304834, "grad_norm": 0.5993476947386401, "learning_rate": 8.28107124521783e-06, "loss": 0.0412, "step": 6337 }, { "epoch": 0.2945167286245353, "grad_norm": 2.9032735341206752, "learning_rate": 8.280503380598354e-06, "loss": 0.0773, "step": 6338 }, { "epoch": 0.2945631970260223, "grad_norm": 1.8265146524645848, "learning_rate": 8.279935441672023e-06, "loss": 0.0894, "step": 6339 }, { "epoch": 0.29460966542750927, "grad_norm": 1.6041496110533635, "learning_rate": 8.279367428451703e-06, "loss": 0.0837, "step": 6340 }, { "epoch": 0.2946561338289963, "grad_norm": 0.9363370822599607, "learning_rate": 8.278799340950257e-06, "loss": 0.0869, "step": 6341 }, { "epoch": 0.2947026022304833, "grad_norm": 0.7847357998706914, "learning_rate": 8.278231179180556e-06, "loss": 0.0557, "step": 6342 }, { "epoch": 0.29474907063197026, "grad_norm": 1.0089812836434098, "learning_rate": 8.277662943155469e-06, "loss": 0.0413, "step": 6343 }, { "epoch": 0.29479553903345723, "grad_norm": 0.8312750137791368, "learning_rate": 8.277094632887866e-06, "loss": 0.0625, "step": 6344 }, { "epoch": 0.29484200743494426, "grad_norm": 1.3234805583228788, "learning_rate": 8.27652624839062e-06, "loss": 0.073, "step": 6345 }, { "epoch": 0.29488847583643124, "grad_norm": 2.8604182886523013, "learning_rate": 8.275957789676607e-06, "loss": 0.0959, "step": 6346 }, { "epoch": 0.2949349442379182, "grad_norm": 1.3433934337343365, "learning_rate": 8.2753892567587e-06, "loss": 0.1213, "step": 6347 }, { "epoch": 0.2949814126394052, "grad_norm": 0.9905663434147388, "learning_rate": 8.274820649649783e-06, "loss": 0.0662, "step": 6348 }, { "epoch": 0.29502788104089217, "grad_norm": 0.7435345271798046, "learning_rate": 8.27425196836273e-06, "loss": 0.0418, "step": 6349 }, { "epoch": 0.2950743494423792, "grad_norm": 1.3013937507084312, "learning_rate": 8.273683212910422e-06, "loss": 0.1259, "step": 6350 }, { "epoch": 0.2951208178438662, "grad_norm": 1.1793013928419758, "learning_rate": 8.273114383305747e-06, "loss": 0.0827, "step": 6351 }, { "epoch": 0.29516728624535316, "grad_norm": 2.317649584567659, "learning_rate": 8.272545479561585e-06, "loss": 0.1018, "step": 6352 }, { "epoch": 0.29521375464684013, "grad_norm": 1.657559975450621, "learning_rate": 8.271976501690822e-06, "loss": 0.0868, "step": 6353 }, { "epoch": 0.29526022304832716, "grad_norm": 1.5805954023846855, "learning_rate": 8.271407449706347e-06, "loss": 0.0755, "step": 6354 }, { "epoch": 0.29530669144981414, "grad_norm": 0.8142670050344742, "learning_rate": 8.270838323621053e-06, "loss": 0.0725, "step": 6355 }, { "epoch": 0.2953531598513011, "grad_norm": 5.600675110602105, "learning_rate": 8.270269123447826e-06, "loss": 0.0876, "step": 6356 }, { "epoch": 0.2953996282527881, "grad_norm": 0.8000202502270813, "learning_rate": 8.269699849199565e-06, "loss": 0.0437, "step": 6357 }, { "epoch": 0.29544609665427507, "grad_norm": 0.8674298311045202, "learning_rate": 8.269130500889158e-06, "loss": 0.0502, "step": 6358 }, { "epoch": 0.2954925650557621, "grad_norm": 1.3071298633913075, "learning_rate": 8.268561078529504e-06, "loss": 0.1035, "step": 6359 }, { "epoch": 0.2955390334572491, "grad_norm": 1.1059505196460337, "learning_rate": 8.267991582133503e-06, "loss": 0.0572, "step": 6360 }, { "epoch": 0.29558550185873605, "grad_norm": 0.8762572515316337, "learning_rate": 8.267422011714054e-06, "loss": 0.0356, "step": 6361 }, { "epoch": 0.29563197026022303, "grad_norm": 0.8541089931856145, "learning_rate": 8.266852367284056e-06, "loss": 0.0658, "step": 6362 }, { "epoch": 0.29567843866171006, "grad_norm": 2.0398589995229055, "learning_rate": 8.266282648856414e-06, "loss": 0.0624, "step": 6363 }, { "epoch": 0.29572490706319704, "grad_norm": 0.5610154040390282, "learning_rate": 8.265712856444033e-06, "loss": 0.048, "step": 6364 }, { "epoch": 0.295771375464684, "grad_norm": 1.9150667494089337, "learning_rate": 8.265142990059818e-06, "loss": 0.07, "step": 6365 }, { "epoch": 0.295817843866171, "grad_norm": 1.0083066226267219, "learning_rate": 8.264573049716678e-06, "loss": 0.0677, "step": 6366 }, { "epoch": 0.29586431226765797, "grad_norm": 1.9635131457842887, "learning_rate": 8.264003035427521e-06, "loss": 0.0837, "step": 6367 }, { "epoch": 0.295910780669145, "grad_norm": 0.7225422843083932, "learning_rate": 8.263432947205263e-06, "loss": 0.0402, "step": 6368 }, { "epoch": 0.295957249070632, "grad_norm": 1.1618830292596538, "learning_rate": 8.26286278506281e-06, "loss": 0.141, "step": 6369 }, { "epoch": 0.29600371747211895, "grad_norm": 0.6348164956771276, "learning_rate": 8.262292549013084e-06, "loss": 0.0338, "step": 6370 }, { "epoch": 0.29605018587360593, "grad_norm": 1.0851293014385752, "learning_rate": 8.261722239068998e-06, "loss": 0.0653, "step": 6371 }, { "epoch": 0.29609665427509296, "grad_norm": 0.7954778238466851, "learning_rate": 8.26115185524347e-06, "loss": 0.0889, "step": 6372 }, { "epoch": 0.29614312267657994, "grad_norm": 1.8602850352413516, "learning_rate": 8.26058139754942e-06, "loss": 0.0785, "step": 6373 }, { "epoch": 0.2961895910780669, "grad_norm": 0.9209533112925632, "learning_rate": 8.26001086599977e-06, "loss": 0.0586, "step": 6374 }, { "epoch": 0.2962360594795539, "grad_norm": 6.10305830066129, "learning_rate": 8.259440260607442e-06, "loss": 0.051, "step": 6375 }, { "epoch": 0.29628252788104087, "grad_norm": 1.3436533255128462, "learning_rate": 8.258869581385361e-06, "loss": 0.0767, "step": 6376 }, { "epoch": 0.2963289962825279, "grad_norm": 1.6851518169218525, "learning_rate": 8.258298828346457e-06, "loss": 0.149, "step": 6377 }, { "epoch": 0.2963754646840149, "grad_norm": 1.0208976317491147, "learning_rate": 8.257728001503653e-06, "loss": 0.0761, "step": 6378 }, { "epoch": 0.29642193308550185, "grad_norm": 1.7984378354537394, "learning_rate": 8.257157100869881e-06, "loss": 0.1171, "step": 6379 }, { "epoch": 0.29646840148698883, "grad_norm": 0.6032444916944256, "learning_rate": 8.256586126458073e-06, "loss": 0.032, "step": 6380 }, { "epoch": 0.29651486988847586, "grad_norm": 8.433842636785728, "learning_rate": 8.256015078281162e-06, "loss": 0.1109, "step": 6381 }, { "epoch": 0.29656133828996284, "grad_norm": 0.8074361804832794, "learning_rate": 8.25544395635208e-06, "loss": 0.0724, "step": 6382 }, { "epoch": 0.2966078066914498, "grad_norm": 1.082658714065157, "learning_rate": 8.25487276068377e-06, "loss": 0.0432, "step": 6383 }, { "epoch": 0.2966542750929368, "grad_norm": 1.6499922979074193, "learning_rate": 8.254301491289165e-06, "loss": 0.0726, "step": 6384 }, { "epoch": 0.29670074349442377, "grad_norm": 1.4140056189793266, "learning_rate": 8.253730148181205e-06, "loss": 0.0989, "step": 6385 }, { "epoch": 0.2967472118959108, "grad_norm": 0.8307505839940594, "learning_rate": 8.253158731372834e-06, "loss": 0.0515, "step": 6386 }, { "epoch": 0.2967936802973978, "grad_norm": 1.230616349720276, "learning_rate": 8.252587240876992e-06, "loss": 0.0556, "step": 6387 }, { "epoch": 0.29684014869888475, "grad_norm": 1.0073125520896842, "learning_rate": 8.252015676706628e-06, "loss": 0.0736, "step": 6388 }, { "epoch": 0.29688661710037173, "grad_norm": 0.6098634570866261, "learning_rate": 8.251444038874685e-06, "loss": 0.052, "step": 6389 }, { "epoch": 0.29693308550185876, "grad_norm": 0.492159444013487, "learning_rate": 8.250872327394112e-06, "loss": 0.0574, "step": 6390 }, { "epoch": 0.29697955390334574, "grad_norm": 2.0667400280005115, "learning_rate": 8.25030054227786e-06, "loss": 0.1531, "step": 6391 }, { "epoch": 0.2970260223048327, "grad_norm": 2.4931721737914283, "learning_rate": 8.24972868353888e-06, "loss": 0.0834, "step": 6392 }, { "epoch": 0.2970724907063197, "grad_norm": 1.771988139521448, "learning_rate": 8.249156751190126e-06, "loss": 0.0643, "step": 6393 }, { "epoch": 0.29711895910780667, "grad_norm": 1.0261342090380583, "learning_rate": 8.24858474524455e-06, "loss": 0.0627, "step": 6394 }, { "epoch": 0.2971654275092937, "grad_norm": 1.057738899961675, "learning_rate": 8.24801266571511e-06, "loss": 0.0794, "step": 6395 }, { "epoch": 0.2972118959107807, "grad_norm": 0.5779078360602435, "learning_rate": 8.247440512614768e-06, "loss": 0.0369, "step": 6396 }, { "epoch": 0.29725836431226765, "grad_norm": 1.3269953019999194, "learning_rate": 8.246868285956478e-06, "loss": 0.0872, "step": 6397 }, { "epoch": 0.29730483271375463, "grad_norm": 0.8034821247001315, "learning_rate": 8.246295985753204e-06, "loss": 0.0568, "step": 6398 }, { "epoch": 0.29735130111524166, "grad_norm": 0.6710172213981003, "learning_rate": 8.24572361201791e-06, "loss": 0.0574, "step": 6399 }, { "epoch": 0.29739776951672864, "grad_norm": 1.045279922310942, "learning_rate": 8.24515116476356e-06, "loss": 0.0741, "step": 6400 }, { "epoch": 0.2974442379182156, "grad_norm": 1.4139920252593432, "learning_rate": 8.244578644003121e-06, "loss": 0.0714, "step": 6401 }, { "epoch": 0.2974907063197026, "grad_norm": 0.8568536690246121, "learning_rate": 8.24400604974956e-06, "loss": 0.056, "step": 6402 }, { "epoch": 0.29753717472118957, "grad_norm": 0.9545544921409695, "learning_rate": 8.243433382015849e-06, "loss": 0.052, "step": 6403 }, { "epoch": 0.2975836431226766, "grad_norm": 1.3353801724897763, "learning_rate": 8.242860640814956e-06, "loss": 0.0574, "step": 6404 }, { "epoch": 0.2976301115241636, "grad_norm": 0.6370688439920127, "learning_rate": 8.242287826159858e-06, "loss": 0.03, "step": 6405 }, { "epoch": 0.29767657992565055, "grad_norm": 5.29684861017642, "learning_rate": 8.241714938063528e-06, "loss": 0.0971, "step": 6406 }, { "epoch": 0.29772304832713753, "grad_norm": 1.3447741609637696, "learning_rate": 8.241141976538944e-06, "loss": 0.0496, "step": 6407 }, { "epoch": 0.29776951672862456, "grad_norm": 0.7135952782612652, "learning_rate": 8.240568941599081e-06, "loss": 0.0544, "step": 6408 }, { "epoch": 0.29781598513011154, "grad_norm": 2.332359722439175, "learning_rate": 8.239995833256922e-06, "loss": 0.113, "step": 6409 }, { "epoch": 0.2978624535315985, "grad_norm": 3.247009736967328, "learning_rate": 8.239422651525445e-06, "loss": 0.1082, "step": 6410 }, { "epoch": 0.2979089219330855, "grad_norm": 1.1806129333599003, "learning_rate": 8.238849396417639e-06, "loss": 0.0825, "step": 6411 }, { "epoch": 0.29795539033457247, "grad_norm": 3.047190431198511, "learning_rate": 8.238276067946484e-06, "loss": 0.1019, "step": 6412 }, { "epoch": 0.2980018587360595, "grad_norm": 0.7134754257579167, "learning_rate": 8.237702666124966e-06, "loss": 0.0528, "step": 6413 }, { "epoch": 0.2980483271375465, "grad_norm": 4.499909417797654, "learning_rate": 8.237129190966077e-06, "loss": 0.0978, "step": 6414 }, { "epoch": 0.29809479553903345, "grad_norm": 0.921091551506236, "learning_rate": 8.236555642482807e-06, "loss": 0.0827, "step": 6415 }, { "epoch": 0.29814126394052043, "grad_norm": 2.5375382059121843, "learning_rate": 8.23598202068814e-06, "loss": 0.0559, "step": 6416 }, { "epoch": 0.29818773234200746, "grad_norm": 0.6585930308142449, "learning_rate": 8.235408325595078e-06, "loss": 0.0604, "step": 6417 }, { "epoch": 0.29823420074349444, "grad_norm": 1.115641399782438, "learning_rate": 8.23483455721661e-06, "loss": 0.0673, "step": 6418 }, { "epoch": 0.2982806691449814, "grad_norm": 0.824920672904853, "learning_rate": 8.234260715565738e-06, "loss": 0.0378, "step": 6419 }, { "epoch": 0.2983271375464684, "grad_norm": 0.7314986065213447, "learning_rate": 8.233686800655456e-06, "loss": 0.0402, "step": 6420 }, { "epoch": 0.29837360594795537, "grad_norm": 1.3580315049328198, "learning_rate": 8.233112812498762e-06, "loss": 0.0768, "step": 6421 }, { "epoch": 0.2984200743494424, "grad_norm": 0.9727455251819609, "learning_rate": 8.232538751108663e-06, "loss": 0.099, "step": 6422 }, { "epoch": 0.2984665427509294, "grad_norm": 0.9520067317893154, "learning_rate": 8.231964616498157e-06, "loss": 0.0538, "step": 6423 }, { "epoch": 0.29851301115241635, "grad_norm": 0.6651867366869353, "learning_rate": 8.231390408680251e-06, "loss": 0.0336, "step": 6424 }, { "epoch": 0.29855947955390333, "grad_norm": 3.4423779984272516, "learning_rate": 8.230816127667951e-06, "loss": 0.112, "step": 6425 }, { "epoch": 0.29860594795539036, "grad_norm": 0.9204317839597235, "learning_rate": 8.230241773474267e-06, "loss": 0.0701, "step": 6426 }, { "epoch": 0.29865241635687734, "grad_norm": 1.416185755255962, "learning_rate": 8.229667346112207e-06, "loss": 0.1018, "step": 6427 }, { "epoch": 0.2986988847583643, "grad_norm": 1.4269591740394103, "learning_rate": 8.22909284559478e-06, "loss": 0.0829, "step": 6428 }, { "epoch": 0.2987453531598513, "grad_norm": 0.8155296565649418, "learning_rate": 8.228518271935005e-06, "loss": 0.0587, "step": 6429 }, { "epoch": 0.29879182156133827, "grad_norm": 1.4162453918961762, "learning_rate": 8.22794362514589e-06, "loss": 0.0935, "step": 6430 }, { "epoch": 0.2988382899628253, "grad_norm": 1.853664983965679, "learning_rate": 8.227368905240456e-06, "loss": 0.1114, "step": 6431 }, { "epoch": 0.2988847583643123, "grad_norm": 1.1940776805886884, "learning_rate": 8.226794112231719e-06, "loss": 0.0663, "step": 6432 }, { "epoch": 0.29893122676579925, "grad_norm": 1.260331754872845, "learning_rate": 8.2262192461327e-06, "loss": 0.0759, "step": 6433 }, { "epoch": 0.29897769516728623, "grad_norm": 1.7632725891895096, "learning_rate": 8.225644306956419e-06, "loss": 0.0655, "step": 6434 }, { "epoch": 0.29902416356877326, "grad_norm": 1.4092371555137615, "learning_rate": 8.225069294715899e-06, "loss": 0.0974, "step": 6435 }, { "epoch": 0.29907063197026024, "grad_norm": 1.6812193528951587, "learning_rate": 8.224494209424166e-06, "loss": 0.0804, "step": 6436 }, { "epoch": 0.2991171003717472, "grad_norm": 1.4015431477056155, "learning_rate": 8.223919051094245e-06, "loss": 0.0754, "step": 6437 }, { "epoch": 0.2991635687732342, "grad_norm": 1.5064477536903789, "learning_rate": 8.223343819739164e-06, "loss": 0.0998, "step": 6438 }, { "epoch": 0.29921003717472117, "grad_norm": 1.087316167268304, "learning_rate": 8.222768515371955e-06, "loss": 0.0659, "step": 6439 }, { "epoch": 0.2992565055762082, "grad_norm": 0.6905623145063825, "learning_rate": 8.222193138005645e-06, "loss": 0.0644, "step": 6440 }, { "epoch": 0.2993029739776952, "grad_norm": 2.755282091401543, "learning_rate": 8.221617687653269e-06, "loss": 0.0835, "step": 6441 }, { "epoch": 0.29934944237918215, "grad_norm": 1.4514082813651177, "learning_rate": 8.221042164327864e-06, "loss": 0.0574, "step": 6442 }, { "epoch": 0.29939591078066913, "grad_norm": 0.8315769953444395, "learning_rate": 8.220466568042462e-06, "loss": 0.0636, "step": 6443 }, { "epoch": 0.29944237918215616, "grad_norm": 1.1810981344025087, "learning_rate": 8.219890898810103e-06, "loss": 0.0693, "step": 6444 }, { "epoch": 0.29948884758364314, "grad_norm": 1.15222906965011, "learning_rate": 8.219315156643826e-06, "loss": 0.0913, "step": 6445 }, { "epoch": 0.2995353159851301, "grad_norm": 2.589554039565773, "learning_rate": 8.218739341556674e-06, "loss": 0.0881, "step": 6446 }, { "epoch": 0.2995817843866171, "grad_norm": 2.157278828359645, "learning_rate": 8.218163453561687e-06, "loss": 0.1045, "step": 6447 }, { "epoch": 0.29962825278810407, "grad_norm": 1.079544937572353, "learning_rate": 8.21758749267191e-06, "loss": 0.0606, "step": 6448 }, { "epoch": 0.2996747211895911, "grad_norm": 0.7914175402869074, "learning_rate": 8.217011458900392e-06, "loss": 0.0598, "step": 6449 }, { "epoch": 0.2997211895910781, "grad_norm": 1.7264941246684098, "learning_rate": 8.216435352260176e-06, "loss": 0.0824, "step": 6450 }, { "epoch": 0.29976765799256505, "grad_norm": 1.3018828248457324, "learning_rate": 8.215859172764316e-06, "loss": 0.0805, "step": 6451 }, { "epoch": 0.299814126394052, "grad_norm": 0.820336309028662, "learning_rate": 8.215282920425862e-06, "loss": 0.0705, "step": 6452 }, { "epoch": 0.29986059479553906, "grad_norm": 1.663025448685554, "learning_rate": 8.214706595257863e-06, "loss": 0.0933, "step": 6453 }, { "epoch": 0.29990706319702604, "grad_norm": 0.7219321218612043, "learning_rate": 8.214130197273377e-06, "loss": 0.0588, "step": 6454 }, { "epoch": 0.299953531598513, "grad_norm": 4.838486700391151, "learning_rate": 8.213553726485461e-06, "loss": 0.1501, "step": 6455 }, { "epoch": 0.3, "grad_norm": 1.0432180024606206, "learning_rate": 8.21297718290717e-06, "loss": 0.0537, "step": 6456 }, { "epoch": 0.30004646840148697, "grad_norm": 0.9340554508262975, "learning_rate": 8.212400566551563e-06, "loss": 0.0639, "step": 6457 }, { "epoch": 0.300092936802974, "grad_norm": 2.2784143103972174, "learning_rate": 8.211823877431704e-06, "loss": 0.0745, "step": 6458 }, { "epoch": 0.300139405204461, "grad_norm": 0.6042069345399633, "learning_rate": 8.211247115560651e-06, "loss": 0.0397, "step": 6459 }, { "epoch": 0.30018587360594795, "grad_norm": 1.1076527071912625, "learning_rate": 8.210670280951474e-06, "loss": 0.0818, "step": 6460 }, { "epoch": 0.3002323420074349, "grad_norm": 1.0896231159296754, "learning_rate": 8.210093373617235e-06, "loss": 0.074, "step": 6461 }, { "epoch": 0.30027881040892196, "grad_norm": 0.7348752874887078, "learning_rate": 8.209516393571001e-06, "loss": 0.0566, "step": 6462 }, { "epoch": 0.30032527881040894, "grad_norm": 1.316305908720487, "learning_rate": 8.208939340825845e-06, "loss": 0.0864, "step": 6463 }, { "epoch": 0.3003717472118959, "grad_norm": 2.488976954770375, "learning_rate": 8.208362215394836e-06, "loss": 0.069, "step": 6464 }, { "epoch": 0.3004182156133829, "grad_norm": 1.1447647209736846, "learning_rate": 8.207785017291043e-06, "loss": 0.0653, "step": 6465 }, { "epoch": 0.30046468401486986, "grad_norm": 1.649000828390926, "learning_rate": 8.207207746527546e-06, "loss": 0.1042, "step": 6466 }, { "epoch": 0.3005111524163569, "grad_norm": 0.7430960984223018, "learning_rate": 8.206630403117417e-06, "loss": 0.0587, "step": 6467 }, { "epoch": 0.3005576208178439, "grad_norm": 2.4930167922100974, "learning_rate": 8.206052987073734e-06, "loss": 0.0871, "step": 6468 }, { "epoch": 0.30060408921933085, "grad_norm": 1.2139862322248631, "learning_rate": 8.205475498409578e-06, "loss": 0.0861, "step": 6469 }, { "epoch": 0.3006505576208178, "grad_norm": 1.2373024055846271, "learning_rate": 8.204897937138025e-06, "loss": 0.0708, "step": 6470 }, { "epoch": 0.30069702602230486, "grad_norm": 2.106879837730833, "learning_rate": 8.204320303272164e-06, "loss": 0.0477, "step": 6471 }, { "epoch": 0.30074349442379184, "grad_norm": 0.83005876189433, "learning_rate": 8.203742596825075e-06, "loss": 0.0546, "step": 6472 }, { "epoch": 0.3007899628252788, "grad_norm": 1.0825158714593257, "learning_rate": 8.203164817809843e-06, "loss": 0.0902, "step": 6473 }, { "epoch": 0.3008364312267658, "grad_norm": 1.8659241462601224, "learning_rate": 8.202586966239558e-06, "loss": 0.0907, "step": 6474 }, { "epoch": 0.30088289962825276, "grad_norm": 1.9043185799529692, "learning_rate": 8.202009042127307e-06, "loss": 0.1043, "step": 6475 }, { "epoch": 0.3009293680297398, "grad_norm": 0.6435360492079586, "learning_rate": 8.201431045486181e-06, "loss": 0.0441, "step": 6476 }, { "epoch": 0.3009758364312268, "grad_norm": 0.8671766990296872, "learning_rate": 8.200852976329272e-06, "loss": 0.0645, "step": 6477 }, { "epoch": 0.30102230483271375, "grad_norm": 5.296877651817096, "learning_rate": 8.200274834669675e-06, "loss": 0.0498, "step": 6478 }, { "epoch": 0.3010687732342007, "grad_norm": 1.184813574259628, "learning_rate": 8.199696620520485e-06, "loss": 0.0825, "step": 6479 }, { "epoch": 0.30111524163568776, "grad_norm": 2.4815248417056535, "learning_rate": 8.199118333894797e-06, "loss": 0.1001, "step": 6480 }, { "epoch": 0.30116171003717473, "grad_norm": 0.8104271669933637, "learning_rate": 8.198539974805715e-06, "loss": 0.0596, "step": 6481 }, { "epoch": 0.3012081784386617, "grad_norm": 1.3881628393440113, "learning_rate": 8.197961543266334e-06, "loss": 0.0708, "step": 6482 }, { "epoch": 0.3012546468401487, "grad_norm": 1.0758408528497987, "learning_rate": 8.19738303928976e-06, "loss": 0.0868, "step": 6483 }, { "epoch": 0.30130111524163566, "grad_norm": 1.4839128772407757, "learning_rate": 8.196804462889093e-06, "loss": 0.079, "step": 6484 }, { "epoch": 0.3013475836431227, "grad_norm": 1.0890020564663263, "learning_rate": 8.196225814077441e-06, "loss": 0.0634, "step": 6485 }, { "epoch": 0.3013940520446097, "grad_norm": 3.1014366065801027, "learning_rate": 8.195647092867908e-06, "loss": 0.0934, "step": 6486 }, { "epoch": 0.30144052044609665, "grad_norm": 2.3654507652906216, "learning_rate": 8.19506829927361e-06, "loss": 0.0731, "step": 6487 }, { "epoch": 0.3014869888475836, "grad_norm": 1.3187412299785708, "learning_rate": 8.194489433307648e-06, "loss": 0.0728, "step": 6488 }, { "epoch": 0.30153345724907066, "grad_norm": 1.167843203858547, "learning_rate": 8.193910494983139e-06, "loss": 0.0389, "step": 6489 }, { "epoch": 0.30157992565055763, "grad_norm": 1.1665626313557855, "learning_rate": 8.193331484313198e-06, "loss": 0.0763, "step": 6490 }, { "epoch": 0.3016263940520446, "grad_norm": 0.7263755986827647, "learning_rate": 8.192752401310937e-06, "loss": 0.0587, "step": 6491 }, { "epoch": 0.3016728624535316, "grad_norm": 7.514725171801346, "learning_rate": 8.192173245989473e-06, "loss": 0.0638, "step": 6492 }, { "epoch": 0.30171933085501856, "grad_norm": 0.9156624653718548, "learning_rate": 8.191594018361924e-06, "loss": 0.044, "step": 6493 }, { "epoch": 0.3017657992565056, "grad_norm": 1.1733341182171635, "learning_rate": 8.191014718441413e-06, "loss": 0.0714, "step": 6494 }, { "epoch": 0.3018122676579926, "grad_norm": 1.196735016733782, "learning_rate": 8.190435346241061e-06, "loss": 0.0622, "step": 6495 }, { "epoch": 0.30185873605947955, "grad_norm": 3.8049018577663287, "learning_rate": 8.18985590177399e-06, "loss": 0.0723, "step": 6496 }, { "epoch": 0.3019052044609665, "grad_norm": 1.4374763804245985, "learning_rate": 8.189276385053325e-06, "loss": 0.0905, "step": 6497 }, { "epoch": 0.30195167286245356, "grad_norm": 0.9571501597800387, "learning_rate": 8.188696796092192e-06, "loss": 0.0465, "step": 6498 }, { "epoch": 0.30199814126394053, "grad_norm": 0.5165699298015665, "learning_rate": 8.188117134903722e-06, "loss": 0.0576, "step": 6499 }, { "epoch": 0.3020446096654275, "grad_norm": 0.9809044165992863, "learning_rate": 8.187537401501044e-06, "loss": 0.0737, "step": 6500 }, { "epoch": 0.3020910780669145, "grad_norm": 1.4929215147982304, "learning_rate": 8.186957595897288e-06, "loss": 0.0435, "step": 6501 }, { "epoch": 0.30213754646840146, "grad_norm": 1.6274763233340181, "learning_rate": 8.186377718105589e-06, "loss": 0.0825, "step": 6502 }, { "epoch": 0.3021840148698885, "grad_norm": 1.3269624803642073, "learning_rate": 8.185797768139081e-06, "loss": 0.0762, "step": 6503 }, { "epoch": 0.30223048327137547, "grad_norm": 2.9570422203836846, "learning_rate": 8.1852177460109e-06, "loss": 0.0555, "step": 6504 }, { "epoch": 0.30227695167286245, "grad_norm": 1.2179738563529545, "learning_rate": 8.184637651734184e-06, "loss": 0.0552, "step": 6505 }, { "epoch": 0.3023234200743494, "grad_norm": 3.6992510117013215, "learning_rate": 8.184057485322075e-06, "loss": 0.0495, "step": 6506 }, { "epoch": 0.30236988847583646, "grad_norm": 1.6169971245080224, "learning_rate": 8.183477246787712e-06, "loss": 0.0807, "step": 6507 }, { "epoch": 0.30241635687732343, "grad_norm": 1.2083084184926092, "learning_rate": 8.18289693614424e-06, "loss": 0.0558, "step": 6508 }, { "epoch": 0.3024628252788104, "grad_norm": 0.9638547663121988, "learning_rate": 8.1823165534048e-06, "loss": 0.0511, "step": 6509 }, { "epoch": 0.3025092936802974, "grad_norm": 4.605507562776964, "learning_rate": 8.181736098582543e-06, "loss": 0.0811, "step": 6510 }, { "epoch": 0.30255576208178436, "grad_norm": 0.9823527405244302, "learning_rate": 8.181155571690613e-06, "loss": 0.0656, "step": 6511 }, { "epoch": 0.3026022304832714, "grad_norm": 0.7604120502915984, "learning_rate": 8.180574972742162e-06, "loss": 0.0521, "step": 6512 }, { "epoch": 0.30264869888475837, "grad_norm": 7.983718999350041, "learning_rate": 8.17999430175034e-06, "loss": 0.1094, "step": 6513 }, { "epoch": 0.30269516728624535, "grad_norm": 0.7796550126395693, "learning_rate": 8.179413558728301e-06, "loss": 0.0577, "step": 6514 }, { "epoch": 0.3027416356877323, "grad_norm": 1.6070988672955988, "learning_rate": 8.1788327436892e-06, "loss": 0.0894, "step": 6515 }, { "epoch": 0.30278810408921936, "grad_norm": 1.7589288570762853, "learning_rate": 8.178251856646189e-06, "loss": 0.0816, "step": 6516 }, { "epoch": 0.30283457249070633, "grad_norm": 3.0117903258902197, "learning_rate": 8.177670897612428e-06, "loss": 0.1119, "step": 6517 }, { "epoch": 0.3028810408921933, "grad_norm": 1.0865356942763553, "learning_rate": 8.177089866601079e-06, "loss": 0.0775, "step": 6518 }, { "epoch": 0.3029275092936803, "grad_norm": 1.7196367790788565, "learning_rate": 8.1765087636253e-06, "loss": 0.0659, "step": 6519 }, { "epoch": 0.30297397769516726, "grad_norm": 1.777944578161146, "learning_rate": 8.175927588698254e-06, "loss": 0.0826, "step": 6520 }, { "epoch": 0.3030204460966543, "grad_norm": 0.7743344362697195, "learning_rate": 8.175346341833106e-06, "loss": 0.033, "step": 6521 }, { "epoch": 0.30306691449814127, "grad_norm": 1.2222371983503126, "learning_rate": 8.17476502304302e-06, "loss": 0.0812, "step": 6522 }, { "epoch": 0.30311338289962825, "grad_norm": 0.4995522136733997, "learning_rate": 8.174183632341167e-06, "loss": 0.0557, "step": 6523 }, { "epoch": 0.3031598513011152, "grad_norm": 0.5434389387361577, "learning_rate": 8.173602169740711e-06, "loss": 0.0398, "step": 6524 }, { "epoch": 0.30320631970260226, "grad_norm": 0.4752061773295316, "learning_rate": 8.17302063525483e-06, "loss": 0.0323, "step": 6525 }, { "epoch": 0.30325278810408923, "grad_norm": 1.6605314677538905, "learning_rate": 8.172439028896688e-06, "loss": 0.0704, "step": 6526 }, { "epoch": 0.3032992565055762, "grad_norm": 4.106030847694552, "learning_rate": 8.171857350679465e-06, "loss": 0.0948, "step": 6527 }, { "epoch": 0.3033457249070632, "grad_norm": 1.0903459902172457, "learning_rate": 8.171275600616334e-06, "loss": 0.1073, "step": 6528 }, { "epoch": 0.30339219330855016, "grad_norm": 0.772911554254679, "learning_rate": 8.170693778720474e-06, "loss": 0.0644, "step": 6529 }, { "epoch": 0.3034386617100372, "grad_norm": 1.5426195526325641, "learning_rate": 8.170111885005062e-06, "loss": 0.1016, "step": 6530 }, { "epoch": 0.30348513011152417, "grad_norm": 2.1426945836207567, "learning_rate": 8.169529919483278e-06, "loss": 0.124, "step": 6531 }, { "epoch": 0.30353159851301115, "grad_norm": 1.0630242898874753, "learning_rate": 8.168947882168307e-06, "loss": 0.0855, "step": 6532 }, { "epoch": 0.3035780669144981, "grad_norm": 5.249412227094485, "learning_rate": 8.168365773073331e-06, "loss": 0.0784, "step": 6533 }, { "epoch": 0.30362453531598516, "grad_norm": 1.0881755036654137, "learning_rate": 8.167783592211535e-06, "loss": 0.08, "step": 6534 }, { "epoch": 0.30367100371747213, "grad_norm": 0.946483978455298, "learning_rate": 8.167201339596106e-06, "loss": 0.0602, "step": 6535 }, { "epoch": 0.3037174721189591, "grad_norm": 0.5704185957406063, "learning_rate": 8.166619015240236e-06, "loss": 0.0543, "step": 6536 }, { "epoch": 0.3037639405204461, "grad_norm": 1.1878178576986134, "learning_rate": 8.166036619157111e-06, "loss": 0.0539, "step": 6537 }, { "epoch": 0.30381040892193306, "grad_norm": 0.7778460249257585, "learning_rate": 8.165454151359923e-06, "loss": 0.03, "step": 6538 }, { "epoch": 0.3038568773234201, "grad_norm": 2.577907719011537, "learning_rate": 8.164871611861869e-06, "loss": 0.0801, "step": 6539 }, { "epoch": 0.30390334572490707, "grad_norm": 1.6487736943157822, "learning_rate": 8.164289000676141e-06, "loss": 0.0797, "step": 6540 }, { "epoch": 0.30394981412639405, "grad_norm": 1.3909813783130314, "learning_rate": 8.163706317815938e-06, "loss": 0.0551, "step": 6541 }, { "epoch": 0.303996282527881, "grad_norm": 0.6694589267112762, "learning_rate": 8.163123563294455e-06, "loss": 0.048, "step": 6542 }, { "epoch": 0.30404275092936806, "grad_norm": 0.8204273521022089, "learning_rate": 8.162540737124896e-06, "loss": 0.0546, "step": 6543 }, { "epoch": 0.30408921933085503, "grad_norm": 2.861090248889288, "learning_rate": 8.16195783932046e-06, "loss": 0.0642, "step": 6544 }, { "epoch": 0.304135687732342, "grad_norm": 2.277802590200837, "learning_rate": 8.161374869894351e-06, "loss": 0.0734, "step": 6545 }, { "epoch": 0.304182156133829, "grad_norm": 1.657928591599131, "learning_rate": 8.160791828859774e-06, "loss": 0.1025, "step": 6546 }, { "epoch": 0.30422862453531596, "grad_norm": 3.831686593912329, "learning_rate": 8.160208716229936e-06, "loss": 0.0682, "step": 6547 }, { "epoch": 0.304275092936803, "grad_norm": 2.2600100382738124, "learning_rate": 8.159625532018042e-06, "loss": 0.1327, "step": 6548 }, { "epoch": 0.30432156133828997, "grad_norm": 2.398276359375918, "learning_rate": 8.159042276237308e-06, "loss": 0.0804, "step": 6549 }, { "epoch": 0.30436802973977695, "grad_norm": 1.1960367285755258, "learning_rate": 8.15845894890094e-06, "loss": 0.0518, "step": 6550 }, { "epoch": 0.3044144981412639, "grad_norm": 1.3639050657199603, "learning_rate": 8.157875550022152e-06, "loss": 0.0924, "step": 6551 }, { "epoch": 0.30446096654275095, "grad_norm": 4.796078367773688, "learning_rate": 8.15729207961416e-06, "loss": 0.0758, "step": 6552 }, { "epoch": 0.30450743494423793, "grad_norm": 0.8758175977067595, "learning_rate": 8.156708537690179e-06, "loss": 0.0625, "step": 6553 }, { "epoch": 0.3045539033457249, "grad_norm": 0.9730419317255952, "learning_rate": 8.156124924263426e-06, "loss": 0.0373, "step": 6554 }, { "epoch": 0.3046003717472119, "grad_norm": 2.2686062948084174, "learning_rate": 8.155541239347124e-06, "loss": 0.08, "step": 6555 }, { "epoch": 0.30464684014869886, "grad_norm": 1.2312117442207051, "learning_rate": 8.154957482954491e-06, "loss": 0.069, "step": 6556 }, { "epoch": 0.3046933085501859, "grad_norm": 1.2061179634459749, "learning_rate": 8.15437365509875e-06, "loss": 0.0701, "step": 6557 }, { "epoch": 0.30473977695167287, "grad_norm": 2.204783921502486, "learning_rate": 8.153789755793126e-06, "loss": 0.0782, "step": 6558 }, { "epoch": 0.30478624535315985, "grad_norm": 2.585444365366914, "learning_rate": 8.153205785050844e-06, "loss": 0.0666, "step": 6559 }, { "epoch": 0.3048327137546468, "grad_norm": 1.218671104892257, "learning_rate": 8.152621742885133e-06, "loss": 0.0595, "step": 6560 }, { "epoch": 0.30487918215613385, "grad_norm": 0.6392834350605808, "learning_rate": 8.152037629309222e-06, "loss": 0.048, "step": 6561 }, { "epoch": 0.30492565055762083, "grad_norm": 2.239722985150496, "learning_rate": 8.151453444336341e-06, "loss": 0.1108, "step": 6562 }, { "epoch": 0.3049721189591078, "grad_norm": 0.9006441092756865, "learning_rate": 8.150869187979722e-06, "loss": 0.0653, "step": 6563 }, { "epoch": 0.3050185873605948, "grad_norm": 1.8727276384021558, "learning_rate": 8.150284860252599e-06, "loss": 0.1042, "step": 6564 }, { "epoch": 0.30506505576208176, "grad_norm": 1.3404535023714836, "learning_rate": 8.14970046116821e-06, "loss": 0.0847, "step": 6565 }, { "epoch": 0.3051115241635688, "grad_norm": 1.2615715203199982, "learning_rate": 8.14911599073979e-06, "loss": 0.0594, "step": 6566 }, { "epoch": 0.30515799256505577, "grad_norm": 1.2495080732686685, "learning_rate": 8.14853144898058e-06, "loss": 0.0634, "step": 6567 }, { "epoch": 0.30520446096654275, "grad_norm": 1.6439084375772695, "learning_rate": 8.147946835903815e-06, "loss": 0.0565, "step": 6568 }, { "epoch": 0.3052509293680297, "grad_norm": 1.4435756873598034, "learning_rate": 8.147362151522744e-06, "loss": 0.0752, "step": 6569 }, { "epoch": 0.30529739776951675, "grad_norm": 1.0959582831526982, "learning_rate": 8.146777395850607e-06, "loss": 0.0487, "step": 6570 }, { "epoch": 0.30534386617100373, "grad_norm": 1.4019820756104728, "learning_rate": 8.146192568900651e-06, "loss": 0.0918, "step": 6571 }, { "epoch": 0.3053903345724907, "grad_norm": 1.6658282629206624, "learning_rate": 8.14560767068612e-06, "loss": 0.0776, "step": 6572 }, { "epoch": 0.3054368029739777, "grad_norm": 1.3737992236084018, "learning_rate": 8.145022701220267e-06, "loss": 0.0772, "step": 6573 }, { "epoch": 0.30548327137546466, "grad_norm": 1.2411915041747086, "learning_rate": 8.144437660516337e-06, "loss": 0.0643, "step": 6574 }, { "epoch": 0.3055297397769517, "grad_norm": 1.2890108126721906, "learning_rate": 8.143852548587585e-06, "loss": 0.1084, "step": 6575 }, { "epoch": 0.30557620817843867, "grad_norm": 3.3437837204448027, "learning_rate": 8.143267365447264e-06, "loss": 0.0857, "step": 6576 }, { "epoch": 0.30562267657992565, "grad_norm": 3.4295649663245857, "learning_rate": 8.14268211110863e-06, "loss": 0.0938, "step": 6577 }, { "epoch": 0.3056691449814126, "grad_norm": 1.0027921273568232, "learning_rate": 8.142096785584939e-06, "loss": 0.0837, "step": 6578 }, { "epoch": 0.30571561338289965, "grad_norm": 0.8676385318200239, "learning_rate": 8.14151138888945e-06, "loss": 0.0532, "step": 6579 }, { "epoch": 0.30576208178438663, "grad_norm": 0.6316089439653941, "learning_rate": 8.14092592103542e-06, "loss": 0.0524, "step": 6580 }, { "epoch": 0.3058085501858736, "grad_norm": 1.3882775023854834, "learning_rate": 8.14034038203611e-06, "loss": 0.0917, "step": 6581 }, { "epoch": 0.3058550185873606, "grad_norm": 0.9551844190292974, "learning_rate": 8.139754771904788e-06, "loss": 0.0672, "step": 6582 }, { "epoch": 0.30590148698884756, "grad_norm": 0.8675771019749876, "learning_rate": 8.139169090654714e-06, "loss": 0.0604, "step": 6583 }, { "epoch": 0.3059479553903346, "grad_norm": 1.2244723023164636, "learning_rate": 8.138583338299157e-06, "loss": 0.0751, "step": 6584 }, { "epoch": 0.30599442379182157, "grad_norm": 0.9941639286211932, "learning_rate": 8.137997514851385e-06, "loss": 0.0707, "step": 6585 }, { "epoch": 0.30604089219330854, "grad_norm": 1.1130580968633472, "learning_rate": 8.137411620324665e-06, "loss": 0.0549, "step": 6586 }, { "epoch": 0.3060873605947955, "grad_norm": 0.9041956620051493, "learning_rate": 8.136825654732272e-06, "loss": 0.0886, "step": 6587 }, { "epoch": 0.30613382899628255, "grad_norm": 1.6755723394719424, "learning_rate": 8.136239618087474e-06, "loss": 0.1075, "step": 6588 }, { "epoch": 0.30618029739776953, "grad_norm": 1.6731435005485693, "learning_rate": 8.13565351040355e-06, "loss": 0.0859, "step": 6589 }, { "epoch": 0.3062267657992565, "grad_norm": 1.5449578075873456, "learning_rate": 8.135067331693773e-06, "loss": 0.0628, "step": 6590 }, { "epoch": 0.3062732342007435, "grad_norm": 0.5491148326898515, "learning_rate": 8.13448108197142e-06, "loss": 0.0362, "step": 6591 }, { "epoch": 0.30631970260223046, "grad_norm": 1.2047260373864392, "learning_rate": 8.133894761249772e-06, "loss": 0.0648, "step": 6592 }, { "epoch": 0.3063661710037175, "grad_norm": 1.128072863826617, "learning_rate": 8.13330836954211e-06, "loss": 0.0987, "step": 6593 }, { "epoch": 0.30641263940520447, "grad_norm": 1.619401619371025, "learning_rate": 8.132721906861715e-06, "loss": 0.0745, "step": 6594 }, { "epoch": 0.30645910780669144, "grad_norm": 2.70128644368497, "learning_rate": 8.132135373221872e-06, "loss": 0.1069, "step": 6595 }, { "epoch": 0.3065055762081784, "grad_norm": 1.0736456103802006, "learning_rate": 8.131548768635867e-06, "loss": 0.0688, "step": 6596 }, { "epoch": 0.30655204460966545, "grad_norm": 1.046427146103279, "learning_rate": 8.130962093116985e-06, "loss": 0.1059, "step": 6597 }, { "epoch": 0.30659851301115243, "grad_norm": 1.8333726303232454, "learning_rate": 8.130375346678518e-06, "loss": 0.0852, "step": 6598 }, { "epoch": 0.3066449814126394, "grad_norm": 0.642453361248382, "learning_rate": 8.129788529333755e-06, "loss": 0.0436, "step": 6599 }, { "epoch": 0.3066914498141264, "grad_norm": 1.5194264941222155, "learning_rate": 8.129201641095986e-06, "loss": 0.0824, "step": 6600 }, { "epoch": 0.30673791821561336, "grad_norm": 1.7085500867772112, "learning_rate": 8.128614681978507e-06, "loss": 0.0987, "step": 6601 }, { "epoch": 0.3067843866171004, "grad_norm": 0.9724503229251131, "learning_rate": 8.128027651994612e-06, "loss": 0.0783, "step": 6602 }, { "epoch": 0.30683085501858737, "grad_norm": 1.5756165086183391, "learning_rate": 8.127440551157599e-06, "loss": 0.0707, "step": 6603 }, { "epoch": 0.30687732342007434, "grad_norm": 0.8133056510703429, "learning_rate": 8.126853379480767e-06, "loss": 0.0747, "step": 6604 }, { "epoch": 0.3069237918215613, "grad_norm": 0.860908446444308, "learning_rate": 8.126266136977414e-06, "loss": 0.0653, "step": 6605 }, { "epoch": 0.30697026022304835, "grad_norm": 2.8898508285090654, "learning_rate": 8.125678823660842e-06, "loss": 0.084, "step": 6606 }, { "epoch": 0.30701672862453533, "grad_norm": 1.0487539129555372, "learning_rate": 8.125091439544356e-06, "loss": 0.0627, "step": 6607 }, { "epoch": 0.3070631970260223, "grad_norm": 0.7413812664681656, "learning_rate": 8.12450398464126e-06, "loss": 0.0568, "step": 6608 }, { "epoch": 0.3071096654275093, "grad_norm": 1.3829152277985568, "learning_rate": 8.12391645896486e-06, "loss": 0.0636, "step": 6609 }, { "epoch": 0.30715613382899626, "grad_norm": 0.83303374912383, "learning_rate": 8.123328862528465e-06, "loss": 0.0634, "step": 6610 }, { "epoch": 0.3072026022304833, "grad_norm": 1.2251195521901896, "learning_rate": 8.122741195345384e-06, "loss": 0.0659, "step": 6611 }, { "epoch": 0.30724907063197027, "grad_norm": 0.6699821176443682, "learning_rate": 8.122153457428928e-06, "loss": 0.0439, "step": 6612 }, { "epoch": 0.30729553903345724, "grad_norm": 2.2669928085322626, "learning_rate": 8.121565648792409e-06, "loss": 0.094, "step": 6613 }, { "epoch": 0.3073420074349442, "grad_norm": 0.8822550871901953, "learning_rate": 8.120977769449143e-06, "loss": 0.0654, "step": 6614 }, { "epoch": 0.30738847583643125, "grad_norm": 1.056917718883343, "learning_rate": 8.120389819412447e-06, "loss": 0.046, "step": 6615 }, { "epoch": 0.30743494423791823, "grad_norm": 0.9085995857080456, "learning_rate": 8.119801798695636e-06, "loss": 0.073, "step": 6616 }, { "epoch": 0.3074814126394052, "grad_norm": 0.9033401356997395, "learning_rate": 8.11921370731203e-06, "loss": 0.0608, "step": 6617 }, { "epoch": 0.3075278810408922, "grad_norm": 0.9209267910838275, "learning_rate": 8.118625545274953e-06, "loss": 0.1224, "step": 6618 }, { "epoch": 0.30757434944237916, "grad_norm": 0.845204260988233, "learning_rate": 8.118037312597725e-06, "loss": 0.0418, "step": 6619 }, { "epoch": 0.3076208178438662, "grad_norm": 0.7973813293115006, "learning_rate": 8.117449009293668e-06, "loss": 0.0536, "step": 6620 }, { "epoch": 0.30766728624535317, "grad_norm": 0.9661999519440541, "learning_rate": 8.116860635376111e-06, "loss": 0.0579, "step": 6621 }, { "epoch": 0.30771375464684014, "grad_norm": 1.6547628048179694, "learning_rate": 8.11627219085838e-06, "loss": 0.0867, "step": 6622 }, { "epoch": 0.3077602230483271, "grad_norm": 1.0100653914831226, "learning_rate": 8.115683675753805e-06, "loss": 0.0605, "step": 6623 }, { "epoch": 0.30780669144981415, "grad_norm": 2.2835042747918592, "learning_rate": 8.115095090075716e-06, "loss": 0.0943, "step": 6624 }, { "epoch": 0.30785315985130113, "grad_norm": 2.143105483050944, "learning_rate": 8.114506433837443e-06, "loss": 0.0728, "step": 6625 }, { "epoch": 0.3078996282527881, "grad_norm": 0.69962982489456, "learning_rate": 8.113917707052322e-06, "loss": 0.0486, "step": 6626 }, { "epoch": 0.3079460966542751, "grad_norm": 0.9096655410897679, "learning_rate": 8.113328909733686e-06, "loss": 0.073, "step": 6627 }, { "epoch": 0.30799256505576206, "grad_norm": 1.9200895249714733, "learning_rate": 8.112740041894875e-06, "loss": 0.0686, "step": 6628 }, { "epoch": 0.3080390334572491, "grad_norm": 2.1035259511333835, "learning_rate": 8.112151103549227e-06, "loss": 0.1019, "step": 6629 }, { "epoch": 0.30808550185873607, "grad_norm": 0.8932428051046734, "learning_rate": 8.111562094710079e-06, "loss": 0.0499, "step": 6630 }, { "epoch": 0.30813197026022304, "grad_norm": 0.9498465476334612, "learning_rate": 8.110973015390777e-06, "loss": 0.0949, "step": 6631 }, { "epoch": 0.30817843866171, "grad_norm": 1.4084026976390354, "learning_rate": 8.11038386560466e-06, "loss": 0.0973, "step": 6632 }, { "epoch": 0.30822490706319705, "grad_norm": 1.0259059917093694, "learning_rate": 8.109794645365075e-06, "loss": 0.0672, "step": 6633 }, { "epoch": 0.308271375464684, "grad_norm": 0.929634508821098, "learning_rate": 8.109205354685367e-06, "loss": 0.0795, "step": 6634 }, { "epoch": 0.308317843866171, "grad_norm": 1.181144617302556, "learning_rate": 8.108615993578888e-06, "loss": 0.0655, "step": 6635 }, { "epoch": 0.308364312267658, "grad_norm": 1.0950714556736203, "learning_rate": 8.108026562058985e-06, "loss": 0.0523, "step": 6636 }, { "epoch": 0.30841078066914496, "grad_norm": 1.0529217336713497, "learning_rate": 8.107437060139008e-06, "loss": 0.072, "step": 6637 }, { "epoch": 0.308457249070632, "grad_norm": 1.0470530848454822, "learning_rate": 8.10684748783231e-06, "loss": 0.0684, "step": 6638 }, { "epoch": 0.30850371747211897, "grad_norm": 1.1672951594856154, "learning_rate": 8.10625784515225e-06, "loss": 0.0719, "step": 6639 }, { "epoch": 0.30855018587360594, "grad_norm": 1.5080746977708606, "learning_rate": 8.105668132112176e-06, "loss": 0.0759, "step": 6640 }, { "epoch": 0.3085966542750929, "grad_norm": 0.9835735182584077, "learning_rate": 8.105078348725454e-06, "loss": 0.076, "step": 6641 }, { "epoch": 0.30864312267657995, "grad_norm": 0.6201426750544669, "learning_rate": 8.104488495005438e-06, "loss": 0.0552, "step": 6642 }, { "epoch": 0.3086895910780669, "grad_norm": 0.5741749680370325, "learning_rate": 8.103898570965489e-06, "loss": 0.0547, "step": 6643 }, { "epoch": 0.3087360594795539, "grad_norm": 1.0561434751980399, "learning_rate": 8.103308576618972e-06, "loss": 0.0556, "step": 6644 }, { "epoch": 0.3087825278810409, "grad_norm": 1.1849302256255447, "learning_rate": 8.10271851197925e-06, "loss": 0.0441, "step": 6645 }, { "epoch": 0.30882899628252786, "grad_norm": 1.5189808562123857, "learning_rate": 8.102128377059688e-06, "loss": 0.064, "step": 6646 }, { "epoch": 0.3088754646840149, "grad_norm": 2.255505845751216, "learning_rate": 8.101538171873652e-06, "loss": 0.072, "step": 6647 }, { "epoch": 0.30892193308550187, "grad_norm": 0.8306870991484827, "learning_rate": 8.100947896434512e-06, "loss": 0.0487, "step": 6648 }, { "epoch": 0.30896840148698884, "grad_norm": 0.7597551443897326, "learning_rate": 8.10035755075564e-06, "loss": 0.0641, "step": 6649 }, { "epoch": 0.3090148698884758, "grad_norm": 1.447898633696828, "learning_rate": 8.099767134850404e-06, "loss": 0.0877, "step": 6650 }, { "epoch": 0.30906133828996285, "grad_norm": 0.7608267004862159, "learning_rate": 8.099176648732182e-06, "loss": 0.05, "step": 6651 }, { "epoch": 0.3091078066914498, "grad_norm": 0.9343187543779969, "learning_rate": 8.098586092414346e-06, "loss": 0.0732, "step": 6652 }, { "epoch": 0.3091542750929368, "grad_norm": 0.89999427818674, "learning_rate": 8.097995465910275e-06, "loss": 0.0528, "step": 6653 }, { "epoch": 0.3092007434944238, "grad_norm": 0.8187708538571332, "learning_rate": 8.097404769233346e-06, "loss": 0.0715, "step": 6654 }, { "epoch": 0.30924721189591076, "grad_norm": 0.9968215946667434, "learning_rate": 8.096814002396939e-06, "loss": 0.0851, "step": 6655 }, { "epoch": 0.3092936802973978, "grad_norm": 1.0674672080823355, "learning_rate": 8.096223165414433e-06, "loss": 0.1052, "step": 6656 }, { "epoch": 0.30934014869888476, "grad_norm": 0.7748121322211962, "learning_rate": 8.095632258299218e-06, "loss": 0.0571, "step": 6657 }, { "epoch": 0.30938661710037174, "grad_norm": 0.9709528237642168, "learning_rate": 8.095041281064672e-06, "loss": 0.0784, "step": 6658 }, { "epoch": 0.3094330855018587, "grad_norm": 0.5012177299905156, "learning_rate": 8.094450233724183e-06, "loss": 0.0117, "step": 6659 }, { "epoch": 0.30947955390334575, "grad_norm": 0.5351513269223143, "learning_rate": 8.093859116291139e-06, "loss": 0.0492, "step": 6660 }, { "epoch": 0.3095260223048327, "grad_norm": 0.6935979968245315, "learning_rate": 8.093267928778932e-06, "loss": 0.0242, "step": 6661 }, { "epoch": 0.3095724907063197, "grad_norm": 1.2474143970544767, "learning_rate": 8.092676671200948e-06, "loss": 0.083, "step": 6662 }, { "epoch": 0.3096189591078067, "grad_norm": 0.8735383455093386, "learning_rate": 8.092085343570584e-06, "loss": 0.0615, "step": 6663 }, { "epoch": 0.30966542750929366, "grad_norm": 0.8942984750354385, "learning_rate": 8.091493945901233e-06, "loss": 0.0767, "step": 6664 }, { "epoch": 0.3097118959107807, "grad_norm": 0.8464869797842776, "learning_rate": 8.090902478206289e-06, "loss": 0.0582, "step": 6665 }, { "epoch": 0.30975836431226766, "grad_norm": 2.0326982491284196, "learning_rate": 8.090310940499152e-06, "loss": 0.0983, "step": 6666 }, { "epoch": 0.30980483271375464, "grad_norm": 1.104602918826988, "learning_rate": 8.08971933279322e-06, "loss": 0.0842, "step": 6667 }, { "epoch": 0.3098513011152416, "grad_norm": 0.9722906666504416, "learning_rate": 8.089127655101892e-06, "loss": 0.0773, "step": 6668 }, { "epoch": 0.30989776951672865, "grad_norm": 1.8153711453901487, "learning_rate": 8.088535907438571e-06, "loss": 0.0721, "step": 6669 }, { "epoch": 0.3099442379182156, "grad_norm": 1.5956747276279224, "learning_rate": 8.087944089816663e-06, "loss": 0.0892, "step": 6670 }, { "epoch": 0.3099907063197026, "grad_norm": 1.9644730805211246, "learning_rate": 8.08735220224957e-06, "loss": 0.1217, "step": 6671 }, { "epoch": 0.3100371747211896, "grad_norm": 1.112553152443814, "learning_rate": 8.0867602447507e-06, "loss": 0.0651, "step": 6672 }, { "epoch": 0.31008364312267656, "grad_norm": 1.1505328654345395, "learning_rate": 8.086168217333463e-06, "loss": 0.0565, "step": 6673 }, { "epoch": 0.3101301115241636, "grad_norm": 1.6555462774844127, "learning_rate": 8.085576120011267e-06, "loss": 0.0781, "step": 6674 }, { "epoch": 0.31017657992565056, "grad_norm": 0.49003127445636413, "learning_rate": 8.084983952797525e-06, "loss": 0.0448, "step": 6675 }, { "epoch": 0.31022304832713754, "grad_norm": 1.185278105388994, "learning_rate": 8.084391715705648e-06, "loss": 0.0709, "step": 6676 }, { "epoch": 0.3102695167286245, "grad_norm": 1.1584831669968538, "learning_rate": 8.083799408749053e-06, "loss": 0.0801, "step": 6677 }, { "epoch": 0.31031598513011155, "grad_norm": 0.7043305635698281, "learning_rate": 8.083207031941157e-06, "loss": 0.0525, "step": 6678 }, { "epoch": 0.3103624535315985, "grad_norm": 2.901072530288602, "learning_rate": 8.082614585295375e-06, "loss": 0.0816, "step": 6679 }, { "epoch": 0.3104089219330855, "grad_norm": 1.37826170263759, "learning_rate": 8.082022068825129e-06, "loss": 0.059, "step": 6680 }, { "epoch": 0.3104553903345725, "grad_norm": 1.089407013140547, "learning_rate": 8.08142948254384e-06, "loss": 0.0711, "step": 6681 }, { "epoch": 0.31050185873605946, "grad_norm": 0.9881822458502411, "learning_rate": 8.080836826464928e-06, "loss": 0.0715, "step": 6682 }, { "epoch": 0.3105483271375465, "grad_norm": 1.9087530666845467, "learning_rate": 8.080244100601822e-06, "loss": 0.0901, "step": 6683 }, { "epoch": 0.31059479553903346, "grad_norm": 0.8366252368091289, "learning_rate": 8.079651304967944e-06, "loss": 0.0588, "step": 6684 }, { "epoch": 0.31064126394052044, "grad_norm": 0.4850935451284873, "learning_rate": 8.079058439576722e-06, "loss": 0.0435, "step": 6685 }, { "epoch": 0.3106877323420074, "grad_norm": 1.441435818099974, "learning_rate": 8.078465504441586e-06, "loss": 0.083, "step": 6686 }, { "epoch": 0.31073420074349445, "grad_norm": 1.8268895727294872, "learning_rate": 8.077872499575968e-06, "loss": 0.1006, "step": 6687 }, { "epoch": 0.3107806691449814, "grad_norm": 3.994961364513022, "learning_rate": 8.077279424993297e-06, "loss": 0.0647, "step": 6688 }, { "epoch": 0.3108271375464684, "grad_norm": 1.501311860690791, "learning_rate": 8.076686280707009e-06, "loss": 0.0623, "step": 6689 }, { "epoch": 0.3108736059479554, "grad_norm": 0.5658805320293689, "learning_rate": 8.076093066730538e-06, "loss": 0.0488, "step": 6690 }, { "epoch": 0.31092007434944235, "grad_norm": 2.526332406424271, "learning_rate": 8.075499783077321e-06, "loss": 0.0851, "step": 6691 }, { "epoch": 0.3109665427509294, "grad_norm": 0.9165436830601622, "learning_rate": 8.074906429760796e-06, "loss": 0.0729, "step": 6692 }, { "epoch": 0.31101301115241636, "grad_norm": 0.9497211949089077, "learning_rate": 8.074313006794406e-06, "loss": 0.0703, "step": 6693 }, { "epoch": 0.31105947955390334, "grad_norm": 1.632127559043751, "learning_rate": 8.073719514191592e-06, "loss": 0.0735, "step": 6694 }, { "epoch": 0.3111059479553903, "grad_norm": 1.358898181648441, "learning_rate": 8.073125951965793e-06, "loss": 0.093, "step": 6695 }, { "epoch": 0.31115241635687735, "grad_norm": 2.183162377040696, "learning_rate": 8.072532320130458e-06, "loss": 0.1155, "step": 6696 }, { "epoch": 0.3111988847583643, "grad_norm": 0.6227938761074837, "learning_rate": 8.071938618699033e-06, "loss": 0.0515, "step": 6697 }, { "epoch": 0.3112453531598513, "grad_norm": 0.8301811770543027, "learning_rate": 8.071344847684965e-06, "loss": 0.0604, "step": 6698 }, { "epoch": 0.3112918215613383, "grad_norm": 1.353529108337013, "learning_rate": 8.070751007101702e-06, "loss": 0.0511, "step": 6699 }, { "epoch": 0.31133828996282525, "grad_norm": 1.1225917638657998, "learning_rate": 8.0701570969627e-06, "loss": 0.0926, "step": 6700 }, { "epoch": 0.3113847583643123, "grad_norm": 1.7520982991665521, "learning_rate": 8.069563117281405e-06, "loss": 0.1467, "step": 6701 }, { "epoch": 0.31143122676579926, "grad_norm": 4.317090316027362, "learning_rate": 8.068969068071275e-06, "loss": 0.0406, "step": 6702 }, { "epoch": 0.31147769516728624, "grad_norm": 1.0868948028240246, "learning_rate": 8.068374949345767e-06, "loss": 0.0561, "step": 6703 }, { "epoch": 0.3115241635687732, "grad_norm": 1.26145692687464, "learning_rate": 8.067780761118337e-06, "loss": 0.1105, "step": 6704 }, { "epoch": 0.31157063197026025, "grad_norm": 0.8742817470878844, "learning_rate": 8.067186503402444e-06, "loss": 0.0533, "step": 6705 }, { "epoch": 0.3116171003717472, "grad_norm": 1.0125479843916, "learning_rate": 8.066592176211547e-06, "loss": 0.0602, "step": 6706 }, { "epoch": 0.3116635687732342, "grad_norm": 1.6714161072594826, "learning_rate": 8.065997779559111e-06, "loss": 0.0997, "step": 6707 }, { "epoch": 0.3117100371747212, "grad_norm": 1.9817031971369377, "learning_rate": 8.0654033134586e-06, "loss": 0.1138, "step": 6708 }, { "epoch": 0.31175650557620815, "grad_norm": 1.6481102322699641, "learning_rate": 8.064808777923475e-06, "loss": 0.0832, "step": 6709 }, { "epoch": 0.3118029739776952, "grad_norm": 2.2207057576208977, "learning_rate": 8.064214172967206e-06, "loss": 0.0763, "step": 6710 }, { "epoch": 0.31184944237918216, "grad_norm": 1.733554040608124, "learning_rate": 8.063619498603261e-06, "loss": 0.0584, "step": 6711 }, { "epoch": 0.31189591078066914, "grad_norm": 1.4224122175164464, "learning_rate": 8.063024754845111e-06, "loss": 0.0483, "step": 6712 }, { "epoch": 0.3119423791821561, "grad_norm": 1.7794965908031457, "learning_rate": 8.062429941706225e-06, "loss": 0.0791, "step": 6713 }, { "epoch": 0.31198884758364315, "grad_norm": 1.222072655882278, "learning_rate": 8.06183505920008e-06, "loss": 0.1213, "step": 6714 }, { "epoch": 0.3120353159851301, "grad_norm": 0.9383805390800996, "learning_rate": 8.061240107340146e-06, "loss": 0.0776, "step": 6715 }, { "epoch": 0.3120817843866171, "grad_norm": 1.165794953921632, "learning_rate": 8.060645086139902e-06, "loss": 0.0953, "step": 6716 }, { "epoch": 0.3121282527881041, "grad_norm": 2.5110248791052823, "learning_rate": 8.060049995612828e-06, "loss": 0.0668, "step": 6717 }, { "epoch": 0.31217472118959105, "grad_norm": 1.794235623762947, "learning_rate": 8.059454835772399e-06, "loss": 0.0727, "step": 6718 }, { "epoch": 0.3122211895910781, "grad_norm": 0.5058349866393231, "learning_rate": 8.058859606632096e-06, "loss": 0.0542, "step": 6719 }, { "epoch": 0.31226765799256506, "grad_norm": 1.83742648188307, "learning_rate": 8.058264308205406e-06, "loss": 0.079, "step": 6720 }, { "epoch": 0.31231412639405204, "grad_norm": 1.4842883533170743, "learning_rate": 8.057668940505811e-06, "loss": 0.0707, "step": 6721 }, { "epoch": 0.312360594795539, "grad_norm": 1.1066621331297424, "learning_rate": 8.057073503546795e-06, "loss": 0.0616, "step": 6722 }, { "epoch": 0.31240706319702605, "grad_norm": 1.9015966865697824, "learning_rate": 8.056477997341848e-06, "loss": 0.0745, "step": 6723 }, { "epoch": 0.312453531598513, "grad_norm": 1.9190544666341631, "learning_rate": 8.055882421904456e-06, "loss": 0.0678, "step": 6724 }, { "epoch": 0.3125, "grad_norm": 1.2288539057122543, "learning_rate": 8.055286777248112e-06, "loss": 0.0568, "step": 6725 }, { "epoch": 0.312546468401487, "grad_norm": 2.1539092597510248, "learning_rate": 8.054691063386306e-06, "loss": 0.0804, "step": 6726 }, { "epoch": 0.31259293680297395, "grad_norm": 0.663677841704123, "learning_rate": 8.054095280332533e-06, "loss": 0.0456, "step": 6727 }, { "epoch": 0.312639405204461, "grad_norm": 1.5038946997318254, "learning_rate": 8.053499428100286e-06, "loss": 0.0753, "step": 6728 }, { "epoch": 0.31268587360594796, "grad_norm": 1.5584340213828018, "learning_rate": 8.052903506703066e-06, "loss": 0.0848, "step": 6729 }, { "epoch": 0.31273234200743494, "grad_norm": 2.513709464682219, "learning_rate": 8.052307516154368e-06, "loss": 0.0841, "step": 6730 }, { "epoch": 0.3127788104089219, "grad_norm": 1.7281226091989923, "learning_rate": 8.05171145646769e-06, "loss": 0.0679, "step": 6731 }, { "epoch": 0.31282527881040895, "grad_norm": 1.4523410991992758, "learning_rate": 8.051115327656538e-06, "loss": 0.0814, "step": 6732 }, { "epoch": 0.3128717472118959, "grad_norm": 1.183083599710745, "learning_rate": 8.050519129734411e-06, "loss": 0.0637, "step": 6733 }, { "epoch": 0.3129182156133829, "grad_norm": 0.6182210278702397, "learning_rate": 8.049922862714814e-06, "loss": 0.0346, "step": 6734 }, { "epoch": 0.3129646840148699, "grad_norm": 0.7372835950415243, "learning_rate": 8.049326526611258e-06, "loss": 0.0649, "step": 6735 }, { "epoch": 0.31301115241635685, "grad_norm": 1.0842929091653548, "learning_rate": 8.048730121437244e-06, "loss": 0.0628, "step": 6736 }, { "epoch": 0.3130576208178439, "grad_norm": 0.7121879546395189, "learning_rate": 8.048133647206285e-06, "loss": 0.0695, "step": 6737 }, { "epoch": 0.31310408921933086, "grad_norm": 0.9351838717165526, "learning_rate": 8.047537103931891e-06, "loss": 0.0554, "step": 6738 }, { "epoch": 0.31315055762081784, "grad_norm": 1.0629754939547518, "learning_rate": 8.046940491627572e-06, "loss": 0.085, "step": 6739 }, { "epoch": 0.3131970260223048, "grad_norm": 0.9665988041222408, "learning_rate": 8.046343810306844e-06, "loss": 0.1091, "step": 6740 }, { "epoch": 0.31324349442379185, "grad_norm": 1.3258641182257367, "learning_rate": 8.045747059983226e-06, "loss": 0.0694, "step": 6741 }, { "epoch": 0.3132899628252788, "grad_norm": 1.049578804961969, "learning_rate": 8.04515024067023e-06, "loss": 0.067, "step": 6742 }, { "epoch": 0.3133364312267658, "grad_norm": 0.8275955379349862, "learning_rate": 8.044553352381375e-06, "loss": 0.0523, "step": 6743 }, { "epoch": 0.3133828996282528, "grad_norm": 1.0004227684413907, "learning_rate": 8.043956395130183e-06, "loss": 0.081, "step": 6744 }, { "epoch": 0.31342936802973975, "grad_norm": 1.41426472080982, "learning_rate": 8.043359368930175e-06, "loss": 0.0953, "step": 6745 }, { "epoch": 0.3134758364312268, "grad_norm": 0.8186981739111512, "learning_rate": 8.042762273794874e-06, "loss": 0.0613, "step": 6746 }, { "epoch": 0.31352230483271376, "grad_norm": 3.0934830074016904, "learning_rate": 8.042165109737806e-06, "loss": 0.0843, "step": 6747 }, { "epoch": 0.31356877323420074, "grad_norm": 0.6133390552722423, "learning_rate": 8.041567876772493e-06, "loss": 0.0471, "step": 6748 }, { "epoch": 0.3136152416356877, "grad_norm": 1.577823475245816, "learning_rate": 8.04097057491247e-06, "loss": 0.0873, "step": 6749 }, { "epoch": 0.31366171003717475, "grad_norm": 1.2942773149207631, "learning_rate": 8.04037320417126e-06, "loss": 0.0823, "step": 6750 }, { "epoch": 0.3137081784386617, "grad_norm": 1.6398700980352032, "learning_rate": 8.039775764562401e-06, "loss": 0.1193, "step": 6751 }, { "epoch": 0.3137546468401487, "grad_norm": 1.769208496566225, "learning_rate": 8.039178256099418e-06, "loss": 0.0742, "step": 6752 }, { "epoch": 0.3138011152416357, "grad_norm": 0.7685386060696525, "learning_rate": 8.038580678795851e-06, "loss": 0.0562, "step": 6753 }, { "epoch": 0.31384758364312265, "grad_norm": 1.9802948549084853, "learning_rate": 8.037983032665233e-06, "loss": 0.071, "step": 6754 }, { "epoch": 0.3138940520446097, "grad_norm": 3.6804417940403122, "learning_rate": 8.0373853177211e-06, "loss": 0.1409, "step": 6755 }, { "epoch": 0.31394052044609666, "grad_norm": 1.2046915129639197, "learning_rate": 8.036787533976997e-06, "loss": 0.0619, "step": 6756 }, { "epoch": 0.31398698884758364, "grad_norm": 0.9793019483524672, "learning_rate": 8.036189681446457e-06, "loss": 0.0606, "step": 6757 }, { "epoch": 0.3140334572490706, "grad_norm": 2.273387353541432, "learning_rate": 8.035591760143025e-06, "loss": 0.0804, "step": 6758 }, { "epoch": 0.31407992565055765, "grad_norm": 1.5489723964772273, "learning_rate": 8.034993770080246e-06, "loss": 0.1007, "step": 6759 }, { "epoch": 0.3141263940520446, "grad_norm": 1.4693063741291545, "learning_rate": 8.034395711271664e-06, "loss": 0.0711, "step": 6760 }, { "epoch": 0.3141728624535316, "grad_norm": 3.8975209419375543, "learning_rate": 8.033797583730825e-06, "loss": 0.077, "step": 6761 }, { "epoch": 0.3142193308550186, "grad_norm": 0.6652855454973355, "learning_rate": 8.033199387471278e-06, "loss": 0.0553, "step": 6762 }, { "epoch": 0.31426579925650555, "grad_norm": 1.2729718094408489, "learning_rate": 8.032601122506571e-06, "loss": 0.0693, "step": 6763 }, { "epoch": 0.3143122676579926, "grad_norm": 1.8059184874687786, "learning_rate": 8.03200278885026e-06, "loss": 0.0711, "step": 6764 }, { "epoch": 0.31435873605947956, "grad_norm": 0.8898863122562471, "learning_rate": 8.03140438651589e-06, "loss": 0.0681, "step": 6765 }, { "epoch": 0.31440520446096654, "grad_norm": 1.315519751825964, "learning_rate": 8.030805915517024e-06, "loss": 0.047, "step": 6766 }, { "epoch": 0.3144516728624535, "grad_norm": 1.0289821374395658, "learning_rate": 8.030207375867213e-06, "loss": 0.0712, "step": 6767 }, { "epoch": 0.31449814126394054, "grad_norm": 0.8197595843101131, "learning_rate": 8.029608767580015e-06, "loss": 0.0541, "step": 6768 }, { "epoch": 0.3145446096654275, "grad_norm": 0.912693576152326, "learning_rate": 8.02901009066899e-06, "loss": 0.0317, "step": 6769 }, { "epoch": 0.3145910780669145, "grad_norm": 1.258311211983334, "learning_rate": 8.028411345147699e-06, "loss": 0.0932, "step": 6770 }, { "epoch": 0.3146375464684015, "grad_norm": 0.9832035808998172, "learning_rate": 8.027812531029704e-06, "loss": 0.0802, "step": 6771 }, { "epoch": 0.31468401486988845, "grad_norm": 1.7886425633357448, "learning_rate": 8.02721364832857e-06, "loss": 0.0913, "step": 6772 }, { "epoch": 0.3147304832713755, "grad_norm": 1.5784947239719431, "learning_rate": 8.026614697057858e-06, "loss": 0.0735, "step": 6773 }, { "epoch": 0.31477695167286246, "grad_norm": 1.4869574621995092, "learning_rate": 8.026015677231137e-06, "loss": 0.1051, "step": 6774 }, { "epoch": 0.31482342007434944, "grad_norm": 1.2419087960921416, "learning_rate": 8.025416588861979e-06, "loss": 0.0731, "step": 6775 }, { "epoch": 0.3148698884758364, "grad_norm": 0.8150462778912393, "learning_rate": 8.02481743196395e-06, "loss": 0.0673, "step": 6776 }, { "epoch": 0.31491635687732344, "grad_norm": 0.8611571363632554, "learning_rate": 8.024218206550622e-06, "loss": 0.0801, "step": 6777 }, { "epoch": 0.3149628252788104, "grad_norm": 1.6051635201424124, "learning_rate": 8.02361891263557e-06, "loss": 0.0736, "step": 6778 }, { "epoch": 0.3150092936802974, "grad_norm": 1.0380615259562151, "learning_rate": 8.023019550232365e-06, "loss": 0.0655, "step": 6779 }, { "epoch": 0.3150557620817844, "grad_norm": 1.1403188668142947, "learning_rate": 8.022420119354585e-06, "loss": 0.0908, "step": 6780 }, { "epoch": 0.31510223048327135, "grad_norm": 3.5761272441890055, "learning_rate": 8.021820620015811e-06, "loss": 0.0699, "step": 6781 }, { "epoch": 0.3151486988847584, "grad_norm": 1.174436124147293, "learning_rate": 8.02122105222962e-06, "loss": 0.0845, "step": 6782 }, { "epoch": 0.31519516728624536, "grad_norm": 0.8340547766423531, "learning_rate": 8.020621416009591e-06, "loss": 0.0681, "step": 6783 }, { "epoch": 0.31524163568773234, "grad_norm": 0.5720148540046202, "learning_rate": 8.020021711369306e-06, "loss": 0.0121, "step": 6784 }, { "epoch": 0.3152881040892193, "grad_norm": 0.6768081956319746, "learning_rate": 8.019421938322354e-06, "loss": 0.0693, "step": 6785 }, { "epoch": 0.31533457249070634, "grad_norm": 1.9990969392915308, "learning_rate": 8.018822096882314e-06, "loss": 0.0824, "step": 6786 }, { "epoch": 0.3153810408921933, "grad_norm": 1.1500036264130216, "learning_rate": 8.018222187062779e-06, "loss": 0.0685, "step": 6787 }, { "epoch": 0.3154275092936803, "grad_norm": 0.6915190220009784, "learning_rate": 8.017622208877333e-06, "loss": 0.0402, "step": 6788 }, { "epoch": 0.3154739776951673, "grad_norm": 0.6146482268304736, "learning_rate": 8.01702216233957e-06, "loss": 0.0572, "step": 6789 }, { "epoch": 0.31552044609665425, "grad_norm": 1.2648006730205303, "learning_rate": 8.016422047463078e-06, "loss": 0.0558, "step": 6790 }, { "epoch": 0.3155669144981413, "grad_norm": 1.0156102027151792, "learning_rate": 8.015821864261453e-06, "loss": 0.0897, "step": 6791 }, { "epoch": 0.31561338289962826, "grad_norm": 0.771024302404767, "learning_rate": 8.015221612748288e-06, "loss": 0.0581, "step": 6792 }, { "epoch": 0.31565985130111524, "grad_norm": 1.367707362666413, "learning_rate": 8.014621292937178e-06, "loss": 0.0963, "step": 6793 }, { "epoch": 0.3157063197026022, "grad_norm": 1.8386918722017114, "learning_rate": 8.014020904841727e-06, "loss": 0.0921, "step": 6794 }, { "epoch": 0.31575278810408924, "grad_norm": 0.9205441164000676, "learning_rate": 8.01342044847553e-06, "loss": 0.0684, "step": 6795 }, { "epoch": 0.3157992565055762, "grad_norm": 2.086548001109451, "learning_rate": 8.012819923852184e-06, "loss": 0.0628, "step": 6796 }, { "epoch": 0.3158457249070632, "grad_norm": 1.791213378430315, "learning_rate": 8.012219330985298e-06, "loss": 0.0518, "step": 6797 }, { "epoch": 0.3158921933085502, "grad_norm": 1.1743883949399552, "learning_rate": 8.011618669888475e-06, "loss": 0.0729, "step": 6798 }, { "epoch": 0.31593866171003715, "grad_norm": 1.192855344904955, "learning_rate": 8.011017940575319e-06, "loss": 0.0568, "step": 6799 }, { "epoch": 0.3159851301115242, "grad_norm": 0.6455847282999979, "learning_rate": 8.010417143059437e-06, "loss": 0.0619, "step": 6800 }, { "epoch": 0.31603159851301116, "grad_norm": 1.0161780325307566, "learning_rate": 8.009816277354438e-06, "loss": 0.0513, "step": 6801 }, { "epoch": 0.31607806691449813, "grad_norm": 1.2886817045526726, "learning_rate": 8.009215343473932e-06, "loss": 0.0534, "step": 6802 }, { "epoch": 0.3161245353159851, "grad_norm": 0.922273996062943, "learning_rate": 8.008614341431534e-06, "loss": 0.0745, "step": 6803 }, { "epoch": 0.31617100371747214, "grad_norm": 1.7208146718566701, "learning_rate": 8.008013271240854e-06, "loss": 0.1031, "step": 6804 }, { "epoch": 0.3162174721189591, "grad_norm": 1.4664635565700677, "learning_rate": 8.007412132915504e-06, "loss": 0.1254, "step": 6805 }, { "epoch": 0.3162639405204461, "grad_norm": 0.9705816975785952, "learning_rate": 8.006810926469108e-06, "loss": 0.0876, "step": 6806 }, { "epoch": 0.3163104089219331, "grad_norm": 1.0721955471858287, "learning_rate": 8.006209651915278e-06, "loss": 0.0828, "step": 6807 }, { "epoch": 0.31635687732342005, "grad_norm": 0.8342240251280475, "learning_rate": 8.005608309267637e-06, "loss": 0.0402, "step": 6808 }, { "epoch": 0.3164033457249071, "grad_norm": 1.0144268233106737, "learning_rate": 8.005006898539803e-06, "loss": 0.082, "step": 6809 }, { "epoch": 0.31644981412639406, "grad_norm": 0.8160925035455976, "learning_rate": 8.0044054197454e-06, "loss": 0.0661, "step": 6810 }, { "epoch": 0.31649628252788103, "grad_norm": 1.3620910810815565, "learning_rate": 8.003803872898051e-06, "loss": 0.0592, "step": 6811 }, { "epoch": 0.316542750929368, "grad_norm": 1.278361415211908, "learning_rate": 8.003202258011386e-06, "loss": 0.0732, "step": 6812 }, { "epoch": 0.31658921933085504, "grad_norm": 1.149668748157094, "learning_rate": 8.002600575099027e-06, "loss": 0.0989, "step": 6813 }, { "epoch": 0.316635687732342, "grad_norm": 0.4626846703147, "learning_rate": 8.001998824174605e-06, "loss": 0.036, "step": 6814 }, { "epoch": 0.316682156133829, "grad_norm": 0.642672519658021, "learning_rate": 8.00139700525175e-06, "loss": 0.064, "step": 6815 }, { "epoch": 0.316728624535316, "grad_norm": 1.4489043070855427, "learning_rate": 8.000795118344094e-06, "loss": 0.0676, "step": 6816 }, { "epoch": 0.31677509293680295, "grad_norm": 0.7640520751788087, "learning_rate": 8.000193163465271e-06, "loss": 0.0706, "step": 6817 }, { "epoch": 0.31682156133829, "grad_norm": 1.3706254425577868, "learning_rate": 7.999591140628915e-06, "loss": 0.0765, "step": 6818 }, { "epoch": 0.31686802973977696, "grad_norm": 3.918404490864476, "learning_rate": 7.99898904984866e-06, "loss": 0.0926, "step": 6819 }, { "epoch": 0.31691449814126393, "grad_norm": 1.5155905106450276, "learning_rate": 7.99838689113815e-06, "loss": 0.0729, "step": 6820 }, { "epoch": 0.3169609665427509, "grad_norm": 2.0068561373650127, "learning_rate": 7.997784664511021e-06, "loss": 0.0889, "step": 6821 }, { "epoch": 0.31700743494423794, "grad_norm": 0.9060674478852498, "learning_rate": 7.997182369980914e-06, "loss": 0.0739, "step": 6822 }, { "epoch": 0.3170539033457249, "grad_norm": 1.6249380188241391, "learning_rate": 7.99658000756147e-06, "loss": 0.0567, "step": 6823 }, { "epoch": 0.3171003717472119, "grad_norm": 1.5739263045455871, "learning_rate": 7.995977577266336e-06, "loss": 0.1052, "step": 6824 }, { "epoch": 0.31714684014869887, "grad_norm": 0.8542628712547436, "learning_rate": 7.995375079109157e-06, "loss": 0.0743, "step": 6825 }, { "epoch": 0.31719330855018585, "grad_norm": 0.774916794372586, "learning_rate": 7.994772513103581e-06, "loss": 0.0692, "step": 6826 }, { "epoch": 0.3172397769516729, "grad_norm": 0.8416957459394503, "learning_rate": 7.994169879263255e-06, "loss": 0.057, "step": 6827 }, { "epoch": 0.31728624535315986, "grad_norm": 1.1364113250963992, "learning_rate": 7.993567177601827e-06, "loss": 0.0934, "step": 6828 }, { "epoch": 0.31733271375464683, "grad_norm": 2.3399234846137413, "learning_rate": 7.992964408132955e-06, "loss": 0.0919, "step": 6829 }, { "epoch": 0.3173791821561338, "grad_norm": 0.3722496894088245, "learning_rate": 7.992361570870289e-06, "loss": 0.0244, "step": 6830 }, { "epoch": 0.31742565055762084, "grad_norm": 0.4780940619974697, "learning_rate": 7.991758665827482e-06, "loss": 0.0311, "step": 6831 }, { "epoch": 0.3174721189591078, "grad_norm": 0.9956702655270172, "learning_rate": 7.991155693018193e-06, "loss": 0.0713, "step": 6832 }, { "epoch": 0.3175185873605948, "grad_norm": 3.2051588995728606, "learning_rate": 7.99055265245608e-06, "loss": 0.0968, "step": 6833 }, { "epoch": 0.31756505576208177, "grad_norm": 1.751460038458329, "learning_rate": 7.989949544154803e-06, "loss": 0.0716, "step": 6834 }, { "epoch": 0.31761152416356875, "grad_norm": 1.1101466767333614, "learning_rate": 7.989346368128021e-06, "loss": 0.0508, "step": 6835 }, { "epoch": 0.3176579925650558, "grad_norm": 1.4298679979727726, "learning_rate": 7.988743124389398e-06, "loss": 0.0656, "step": 6836 }, { "epoch": 0.31770446096654276, "grad_norm": 1.3521919055134666, "learning_rate": 7.988139812952596e-06, "loss": 0.0869, "step": 6837 }, { "epoch": 0.31775092936802973, "grad_norm": 1.3554120140137802, "learning_rate": 7.987536433831285e-06, "loss": 0.0698, "step": 6838 }, { "epoch": 0.3177973977695167, "grad_norm": 1.0708341379413826, "learning_rate": 7.986932987039127e-06, "loss": 0.0519, "step": 6839 }, { "epoch": 0.31784386617100374, "grad_norm": 1.2132347930188854, "learning_rate": 7.986329472589793e-06, "loss": 0.0467, "step": 6840 }, { "epoch": 0.3178903345724907, "grad_norm": 1.5889205309334484, "learning_rate": 7.985725890496954e-06, "loss": 0.0809, "step": 6841 }, { "epoch": 0.3179368029739777, "grad_norm": 1.4493818840109691, "learning_rate": 7.985122240774281e-06, "loss": 0.0821, "step": 6842 }, { "epoch": 0.31798327137546467, "grad_norm": 1.1810769231597404, "learning_rate": 7.98451852343545e-06, "loss": 0.0788, "step": 6843 }, { "epoch": 0.31802973977695165, "grad_norm": 2.2638300003699046, "learning_rate": 7.98391473849413e-06, "loss": 0.079, "step": 6844 }, { "epoch": 0.3180762081784387, "grad_norm": 0.864304886798276, "learning_rate": 7.983310885964002e-06, "loss": 0.1083, "step": 6845 }, { "epoch": 0.31812267657992566, "grad_norm": 1.3911003753840114, "learning_rate": 7.982706965858743e-06, "loss": 0.049, "step": 6846 }, { "epoch": 0.31816914498141263, "grad_norm": 0.978853954262455, "learning_rate": 7.982102978192031e-06, "loss": 0.0588, "step": 6847 }, { "epoch": 0.3182156133828996, "grad_norm": 0.6574989797637639, "learning_rate": 7.981498922977548e-06, "loss": 0.0584, "step": 6848 }, { "epoch": 0.31826208178438664, "grad_norm": 1.0657945677934195, "learning_rate": 7.980894800228978e-06, "loss": 0.0694, "step": 6849 }, { "epoch": 0.3183085501858736, "grad_norm": 1.2694158254739554, "learning_rate": 7.980290609960002e-06, "loss": 0.0617, "step": 6850 }, { "epoch": 0.3183550185873606, "grad_norm": 0.9778940185026648, "learning_rate": 7.979686352184306e-06, "loss": 0.0597, "step": 6851 }, { "epoch": 0.31840148698884757, "grad_norm": 0.8413709047738159, "learning_rate": 7.979082026915582e-06, "loss": 0.0499, "step": 6852 }, { "epoch": 0.31844795539033455, "grad_norm": 1.622787918679914, "learning_rate": 7.97847763416751e-06, "loss": 0.1022, "step": 6853 }, { "epoch": 0.3184944237918216, "grad_norm": 1.2882303536862425, "learning_rate": 7.977873173953787e-06, "loss": 0.0747, "step": 6854 }, { "epoch": 0.31854089219330856, "grad_norm": 0.8944589470605186, "learning_rate": 7.977268646288102e-06, "loss": 0.0842, "step": 6855 }, { "epoch": 0.31858736059479553, "grad_norm": 1.1044882496689055, "learning_rate": 7.97666405118415e-06, "loss": 0.0871, "step": 6856 }, { "epoch": 0.3186338289962825, "grad_norm": 1.1654350819171477, "learning_rate": 7.976059388655622e-06, "loss": 0.107, "step": 6857 }, { "epoch": 0.31868029739776954, "grad_norm": 0.9830444276836112, "learning_rate": 7.975454658716218e-06, "loss": 0.0971, "step": 6858 }, { "epoch": 0.3187267657992565, "grad_norm": 1.2683355006338213, "learning_rate": 7.974849861379635e-06, "loss": 0.0408, "step": 6859 }, { "epoch": 0.3187732342007435, "grad_norm": 1.3808360665813533, "learning_rate": 7.97424499665957e-06, "loss": 0.056, "step": 6860 }, { "epoch": 0.31881970260223047, "grad_norm": 0.5893565885910902, "learning_rate": 7.973640064569727e-06, "loss": 0.0518, "step": 6861 }, { "epoch": 0.31886617100371745, "grad_norm": 0.7844797383749668, "learning_rate": 7.973035065123807e-06, "loss": 0.0433, "step": 6862 }, { "epoch": 0.3189126394052045, "grad_norm": 0.49729868918558945, "learning_rate": 7.972429998335511e-06, "loss": 0.0412, "step": 6863 }, { "epoch": 0.31895910780669146, "grad_norm": 1.1545248754116475, "learning_rate": 7.971824864218549e-06, "loss": 0.0468, "step": 6864 }, { "epoch": 0.31900557620817843, "grad_norm": 1.0742347667496583, "learning_rate": 7.971219662786626e-06, "loss": 0.0503, "step": 6865 }, { "epoch": 0.3190520446096654, "grad_norm": 0.6974293314505927, "learning_rate": 7.97061439405345e-06, "loss": 0.077, "step": 6866 }, { "epoch": 0.31909851301115244, "grad_norm": 3.094119533121921, "learning_rate": 7.970009058032732e-06, "loss": 0.1101, "step": 6867 }, { "epoch": 0.3191449814126394, "grad_norm": 1.1763811282434122, "learning_rate": 7.969403654738184e-06, "loss": 0.0459, "step": 6868 }, { "epoch": 0.3191914498141264, "grad_norm": 0.5807785435136505, "learning_rate": 7.968798184183516e-06, "loss": 0.0583, "step": 6869 }, { "epoch": 0.31923791821561337, "grad_norm": 0.5793327483300952, "learning_rate": 7.968192646382447e-06, "loss": 0.0564, "step": 6870 }, { "epoch": 0.31928438661710035, "grad_norm": 0.9173360982963251, "learning_rate": 7.96758704134869e-06, "loss": 0.0501, "step": 6871 }, { "epoch": 0.3193308550185874, "grad_norm": 1.4034940510861054, "learning_rate": 7.96698136909596e-06, "loss": 0.0962, "step": 6872 }, { "epoch": 0.31937732342007435, "grad_norm": 1.9564022731967587, "learning_rate": 7.966375629637983e-06, "loss": 0.0945, "step": 6873 }, { "epoch": 0.31942379182156133, "grad_norm": 2.0745934079708586, "learning_rate": 7.965769822988475e-06, "loss": 0.0663, "step": 6874 }, { "epoch": 0.3194702602230483, "grad_norm": 1.3090287147947615, "learning_rate": 7.965163949161158e-06, "loss": 0.0799, "step": 6875 }, { "epoch": 0.31951672862453534, "grad_norm": 1.0395759884493148, "learning_rate": 7.964558008169756e-06, "loss": 0.0656, "step": 6876 }, { "epoch": 0.3195631970260223, "grad_norm": 3.622964803333645, "learning_rate": 7.963952000027999e-06, "loss": 0.0886, "step": 6877 }, { "epoch": 0.3196096654275093, "grad_norm": 4.30068106406489, "learning_rate": 7.963345924749606e-06, "loss": 0.1022, "step": 6878 }, { "epoch": 0.31965613382899627, "grad_norm": 1.0104812445844435, "learning_rate": 7.96273978234831e-06, "loss": 0.064, "step": 6879 }, { "epoch": 0.31970260223048325, "grad_norm": 1.279150996559114, "learning_rate": 7.96213357283784e-06, "loss": 0.0473, "step": 6880 }, { "epoch": 0.3197490706319703, "grad_norm": 1.183699724823256, "learning_rate": 7.961527296231925e-06, "loss": 0.121, "step": 6881 }, { "epoch": 0.31979553903345725, "grad_norm": 2.3379783308619677, "learning_rate": 7.960920952544303e-06, "loss": 0.1053, "step": 6882 }, { "epoch": 0.31984200743494423, "grad_norm": 1.1229246191666489, "learning_rate": 7.960314541788703e-06, "loss": 0.0487, "step": 6883 }, { "epoch": 0.3198884758364312, "grad_norm": 1.2835022610798956, "learning_rate": 7.959708063978863e-06, "loss": 0.1058, "step": 6884 }, { "epoch": 0.31993494423791824, "grad_norm": 0.8499976238686756, "learning_rate": 7.959101519128522e-06, "loss": 0.0549, "step": 6885 }, { "epoch": 0.3199814126394052, "grad_norm": 0.64489089014012, "learning_rate": 7.958494907251415e-06, "loss": 0.0717, "step": 6886 }, { "epoch": 0.3200278810408922, "grad_norm": 0.546957116005972, "learning_rate": 7.957888228361286e-06, "loss": 0.0617, "step": 6887 }, { "epoch": 0.32007434944237917, "grad_norm": 1.0158332942126393, "learning_rate": 7.957281482471875e-06, "loss": 0.0658, "step": 6888 }, { "epoch": 0.32012081784386615, "grad_norm": 0.9242543415409955, "learning_rate": 7.956674669596926e-06, "loss": 0.0583, "step": 6889 }, { "epoch": 0.3201672862453532, "grad_norm": 1.5597695069404338, "learning_rate": 7.956067789750184e-06, "loss": 0.0804, "step": 6890 }, { "epoch": 0.32021375464684015, "grad_norm": 1.2622649202493483, "learning_rate": 7.955460842945396e-06, "loss": 0.0866, "step": 6891 }, { "epoch": 0.32026022304832713, "grad_norm": 3.2626833486301487, "learning_rate": 7.95485382919631e-06, "loss": 0.1006, "step": 6892 }, { "epoch": 0.3203066914498141, "grad_norm": 1.1092128992746846, "learning_rate": 7.954246748516671e-06, "loss": 0.0672, "step": 6893 }, { "epoch": 0.32035315985130114, "grad_norm": 1.8137110347869863, "learning_rate": 7.953639600920238e-06, "loss": 0.0667, "step": 6894 }, { "epoch": 0.3203996282527881, "grad_norm": 2.0491589984656295, "learning_rate": 7.953032386420756e-06, "loss": 0.1042, "step": 6895 }, { "epoch": 0.3204460966542751, "grad_norm": 1.0415603497862143, "learning_rate": 7.952425105031984e-06, "loss": 0.0745, "step": 6896 }, { "epoch": 0.32049256505576207, "grad_norm": 0.9308409227304957, "learning_rate": 7.951817756767677e-06, "loss": 0.0836, "step": 6897 }, { "epoch": 0.32053903345724905, "grad_norm": 1.2270263941951198, "learning_rate": 7.951210341641591e-06, "loss": 0.0908, "step": 6898 }, { "epoch": 0.3205855018587361, "grad_norm": 1.0286987258742752, "learning_rate": 7.950602859667484e-06, "loss": 0.0631, "step": 6899 }, { "epoch": 0.32063197026022305, "grad_norm": 0.4984774013387519, "learning_rate": 7.949995310859116e-06, "loss": 0.0359, "step": 6900 }, { "epoch": 0.32067843866171003, "grad_norm": 0.8993008141570449, "learning_rate": 7.94938769523025e-06, "loss": 0.0625, "step": 6901 }, { "epoch": 0.320724907063197, "grad_norm": 1.034561204930276, "learning_rate": 7.948780012794647e-06, "loss": 0.0414, "step": 6902 }, { "epoch": 0.32077137546468404, "grad_norm": 1.1090749985214754, "learning_rate": 7.948172263566077e-06, "loss": 0.0725, "step": 6903 }, { "epoch": 0.320817843866171, "grad_norm": 0.7719081565067368, "learning_rate": 7.9475644475583e-06, "loss": 0.0729, "step": 6904 }, { "epoch": 0.320864312267658, "grad_norm": 0.8112415185613936, "learning_rate": 7.946956564785086e-06, "loss": 0.0669, "step": 6905 }, { "epoch": 0.32091078066914497, "grad_norm": 0.6319633523562781, "learning_rate": 7.946348615260205e-06, "loss": 0.0517, "step": 6906 }, { "epoch": 0.32095724907063194, "grad_norm": 0.6831473035180847, "learning_rate": 7.945740598997425e-06, "loss": 0.0601, "step": 6907 }, { "epoch": 0.321003717472119, "grad_norm": 0.5280589804571666, "learning_rate": 7.94513251601052e-06, "loss": 0.0483, "step": 6908 }, { "epoch": 0.32105018587360595, "grad_norm": 0.7952608840288525, "learning_rate": 7.944524366313266e-06, "loss": 0.0533, "step": 6909 }, { "epoch": 0.32109665427509293, "grad_norm": 1.120756582754203, "learning_rate": 7.943916149919437e-06, "loss": 0.0548, "step": 6910 }, { "epoch": 0.3211431226765799, "grad_norm": 0.8739804924964035, "learning_rate": 7.943307866842806e-06, "loss": 0.0605, "step": 6911 }, { "epoch": 0.32118959107806694, "grad_norm": 0.49446138637703546, "learning_rate": 7.942699517097155e-06, "loss": 0.0334, "step": 6912 }, { "epoch": 0.3212360594795539, "grad_norm": 0.9256676625269691, "learning_rate": 7.942091100696263e-06, "loss": 0.0761, "step": 6913 }, { "epoch": 0.3212825278810409, "grad_norm": 1.4439460898917338, "learning_rate": 7.94148261765391e-06, "loss": 0.0695, "step": 6914 }, { "epoch": 0.32132899628252787, "grad_norm": 1.111415590045081, "learning_rate": 7.940874067983882e-06, "loss": 0.0432, "step": 6915 }, { "epoch": 0.32137546468401484, "grad_norm": 0.5861595937123094, "learning_rate": 7.94026545169996e-06, "loss": 0.0528, "step": 6916 }, { "epoch": 0.3214219330855019, "grad_norm": 0.565935138763566, "learning_rate": 7.93965676881593e-06, "loss": 0.0569, "step": 6917 }, { "epoch": 0.32146840148698885, "grad_norm": 1.3594366217954625, "learning_rate": 7.939048019345581e-06, "loss": 0.1009, "step": 6918 }, { "epoch": 0.32151486988847583, "grad_norm": 1.656506132001721, "learning_rate": 7.938439203302702e-06, "loss": 0.0794, "step": 6919 }, { "epoch": 0.3215613382899628, "grad_norm": 2.0001313465556785, "learning_rate": 7.937830320701081e-06, "loss": 0.0736, "step": 6920 }, { "epoch": 0.32160780669144984, "grad_norm": 0.865663629284732, "learning_rate": 7.937221371554512e-06, "loss": 0.0824, "step": 6921 }, { "epoch": 0.3216542750929368, "grad_norm": 0.9767264899296044, "learning_rate": 7.936612355876787e-06, "loss": 0.0884, "step": 6922 }, { "epoch": 0.3217007434944238, "grad_norm": 0.8406875036967648, "learning_rate": 7.936003273681702e-06, "loss": 0.0449, "step": 6923 }, { "epoch": 0.32174721189591077, "grad_norm": 1.2976701792846972, "learning_rate": 7.935394124983053e-06, "loss": 0.0509, "step": 6924 }, { "epoch": 0.32179368029739774, "grad_norm": 1.288073933195133, "learning_rate": 7.934784909794636e-06, "loss": 0.0863, "step": 6925 }, { "epoch": 0.3218401486988848, "grad_norm": 3.191000511051748, "learning_rate": 7.934175628130253e-06, "loss": 0.0752, "step": 6926 }, { "epoch": 0.32188661710037175, "grad_norm": 1.172090155536449, "learning_rate": 7.933566280003705e-06, "loss": 0.0852, "step": 6927 }, { "epoch": 0.32193308550185873, "grad_norm": 1.3591572650864734, "learning_rate": 7.932956865428792e-06, "loss": 0.05, "step": 6928 }, { "epoch": 0.3219795539033457, "grad_norm": 1.2560126933121871, "learning_rate": 7.932347384419319e-06, "loss": 0.0694, "step": 6929 }, { "epoch": 0.32202602230483274, "grad_norm": 1.1235935883021981, "learning_rate": 7.931737836989092e-06, "loss": 0.055, "step": 6930 }, { "epoch": 0.3220724907063197, "grad_norm": 1.0948310032868707, "learning_rate": 7.931128223151917e-06, "loss": 0.065, "step": 6931 }, { "epoch": 0.3221189591078067, "grad_norm": 0.9753604678479743, "learning_rate": 7.930518542921604e-06, "loss": 0.0873, "step": 6932 }, { "epoch": 0.32216542750929367, "grad_norm": 1.2689229365052566, "learning_rate": 7.92990879631196e-06, "loss": 0.0698, "step": 6933 }, { "epoch": 0.32221189591078064, "grad_norm": 2.7882586990374247, "learning_rate": 7.9292989833368e-06, "loss": 0.0637, "step": 6934 }, { "epoch": 0.3222583643122677, "grad_norm": 0.5974053205749396, "learning_rate": 7.928689104009933e-06, "loss": 0.0424, "step": 6935 }, { "epoch": 0.32230483271375465, "grad_norm": 1.1530950672310547, "learning_rate": 7.928079158345175e-06, "loss": 0.0443, "step": 6936 }, { "epoch": 0.32235130111524163, "grad_norm": 3.1221020823532455, "learning_rate": 7.927469146356343e-06, "loss": 0.0854, "step": 6937 }, { "epoch": 0.3223977695167286, "grad_norm": 0.6558390669239219, "learning_rate": 7.926859068057256e-06, "loss": 0.0547, "step": 6938 }, { "epoch": 0.32244423791821564, "grad_norm": 0.9891449064894958, "learning_rate": 7.926248923461728e-06, "loss": 0.0601, "step": 6939 }, { "epoch": 0.3224907063197026, "grad_norm": 1.1485101608315882, "learning_rate": 7.925638712583583e-06, "loss": 0.0599, "step": 6940 }, { "epoch": 0.3225371747211896, "grad_norm": 1.6961044796939952, "learning_rate": 7.925028435436642e-06, "loss": 0.0421, "step": 6941 }, { "epoch": 0.32258364312267657, "grad_norm": 4.430514260721786, "learning_rate": 7.92441809203473e-06, "loss": 0.0743, "step": 6942 }, { "epoch": 0.32263011152416354, "grad_norm": 1.621574781974699, "learning_rate": 7.923807682391666e-06, "loss": 0.0665, "step": 6943 }, { "epoch": 0.3226765799256506, "grad_norm": 3.612348816957225, "learning_rate": 7.923197206521285e-06, "loss": 0.0789, "step": 6944 }, { "epoch": 0.32272304832713755, "grad_norm": 0.7883929715526249, "learning_rate": 7.922586664437409e-06, "loss": 0.0721, "step": 6945 }, { "epoch": 0.32276951672862453, "grad_norm": 0.7880760850695868, "learning_rate": 7.921976056153869e-06, "loss": 0.0506, "step": 6946 }, { "epoch": 0.3228159851301115, "grad_norm": 0.7491687802377388, "learning_rate": 7.921365381684496e-06, "loss": 0.047, "step": 6947 }, { "epoch": 0.32286245353159854, "grad_norm": 1.0544175073410527, "learning_rate": 7.920754641043123e-06, "loss": 0.0643, "step": 6948 }, { "epoch": 0.3229089219330855, "grad_norm": 2.1631711279832624, "learning_rate": 7.920143834243581e-06, "loss": 0.0383, "step": 6949 }, { "epoch": 0.3229553903345725, "grad_norm": 1.7181859198378584, "learning_rate": 7.919532961299711e-06, "loss": 0.0806, "step": 6950 }, { "epoch": 0.32300185873605947, "grad_norm": 1.0307078751893808, "learning_rate": 7.918922022225347e-06, "loss": 0.0676, "step": 6951 }, { "epoch": 0.32304832713754644, "grad_norm": 0.822644291793466, "learning_rate": 7.918311017034324e-06, "loss": 0.0587, "step": 6952 }, { "epoch": 0.3230947955390335, "grad_norm": 1.970440166499329, "learning_rate": 7.917699945740486e-06, "loss": 0.0805, "step": 6953 }, { "epoch": 0.32314126394052045, "grad_norm": 2.1958459272383335, "learning_rate": 7.917088808357673e-06, "loss": 0.0595, "step": 6954 }, { "epoch": 0.3231877323420074, "grad_norm": 2.198000343172353, "learning_rate": 7.916477604899731e-06, "loss": 0.073, "step": 6955 }, { "epoch": 0.3232342007434944, "grad_norm": 1.4804900714095481, "learning_rate": 7.9158663353805e-06, "loss": 0.0948, "step": 6956 }, { "epoch": 0.32328066914498144, "grad_norm": 1.178242604675689, "learning_rate": 7.915254999813828e-06, "loss": 0.0761, "step": 6957 }, { "epoch": 0.3233271375464684, "grad_norm": 2.1677383955320186, "learning_rate": 7.914643598213561e-06, "loss": 0.0796, "step": 6958 }, { "epoch": 0.3233736059479554, "grad_norm": 0.9470886305975453, "learning_rate": 7.91403213059355e-06, "loss": 0.0841, "step": 6959 }, { "epoch": 0.32342007434944237, "grad_norm": 3.3845079638617794, "learning_rate": 7.913420596967643e-06, "loss": 0.0573, "step": 6960 }, { "epoch": 0.32346654275092934, "grad_norm": 0.8617094838290075, "learning_rate": 7.912808997349693e-06, "loss": 0.0631, "step": 6961 }, { "epoch": 0.3235130111524164, "grad_norm": 0.9160455926478055, "learning_rate": 7.912197331753554e-06, "loss": 0.0751, "step": 6962 }, { "epoch": 0.32355947955390335, "grad_norm": 1.3945732567894573, "learning_rate": 7.911585600193082e-06, "loss": 0.1063, "step": 6963 }, { "epoch": 0.3236059479553903, "grad_norm": 1.6227588480661723, "learning_rate": 7.91097380268213e-06, "loss": 0.0651, "step": 6964 }, { "epoch": 0.3236524163568773, "grad_norm": 0.7037879579084522, "learning_rate": 7.910361939234558e-06, "loss": 0.0585, "step": 6965 }, { "epoch": 0.32369888475836434, "grad_norm": 2.277975789587386, "learning_rate": 7.909750009864225e-06, "loss": 0.1282, "step": 6966 }, { "epoch": 0.3237453531598513, "grad_norm": 0.8456382570883949, "learning_rate": 7.909138014584992e-06, "loss": 0.0966, "step": 6967 }, { "epoch": 0.3237918215613383, "grad_norm": 1.1078805588862959, "learning_rate": 7.908525953410721e-06, "loss": 0.0438, "step": 6968 }, { "epoch": 0.32383828996282527, "grad_norm": 0.8555116899440319, "learning_rate": 7.907913826355275e-06, "loss": 0.0583, "step": 6969 }, { "epoch": 0.32388475836431224, "grad_norm": 1.7036450712796343, "learning_rate": 7.907301633432523e-06, "loss": 0.0708, "step": 6970 }, { "epoch": 0.3239312267657993, "grad_norm": 1.4782960030704748, "learning_rate": 7.906689374656327e-06, "loss": 0.0769, "step": 6971 }, { "epoch": 0.32397769516728625, "grad_norm": 1.135230734416558, "learning_rate": 7.906077050040558e-06, "loss": 0.0685, "step": 6972 }, { "epoch": 0.3240241635687732, "grad_norm": 0.6562434922421647, "learning_rate": 7.905464659599085e-06, "loss": 0.0662, "step": 6973 }, { "epoch": 0.3240706319702602, "grad_norm": 0.8457771246828648, "learning_rate": 7.904852203345782e-06, "loss": 0.0668, "step": 6974 }, { "epoch": 0.32411710037174724, "grad_norm": 2.0781368341206776, "learning_rate": 7.904239681294515e-06, "loss": 0.0705, "step": 6975 }, { "epoch": 0.3241635687732342, "grad_norm": 1.0107362589794893, "learning_rate": 7.903627093459165e-06, "loss": 0.0629, "step": 6976 }, { "epoch": 0.3242100371747212, "grad_norm": 3.6988997107637647, "learning_rate": 7.903014439853605e-06, "loss": 0.0651, "step": 6977 }, { "epoch": 0.32425650557620816, "grad_norm": 0.868848215740779, "learning_rate": 7.902401720491712e-06, "loss": 0.0779, "step": 6978 }, { "epoch": 0.32430297397769514, "grad_norm": 0.5614541995956694, "learning_rate": 7.901788935387364e-06, "loss": 0.0547, "step": 6979 }, { "epoch": 0.3243494423791822, "grad_norm": 1.3916625649049648, "learning_rate": 7.901176084554446e-06, "loss": 0.0631, "step": 6980 }, { "epoch": 0.32439591078066915, "grad_norm": 1.532402942578459, "learning_rate": 7.900563168006834e-06, "loss": 0.0677, "step": 6981 }, { "epoch": 0.3244423791821561, "grad_norm": 1.7712847258511557, "learning_rate": 7.899950185758414e-06, "loss": 0.0795, "step": 6982 }, { "epoch": 0.3244888475836431, "grad_norm": 1.057400374801689, "learning_rate": 7.89933713782307e-06, "loss": 0.0646, "step": 6983 }, { "epoch": 0.32453531598513014, "grad_norm": 1.969052635111559, "learning_rate": 7.898724024214688e-06, "loss": 0.0506, "step": 6984 }, { "epoch": 0.3245817843866171, "grad_norm": 0.7305235044391548, "learning_rate": 7.898110844947158e-06, "loss": 0.0613, "step": 6985 }, { "epoch": 0.3246282527881041, "grad_norm": 3.665176094424777, "learning_rate": 7.897497600034366e-06, "loss": 0.1063, "step": 6986 }, { "epoch": 0.32467472118959106, "grad_norm": 1.0888066352429835, "learning_rate": 7.896884289490203e-06, "loss": 0.0729, "step": 6987 }, { "epoch": 0.32472118959107804, "grad_norm": 3.5471204992917693, "learning_rate": 7.896270913328561e-06, "loss": 0.038, "step": 6988 }, { "epoch": 0.3247676579925651, "grad_norm": 1.0337305673373671, "learning_rate": 7.895657471563335e-06, "loss": 0.0614, "step": 6989 }, { "epoch": 0.32481412639405205, "grad_norm": 0.4879303979895426, "learning_rate": 7.895043964208421e-06, "loss": 0.0384, "step": 6990 }, { "epoch": 0.324860594795539, "grad_norm": 0.8341418148428206, "learning_rate": 7.894430391277714e-06, "loss": 0.0483, "step": 6991 }, { "epoch": 0.324907063197026, "grad_norm": 0.8785954473160414, "learning_rate": 7.893816752785113e-06, "loss": 0.1077, "step": 6992 }, { "epoch": 0.32495353159851303, "grad_norm": 1.2744916979586876, "learning_rate": 7.893203048744516e-06, "loss": 0.072, "step": 6993 }, { "epoch": 0.325, "grad_norm": 6.842994376447199, "learning_rate": 7.892589279169824e-06, "loss": 0.1625, "step": 6994 }, { "epoch": 0.325046468401487, "grad_norm": 2.2636936983521805, "learning_rate": 7.891975444074942e-06, "loss": 0.1034, "step": 6995 }, { "epoch": 0.32509293680297396, "grad_norm": 1.5046483677076277, "learning_rate": 7.891361543473773e-06, "loss": 0.0595, "step": 6996 }, { "epoch": 0.32513940520446094, "grad_norm": 1.4016274456547546, "learning_rate": 7.89074757738022e-06, "loss": 0.0734, "step": 6997 }, { "epoch": 0.325185873605948, "grad_norm": 3.082817519093343, "learning_rate": 7.890133545808191e-06, "loss": 0.1388, "step": 6998 }, { "epoch": 0.32523234200743495, "grad_norm": 1.580866748631769, "learning_rate": 7.8895194487716e-06, "loss": 0.0538, "step": 6999 }, { "epoch": 0.3252788104089219, "grad_norm": 1.3432246180928686, "learning_rate": 7.888905286284348e-06, "loss": 0.0801, "step": 7000 }, { "epoch": 0.3253252788104089, "grad_norm": 0.7391550894680373, "learning_rate": 7.88829105836035e-06, "loss": 0.0417, "step": 7001 }, { "epoch": 0.32537174721189593, "grad_norm": 1.2507846544480203, "learning_rate": 7.887676765013524e-06, "loss": 0.0478, "step": 7002 }, { "epoch": 0.3254182156133829, "grad_norm": 2.2151031722638033, "learning_rate": 7.887062406257778e-06, "loss": 0.0826, "step": 7003 }, { "epoch": 0.3254646840148699, "grad_norm": 3.151376920606423, "learning_rate": 7.886447982107028e-06, "loss": 0.0641, "step": 7004 }, { "epoch": 0.32551115241635686, "grad_norm": 1.37184111298863, "learning_rate": 7.885833492575194e-06, "loss": 0.1074, "step": 7005 }, { "epoch": 0.32555762081784384, "grad_norm": 2.0742775049432227, "learning_rate": 7.885218937676196e-06, "loss": 0.0684, "step": 7006 }, { "epoch": 0.32560408921933087, "grad_norm": 2.411426452762556, "learning_rate": 7.884604317423952e-06, "loss": 0.0553, "step": 7007 }, { "epoch": 0.32565055762081785, "grad_norm": 1.2333643466954989, "learning_rate": 7.883989631832383e-06, "loss": 0.0503, "step": 7008 }, { "epoch": 0.3256970260223048, "grad_norm": 2.883159947326353, "learning_rate": 7.883374880915416e-06, "loss": 0.0977, "step": 7009 }, { "epoch": 0.3257434944237918, "grad_norm": 1.8111266793109062, "learning_rate": 7.88276006468697e-06, "loss": 0.1153, "step": 7010 }, { "epoch": 0.32578996282527883, "grad_norm": 1.8144948181940823, "learning_rate": 7.882145183160976e-06, "loss": 0.0756, "step": 7011 }, { "epoch": 0.3258364312267658, "grad_norm": 19.911158482878125, "learning_rate": 7.88153023635136e-06, "loss": 0.0856, "step": 7012 }, { "epoch": 0.3258828996282528, "grad_norm": 1.7882274579617372, "learning_rate": 7.880915224272051e-06, "loss": 0.1007, "step": 7013 }, { "epoch": 0.32592936802973976, "grad_norm": 1.4553639735070352, "learning_rate": 7.88030014693698e-06, "loss": 0.0539, "step": 7014 }, { "epoch": 0.32597583643122674, "grad_norm": 1.783170946660996, "learning_rate": 7.87968500436008e-06, "loss": 0.0714, "step": 7015 }, { "epoch": 0.32602230483271377, "grad_norm": 2.8184293569859973, "learning_rate": 7.879069796555285e-06, "loss": 0.0922, "step": 7016 }, { "epoch": 0.32606877323420075, "grad_norm": 1.9482521603167389, "learning_rate": 7.878454523536528e-06, "loss": 0.0943, "step": 7017 }, { "epoch": 0.3261152416356877, "grad_norm": 2.684224065266787, "learning_rate": 7.877839185317745e-06, "loss": 0.0715, "step": 7018 }, { "epoch": 0.3261617100371747, "grad_norm": 5.1455142342722935, "learning_rate": 7.877223781912877e-06, "loss": 0.086, "step": 7019 }, { "epoch": 0.32620817843866173, "grad_norm": 2.1205814167607424, "learning_rate": 7.876608313335861e-06, "loss": 0.0785, "step": 7020 }, { "epoch": 0.3262546468401487, "grad_norm": 3.0578946480086455, "learning_rate": 7.875992779600641e-06, "loss": 0.0851, "step": 7021 }, { "epoch": 0.3263011152416357, "grad_norm": 1.1249369469932051, "learning_rate": 7.875377180721155e-06, "loss": 0.0939, "step": 7022 }, { "epoch": 0.32634758364312266, "grad_norm": 0.7640979287832788, "learning_rate": 7.874761516711351e-06, "loss": 0.0465, "step": 7023 }, { "epoch": 0.32639405204460964, "grad_norm": 3.7644029988642256, "learning_rate": 7.874145787585174e-06, "loss": 0.0645, "step": 7024 }, { "epoch": 0.32644052044609667, "grad_norm": 1.700260040862967, "learning_rate": 7.873529993356569e-06, "loss": 0.0807, "step": 7025 }, { "epoch": 0.32648698884758365, "grad_norm": 0.9581676615182378, "learning_rate": 7.872914134039485e-06, "loss": 0.08, "step": 7026 }, { "epoch": 0.3265334572490706, "grad_norm": 1.12836828201484, "learning_rate": 7.872298209647873e-06, "loss": 0.0833, "step": 7027 }, { "epoch": 0.3265799256505576, "grad_norm": 2.1107525721959868, "learning_rate": 7.871682220195681e-06, "loss": 0.0846, "step": 7028 }, { "epoch": 0.32662639405204463, "grad_norm": 0.8290580331558639, "learning_rate": 7.871066165696868e-06, "loss": 0.0626, "step": 7029 }, { "epoch": 0.3266728624535316, "grad_norm": 6.854830080762693, "learning_rate": 7.870450046165383e-06, "loss": 0.0444, "step": 7030 }, { "epoch": 0.3267193308550186, "grad_norm": 0.7434663992667743, "learning_rate": 7.869833861615184e-06, "loss": 0.0677, "step": 7031 }, { "epoch": 0.32676579925650556, "grad_norm": 1.379547986827645, "learning_rate": 7.869217612060226e-06, "loss": 0.0733, "step": 7032 }, { "epoch": 0.32681226765799254, "grad_norm": 2.886071552194487, "learning_rate": 7.86860129751447e-06, "loss": 0.0864, "step": 7033 }, { "epoch": 0.32685873605947957, "grad_norm": 1.4331154218885112, "learning_rate": 7.867984917991875e-06, "loss": 0.0571, "step": 7034 }, { "epoch": 0.32690520446096655, "grad_norm": 1.8969016008547506, "learning_rate": 7.867368473506404e-06, "loss": 0.0796, "step": 7035 }, { "epoch": 0.3269516728624535, "grad_norm": 2.373110175276219, "learning_rate": 7.86675196407202e-06, "loss": 0.075, "step": 7036 }, { "epoch": 0.3269981412639405, "grad_norm": 4.025237526321896, "learning_rate": 7.866135389702684e-06, "loss": 0.1214, "step": 7037 }, { "epoch": 0.32704460966542753, "grad_norm": 4.16846466974952, "learning_rate": 7.865518750412369e-06, "loss": 0.0888, "step": 7038 }, { "epoch": 0.3270910780669145, "grad_norm": 2.6321953988029363, "learning_rate": 7.864902046215034e-06, "loss": 0.0946, "step": 7039 }, { "epoch": 0.3271375464684015, "grad_norm": 2.531216322641127, "learning_rate": 7.864285277124656e-06, "loss": 0.084, "step": 7040 }, { "epoch": 0.32718401486988846, "grad_norm": 2.9589226151744583, "learning_rate": 7.8636684431552e-06, "loss": 0.0862, "step": 7041 }, { "epoch": 0.32723048327137544, "grad_norm": 1.8433191352590519, "learning_rate": 7.86305154432064e-06, "loss": 0.079, "step": 7042 }, { "epoch": 0.32727695167286247, "grad_norm": 3.679962882254671, "learning_rate": 7.86243458063495e-06, "loss": 0.0652, "step": 7043 }, { "epoch": 0.32732342007434945, "grad_norm": 2.0790642245056237, "learning_rate": 7.861817552112104e-06, "loss": 0.0746, "step": 7044 }, { "epoch": 0.3273698884758364, "grad_norm": 2.362339501004012, "learning_rate": 7.861200458766077e-06, "loss": 0.0913, "step": 7045 }, { "epoch": 0.3274163568773234, "grad_norm": 0.8538298668551294, "learning_rate": 7.860583300610849e-06, "loss": 0.0508, "step": 7046 }, { "epoch": 0.32746282527881043, "grad_norm": 1.2135923498750716, "learning_rate": 7.859966077660398e-06, "loss": 0.0371, "step": 7047 }, { "epoch": 0.3275092936802974, "grad_norm": 1.946139974163504, "learning_rate": 7.859348789928707e-06, "loss": 0.0986, "step": 7048 }, { "epoch": 0.3275557620817844, "grad_norm": 4.580149233265066, "learning_rate": 7.858731437429755e-06, "loss": 0.077, "step": 7049 }, { "epoch": 0.32760223048327136, "grad_norm": 1.4320207631517325, "learning_rate": 7.858114020177528e-06, "loss": 0.1561, "step": 7050 }, { "epoch": 0.32764869888475834, "grad_norm": 2.708350780386967, "learning_rate": 7.85749653818601e-06, "loss": 0.1029, "step": 7051 }, { "epoch": 0.32769516728624537, "grad_norm": 1.990455354609328, "learning_rate": 7.856878991469187e-06, "loss": 0.0962, "step": 7052 }, { "epoch": 0.32774163568773235, "grad_norm": 1.670147026601788, "learning_rate": 7.856261380041049e-06, "loss": 0.091, "step": 7053 }, { "epoch": 0.3277881040892193, "grad_norm": 0.7882033521081977, "learning_rate": 7.855643703915585e-06, "loss": 0.0694, "step": 7054 }, { "epoch": 0.3278345724907063, "grad_norm": 0.6939493421121007, "learning_rate": 7.855025963106785e-06, "loss": 0.0609, "step": 7055 }, { "epoch": 0.32788104089219333, "grad_norm": 1.239972409890148, "learning_rate": 7.854408157628643e-06, "loss": 0.093, "step": 7056 }, { "epoch": 0.3279275092936803, "grad_norm": 1.6459316095325736, "learning_rate": 7.853790287495151e-06, "loss": 0.0672, "step": 7057 }, { "epoch": 0.3279739776951673, "grad_norm": 4.5658045655068955, "learning_rate": 7.853172352720304e-06, "loss": 0.0777, "step": 7058 }, { "epoch": 0.32802044609665426, "grad_norm": 1.2468085775775861, "learning_rate": 7.852554353318104e-06, "loss": 0.0804, "step": 7059 }, { "epoch": 0.32806691449814124, "grad_norm": 1.5858425094454751, "learning_rate": 7.851936289302543e-06, "loss": 0.0636, "step": 7060 }, { "epoch": 0.32811338289962827, "grad_norm": 1.2451641637067077, "learning_rate": 7.851318160687625e-06, "loss": 0.0561, "step": 7061 }, { "epoch": 0.32815985130111525, "grad_norm": 66.8998698085003, "learning_rate": 7.850699967487348e-06, "loss": 0.1239, "step": 7062 }, { "epoch": 0.3282063197026022, "grad_norm": 1.473435239673106, "learning_rate": 7.850081709715717e-06, "loss": 0.1106, "step": 7063 }, { "epoch": 0.3282527881040892, "grad_norm": 0.7923917682230782, "learning_rate": 7.849463387386736e-06, "loss": 0.0482, "step": 7064 }, { "epoch": 0.32829925650557623, "grad_norm": 3.1564353601720576, "learning_rate": 7.84884500051441e-06, "loss": 0.0687, "step": 7065 }, { "epoch": 0.3283457249070632, "grad_norm": 4.6264578643526875, "learning_rate": 7.848226549112745e-06, "loss": 0.306, "step": 7066 }, { "epoch": 0.3283921933085502, "grad_norm": 6.083662531190567, "learning_rate": 7.847608033195753e-06, "loss": 0.2768, "step": 7067 }, { "epoch": 0.32843866171003716, "grad_norm": 6.569634443423151, "learning_rate": 7.846989452777443e-06, "loss": 0.3607, "step": 7068 }, { "epoch": 0.32848513011152414, "grad_norm": 3.895262624983083, "learning_rate": 7.846370807871822e-06, "loss": 0.1797, "step": 7069 }, { "epoch": 0.32853159851301117, "grad_norm": 8.859805526387339, "learning_rate": 7.845752098492909e-06, "loss": 0.2027, "step": 7070 }, { "epoch": 0.32857806691449815, "grad_norm": 1.17329566284285, "learning_rate": 7.845133324654714e-06, "loss": 0.104, "step": 7071 }, { "epoch": 0.3286245353159851, "grad_norm": 10.877574129844549, "learning_rate": 7.844514486371257e-06, "loss": 0.2569, "step": 7072 }, { "epoch": 0.3286710037174721, "grad_norm": 2.6876838082634573, "learning_rate": 7.84389558365655e-06, "loss": 0.2032, "step": 7073 }, { "epoch": 0.32871747211895913, "grad_norm": 2.156989283590995, "learning_rate": 7.843276616524619e-06, "loss": 0.1026, "step": 7074 }, { "epoch": 0.3287639405204461, "grad_norm": 4.602794639396383, "learning_rate": 7.842657584989477e-06, "loss": 0.1126, "step": 7075 }, { "epoch": 0.3288104089219331, "grad_norm": 3.583414660516552, "learning_rate": 7.842038489065151e-06, "loss": 0.0971, "step": 7076 }, { "epoch": 0.32885687732342006, "grad_norm": 2.867240867680583, "learning_rate": 7.84141932876566e-06, "loss": 0.1658, "step": 7077 }, { "epoch": 0.32890334572490704, "grad_norm": 4.270553904886326, "learning_rate": 7.840800104105033e-06, "loss": 0.1917, "step": 7078 }, { "epoch": 0.32894981412639407, "grad_norm": 2.320075405903987, "learning_rate": 7.840180815097291e-06, "loss": 0.1489, "step": 7079 }, { "epoch": 0.32899628252788105, "grad_norm": 2.3969251701155385, "learning_rate": 7.839561461756466e-06, "loss": 0.1569, "step": 7080 }, { "epoch": 0.329042750929368, "grad_norm": 0.8861305828946497, "learning_rate": 7.838942044096585e-06, "loss": 0.0853, "step": 7081 }, { "epoch": 0.329089219330855, "grad_norm": 4.9705693233298565, "learning_rate": 7.83832256213168e-06, "loss": 0.2734, "step": 7082 }, { "epoch": 0.32913568773234203, "grad_norm": 1.6741844747122099, "learning_rate": 7.837703015875779e-06, "loss": 0.0982, "step": 7083 }, { "epoch": 0.329182156133829, "grad_norm": 4.094309195574908, "learning_rate": 7.83708340534292e-06, "loss": 0.0994, "step": 7084 }, { "epoch": 0.329228624535316, "grad_norm": 1.2872883015451595, "learning_rate": 7.836463730547134e-06, "loss": 0.1039, "step": 7085 }, { "epoch": 0.32927509293680296, "grad_norm": 1.7318798697074227, "learning_rate": 7.835843991502461e-06, "loss": 0.0738, "step": 7086 }, { "epoch": 0.32932156133828994, "grad_norm": 1.7627648086022978, "learning_rate": 7.835224188222937e-06, "loss": 0.0777, "step": 7087 }, { "epoch": 0.32936802973977697, "grad_norm": 1.4404746433658118, "learning_rate": 7.8346043207226e-06, "loss": 0.0731, "step": 7088 }, { "epoch": 0.32941449814126395, "grad_norm": 1.2109696578291411, "learning_rate": 7.833984389015495e-06, "loss": 0.0673, "step": 7089 }, { "epoch": 0.3294609665427509, "grad_norm": 3.6619954791279232, "learning_rate": 7.833364393115657e-06, "loss": 0.132, "step": 7090 }, { "epoch": 0.3295074349442379, "grad_norm": 1.6456841714486372, "learning_rate": 7.832744333037135e-06, "loss": 0.1582, "step": 7091 }, { "epoch": 0.32955390334572493, "grad_norm": 1.8480681312812643, "learning_rate": 7.832124208793972e-06, "loss": 0.0854, "step": 7092 }, { "epoch": 0.3296003717472119, "grad_norm": 2.142308476325913, "learning_rate": 7.831504020400213e-06, "loss": 0.1323, "step": 7093 }, { "epoch": 0.3296468401486989, "grad_norm": 2.9826274247929367, "learning_rate": 7.83088376786991e-06, "loss": 0.116, "step": 7094 }, { "epoch": 0.32969330855018586, "grad_norm": 1.9678426945891196, "learning_rate": 7.830263451217111e-06, "loss": 0.0908, "step": 7095 }, { "epoch": 0.32973977695167284, "grad_norm": 4.560121487528582, "learning_rate": 7.829643070455866e-06, "loss": 0.0648, "step": 7096 }, { "epoch": 0.32978624535315987, "grad_norm": 0.8629237499468301, "learning_rate": 7.829022625600226e-06, "loss": 0.1065, "step": 7097 }, { "epoch": 0.32983271375464684, "grad_norm": 1.0433426046804222, "learning_rate": 7.828402116664247e-06, "loss": 0.0744, "step": 7098 }, { "epoch": 0.3298791821561338, "grad_norm": 1.4621349673149973, "learning_rate": 7.827781543661981e-06, "loss": 0.1221, "step": 7099 }, { "epoch": 0.3299256505576208, "grad_norm": 2.710169329821063, "learning_rate": 7.827160906607489e-06, "loss": 0.2365, "step": 7100 }, { "epoch": 0.32997211895910783, "grad_norm": 3.1101153094808995, "learning_rate": 7.826540205514828e-06, "loss": 0.1086, "step": 7101 }, { "epoch": 0.3300185873605948, "grad_norm": 1.7802268709420337, "learning_rate": 7.825919440398053e-06, "loss": 0.1395, "step": 7102 }, { "epoch": 0.3300650557620818, "grad_norm": 2.6352919998902693, "learning_rate": 7.82529861127123e-06, "loss": 0.0841, "step": 7103 }, { "epoch": 0.33011152416356876, "grad_norm": 1.6632586951586907, "learning_rate": 7.824677718148421e-06, "loss": 0.0981, "step": 7104 }, { "epoch": 0.33015799256505574, "grad_norm": 0.906876602424331, "learning_rate": 7.824056761043689e-06, "loss": 0.0627, "step": 7105 }, { "epoch": 0.33020446096654277, "grad_norm": 2.4978765802482323, "learning_rate": 7.823435739971098e-06, "loss": 0.0749, "step": 7106 }, { "epoch": 0.33025092936802974, "grad_norm": 3.1364169838905847, "learning_rate": 7.822814654944715e-06, "loss": 0.142, "step": 7107 }, { "epoch": 0.3302973977695167, "grad_norm": 1.2868119556604436, "learning_rate": 7.822193505978611e-06, "loss": 0.0926, "step": 7108 }, { "epoch": 0.3303438661710037, "grad_norm": 1.3800890894450906, "learning_rate": 7.821572293086852e-06, "loss": 0.0995, "step": 7109 }, { "epoch": 0.33039033457249073, "grad_norm": 3.9736884169593223, "learning_rate": 7.820951016283512e-06, "loss": 0.0785, "step": 7110 }, { "epoch": 0.3304368029739777, "grad_norm": 3.8838478313348586, "learning_rate": 7.820329675582663e-06, "loss": 0.0908, "step": 7111 }, { "epoch": 0.3304832713754647, "grad_norm": 1.1696808174901019, "learning_rate": 7.819708270998377e-06, "loss": 0.073, "step": 7112 }, { "epoch": 0.33052973977695166, "grad_norm": 2.4994910683058085, "learning_rate": 7.819086802544733e-06, "loss": 0.1216, "step": 7113 }, { "epoch": 0.33057620817843864, "grad_norm": 1.9327906313215317, "learning_rate": 7.818465270235805e-06, "loss": 0.1062, "step": 7114 }, { "epoch": 0.33062267657992567, "grad_norm": 3.0427726728088995, "learning_rate": 7.817843674085671e-06, "loss": 0.0739, "step": 7115 }, { "epoch": 0.33066914498141264, "grad_norm": 0.8014465911538255, "learning_rate": 7.817222014108412e-06, "loss": 0.0494, "step": 7116 }, { "epoch": 0.3307156133828996, "grad_norm": 1.105984166545279, "learning_rate": 7.81660029031811e-06, "loss": 0.0969, "step": 7117 }, { "epoch": 0.3307620817843866, "grad_norm": 2.039550689314718, "learning_rate": 7.81597850272885e-06, "loss": 0.0858, "step": 7118 }, { "epoch": 0.33080855018587363, "grad_norm": 1.8236072256880969, "learning_rate": 7.815356651354709e-06, "loss": 0.0788, "step": 7119 }, { "epoch": 0.3308550185873606, "grad_norm": 1.6840783962234103, "learning_rate": 7.814734736209777e-06, "loss": 0.0743, "step": 7120 }, { "epoch": 0.3309014869888476, "grad_norm": 4.47312357301412, "learning_rate": 7.814112757308143e-06, "loss": 0.0949, "step": 7121 }, { "epoch": 0.33094795539033456, "grad_norm": 1.9158371701419874, "learning_rate": 7.813490714663892e-06, "loss": 0.0745, "step": 7122 }, { "epoch": 0.33099442379182153, "grad_norm": 1.8092434408753808, "learning_rate": 7.812868608291116e-06, "loss": 0.1015, "step": 7123 }, { "epoch": 0.33104089219330857, "grad_norm": 1.1340615494670365, "learning_rate": 7.812246438203905e-06, "loss": 0.0454, "step": 7124 }, { "epoch": 0.33108736059479554, "grad_norm": 1.0489480859678495, "learning_rate": 7.811624204416352e-06, "loss": 0.0789, "step": 7125 }, { "epoch": 0.3311338289962825, "grad_norm": 2.175193117113776, "learning_rate": 7.81100190694255e-06, "loss": 0.1304, "step": 7126 }, { "epoch": 0.3311802973977695, "grad_norm": 0.8910418272349675, "learning_rate": 7.8103795457966e-06, "loss": 0.074, "step": 7127 }, { "epoch": 0.33122676579925653, "grad_norm": 0.8084325517932361, "learning_rate": 7.809757120992592e-06, "loss": 0.0596, "step": 7128 }, { "epoch": 0.3312732342007435, "grad_norm": 2.337738741455802, "learning_rate": 7.809134632544631e-06, "loss": 0.1432, "step": 7129 }, { "epoch": 0.3313197026022305, "grad_norm": 1.2431450028604745, "learning_rate": 7.808512080466813e-06, "loss": 0.0713, "step": 7130 }, { "epoch": 0.33136617100371746, "grad_norm": 2.991019690851461, "learning_rate": 7.807889464773238e-06, "loss": 0.1026, "step": 7131 }, { "epoch": 0.33141263940520443, "grad_norm": 5.123049204084646, "learning_rate": 7.807266785478014e-06, "loss": 0.0783, "step": 7132 }, { "epoch": 0.33145910780669147, "grad_norm": 1.418415097715892, "learning_rate": 7.806644042595242e-06, "loss": 0.09, "step": 7133 }, { "epoch": 0.33150557620817844, "grad_norm": 0.6401128561466991, "learning_rate": 7.806021236139027e-06, "loss": 0.0617, "step": 7134 }, { "epoch": 0.3315520446096654, "grad_norm": 1.4985673331121125, "learning_rate": 7.80539836612348e-06, "loss": 0.0694, "step": 7135 }, { "epoch": 0.3315985130111524, "grad_norm": 0.8937768739339931, "learning_rate": 7.804775432562708e-06, "loss": 0.049, "step": 7136 }, { "epoch": 0.33164498141263943, "grad_norm": 1.7312253565011282, "learning_rate": 7.804152435470818e-06, "loss": 0.085, "step": 7137 }, { "epoch": 0.3316914498141264, "grad_norm": 2.3639140773369562, "learning_rate": 7.803529374861925e-06, "loss": 0.0925, "step": 7138 }, { "epoch": 0.3317379182156134, "grad_norm": 1.33417281477558, "learning_rate": 7.802906250750142e-06, "loss": 0.0708, "step": 7139 }, { "epoch": 0.33178438661710036, "grad_norm": 0.8194634285673597, "learning_rate": 7.80228306314958e-06, "loss": 0.0512, "step": 7140 }, { "epoch": 0.33183085501858733, "grad_norm": 3.2014862121573877, "learning_rate": 7.80165981207436e-06, "loss": 0.122, "step": 7141 }, { "epoch": 0.33187732342007437, "grad_norm": 3.5456931741776176, "learning_rate": 7.801036497538597e-06, "loss": 0.0748, "step": 7142 }, { "epoch": 0.33192379182156134, "grad_norm": 0.8081910473291095, "learning_rate": 7.800413119556406e-06, "loss": 0.0756, "step": 7143 }, { "epoch": 0.3319702602230483, "grad_norm": 6.334594803488664, "learning_rate": 7.799789678141912e-06, "loss": 0.0829, "step": 7144 }, { "epoch": 0.3320167286245353, "grad_norm": 2.315424065152195, "learning_rate": 7.799166173309236e-06, "loss": 0.1035, "step": 7145 }, { "epoch": 0.3320631970260223, "grad_norm": 2.1646533514123925, "learning_rate": 7.798542605072499e-06, "loss": 0.0912, "step": 7146 }, { "epoch": 0.3321096654275093, "grad_norm": 1.3000950922835088, "learning_rate": 7.797918973445826e-06, "loss": 0.0728, "step": 7147 }, { "epoch": 0.3321561338289963, "grad_norm": 1.4806138102880528, "learning_rate": 7.797295278443345e-06, "loss": 0.0809, "step": 7148 }, { "epoch": 0.33220260223048326, "grad_norm": 2.745365625874226, "learning_rate": 7.796671520079182e-06, "loss": 0.0713, "step": 7149 }, { "epoch": 0.33224907063197023, "grad_norm": 2.705630450656405, "learning_rate": 7.796047698367464e-06, "loss": 0.0762, "step": 7150 }, { "epoch": 0.33229553903345727, "grad_norm": 5.898916551156006, "learning_rate": 7.795423813322325e-06, "loss": 0.0754, "step": 7151 }, { "epoch": 0.33234200743494424, "grad_norm": 1.5456538120399075, "learning_rate": 7.794799864957892e-06, "loss": 0.1033, "step": 7152 }, { "epoch": 0.3323884758364312, "grad_norm": 0.6863421647330451, "learning_rate": 7.794175853288301e-06, "loss": 0.0551, "step": 7153 }, { "epoch": 0.3324349442379182, "grad_norm": 1.2196864900431144, "learning_rate": 7.793551778327688e-06, "loss": 0.0701, "step": 7154 }, { "epoch": 0.3324814126394052, "grad_norm": 1.3814708460087117, "learning_rate": 7.792927640090186e-06, "loss": 0.0845, "step": 7155 }, { "epoch": 0.3325278810408922, "grad_norm": 1.3941628249781406, "learning_rate": 7.792303438589933e-06, "loss": 0.0771, "step": 7156 }, { "epoch": 0.3325743494423792, "grad_norm": 1.4039589189773707, "learning_rate": 7.791679173841069e-06, "loss": 0.0802, "step": 7157 }, { "epoch": 0.33262081784386616, "grad_norm": 1.3973189619242736, "learning_rate": 7.791054845857733e-06, "loss": 0.0785, "step": 7158 }, { "epoch": 0.33266728624535313, "grad_norm": 0.9978036840616962, "learning_rate": 7.790430454654068e-06, "loss": 0.033, "step": 7159 }, { "epoch": 0.33271375464684017, "grad_norm": 1.487671280472247, "learning_rate": 7.789806000244216e-06, "loss": 0.1119, "step": 7160 }, { "epoch": 0.33276022304832714, "grad_norm": 1.201966929045359, "learning_rate": 7.78918148264232e-06, "loss": 0.0854, "step": 7161 }, { "epoch": 0.3328066914498141, "grad_norm": 1.8182234118847622, "learning_rate": 7.78855690186253e-06, "loss": 0.1123, "step": 7162 }, { "epoch": 0.3328531598513011, "grad_norm": 1.395074925853279, "learning_rate": 7.787932257918991e-06, "loss": 0.0737, "step": 7163 }, { "epoch": 0.3328996282527881, "grad_norm": 0.805898622671538, "learning_rate": 7.78730755082585e-06, "loss": 0.0575, "step": 7164 }, { "epoch": 0.3329460966542751, "grad_norm": 2.2967146844164756, "learning_rate": 7.786682780597263e-06, "loss": 0.1232, "step": 7165 }, { "epoch": 0.3329925650557621, "grad_norm": 1.5406172424598439, "learning_rate": 7.786057947247376e-06, "loss": 0.0594, "step": 7166 }, { "epoch": 0.33303903345724906, "grad_norm": 1.7359526301454198, "learning_rate": 7.785433050790346e-06, "loss": 0.0749, "step": 7167 }, { "epoch": 0.33308550185873603, "grad_norm": 1.9379178479197536, "learning_rate": 7.784808091240323e-06, "loss": 0.1524, "step": 7168 }, { "epoch": 0.33313197026022306, "grad_norm": 0.7116461487125132, "learning_rate": 7.784183068611469e-06, "loss": 0.0647, "step": 7169 }, { "epoch": 0.33317843866171004, "grad_norm": 1.5150752904467926, "learning_rate": 7.783557982917935e-06, "loss": 0.0848, "step": 7170 }, { "epoch": 0.333224907063197, "grad_norm": 1.1743065281205702, "learning_rate": 7.782932834173885e-06, "loss": 0.0714, "step": 7171 }, { "epoch": 0.333271375464684, "grad_norm": 0.6806091901449746, "learning_rate": 7.782307622393476e-06, "loss": 0.0501, "step": 7172 }, { "epoch": 0.333317843866171, "grad_norm": 0.9695448807438648, "learning_rate": 7.78168234759087e-06, "loss": 0.0444, "step": 7173 }, { "epoch": 0.333364312267658, "grad_norm": 0.6941959356898226, "learning_rate": 7.781057009780235e-06, "loss": 0.0476, "step": 7174 }, { "epoch": 0.333410780669145, "grad_norm": 2.2126493352346945, "learning_rate": 7.780431608975729e-06, "loss": 0.1158, "step": 7175 }, { "epoch": 0.33345724907063196, "grad_norm": 0.8308097609585474, "learning_rate": 7.779806145191521e-06, "loss": 0.054, "step": 7176 }, { "epoch": 0.33350371747211893, "grad_norm": 1.027799546112084, "learning_rate": 7.779180618441779e-06, "loss": 0.0706, "step": 7177 }, { "epoch": 0.33355018587360596, "grad_norm": 1.8313232123251997, "learning_rate": 7.77855502874067e-06, "loss": 0.0946, "step": 7178 }, { "epoch": 0.33359665427509294, "grad_norm": 1.244714791394402, "learning_rate": 7.777929376102366e-06, "loss": 0.0928, "step": 7179 }, { "epoch": 0.3336431226765799, "grad_norm": 1.4895281523230801, "learning_rate": 7.777303660541038e-06, "loss": 0.0804, "step": 7180 }, { "epoch": 0.3336895910780669, "grad_norm": 4.625247047620925, "learning_rate": 7.776677882070858e-06, "loss": 0.0796, "step": 7181 }, { "epoch": 0.3337360594795539, "grad_norm": 0.9684524037743352, "learning_rate": 7.776052040706002e-06, "loss": 0.0468, "step": 7182 }, { "epoch": 0.3337825278810409, "grad_norm": 1.2788049806851571, "learning_rate": 7.775426136460646e-06, "loss": 0.0486, "step": 7183 }, { "epoch": 0.3338289962825279, "grad_norm": 0.9365074353811241, "learning_rate": 7.774800169348966e-06, "loss": 0.0565, "step": 7184 }, { "epoch": 0.33387546468401486, "grad_norm": 1.972713957396484, "learning_rate": 7.774174139385143e-06, "loss": 0.0664, "step": 7185 }, { "epoch": 0.33392193308550183, "grad_norm": 1.237694246884191, "learning_rate": 7.773548046583355e-06, "loss": 0.0924, "step": 7186 }, { "epoch": 0.33396840148698886, "grad_norm": 2.513645479690248, "learning_rate": 7.772921890957786e-06, "loss": 0.089, "step": 7187 }, { "epoch": 0.33401486988847584, "grad_norm": 1.443857170509477, "learning_rate": 7.772295672522615e-06, "loss": 0.097, "step": 7188 }, { "epoch": 0.3340613382899628, "grad_norm": 0.8065172691315463, "learning_rate": 7.77166939129203e-06, "loss": 0.0704, "step": 7189 }, { "epoch": 0.3341078066914498, "grad_norm": 1.0207627936623669, "learning_rate": 7.771043047280219e-06, "loss": 0.0548, "step": 7190 }, { "epoch": 0.3341542750929368, "grad_norm": 0.6141989035601572, "learning_rate": 7.770416640501364e-06, "loss": 0.0608, "step": 7191 }, { "epoch": 0.3342007434944238, "grad_norm": 1.2526987627685244, "learning_rate": 7.769790170969657e-06, "loss": 0.0903, "step": 7192 }, { "epoch": 0.3342472118959108, "grad_norm": 0.5811110850658058, "learning_rate": 7.769163638699286e-06, "loss": 0.0486, "step": 7193 }, { "epoch": 0.33429368029739776, "grad_norm": 0.8838957589679426, "learning_rate": 7.768537043704447e-06, "loss": 0.0609, "step": 7194 }, { "epoch": 0.33434014869888473, "grad_norm": 0.7999654538696096, "learning_rate": 7.767910385999329e-06, "loss": 0.0644, "step": 7195 }, { "epoch": 0.33438661710037176, "grad_norm": 0.93562002853911, "learning_rate": 7.767283665598127e-06, "loss": 0.0766, "step": 7196 }, { "epoch": 0.33443308550185874, "grad_norm": 1.2573192625140412, "learning_rate": 7.766656882515036e-06, "loss": 0.0944, "step": 7197 }, { "epoch": 0.3344795539033457, "grad_norm": 1.837957719813823, "learning_rate": 7.766030036764256e-06, "loss": 0.1009, "step": 7198 }, { "epoch": 0.3345260223048327, "grad_norm": 0.9600334816709978, "learning_rate": 7.765403128359984e-06, "loss": 0.0796, "step": 7199 }, { "epoch": 0.3345724907063197, "grad_norm": 1.6522087306055377, "learning_rate": 7.764776157316422e-06, "loss": 0.067, "step": 7200 }, { "epoch": 0.3346189591078067, "grad_norm": 2.2903793863305166, "learning_rate": 7.76414912364777e-06, "loss": 0.0883, "step": 7201 }, { "epoch": 0.3346654275092937, "grad_norm": 2.4579302323957477, "learning_rate": 7.76352202736823e-06, "loss": 0.0483, "step": 7202 }, { "epoch": 0.33471189591078065, "grad_norm": 1.1285784517828352, "learning_rate": 7.762894868492009e-06, "loss": 0.065, "step": 7203 }, { "epoch": 0.33475836431226763, "grad_norm": 2.360150903591034, "learning_rate": 7.76226764703331e-06, "loss": 0.1438, "step": 7204 }, { "epoch": 0.33480483271375466, "grad_norm": 0.6163150960894475, "learning_rate": 7.761640363006344e-06, "loss": 0.0652, "step": 7205 }, { "epoch": 0.33485130111524164, "grad_norm": 1.2974567760398932, "learning_rate": 7.761013016425314e-06, "loss": 0.0706, "step": 7206 }, { "epoch": 0.3348977695167286, "grad_norm": 2.73169715193357, "learning_rate": 7.760385607304434e-06, "loss": 0.0789, "step": 7207 }, { "epoch": 0.3349442379182156, "grad_norm": 2.1329938080323987, "learning_rate": 7.759758135657917e-06, "loss": 0.0923, "step": 7208 }, { "epoch": 0.3349907063197026, "grad_norm": 0.7994227354402685, "learning_rate": 7.759130601499972e-06, "loss": 0.0488, "step": 7209 }, { "epoch": 0.3350371747211896, "grad_norm": 1.076812780595387, "learning_rate": 7.758503004844815e-06, "loss": 0.0725, "step": 7210 }, { "epoch": 0.3350836431226766, "grad_norm": 2.555973020326835, "learning_rate": 7.75787534570666e-06, "loss": 0.1263, "step": 7211 }, { "epoch": 0.33513011152416355, "grad_norm": 1.115469244393649, "learning_rate": 7.757247624099728e-06, "loss": 0.0825, "step": 7212 }, { "epoch": 0.33517657992565053, "grad_norm": 0.8387427577477797, "learning_rate": 7.756619840038235e-06, "loss": 0.0562, "step": 7213 }, { "epoch": 0.33522304832713756, "grad_norm": 0.9629374482537505, "learning_rate": 7.7559919935364e-06, "loss": 0.0497, "step": 7214 }, { "epoch": 0.33526951672862454, "grad_norm": 0.7572941149922984, "learning_rate": 7.755364084608448e-06, "loss": 0.0513, "step": 7215 }, { "epoch": 0.3353159851301115, "grad_norm": 1.049004801811673, "learning_rate": 7.754736113268599e-06, "loss": 0.0621, "step": 7216 }, { "epoch": 0.3353624535315985, "grad_norm": 0.7713618501908868, "learning_rate": 7.754108079531075e-06, "loss": 0.0527, "step": 7217 }, { "epoch": 0.3354089219330855, "grad_norm": 1.415615427000559, "learning_rate": 7.753479983410107e-06, "loss": 0.0793, "step": 7218 }, { "epoch": 0.3354553903345725, "grad_norm": 1.0300193527006953, "learning_rate": 7.75285182491992e-06, "loss": 0.0617, "step": 7219 }, { "epoch": 0.3355018587360595, "grad_norm": 2.0811241768841535, "learning_rate": 7.75222360407474e-06, "loss": 0.0587, "step": 7220 }, { "epoch": 0.33554832713754645, "grad_norm": 0.723825308893493, "learning_rate": 7.751595320888798e-06, "loss": 0.0649, "step": 7221 }, { "epoch": 0.33559479553903343, "grad_norm": 1.674609408742634, "learning_rate": 7.750966975376328e-06, "loss": 0.0833, "step": 7222 }, { "epoch": 0.33564126394052046, "grad_norm": 1.524670265961882, "learning_rate": 7.750338567551562e-06, "loss": 0.0668, "step": 7223 }, { "epoch": 0.33568773234200744, "grad_norm": 1.6965903086675682, "learning_rate": 7.749710097428728e-06, "loss": 0.0934, "step": 7224 }, { "epoch": 0.3357342007434944, "grad_norm": 2.7487333756912053, "learning_rate": 7.74908156502207e-06, "loss": 0.1015, "step": 7225 }, { "epoch": 0.3357806691449814, "grad_norm": 0.9544733076050227, "learning_rate": 7.74845297034582e-06, "loss": 0.0757, "step": 7226 }, { "epoch": 0.3358271375464684, "grad_norm": 1.4250693465910085, "learning_rate": 7.747824313414218e-06, "loss": 0.0913, "step": 7227 }, { "epoch": 0.3358736059479554, "grad_norm": 0.7737335034057927, "learning_rate": 7.747195594241503e-06, "loss": 0.0539, "step": 7228 }, { "epoch": 0.3359200743494424, "grad_norm": 1.4323445829499173, "learning_rate": 7.746566812841915e-06, "loss": 0.0709, "step": 7229 }, { "epoch": 0.33596654275092935, "grad_norm": 0.5421069820231774, "learning_rate": 7.7459379692297e-06, "loss": 0.0317, "step": 7230 }, { "epoch": 0.33601301115241633, "grad_norm": 0.5628924108448945, "learning_rate": 7.745309063419101e-06, "loss": 0.0435, "step": 7231 }, { "epoch": 0.33605947955390336, "grad_norm": 1.0374588667346678, "learning_rate": 7.74468009542436e-06, "loss": 0.0935, "step": 7232 }, { "epoch": 0.33610594795539034, "grad_norm": 0.6798075460488007, "learning_rate": 7.744051065259728e-06, "loss": 0.0304, "step": 7233 }, { "epoch": 0.3361524163568773, "grad_norm": 1.3301477830725743, "learning_rate": 7.743421972939448e-06, "loss": 0.0709, "step": 7234 }, { "epoch": 0.3361988847583643, "grad_norm": 1.923404714853104, "learning_rate": 7.742792818477776e-06, "loss": 0.0886, "step": 7235 }, { "epoch": 0.3362453531598513, "grad_norm": 1.0237175000477796, "learning_rate": 7.742163601888958e-06, "loss": 0.0584, "step": 7236 }, { "epoch": 0.3362918215613383, "grad_norm": 1.1180331829205212, "learning_rate": 7.74153432318725e-06, "loss": 0.0517, "step": 7237 }, { "epoch": 0.3363382899628253, "grad_norm": 1.428109925346106, "learning_rate": 7.740904982386905e-06, "loss": 0.0552, "step": 7238 }, { "epoch": 0.33638475836431225, "grad_norm": 2.09064061559062, "learning_rate": 7.740275579502174e-06, "loss": 0.0935, "step": 7239 }, { "epoch": 0.33643122676579923, "grad_norm": 0.70642732945331, "learning_rate": 7.739646114547319e-06, "loss": 0.0735, "step": 7240 }, { "epoch": 0.33647769516728626, "grad_norm": 2.9734599979823995, "learning_rate": 7.739016587536596e-06, "loss": 0.0998, "step": 7241 }, { "epoch": 0.33652416356877324, "grad_norm": 1.2003311075775331, "learning_rate": 7.738386998484266e-06, "loss": 0.063, "step": 7242 }, { "epoch": 0.3365706319702602, "grad_norm": 0.9770785908428957, "learning_rate": 7.737757347404586e-06, "loss": 0.0618, "step": 7243 }, { "epoch": 0.3366171003717472, "grad_norm": 0.9030019671865653, "learning_rate": 7.737127634311822e-06, "loss": 0.0656, "step": 7244 }, { "epoch": 0.3366635687732342, "grad_norm": 3.7767083165019018, "learning_rate": 7.736497859220235e-06, "loss": 0.0826, "step": 7245 }, { "epoch": 0.3367100371747212, "grad_norm": 1.388325308503055, "learning_rate": 7.735868022144093e-06, "loss": 0.0755, "step": 7246 }, { "epoch": 0.3367565055762082, "grad_norm": 4.342137023100849, "learning_rate": 7.735238123097659e-06, "loss": 0.1173, "step": 7247 }, { "epoch": 0.33680297397769515, "grad_norm": 0.9222747700842189, "learning_rate": 7.734608162095205e-06, "loss": 0.0609, "step": 7248 }, { "epoch": 0.33684944237918213, "grad_norm": 0.9616399575147384, "learning_rate": 7.733978139150997e-06, "loss": 0.0541, "step": 7249 }, { "epoch": 0.33689591078066916, "grad_norm": 1.0517521097345388, "learning_rate": 7.733348054279307e-06, "loss": 0.0815, "step": 7250 }, { "epoch": 0.33694237918215614, "grad_norm": 1.5065108383197148, "learning_rate": 7.732717907494404e-06, "loss": 0.094, "step": 7251 }, { "epoch": 0.3369888475836431, "grad_norm": 3.0611222077896656, "learning_rate": 7.732087698810568e-06, "loss": 0.0678, "step": 7252 }, { "epoch": 0.3370353159851301, "grad_norm": 1.1137030228675244, "learning_rate": 7.731457428242068e-06, "loss": 0.1066, "step": 7253 }, { "epoch": 0.3370817843866171, "grad_norm": 0.9586097036220105, "learning_rate": 7.730827095803183e-06, "loss": 0.0791, "step": 7254 }, { "epoch": 0.3371282527881041, "grad_norm": 0.6984151377620907, "learning_rate": 7.730196701508191e-06, "loss": 0.0509, "step": 7255 }, { "epoch": 0.3371747211895911, "grad_norm": 1.9698043034932893, "learning_rate": 7.729566245371369e-06, "loss": 0.0598, "step": 7256 }, { "epoch": 0.33722118959107805, "grad_norm": 0.8174356546142514, "learning_rate": 7.728935727407e-06, "loss": 0.055, "step": 7257 }, { "epoch": 0.33726765799256503, "grad_norm": 0.7436839067211279, "learning_rate": 7.728305147629363e-06, "loss": 0.0648, "step": 7258 }, { "epoch": 0.33731412639405206, "grad_norm": 1.9373214385936157, "learning_rate": 7.727674506052744e-06, "loss": 0.095, "step": 7259 }, { "epoch": 0.33736059479553904, "grad_norm": 1.2264186178476364, "learning_rate": 7.727043802691428e-06, "loss": 0.093, "step": 7260 }, { "epoch": 0.337407063197026, "grad_norm": 0.9775749298422766, "learning_rate": 7.726413037559697e-06, "loss": 0.0435, "step": 7261 }, { "epoch": 0.337453531598513, "grad_norm": 0.8389497656254864, "learning_rate": 7.725782210671841e-06, "loss": 0.0472, "step": 7262 }, { "epoch": 0.3375, "grad_norm": 2.191924919449508, "learning_rate": 7.725151322042151e-06, "loss": 0.0671, "step": 7263 }, { "epoch": 0.337546468401487, "grad_norm": 1.7252661175438666, "learning_rate": 7.724520371684915e-06, "loss": 0.0721, "step": 7264 }, { "epoch": 0.337592936802974, "grad_norm": 2.7452776188879193, "learning_rate": 7.723889359614425e-06, "loss": 0.1363, "step": 7265 }, { "epoch": 0.33763940520446095, "grad_norm": 1.695196975649505, "learning_rate": 7.723258285844974e-06, "loss": 0.1149, "step": 7266 }, { "epoch": 0.33768587360594793, "grad_norm": 1.2163535117605375, "learning_rate": 7.722627150390857e-06, "loss": 0.0713, "step": 7267 }, { "epoch": 0.33773234200743496, "grad_norm": 2.084504620810288, "learning_rate": 7.721995953266367e-06, "loss": 0.0573, "step": 7268 }, { "epoch": 0.33777881040892194, "grad_norm": 0.7130292139603909, "learning_rate": 7.721364694485808e-06, "loss": 0.0387, "step": 7269 }, { "epoch": 0.3378252788104089, "grad_norm": 2.7579883825764937, "learning_rate": 7.720733374063472e-06, "loss": 0.0817, "step": 7270 }, { "epoch": 0.3378717472118959, "grad_norm": 1.5579624542897048, "learning_rate": 7.720101992013661e-06, "loss": 0.0632, "step": 7271 }, { "epoch": 0.3379182156133829, "grad_norm": 1.277462738933918, "learning_rate": 7.719470548350679e-06, "loss": 0.0977, "step": 7272 }, { "epoch": 0.3379646840148699, "grad_norm": 0.9310005684159771, "learning_rate": 7.718839043088827e-06, "loss": 0.0772, "step": 7273 }, { "epoch": 0.3380111524163569, "grad_norm": 0.7321932749154495, "learning_rate": 7.718207476242407e-06, "loss": 0.0347, "step": 7274 }, { "epoch": 0.33805762081784385, "grad_norm": 1.6979483394091006, "learning_rate": 7.71757584782573e-06, "loss": 0.1011, "step": 7275 }, { "epoch": 0.33810408921933083, "grad_norm": 1.2630561968053633, "learning_rate": 7.716944157853095e-06, "loss": 0.0731, "step": 7276 }, { "epoch": 0.33815055762081786, "grad_norm": 0.5158967744059443, "learning_rate": 7.716312406338819e-06, "loss": 0.0561, "step": 7277 }, { "epoch": 0.33819702602230484, "grad_norm": 0.8399738129330178, "learning_rate": 7.715680593297208e-06, "loss": 0.0482, "step": 7278 }, { "epoch": 0.3382434944237918, "grad_norm": 1.0699772547200983, "learning_rate": 7.715048718742571e-06, "loss": 0.0608, "step": 7279 }, { "epoch": 0.3382899628252788, "grad_norm": 2.9042402896666712, "learning_rate": 7.714416782689225e-06, "loss": 0.0443, "step": 7280 }, { "epoch": 0.3383364312267658, "grad_norm": 2.1349599603872806, "learning_rate": 7.713784785151482e-06, "loss": 0.1008, "step": 7281 }, { "epoch": 0.3383828996282528, "grad_norm": 2.309898424200752, "learning_rate": 7.713152726143658e-06, "loss": 0.0888, "step": 7282 }, { "epoch": 0.3384293680297398, "grad_norm": 0.9107928155292634, "learning_rate": 7.712520605680066e-06, "loss": 0.061, "step": 7283 }, { "epoch": 0.33847583643122675, "grad_norm": 2.2474610078266237, "learning_rate": 7.71188842377503e-06, "loss": 0.0708, "step": 7284 }, { "epoch": 0.3385223048327137, "grad_norm": 2.6717305196762964, "learning_rate": 7.711256180442868e-06, "loss": 0.0975, "step": 7285 }, { "epoch": 0.33856877323420076, "grad_norm": 0.8024246008731247, "learning_rate": 7.710623875697897e-06, "loss": 0.0742, "step": 7286 }, { "epoch": 0.33861524163568774, "grad_norm": 0.9868730906911957, "learning_rate": 7.709991509554444e-06, "loss": 0.0773, "step": 7287 }, { "epoch": 0.3386617100371747, "grad_norm": 2.30082266959123, "learning_rate": 7.70935908202683e-06, "loss": 0.0923, "step": 7288 }, { "epoch": 0.3387081784386617, "grad_norm": 2.811206623460422, "learning_rate": 7.708726593129382e-06, "loss": 0.1462, "step": 7289 }, { "epoch": 0.3387546468401487, "grad_norm": 0.9158595124936599, "learning_rate": 7.708094042876427e-06, "loss": 0.0602, "step": 7290 }, { "epoch": 0.3388011152416357, "grad_norm": 0.7672680020488444, "learning_rate": 7.70746143128229e-06, "loss": 0.0468, "step": 7291 }, { "epoch": 0.3388475836431227, "grad_norm": 1.0714575551589216, "learning_rate": 7.706828758361304e-06, "loss": 0.0622, "step": 7292 }, { "epoch": 0.33889405204460965, "grad_norm": 2.5995533207965944, "learning_rate": 7.706196024127796e-06, "loss": 0.0409, "step": 7293 }, { "epoch": 0.3389405204460966, "grad_norm": 2.4446435848206574, "learning_rate": 7.7055632285961e-06, "loss": 0.0586, "step": 7294 }, { "epoch": 0.33898698884758366, "grad_norm": 0.8631894188825211, "learning_rate": 7.70493037178055e-06, "loss": 0.0311, "step": 7295 }, { "epoch": 0.33903345724907064, "grad_norm": 0.8962398912515677, "learning_rate": 7.70429745369548e-06, "loss": 0.0718, "step": 7296 }, { "epoch": 0.3390799256505576, "grad_norm": 0.7759938692781546, "learning_rate": 7.703664474355225e-06, "loss": 0.0733, "step": 7297 }, { "epoch": 0.3391263940520446, "grad_norm": 1.3163083870847534, "learning_rate": 7.703031433774126e-06, "loss": 0.0563, "step": 7298 }, { "epoch": 0.3391728624535316, "grad_norm": 0.7284282766122757, "learning_rate": 7.702398331966519e-06, "loss": 0.0648, "step": 7299 }, { "epoch": 0.3392193308550186, "grad_norm": 0.9211539143292058, "learning_rate": 7.701765168946746e-06, "loss": 0.0613, "step": 7300 }, { "epoch": 0.3392657992565056, "grad_norm": 1.5802302419637642, "learning_rate": 7.70113194472915e-06, "loss": 0.0612, "step": 7301 }, { "epoch": 0.33931226765799255, "grad_norm": 1.5398357113911483, "learning_rate": 7.70049865932807e-06, "loss": 0.0787, "step": 7302 }, { "epoch": 0.3393587360594795, "grad_norm": 0.45050361280144, "learning_rate": 7.699865312757854e-06, "loss": 0.0523, "step": 7303 }, { "epoch": 0.33940520446096656, "grad_norm": 3.798764882294702, "learning_rate": 7.699231905032848e-06, "loss": 0.1069, "step": 7304 }, { "epoch": 0.33945167286245354, "grad_norm": 0.4521471993135908, "learning_rate": 7.698598436167398e-06, "loss": 0.0583, "step": 7305 }, { "epoch": 0.3394981412639405, "grad_norm": 1.8539266160421024, "learning_rate": 7.69796490617585e-06, "loss": 0.1101, "step": 7306 }, { "epoch": 0.3395446096654275, "grad_norm": 1.9383947342584695, "learning_rate": 7.697331315072563e-06, "loss": 0.1001, "step": 7307 }, { "epoch": 0.3395910780669145, "grad_norm": 0.8637019779179924, "learning_rate": 7.69669766287188e-06, "loss": 0.0466, "step": 7308 }, { "epoch": 0.3396375464684015, "grad_norm": 0.6406611499075797, "learning_rate": 7.696063949588155e-06, "loss": 0.0879, "step": 7309 }, { "epoch": 0.3396840148698885, "grad_norm": 1.4707615995050651, "learning_rate": 7.695430175235748e-06, "loss": 0.0731, "step": 7310 }, { "epoch": 0.33973048327137545, "grad_norm": 1.6410368022485171, "learning_rate": 7.694796339829007e-06, "loss": 0.0907, "step": 7311 }, { "epoch": 0.3397769516728624, "grad_norm": 1.3659685536352137, "learning_rate": 7.694162443382294e-06, "loss": 0.072, "step": 7312 }, { "epoch": 0.33982342007434946, "grad_norm": 0.9915745926390013, "learning_rate": 7.693528485909966e-06, "loss": 0.0874, "step": 7313 }, { "epoch": 0.33986988847583643, "grad_norm": 0.8510588875177237, "learning_rate": 7.692894467426385e-06, "loss": 0.0601, "step": 7314 }, { "epoch": 0.3399163568773234, "grad_norm": 2.2627902151232675, "learning_rate": 7.692260387945906e-06, "loss": 0.1136, "step": 7315 }, { "epoch": 0.3399628252788104, "grad_norm": 2.6409053689292294, "learning_rate": 7.691626247482897e-06, "loss": 0.08, "step": 7316 }, { "epoch": 0.3400092936802974, "grad_norm": 1.8843518740644025, "learning_rate": 7.690992046051722e-06, "loss": 0.0735, "step": 7317 }, { "epoch": 0.3400557620817844, "grad_norm": 1.3610282281140331, "learning_rate": 7.690357783666744e-06, "loss": 0.0791, "step": 7318 }, { "epoch": 0.3401022304832714, "grad_norm": 2.1747501786659598, "learning_rate": 7.689723460342332e-06, "loss": 0.0801, "step": 7319 }, { "epoch": 0.34014869888475835, "grad_norm": 1.355545555268926, "learning_rate": 7.689089076092851e-06, "loss": 0.0634, "step": 7320 }, { "epoch": 0.3401951672862453, "grad_norm": 1.253419800497227, "learning_rate": 7.688454630932674e-06, "loss": 0.0685, "step": 7321 }, { "epoch": 0.34024163568773236, "grad_norm": 0.9714112934532596, "learning_rate": 7.68782012487617e-06, "loss": 0.071, "step": 7322 }, { "epoch": 0.34028810408921933, "grad_norm": 1.5346832239160564, "learning_rate": 7.68718555793771e-06, "loss": 0.1091, "step": 7323 }, { "epoch": 0.3403345724907063, "grad_norm": 0.9359658730312064, "learning_rate": 7.686550930131667e-06, "loss": 0.0491, "step": 7324 }, { "epoch": 0.3403810408921933, "grad_norm": 1.6637083524051024, "learning_rate": 7.685916241472421e-06, "loss": 0.0537, "step": 7325 }, { "epoch": 0.3404275092936803, "grad_norm": 1.5731287505555964, "learning_rate": 7.685281491974344e-06, "loss": 0.0883, "step": 7326 }, { "epoch": 0.3404739776951673, "grad_norm": 1.204795625719453, "learning_rate": 7.684646681651814e-06, "loss": 0.0581, "step": 7327 }, { "epoch": 0.3405204460966543, "grad_norm": 1.1415237085637708, "learning_rate": 7.684011810519213e-06, "loss": 0.1025, "step": 7328 }, { "epoch": 0.34056691449814125, "grad_norm": 0.9319664077853632, "learning_rate": 7.683376878590921e-06, "loss": 0.0526, "step": 7329 }, { "epoch": 0.3406133828996282, "grad_norm": 1.103446199362086, "learning_rate": 7.682741885881314e-06, "loss": 0.0831, "step": 7330 }, { "epoch": 0.34065985130111526, "grad_norm": 2.2005428875710207, "learning_rate": 7.682106832404785e-06, "loss": 0.1361, "step": 7331 }, { "epoch": 0.34070631970260223, "grad_norm": 0.7154664278657568, "learning_rate": 7.681471718175709e-06, "loss": 0.0406, "step": 7332 }, { "epoch": 0.3407527881040892, "grad_norm": 1.1411366916566748, "learning_rate": 7.680836543208478e-06, "loss": 0.0678, "step": 7333 }, { "epoch": 0.3407992565055762, "grad_norm": 1.6218824309611544, "learning_rate": 7.680201307517479e-06, "loss": 0.1007, "step": 7334 }, { "epoch": 0.3408457249070632, "grad_norm": 3.1144061549865034, "learning_rate": 7.679566011117098e-06, "loss": 0.0729, "step": 7335 }, { "epoch": 0.3408921933085502, "grad_norm": 0.8536106224968611, "learning_rate": 7.678930654021727e-06, "loss": 0.0663, "step": 7336 }, { "epoch": 0.34093866171003717, "grad_norm": 1.5176578300114016, "learning_rate": 7.678295236245758e-06, "loss": 0.0652, "step": 7337 }, { "epoch": 0.34098513011152415, "grad_norm": 0.7467939520336027, "learning_rate": 7.677659757803582e-06, "loss": 0.05, "step": 7338 }, { "epoch": 0.3410315985130111, "grad_norm": 1.7730332632013128, "learning_rate": 7.677024218709594e-06, "loss": 0.0968, "step": 7339 }, { "epoch": 0.34107806691449816, "grad_norm": 0.9432233726222833, "learning_rate": 7.67638861897819e-06, "loss": 0.0679, "step": 7340 }, { "epoch": 0.34112453531598513, "grad_norm": 0.48183921241681205, "learning_rate": 7.675752958623767e-06, "loss": 0.0584, "step": 7341 }, { "epoch": 0.3411710037174721, "grad_norm": 0.8742090129619827, "learning_rate": 7.675117237660725e-06, "loss": 0.0582, "step": 7342 }, { "epoch": 0.3412174721189591, "grad_norm": 1.1259400366176495, "learning_rate": 7.674481456103461e-06, "loss": 0.063, "step": 7343 }, { "epoch": 0.3412639405204461, "grad_norm": 1.4731600912772245, "learning_rate": 7.673845613966376e-06, "loss": 0.1173, "step": 7344 }, { "epoch": 0.3413104089219331, "grad_norm": 1.6133323333370198, "learning_rate": 7.673209711263875e-06, "loss": 0.093, "step": 7345 }, { "epoch": 0.34135687732342007, "grad_norm": 2.0392864868947482, "learning_rate": 7.67257374801036e-06, "loss": 0.0735, "step": 7346 }, { "epoch": 0.34140334572490705, "grad_norm": 1.574962907827116, "learning_rate": 7.671937724220236e-06, "loss": 0.0706, "step": 7347 }, { "epoch": 0.341449814126394, "grad_norm": 2.117553901237959, "learning_rate": 7.671301639907913e-06, "loss": 0.108, "step": 7348 }, { "epoch": 0.34149628252788106, "grad_norm": 1.4233086605493643, "learning_rate": 7.670665495087795e-06, "loss": 0.0582, "step": 7349 }, { "epoch": 0.34154275092936803, "grad_norm": 0.8071760304606056, "learning_rate": 7.670029289774292e-06, "loss": 0.0682, "step": 7350 }, { "epoch": 0.341589219330855, "grad_norm": 1.2035744129213526, "learning_rate": 7.669393023981816e-06, "loss": 0.0766, "step": 7351 }, { "epoch": 0.341635687732342, "grad_norm": 1.5757168908596784, "learning_rate": 7.668756697724777e-06, "loss": 0.0917, "step": 7352 }, { "epoch": 0.341682156133829, "grad_norm": 0.6187359867506583, "learning_rate": 7.668120311017591e-06, "loss": 0.0455, "step": 7353 }, { "epoch": 0.341728624535316, "grad_norm": 0.8713823205115612, "learning_rate": 7.667483863874674e-06, "loss": 0.0731, "step": 7354 }, { "epoch": 0.34177509293680297, "grad_norm": 2.837528083710069, "learning_rate": 7.666847356310438e-06, "loss": 0.0845, "step": 7355 }, { "epoch": 0.34182156133828995, "grad_norm": 1.6539870068754952, "learning_rate": 7.6662107883393e-06, "loss": 0.0932, "step": 7356 }, { "epoch": 0.3418680297397769, "grad_norm": 2.114695009995938, "learning_rate": 7.665574159975686e-06, "loss": 0.0984, "step": 7357 }, { "epoch": 0.34191449814126396, "grad_norm": 0.6178715176172537, "learning_rate": 7.66493747123401e-06, "loss": 0.0685, "step": 7358 }, { "epoch": 0.34196096654275093, "grad_norm": 1.8532350132715478, "learning_rate": 7.664300722128695e-06, "loss": 0.068, "step": 7359 }, { "epoch": 0.3420074349442379, "grad_norm": 0.6290992169786763, "learning_rate": 7.663663912674167e-06, "loss": 0.0555, "step": 7360 }, { "epoch": 0.3420539033457249, "grad_norm": 1.693959874053741, "learning_rate": 7.663027042884846e-06, "loss": 0.0798, "step": 7361 }, { "epoch": 0.3421003717472119, "grad_norm": 1.1257224626777609, "learning_rate": 7.662390112775158e-06, "loss": 0.0878, "step": 7362 }, { "epoch": 0.3421468401486989, "grad_norm": 0.780843208201448, "learning_rate": 7.661753122359535e-06, "loss": 0.0551, "step": 7363 }, { "epoch": 0.34219330855018587, "grad_norm": 1.2093958615557183, "learning_rate": 7.6611160716524e-06, "loss": 0.0929, "step": 7364 }, { "epoch": 0.34223977695167285, "grad_norm": 0.8160905716854661, "learning_rate": 7.660478960668186e-06, "loss": 0.0877, "step": 7365 }, { "epoch": 0.3422862453531598, "grad_norm": 1.094255099376096, "learning_rate": 7.659841789421323e-06, "loss": 0.0593, "step": 7366 }, { "epoch": 0.34233271375464686, "grad_norm": 0.7679889974915035, "learning_rate": 7.659204557926246e-06, "loss": 0.0724, "step": 7367 }, { "epoch": 0.34237918215613383, "grad_norm": 6.961590361773861, "learning_rate": 7.658567266197383e-06, "loss": 0.0959, "step": 7368 }, { "epoch": 0.3424256505576208, "grad_norm": 0.9744387790811806, "learning_rate": 7.657929914249177e-06, "loss": 0.052, "step": 7369 }, { "epoch": 0.3424721189591078, "grad_norm": 1.2443311854244279, "learning_rate": 7.657292502096059e-06, "loss": 0.0391, "step": 7370 }, { "epoch": 0.3425185873605948, "grad_norm": 0.9100424093191247, "learning_rate": 7.65665502975247e-06, "loss": 0.0481, "step": 7371 }, { "epoch": 0.3425650557620818, "grad_norm": 1.4488304752958943, "learning_rate": 7.65601749723285e-06, "loss": 0.0818, "step": 7372 }, { "epoch": 0.34261152416356877, "grad_norm": 1.0009275312747874, "learning_rate": 7.655379904551637e-06, "loss": 0.0411, "step": 7373 }, { "epoch": 0.34265799256505575, "grad_norm": 0.9140492863402234, "learning_rate": 7.654742251723274e-06, "loss": 0.0724, "step": 7374 }, { "epoch": 0.3427044609665427, "grad_norm": 0.8699076265946625, "learning_rate": 7.654104538762204e-06, "loss": 0.0817, "step": 7375 }, { "epoch": 0.34275092936802976, "grad_norm": 1.1969090772553344, "learning_rate": 7.653466765682874e-06, "loss": 0.0614, "step": 7376 }, { "epoch": 0.34279739776951673, "grad_norm": 2.384539408047352, "learning_rate": 7.65282893249973e-06, "loss": 0.1162, "step": 7377 }, { "epoch": 0.3428438661710037, "grad_norm": 2.793653931973786, "learning_rate": 7.652191039227217e-06, "loss": 0.0658, "step": 7378 }, { "epoch": 0.3428903345724907, "grad_norm": 0.984050143959981, "learning_rate": 7.651553085879786e-06, "loss": 0.0845, "step": 7379 }, { "epoch": 0.3429368029739777, "grad_norm": 1.5611146326401182, "learning_rate": 7.650915072471886e-06, "loss": 0.0637, "step": 7380 }, { "epoch": 0.3429832713754647, "grad_norm": 2.483758681455663, "learning_rate": 7.650276999017971e-06, "loss": 0.0729, "step": 7381 }, { "epoch": 0.34302973977695167, "grad_norm": 0.7705961567433524, "learning_rate": 7.649638865532493e-06, "loss": 0.0755, "step": 7382 }, { "epoch": 0.34307620817843865, "grad_norm": 1.149643972433044, "learning_rate": 7.649000672029903e-06, "loss": 0.0834, "step": 7383 }, { "epoch": 0.3431226765799256, "grad_norm": 0.9268170562118851, "learning_rate": 7.648362418524663e-06, "loss": 0.0761, "step": 7384 }, { "epoch": 0.34316914498141265, "grad_norm": 1.2848515613539566, "learning_rate": 7.647724105031227e-06, "loss": 0.0617, "step": 7385 }, { "epoch": 0.34321561338289963, "grad_norm": 1.9170425235946829, "learning_rate": 7.647085731564053e-06, "loss": 0.0911, "step": 7386 }, { "epoch": 0.3432620817843866, "grad_norm": 1.4627755908426665, "learning_rate": 7.646447298137598e-06, "loss": 0.0932, "step": 7387 }, { "epoch": 0.3433085501858736, "grad_norm": 1.6239552652074685, "learning_rate": 7.64580880476633e-06, "loss": 0.0839, "step": 7388 }, { "epoch": 0.3433550185873606, "grad_norm": 1.6907932765698377, "learning_rate": 7.645170251464706e-06, "loss": 0.086, "step": 7389 }, { "epoch": 0.3434014869888476, "grad_norm": 1.2550363886809128, "learning_rate": 7.644531638247192e-06, "loss": 0.0532, "step": 7390 }, { "epoch": 0.34344795539033457, "grad_norm": 2.948817987629196, "learning_rate": 7.643892965128254e-06, "loss": 0.0993, "step": 7391 }, { "epoch": 0.34349442379182155, "grad_norm": 1.3824153973996085, "learning_rate": 7.643254232122357e-06, "loss": 0.0581, "step": 7392 }, { "epoch": 0.3435408921933085, "grad_norm": 2.4729657290496347, "learning_rate": 7.64261543924397e-06, "loss": 0.089, "step": 7393 }, { "epoch": 0.34358736059479555, "grad_norm": 1.6087363952515363, "learning_rate": 7.641976586507563e-06, "loss": 0.0758, "step": 7394 }, { "epoch": 0.34363382899628253, "grad_norm": 1.8034342890107007, "learning_rate": 7.641337673927602e-06, "loss": 0.0899, "step": 7395 }, { "epoch": 0.3436802973977695, "grad_norm": 1.5508612117329033, "learning_rate": 7.640698701518568e-06, "loss": 0.0497, "step": 7396 }, { "epoch": 0.3437267657992565, "grad_norm": 0.9215591080842733, "learning_rate": 7.640059669294924e-06, "loss": 0.0689, "step": 7397 }, { "epoch": 0.3437732342007435, "grad_norm": 0.749793427010664, "learning_rate": 7.639420577271153e-06, "loss": 0.0438, "step": 7398 }, { "epoch": 0.3438197026022305, "grad_norm": 1.1740605386904577, "learning_rate": 7.638781425461727e-06, "loss": 0.0708, "step": 7399 }, { "epoch": 0.34386617100371747, "grad_norm": 1.8112169231378776, "learning_rate": 7.638142213881124e-06, "loss": 0.0756, "step": 7400 }, { "epoch": 0.34391263940520445, "grad_norm": 2.0310857126715343, "learning_rate": 7.637502942543825e-06, "loss": 0.0648, "step": 7401 }, { "epoch": 0.3439591078066915, "grad_norm": 1.9144592814427364, "learning_rate": 7.636863611464306e-06, "loss": 0.077, "step": 7402 }, { "epoch": 0.34400557620817845, "grad_norm": 2.680140305366576, "learning_rate": 7.636224220657054e-06, "loss": 0.0975, "step": 7403 }, { "epoch": 0.34405204460966543, "grad_norm": 2.3172997897289136, "learning_rate": 7.635584770136547e-06, "loss": 0.1191, "step": 7404 }, { "epoch": 0.3440985130111524, "grad_norm": 0.9822838255712868, "learning_rate": 7.634945259917271e-06, "loss": 0.0666, "step": 7405 }, { "epoch": 0.3441449814126394, "grad_norm": 1.019181406385353, "learning_rate": 7.634305690013713e-06, "loss": 0.0753, "step": 7406 }, { "epoch": 0.3441914498141264, "grad_norm": 0.9550627200903399, "learning_rate": 7.633666060440356e-06, "loss": 0.0669, "step": 7407 }, { "epoch": 0.3442379182156134, "grad_norm": 2.0844411985478595, "learning_rate": 7.633026371211694e-06, "loss": 0.0574, "step": 7408 }, { "epoch": 0.34428438661710037, "grad_norm": 1.054834445077173, "learning_rate": 7.63238662234221e-06, "loss": 0.1, "step": 7409 }, { "epoch": 0.34433085501858735, "grad_norm": 1.1106926453830508, "learning_rate": 7.6317468138464e-06, "loss": 0.0797, "step": 7410 }, { "epoch": 0.3443773234200744, "grad_norm": 1.9171799129912408, "learning_rate": 7.631106945738756e-06, "loss": 0.0812, "step": 7411 }, { "epoch": 0.34442379182156135, "grad_norm": 0.529466567906053, "learning_rate": 7.630467018033769e-06, "loss": 0.0607, "step": 7412 }, { "epoch": 0.34447026022304833, "grad_norm": 1.1874633010293083, "learning_rate": 7.629827030745937e-06, "loss": 0.0986, "step": 7413 }, { "epoch": 0.3445167286245353, "grad_norm": 4.4625945882602664, "learning_rate": 7.6291869838897535e-06, "loss": 0.101, "step": 7414 }, { "epoch": 0.3445631970260223, "grad_norm": 0.9046732724170405, "learning_rate": 7.6285468774797175e-06, "loss": 0.0606, "step": 7415 }, { "epoch": 0.3446096654275093, "grad_norm": 7.656277940592699, "learning_rate": 7.62790671153033e-06, "loss": 0.0365, "step": 7416 }, { "epoch": 0.3446561338289963, "grad_norm": 1.2885079140374078, "learning_rate": 7.6272664860560875e-06, "loss": 0.0621, "step": 7417 }, { "epoch": 0.34470260223048327, "grad_norm": 1.9763757357077918, "learning_rate": 7.626626201071494e-06, "loss": 0.1053, "step": 7418 }, { "epoch": 0.34474907063197024, "grad_norm": 1.0098135447358296, "learning_rate": 7.6259858565910535e-06, "loss": 0.0517, "step": 7419 }, { "epoch": 0.3447955390334573, "grad_norm": 1.6228722263425563, "learning_rate": 7.6253454526292716e-06, "loss": 0.068, "step": 7420 }, { "epoch": 0.34484200743494425, "grad_norm": 4.2099787461816645, "learning_rate": 7.624704989200649e-06, "loss": 0.073, "step": 7421 }, { "epoch": 0.34488847583643123, "grad_norm": 1.2164470713684452, "learning_rate": 7.624064466319696e-06, "loss": 0.1008, "step": 7422 }, { "epoch": 0.3449349442379182, "grad_norm": 1.0602991823058265, "learning_rate": 7.623423884000922e-06, "loss": 0.0731, "step": 7423 }, { "epoch": 0.3449814126394052, "grad_norm": 3.4407262072715166, "learning_rate": 7.6227832422588356e-06, "loss": 0.1409, "step": 7424 }, { "epoch": 0.3450278810408922, "grad_norm": 1.246450655882244, "learning_rate": 7.62214254110795e-06, "loss": 0.0575, "step": 7425 }, { "epoch": 0.3450743494423792, "grad_norm": 0.958178516853988, "learning_rate": 7.6215017805627745e-06, "loss": 0.1151, "step": 7426 }, { "epoch": 0.34512081784386617, "grad_norm": 1.7695785008680773, "learning_rate": 7.620860960637826e-06, "loss": 0.0801, "step": 7427 }, { "epoch": 0.34516728624535314, "grad_norm": 0.9206121870985217, "learning_rate": 7.620220081347615e-06, "loss": 0.1028, "step": 7428 }, { "epoch": 0.3452137546468402, "grad_norm": 5.438793745163104, "learning_rate": 7.619579142706664e-06, "loss": 0.1145, "step": 7429 }, { "epoch": 0.34526022304832715, "grad_norm": 0.7628131760966779, "learning_rate": 7.618938144729489e-06, "loss": 0.0403, "step": 7430 }, { "epoch": 0.34530669144981413, "grad_norm": 1.062384797650436, "learning_rate": 7.6182970874306075e-06, "loss": 0.0627, "step": 7431 }, { "epoch": 0.3453531598513011, "grad_norm": 0.7280511797911581, "learning_rate": 7.617655970824543e-06, "loss": 0.0364, "step": 7432 }, { "epoch": 0.3453996282527881, "grad_norm": 0.9572524911213759, "learning_rate": 7.617014794925812e-06, "loss": 0.0891, "step": 7433 }, { "epoch": 0.3454460966542751, "grad_norm": 1.0705013997040242, "learning_rate": 7.616373559748944e-06, "loss": 0.0692, "step": 7434 }, { "epoch": 0.3454925650557621, "grad_norm": 0.7032901625531393, "learning_rate": 7.615732265308461e-06, "loss": 0.0616, "step": 7435 }, { "epoch": 0.34553903345724907, "grad_norm": 0.6810907821509025, "learning_rate": 7.615090911618888e-06, "loss": 0.0515, "step": 7436 }, { "epoch": 0.34558550185873604, "grad_norm": 1.2107028614590165, "learning_rate": 7.614449498694756e-06, "loss": 0.1078, "step": 7437 }, { "epoch": 0.3456319702602231, "grad_norm": 0.78950730099765, "learning_rate": 7.613808026550588e-06, "loss": 0.0339, "step": 7438 }, { "epoch": 0.34567843866171005, "grad_norm": 2.4705869624111525, "learning_rate": 7.61316649520092e-06, "loss": 0.1189, "step": 7439 }, { "epoch": 0.34572490706319703, "grad_norm": 0.6165097429700661, "learning_rate": 7.61252490466028e-06, "loss": 0.0506, "step": 7440 }, { "epoch": 0.345771375464684, "grad_norm": 0.6910408554009081, "learning_rate": 7.6118832549431995e-06, "loss": 0.0584, "step": 7441 }, { "epoch": 0.345817843866171, "grad_norm": 1.779492981929763, "learning_rate": 7.611241546064217e-06, "loss": 0.0851, "step": 7442 }, { "epoch": 0.345864312267658, "grad_norm": 1.334238268286046, "learning_rate": 7.6105997780378635e-06, "loss": 0.0682, "step": 7443 }, { "epoch": 0.345910780669145, "grad_norm": 1.3873895221262447, "learning_rate": 7.6099579508786765e-06, "loss": 0.0882, "step": 7444 }, { "epoch": 0.34595724907063197, "grad_norm": 0.7620782555232385, "learning_rate": 7.609316064601197e-06, "loss": 0.0613, "step": 7445 }, { "epoch": 0.34600371747211894, "grad_norm": 0.9532298886047768, "learning_rate": 7.608674119219961e-06, "loss": 0.0704, "step": 7446 }, { "epoch": 0.346050185873606, "grad_norm": 0.7877695841830129, "learning_rate": 7.608032114749513e-06, "loss": 0.0575, "step": 7447 }, { "epoch": 0.34609665427509295, "grad_norm": 2.9082619199403767, "learning_rate": 7.607390051204389e-06, "loss": 0.1005, "step": 7448 }, { "epoch": 0.34614312267657993, "grad_norm": 1.0786459149346552, "learning_rate": 7.60674792859914e-06, "loss": 0.0566, "step": 7449 }, { "epoch": 0.3461895910780669, "grad_norm": 0.8460523192044093, "learning_rate": 7.6061057469483036e-06, "loss": 0.0737, "step": 7450 }, { "epoch": 0.3462360594795539, "grad_norm": 1.3008458064639052, "learning_rate": 7.605463506266431e-06, "loss": 0.0793, "step": 7451 }, { "epoch": 0.3462825278810409, "grad_norm": 7.4020385215414946, "learning_rate": 7.604821206568067e-06, "loss": 0.0893, "step": 7452 }, { "epoch": 0.3463289962825279, "grad_norm": 1.1485551214789929, "learning_rate": 7.604178847867761e-06, "loss": 0.0677, "step": 7453 }, { "epoch": 0.34637546468401487, "grad_norm": 3.613783970656334, "learning_rate": 7.603536430180063e-06, "loss": 0.0653, "step": 7454 }, { "epoch": 0.34642193308550184, "grad_norm": 1.756345623754672, "learning_rate": 7.602893953519526e-06, "loss": 0.0582, "step": 7455 }, { "epoch": 0.3464684014869889, "grad_norm": 1.4613268436546765, "learning_rate": 7.602251417900701e-06, "loss": 0.0839, "step": 7456 }, { "epoch": 0.34651486988847585, "grad_norm": 0.9409947388751925, "learning_rate": 7.601608823338142e-06, "loss": 0.062, "step": 7457 }, { "epoch": 0.34656133828996283, "grad_norm": 1.8917252313428703, "learning_rate": 7.6009661698464054e-06, "loss": 0.0893, "step": 7458 }, { "epoch": 0.3466078066914498, "grad_norm": 1.502802046817029, "learning_rate": 7.600323457440048e-06, "loss": 0.0618, "step": 7459 }, { "epoch": 0.3466542750929368, "grad_norm": 5.267177993448517, "learning_rate": 7.599680686133626e-06, "loss": 0.0561, "step": 7460 }, { "epoch": 0.3467007434944238, "grad_norm": 1.4800028674593027, "learning_rate": 7.599037855941702e-06, "loss": 0.0956, "step": 7461 }, { "epoch": 0.3467472118959108, "grad_norm": 0.7060320152153587, "learning_rate": 7.5983949668788345e-06, "loss": 0.0566, "step": 7462 }, { "epoch": 0.34679368029739777, "grad_norm": 5.370625566211762, "learning_rate": 7.5977520189595856e-06, "loss": 0.0918, "step": 7463 }, { "epoch": 0.34684014869888474, "grad_norm": 1.5515105707892773, "learning_rate": 7.5971090121985205e-06, "loss": 0.082, "step": 7464 }, { "epoch": 0.3468866171003718, "grad_norm": 0.9564375779366133, "learning_rate": 7.596465946610202e-06, "loss": 0.0441, "step": 7465 }, { "epoch": 0.34693308550185875, "grad_norm": 0.9550095062678132, "learning_rate": 7.595822822209198e-06, "loss": 0.0688, "step": 7466 }, { "epoch": 0.3469795539033457, "grad_norm": 1.334171204954149, "learning_rate": 7.5951796390100755e-06, "loss": 0.0732, "step": 7467 }, { "epoch": 0.3470260223048327, "grad_norm": 1.1930839983471664, "learning_rate": 7.594536397027401e-06, "loss": 0.1069, "step": 7468 }, { "epoch": 0.3470724907063197, "grad_norm": 1.3482127990994432, "learning_rate": 7.593893096275747e-06, "loss": 0.0798, "step": 7469 }, { "epoch": 0.3471189591078067, "grad_norm": 0.8291389110875981, "learning_rate": 7.593249736769685e-06, "loss": 0.0581, "step": 7470 }, { "epoch": 0.3471654275092937, "grad_norm": 0.8472215554903972, "learning_rate": 7.592606318523787e-06, "loss": 0.0577, "step": 7471 }, { "epoch": 0.34721189591078067, "grad_norm": 9.985986270708795, "learning_rate": 7.591962841552627e-06, "loss": 0.1431, "step": 7472 }, { "epoch": 0.34725836431226764, "grad_norm": 0.6561300113443932, "learning_rate": 7.59131930587078e-06, "loss": 0.0587, "step": 7473 }, { "epoch": 0.3473048327137547, "grad_norm": 0.5853647238832137, "learning_rate": 7.590675711492823e-06, "loss": 0.0468, "step": 7474 }, { "epoch": 0.34735130111524165, "grad_norm": 1.4310582243144008, "learning_rate": 7.590032058433336e-06, "loss": 0.0846, "step": 7475 }, { "epoch": 0.3473977695167286, "grad_norm": 1.729449337292609, "learning_rate": 7.589388346706896e-06, "loss": 0.1045, "step": 7476 }, { "epoch": 0.3474442379182156, "grad_norm": 0.7230128952990585, "learning_rate": 7.5887445763280844e-06, "loss": 0.0664, "step": 7477 }, { "epoch": 0.3474907063197026, "grad_norm": 1.3806538348255488, "learning_rate": 7.588100747311485e-06, "loss": 0.086, "step": 7478 }, { "epoch": 0.3475371747211896, "grad_norm": 0.6810206404239272, "learning_rate": 7.587456859671678e-06, "loss": 0.0705, "step": 7479 }, { "epoch": 0.3475836431226766, "grad_norm": 1.0643682535830028, "learning_rate": 7.586812913423252e-06, "loss": 0.0852, "step": 7480 }, { "epoch": 0.34763011152416357, "grad_norm": 1.1132781610665126, "learning_rate": 7.58616890858079e-06, "loss": 0.0888, "step": 7481 }, { "epoch": 0.34767657992565054, "grad_norm": 2.3803316081708776, "learning_rate": 7.5855248451588795e-06, "loss": 0.0855, "step": 7482 }, { "epoch": 0.3477230483271376, "grad_norm": 14.641974656429415, "learning_rate": 7.58488072317211e-06, "loss": 0.1115, "step": 7483 }, { "epoch": 0.34776951672862455, "grad_norm": 1.9492902985407092, "learning_rate": 7.584236542635073e-06, "loss": 0.0628, "step": 7484 }, { "epoch": 0.3478159851301115, "grad_norm": 1.2217803468585475, "learning_rate": 7.583592303562357e-06, "loss": 0.073, "step": 7485 }, { "epoch": 0.3478624535315985, "grad_norm": 1.685846218203322, "learning_rate": 7.582948005968557e-06, "loss": 0.0792, "step": 7486 }, { "epoch": 0.3479089219330855, "grad_norm": 1.6155022058511006, "learning_rate": 7.582303649868266e-06, "loss": 0.1292, "step": 7487 }, { "epoch": 0.3479553903345725, "grad_norm": 0.7671603967254731, "learning_rate": 7.58165923527608e-06, "loss": 0.069, "step": 7488 }, { "epoch": 0.3480018587360595, "grad_norm": 1.6544002705341347, "learning_rate": 7.581014762206593e-06, "loss": 0.0644, "step": 7489 }, { "epoch": 0.34804832713754646, "grad_norm": 4.723588958881206, "learning_rate": 7.580370230674408e-06, "loss": 0.1295, "step": 7490 }, { "epoch": 0.34809479553903344, "grad_norm": 1.0821704828382834, "learning_rate": 7.57972564069412e-06, "loss": 0.1062, "step": 7491 }, { "epoch": 0.3481412639405205, "grad_norm": 1.1671802314786646, "learning_rate": 7.579080992280332e-06, "loss": 0.046, "step": 7492 }, { "epoch": 0.34818773234200745, "grad_norm": 1.013735755092451, "learning_rate": 7.578436285447644e-06, "loss": 0.0756, "step": 7493 }, { "epoch": 0.3482342007434944, "grad_norm": 0.9053921862189034, "learning_rate": 7.57779152021066e-06, "loss": 0.078, "step": 7494 }, { "epoch": 0.3482806691449814, "grad_norm": 2.712042526746993, "learning_rate": 7.577146696583986e-06, "loss": 0.0978, "step": 7495 }, { "epoch": 0.3483271375464684, "grad_norm": 1.1658487063672278, "learning_rate": 7.576501814582227e-06, "loss": 0.0664, "step": 7496 }, { "epoch": 0.3483736059479554, "grad_norm": 1.0198618772713364, "learning_rate": 7.575856874219991e-06, "loss": 0.0495, "step": 7497 }, { "epoch": 0.3484200743494424, "grad_norm": 1.0842119006749062, "learning_rate": 7.575211875511883e-06, "loss": 0.0848, "step": 7498 }, { "epoch": 0.34846654275092936, "grad_norm": 3.234658326682555, "learning_rate": 7.574566818472517e-06, "loss": 0.0883, "step": 7499 }, { "epoch": 0.34851301115241634, "grad_norm": 1.362949748907164, "learning_rate": 7.573921703116504e-06, "loss": 0.0717, "step": 7500 }, { "epoch": 0.3485594795539034, "grad_norm": 1.9710014963918958, "learning_rate": 7.573276529458454e-06, "loss": 0.0953, "step": 7501 }, { "epoch": 0.34860594795539035, "grad_norm": 0.7973462643897332, "learning_rate": 7.572631297512985e-06, "loss": 0.0594, "step": 7502 }, { "epoch": 0.3486524163568773, "grad_norm": 4.358542453525108, "learning_rate": 7.571986007294706e-06, "loss": 0.0405, "step": 7503 }, { "epoch": 0.3486988847583643, "grad_norm": 0.6540725731554744, "learning_rate": 7.571340658818238e-06, "loss": 0.0512, "step": 7504 }, { "epoch": 0.3487453531598513, "grad_norm": 1.0026967018827306, "learning_rate": 7.570695252098199e-06, "loss": 0.0429, "step": 7505 }, { "epoch": 0.3487918215613383, "grad_norm": 1.829226653672768, "learning_rate": 7.5700497871492055e-06, "loss": 0.0797, "step": 7506 }, { "epoch": 0.3488382899628253, "grad_norm": 1.7718018870219865, "learning_rate": 7.56940426398588e-06, "loss": 0.0707, "step": 7507 }, { "epoch": 0.34888475836431226, "grad_norm": 1.7688475443706912, "learning_rate": 7.568758682622845e-06, "loss": 0.0694, "step": 7508 }, { "epoch": 0.34893122676579924, "grad_norm": 0.46760494321763957, "learning_rate": 7.568113043074721e-06, "loss": 0.0485, "step": 7509 }, { "epoch": 0.3489776951672863, "grad_norm": 1.0452143516549819, "learning_rate": 7.567467345356132e-06, "loss": 0.0644, "step": 7510 }, { "epoch": 0.34902416356877325, "grad_norm": 0.6747324406428308, "learning_rate": 7.566821589481708e-06, "loss": 0.0695, "step": 7511 }, { "epoch": 0.3490706319702602, "grad_norm": 1.0775043397748532, "learning_rate": 7.566175775466073e-06, "loss": 0.0642, "step": 7512 }, { "epoch": 0.3491171003717472, "grad_norm": 2.716084531314967, "learning_rate": 7.565529903323854e-06, "loss": 0.0786, "step": 7513 }, { "epoch": 0.3491635687732342, "grad_norm": 1.3347257394663945, "learning_rate": 7.564883973069686e-06, "loss": 0.0488, "step": 7514 }, { "epoch": 0.3492100371747212, "grad_norm": 0.6332323785314194, "learning_rate": 7.5642379847181925e-06, "loss": 0.0637, "step": 7515 }, { "epoch": 0.3492565055762082, "grad_norm": 1.292389259930218, "learning_rate": 7.563591938284012e-06, "loss": 0.049, "step": 7516 }, { "epoch": 0.34930297397769516, "grad_norm": 0.7041340518474833, "learning_rate": 7.5629458337817764e-06, "loss": 0.0509, "step": 7517 }, { "epoch": 0.34934944237918214, "grad_norm": 0.7032762901427909, "learning_rate": 7.562299671226119e-06, "loss": 0.0495, "step": 7518 }, { "epoch": 0.34939591078066917, "grad_norm": 0.48552776234965145, "learning_rate": 7.561653450631678e-06, "loss": 0.0349, "step": 7519 }, { "epoch": 0.34944237918215615, "grad_norm": 0.7336268024229614, "learning_rate": 7.5610071720130895e-06, "loss": 0.0558, "step": 7520 }, { "epoch": 0.3494888475836431, "grad_norm": 1.0290854173229853, "learning_rate": 7.560360835384994e-06, "loss": 0.0711, "step": 7521 }, { "epoch": 0.3495353159851301, "grad_norm": 2.66700354640412, "learning_rate": 7.5597144407620295e-06, "loss": 0.1478, "step": 7522 }, { "epoch": 0.3495817843866171, "grad_norm": 3.365542046194944, "learning_rate": 7.55906798815884e-06, "loss": 0.0765, "step": 7523 }, { "epoch": 0.3496282527881041, "grad_norm": 0.885260156933002, "learning_rate": 7.558421477590068e-06, "loss": 0.0623, "step": 7524 }, { "epoch": 0.3496747211895911, "grad_norm": 1.3170471385930995, "learning_rate": 7.557774909070355e-06, "loss": 0.0693, "step": 7525 }, { "epoch": 0.34972118959107806, "grad_norm": 1.6497610014194166, "learning_rate": 7.5571282826143495e-06, "loss": 0.0788, "step": 7526 }, { "epoch": 0.34976765799256504, "grad_norm": 1.5434188178005317, "learning_rate": 7.5564815982366965e-06, "loss": 0.08, "step": 7527 }, { "epoch": 0.34981412639405207, "grad_norm": 1.116999583625125, "learning_rate": 7.555834855952045e-06, "loss": 0.0867, "step": 7528 }, { "epoch": 0.34986059479553905, "grad_norm": 0.546239966952688, "learning_rate": 7.555188055775045e-06, "loss": 0.0465, "step": 7529 }, { "epoch": 0.349907063197026, "grad_norm": 0.7035228064606224, "learning_rate": 7.5545411977203444e-06, "loss": 0.0394, "step": 7530 }, { "epoch": 0.349953531598513, "grad_norm": 3.3958221750677455, "learning_rate": 7.553894281802598e-06, "loss": 0.0879, "step": 7531 }, { "epoch": 0.35, "grad_norm": 2.0266133585120594, "learning_rate": 7.553247308036458e-06, "loss": 0.0432, "step": 7532 }, { "epoch": 0.350046468401487, "grad_norm": 0.6613259260104895, "learning_rate": 7.5526002764365795e-06, "loss": 0.0433, "step": 7533 }, { "epoch": 0.350092936802974, "grad_norm": 1.0603108171173283, "learning_rate": 7.551953187017619e-06, "loss": 0.0854, "step": 7534 }, { "epoch": 0.35013940520446096, "grad_norm": 0.7383523346457469, "learning_rate": 7.55130603979423e-06, "loss": 0.0573, "step": 7535 }, { "epoch": 0.35018587360594794, "grad_norm": 1.0902972341394463, "learning_rate": 7.550658834781077e-06, "loss": 0.0584, "step": 7536 }, { "epoch": 0.35023234200743497, "grad_norm": 1.0628247443318126, "learning_rate": 7.550011571992816e-06, "loss": 0.0958, "step": 7537 }, { "epoch": 0.35027881040892195, "grad_norm": 1.3758187957408226, "learning_rate": 7.549364251444109e-06, "loss": 0.0834, "step": 7538 }, { "epoch": 0.3503252788104089, "grad_norm": 2.566379337564209, "learning_rate": 7.548716873149617e-06, "loss": 0.0889, "step": 7539 }, { "epoch": 0.3503717472118959, "grad_norm": 0.6726135792556835, "learning_rate": 7.548069437124008e-06, "loss": 0.0462, "step": 7540 }, { "epoch": 0.3504182156133829, "grad_norm": 0.5042120111357034, "learning_rate": 7.547421943381943e-06, "loss": 0.0628, "step": 7541 }, { "epoch": 0.3504646840148699, "grad_norm": 0.8919947663544748, "learning_rate": 7.54677439193809e-06, "loss": 0.0612, "step": 7542 }, { "epoch": 0.3505111524163569, "grad_norm": 0.9302109288880048, "learning_rate": 7.546126782807117e-06, "loss": 0.0929, "step": 7543 }, { "epoch": 0.35055762081784386, "grad_norm": 1.305681488801315, "learning_rate": 7.5454791160036914e-06, "loss": 0.0991, "step": 7544 }, { "epoch": 0.35060408921933084, "grad_norm": 1.402010085529745, "learning_rate": 7.544831391542485e-06, "loss": 0.0726, "step": 7545 }, { "epoch": 0.35065055762081787, "grad_norm": 0.6919335686932533, "learning_rate": 7.544183609438172e-06, "loss": 0.0352, "step": 7546 }, { "epoch": 0.35069702602230485, "grad_norm": 1.225759975318176, "learning_rate": 7.543535769705419e-06, "loss": 0.0464, "step": 7547 }, { "epoch": 0.3507434944237918, "grad_norm": 0.6504293218130057, "learning_rate": 7.542887872358906e-06, "loss": 0.0472, "step": 7548 }, { "epoch": 0.3507899628252788, "grad_norm": 0.8151438400825937, "learning_rate": 7.542239917413306e-06, "loss": 0.0669, "step": 7549 }, { "epoch": 0.3508364312267658, "grad_norm": 0.8634262410787252, "learning_rate": 7.541591904883294e-06, "loss": 0.081, "step": 7550 }, { "epoch": 0.3508828996282528, "grad_norm": 0.8570844184945456, "learning_rate": 7.540943834783552e-06, "loss": 0.0423, "step": 7551 }, { "epoch": 0.3509293680297398, "grad_norm": 0.655851452592393, "learning_rate": 7.5402957071287586e-06, "loss": 0.064, "step": 7552 }, { "epoch": 0.35097583643122676, "grad_norm": 1.8273340389778099, "learning_rate": 7.539647521933593e-06, "loss": 0.1115, "step": 7553 }, { "epoch": 0.35102230483271374, "grad_norm": 0.9330933380277339, "learning_rate": 7.538999279212737e-06, "loss": 0.0602, "step": 7554 }, { "epoch": 0.35106877323420077, "grad_norm": 1.1112103879344668, "learning_rate": 7.538350978980877e-06, "loss": 0.1087, "step": 7555 }, { "epoch": 0.35111524163568775, "grad_norm": 3.2013538057100575, "learning_rate": 7.537702621252696e-06, "loss": 0.0722, "step": 7556 }, { "epoch": 0.3511617100371747, "grad_norm": 2.3114852869816906, "learning_rate": 7.537054206042877e-06, "loss": 0.0984, "step": 7557 }, { "epoch": 0.3512081784386617, "grad_norm": 1.2487666532992647, "learning_rate": 7.536405733366115e-06, "loss": 0.0575, "step": 7558 }, { "epoch": 0.3512546468401487, "grad_norm": 1.2195262099544943, "learning_rate": 7.5357572032370885e-06, "loss": 0.0409, "step": 7559 }, { "epoch": 0.3513011152416357, "grad_norm": 1.0146196310557487, "learning_rate": 7.535108615670495e-06, "loss": 0.1002, "step": 7560 }, { "epoch": 0.3513475836431227, "grad_norm": 0.9196749711391011, "learning_rate": 7.534459970681023e-06, "loss": 0.0762, "step": 7561 }, { "epoch": 0.35139405204460966, "grad_norm": 1.5808515775361398, "learning_rate": 7.533811268283364e-06, "loss": 0.1028, "step": 7562 }, { "epoch": 0.35144052044609664, "grad_norm": 1.133129265668072, "learning_rate": 7.533162508492214e-06, "loss": 0.0785, "step": 7563 }, { "epoch": 0.35148698884758367, "grad_norm": 0.5852026720307014, "learning_rate": 7.532513691322267e-06, "loss": 0.0673, "step": 7564 }, { "epoch": 0.35153345724907065, "grad_norm": 1.0946326059572404, "learning_rate": 7.53186481678822e-06, "loss": 0.064, "step": 7565 }, { "epoch": 0.3515799256505576, "grad_norm": 1.0862669089569628, "learning_rate": 7.53121588490477e-06, "loss": 0.0611, "step": 7566 }, { "epoch": 0.3516263940520446, "grad_norm": 0.8465342697536719, "learning_rate": 7.530566895686615e-06, "loss": 0.1029, "step": 7567 }, { "epoch": 0.3516728624535316, "grad_norm": 0.6122929950300866, "learning_rate": 7.529917849148458e-06, "loss": 0.0439, "step": 7568 }, { "epoch": 0.3517193308550186, "grad_norm": 1.0991865035489032, "learning_rate": 7.529268745304998e-06, "loss": 0.0562, "step": 7569 }, { "epoch": 0.3517657992565056, "grad_norm": 0.281570249050212, "learning_rate": 7.52861958417094e-06, "loss": 0.0212, "step": 7570 }, { "epoch": 0.35181226765799256, "grad_norm": 1.8503222377382251, "learning_rate": 7.527970365760985e-06, "loss": 0.0815, "step": 7571 }, { "epoch": 0.35185873605947954, "grad_norm": 1.7538739466860869, "learning_rate": 7.527321090089842e-06, "loss": 0.0987, "step": 7572 }, { "epoch": 0.35190520446096657, "grad_norm": 1.2303959218499632, "learning_rate": 7.526671757172217e-06, "loss": 0.079, "step": 7573 }, { "epoch": 0.35195167286245355, "grad_norm": 1.7483992259290557, "learning_rate": 7.5260223670228165e-06, "loss": 0.0554, "step": 7574 }, { "epoch": 0.3519981412639405, "grad_norm": 0.8162384873222328, "learning_rate": 7.525372919656349e-06, "loss": 0.0525, "step": 7575 }, { "epoch": 0.3520446096654275, "grad_norm": 2.9457109719157537, "learning_rate": 7.524723415087529e-06, "loss": 0.0572, "step": 7576 }, { "epoch": 0.3520910780669145, "grad_norm": 2.1288015600596695, "learning_rate": 7.5240738533310675e-06, "loss": 0.0947, "step": 7577 }, { "epoch": 0.3521375464684015, "grad_norm": 2.4653612302852825, "learning_rate": 7.523424234401675e-06, "loss": 0.0732, "step": 7578 }, { "epoch": 0.3521840148698885, "grad_norm": 1.3791631021646604, "learning_rate": 7.5227745583140676e-06, "loss": 0.0667, "step": 7579 }, { "epoch": 0.35223048327137546, "grad_norm": 1.3791631021646604, "learning_rate": 7.5227745583140676e-06, "loss": 0.074, "step": 7580 }, { "epoch": 0.35227695167286244, "grad_norm": 1.3274921967036095, "learning_rate": 7.522124825082963e-06, "loss": 0.0755, "step": 7581 }, { "epoch": 0.35232342007434947, "grad_norm": 1.3067168762173569, "learning_rate": 7.521475034723074e-06, "loss": 0.0877, "step": 7582 }, { "epoch": 0.35236988847583645, "grad_norm": 0.856962188048072, "learning_rate": 7.520825187249124e-06, "loss": 0.0641, "step": 7583 }, { "epoch": 0.3524163568773234, "grad_norm": 0.6908324635678987, "learning_rate": 7.52017528267583e-06, "loss": 0.0509, "step": 7584 }, { "epoch": 0.3524628252788104, "grad_norm": 1.5757780309572165, "learning_rate": 7.5195253210179135e-06, "loss": 0.0475, "step": 7585 }, { "epoch": 0.3525092936802974, "grad_norm": 1.1359356612282416, "learning_rate": 7.518875302290097e-06, "loss": 0.062, "step": 7586 }, { "epoch": 0.3525557620817844, "grad_norm": 1.1359356612282416, "learning_rate": 7.518875302290097e-06, "loss": 0.0972, "step": 7587 }, { "epoch": 0.3526022304832714, "grad_norm": 1.0818730554555431, "learning_rate": 7.518225226507101e-06, "loss": 0.0756, "step": 7588 }, { "epoch": 0.35264869888475836, "grad_norm": 1.0149841777363457, "learning_rate": 7.5175750936836555e-06, "loss": 0.061, "step": 7589 }, { "epoch": 0.35269516728624534, "grad_norm": 0.7202955818727836, "learning_rate": 7.5169249038344835e-06, "loss": 0.0569, "step": 7590 }, { "epoch": 0.35274163568773237, "grad_norm": 1.097355569018197, "learning_rate": 7.516274656974315e-06, "loss": 0.0695, "step": 7591 }, { "epoch": 0.35278810408921935, "grad_norm": 0.9352625130446076, "learning_rate": 7.515624353117876e-06, "loss": 0.0696, "step": 7592 }, { "epoch": 0.3528345724907063, "grad_norm": 0.9158853973597169, "learning_rate": 7.514973992279898e-06, "loss": 0.0416, "step": 7593 }, { "epoch": 0.3528810408921933, "grad_norm": 0.6574059069292237, "learning_rate": 7.514323574475113e-06, "loss": 0.0502, "step": 7594 }, { "epoch": 0.3529275092936803, "grad_norm": 1.320491516243096, "learning_rate": 7.513673099718251e-06, "loss": 0.0963, "step": 7595 }, { "epoch": 0.3529739776951673, "grad_norm": 0.9395661067696393, "learning_rate": 7.5130225680240486e-06, "loss": 0.0701, "step": 7596 }, { "epoch": 0.3530204460966543, "grad_norm": 0.6226263033631779, "learning_rate": 7.51237197940724e-06, "loss": 0.049, "step": 7597 }, { "epoch": 0.35306691449814126, "grad_norm": 1.3272425210810401, "learning_rate": 7.511721333882562e-06, "loss": 0.0649, "step": 7598 }, { "epoch": 0.35311338289962824, "grad_norm": 1.6878384382394764, "learning_rate": 7.5110706314647516e-06, "loss": 0.0673, "step": 7599 }, { "epoch": 0.35315985130111527, "grad_norm": 0.9384028828696779, "learning_rate": 7.51041987216855e-06, "loss": 0.057, "step": 7600 }, { "epoch": 0.35320631970260224, "grad_norm": 0.9415659867845448, "learning_rate": 7.509769056008696e-06, "loss": 0.0834, "step": 7601 }, { "epoch": 0.3532527881040892, "grad_norm": 1.328192037084293, "learning_rate": 7.509118182999931e-06, "loss": 0.0721, "step": 7602 }, { "epoch": 0.3532992565055762, "grad_norm": 1.0190215440360186, "learning_rate": 7.508467253156997e-06, "loss": 0.0677, "step": 7603 }, { "epoch": 0.3533457249070632, "grad_norm": 0.9319106291260718, "learning_rate": 7.507816266494641e-06, "loss": 0.0782, "step": 7604 }, { "epoch": 0.3533921933085502, "grad_norm": 1.2092007786888277, "learning_rate": 7.507165223027606e-06, "loss": 0.107, "step": 7605 }, { "epoch": 0.3534386617100372, "grad_norm": 0.8947819202234422, "learning_rate": 7.506514122770641e-06, "loss": 0.0768, "step": 7606 }, { "epoch": 0.35348513011152416, "grad_norm": 0.9169118150920939, "learning_rate": 7.505862965738493e-06, "loss": 0.0536, "step": 7607 }, { "epoch": 0.35353159851301114, "grad_norm": 0.7914528657967163, "learning_rate": 7.505211751945911e-06, "loss": 0.0752, "step": 7608 }, { "epoch": 0.35357806691449817, "grad_norm": 0.8584040957861985, "learning_rate": 7.504560481407646e-06, "loss": 0.0616, "step": 7609 }, { "epoch": 0.35362453531598514, "grad_norm": 3.3045080524885564, "learning_rate": 7.503909154138451e-06, "loss": 0.0865, "step": 7610 }, { "epoch": 0.3536710037174721, "grad_norm": 1.0904345225321102, "learning_rate": 7.503257770153078e-06, "loss": 0.0549, "step": 7611 }, { "epoch": 0.3537174721189591, "grad_norm": 1.0608746565170744, "learning_rate": 7.502606329466281e-06, "loss": 0.0524, "step": 7612 }, { "epoch": 0.3537639405204461, "grad_norm": 1.664641156790412, "learning_rate": 7.5019548320928195e-06, "loss": 0.0612, "step": 7613 }, { "epoch": 0.3538104089219331, "grad_norm": 0.7633234219618649, "learning_rate": 7.501303278047445e-06, "loss": 0.0611, "step": 7614 }, { "epoch": 0.3538568773234201, "grad_norm": 1.221649327837306, "learning_rate": 7.500651667344918e-06, "loss": 0.079, "step": 7615 }, { "epoch": 0.35390334572490706, "grad_norm": 0.7162515975147028, "learning_rate": 7.500000000000001e-06, "loss": 0.0718, "step": 7616 }, { "epoch": 0.35394981412639404, "grad_norm": 1.1096533838089055, "learning_rate": 7.499348276027452e-06, "loss": 0.0649, "step": 7617 }, { "epoch": 0.35399628252788107, "grad_norm": 0.9683525975879144, "learning_rate": 7.498696495442034e-06, "loss": 0.1064, "step": 7618 }, { "epoch": 0.35404275092936804, "grad_norm": 0.9421544655248131, "learning_rate": 7.498044658258509e-06, "loss": 0.0592, "step": 7619 }, { "epoch": 0.354089219330855, "grad_norm": 0.9643879514023572, "learning_rate": 7.497392764491647e-06, "loss": 0.0609, "step": 7620 }, { "epoch": 0.354135687732342, "grad_norm": 0.6461933065388851, "learning_rate": 7.496740814156207e-06, "loss": 0.0593, "step": 7621 }, { "epoch": 0.354182156133829, "grad_norm": 0.6646226302984363, "learning_rate": 7.496088807266961e-06, "loss": 0.047, "step": 7622 }, { "epoch": 0.354228624535316, "grad_norm": 1.1329824813280374, "learning_rate": 7.495436743838678e-06, "loss": 0.0726, "step": 7623 }, { "epoch": 0.354275092936803, "grad_norm": 1.5938603916984604, "learning_rate": 7.494784623886123e-06, "loss": 0.0646, "step": 7624 }, { "epoch": 0.35432156133828996, "grad_norm": 1.7490920503642964, "learning_rate": 7.494132447424074e-06, "loss": 0.0787, "step": 7625 }, { "epoch": 0.35436802973977694, "grad_norm": 0.6049424229696462, "learning_rate": 7.493480214467298e-06, "loss": 0.0383, "step": 7626 }, { "epoch": 0.35441449814126397, "grad_norm": 0.6236774228892256, "learning_rate": 7.492827925030571e-06, "loss": 0.0534, "step": 7627 }, { "epoch": 0.35446096654275094, "grad_norm": 1.3689694540536654, "learning_rate": 7.492175579128669e-06, "loss": 0.0909, "step": 7628 }, { "epoch": 0.3545074349442379, "grad_norm": 0.8101969398214429, "learning_rate": 7.491523176776365e-06, "loss": 0.0818, "step": 7629 }, { "epoch": 0.3545539033457249, "grad_norm": 1.4608594683508735, "learning_rate": 7.490870717988441e-06, "loss": 0.0692, "step": 7630 }, { "epoch": 0.3546003717472119, "grad_norm": 1.1518270562536401, "learning_rate": 7.490218202779673e-06, "loss": 0.053, "step": 7631 }, { "epoch": 0.3546468401486989, "grad_norm": 0.9449813067635848, "learning_rate": 7.489565631164841e-06, "loss": 0.0657, "step": 7632 }, { "epoch": 0.3546933085501859, "grad_norm": 1.3708677630364292, "learning_rate": 7.4889130031587275e-06, "loss": 0.0527, "step": 7633 }, { "epoch": 0.35473977695167286, "grad_norm": 0.9533577742382582, "learning_rate": 7.488260318776116e-06, "loss": 0.0688, "step": 7634 }, { "epoch": 0.35478624535315983, "grad_norm": 0.8185313159793807, "learning_rate": 7.487607578031788e-06, "loss": 0.0538, "step": 7635 }, { "epoch": 0.35483271375464687, "grad_norm": 0.545552376673886, "learning_rate": 7.4869547809405305e-06, "loss": 0.049, "step": 7636 }, { "epoch": 0.35487918215613384, "grad_norm": 2.441439432201404, "learning_rate": 7.486301927517131e-06, "loss": 0.0601, "step": 7637 }, { "epoch": 0.3549256505576208, "grad_norm": 0.5504412914698561, "learning_rate": 7.485649017776375e-06, "loss": 0.0427, "step": 7638 }, { "epoch": 0.3549721189591078, "grad_norm": 0.713221131638914, "learning_rate": 7.484996051733053e-06, "loss": 0.0578, "step": 7639 }, { "epoch": 0.3550185873605948, "grad_norm": 0.7391866265775036, "learning_rate": 7.484343029401955e-06, "loss": 0.0828, "step": 7640 }, { "epoch": 0.3550650557620818, "grad_norm": 0.8661509135898153, "learning_rate": 7.483689950797872e-06, "loss": 0.0616, "step": 7641 }, { "epoch": 0.3551115241635688, "grad_norm": 2.6037248489985756, "learning_rate": 7.483036815935598e-06, "loss": 0.0978, "step": 7642 }, { "epoch": 0.35515799256505576, "grad_norm": 1.7193083985963196, "learning_rate": 7.4823836248299276e-06, "loss": 0.0898, "step": 7643 }, { "epoch": 0.35520446096654273, "grad_norm": 1.1839049707042484, "learning_rate": 7.481730377495654e-06, "loss": 0.0543, "step": 7644 }, { "epoch": 0.35525092936802977, "grad_norm": 0.6200581643102454, "learning_rate": 7.481077073947576e-06, "loss": 0.0429, "step": 7645 }, { "epoch": 0.35529739776951674, "grad_norm": 1.410932935907303, "learning_rate": 7.480423714200491e-06, "loss": 0.0909, "step": 7646 }, { "epoch": 0.3553438661710037, "grad_norm": 0.8659548933075918, "learning_rate": 7.479770298269198e-06, "loss": 0.1064, "step": 7647 }, { "epoch": 0.3553903345724907, "grad_norm": 0.8316256828524698, "learning_rate": 7.4791168261684985e-06, "loss": 0.0685, "step": 7648 }, { "epoch": 0.3554368029739777, "grad_norm": 1.6533923024410848, "learning_rate": 7.4784632979131945e-06, "loss": 0.0672, "step": 7649 }, { "epoch": 0.3554832713754647, "grad_norm": 1.465536785756869, "learning_rate": 7.477809713518085e-06, "loss": 0.0814, "step": 7650 }, { "epoch": 0.3555297397769517, "grad_norm": 1.05974556466046, "learning_rate": 7.477156072997981e-06, "loss": 0.0785, "step": 7651 }, { "epoch": 0.35557620817843866, "grad_norm": 1.3071688273574786, "learning_rate": 7.476502376367684e-06, "loss": 0.0544, "step": 7652 }, { "epoch": 0.35562267657992563, "grad_norm": 1.3516791460141566, "learning_rate": 7.475848623641999e-06, "loss": 0.0626, "step": 7653 }, { "epoch": 0.35566914498141267, "grad_norm": 1.0698962860986783, "learning_rate": 7.475194814835741e-06, "loss": 0.064, "step": 7654 }, { "epoch": 0.35571561338289964, "grad_norm": 0.4971116590820499, "learning_rate": 7.474540949963713e-06, "loss": 0.0597, "step": 7655 }, { "epoch": 0.3557620817843866, "grad_norm": 3.0565220079107, "learning_rate": 7.473887029040729e-06, "loss": 0.1158, "step": 7656 }, { "epoch": 0.3558085501858736, "grad_norm": 1.1332258539597537, "learning_rate": 7.473233052081599e-06, "loss": 0.0697, "step": 7657 }, { "epoch": 0.35585501858736057, "grad_norm": 1.2100202434073604, "learning_rate": 7.472579019101137e-06, "loss": 0.0715, "step": 7658 }, { "epoch": 0.3559014869888476, "grad_norm": 0.6287520462299769, "learning_rate": 7.47192493011416e-06, "loss": 0.0448, "step": 7659 }, { "epoch": 0.3559479553903346, "grad_norm": 0.7079268518202672, "learning_rate": 7.4712707851354805e-06, "loss": 0.0538, "step": 7660 }, { "epoch": 0.35599442379182156, "grad_norm": 1.6971629010587563, "learning_rate": 7.4706165841799175e-06, "loss": 0.093, "step": 7661 }, { "epoch": 0.35604089219330853, "grad_norm": 1.1175880608333777, "learning_rate": 7.469962327262285e-06, "loss": 0.0625, "step": 7662 }, { "epoch": 0.35608736059479557, "grad_norm": 2.967687401923668, "learning_rate": 7.46930801439741e-06, "loss": 0.087, "step": 7663 }, { "epoch": 0.35613382899628254, "grad_norm": 0.5988025201380018, "learning_rate": 7.468653645600109e-06, "loss": 0.0359, "step": 7664 }, { "epoch": 0.3561802973977695, "grad_norm": 1.4070689247854398, "learning_rate": 7.467999220885204e-06, "loss": 0.0698, "step": 7665 }, { "epoch": 0.3562267657992565, "grad_norm": 0.863956840166477, "learning_rate": 7.467344740267521e-06, "loss": 0.066, "step": 7666 }, { "epoch": 0.35627323420074347, "grad_norm": 1.146699402735085, "learning_rate": 7.466690203761879e-06, "loss": 0.0643, "step": 7667 }, { "epoch": 0.3563197026022305, "grad_norm": 0.9438492104179657, "learning_rate": 7.466035611383111e-06, "loss": 0.0809, "step": 7668 }, { "epoch": 0.3563661710037175, "grad_norm": 2.445912837840837, "learning_rate": 7.4653809631460406e-06, "loss": 0.0703, "step": 7669 }, { "epoch": 0.35641263940520446, "grad_norm": 0.689593075685702, "learning_rate": 7.4647262590654955e-06, "loss": 0.0548, "step": 7670 }, { "epoch": 0.35645910780669143, "grad_norm": 0.5615812709458452, "learning_rate": 7.464071499156307e-06, "loss": 0.0443, "step": 7671 }, { "epoch": 0.35650557620817847, "grad_norm": 0.8533969112209864, "learning_rate": 7.463416683433306e-06, "loss": 0.0487, "step": 7672 }, { "epoch": 0.35655204460966544, "grad_norm": 0.6812871755912596, "learning_rate": 7.462761811911323e-06, "loss": 0.054, "step": 7673 }, { "epoch": 0.3565985130111524, "grad_norm": 1.2818676049571471, "learning_rate": 7.462106884605193e-06, "loss": 0.0968, "step": 7674 }, { "epoch": 0.3566449814126394, "grad_norm": 0.7089320225626387, "learning_rate": 7.4614519015297526e-06, "loss": 0.0684, "step": 7675 }, { "epoch": 0.35669144981412637, "grad_norm": 0.7315883605734563, "learning_rate": 7.460796862699836e-06, "loss": 0.0598, "step": 7676 }, { "epoch": 0.3567379182156134, "grad_norm": 1.3676594104170605, "learning_rate": 7.4601417681302794e-06, "loss": 0.0712, "step": 7677 }, { "epoch": 0.3567843866171004, "grad_norm": 0.5260974259847355, "learning_rate": 7.459486617835924e-06, "loss": 0.0628, "step": 7678 }, { "epoch": 0.35683085501858736, "grad_norm": 0.632984286186742, "learning_rate": 7.458831411831606e-06, "loss": 0.0501, "step": 7679 }, { "epoch": 0.35687732342007433, "grad_norm": 0.7028248370958141, "learning_rate": 7.4581761501321694e-06, "loss": 0.0681, "step": 7680 }, { "epoch": 0.35692379182156136, "grad_norm": 0.5171605254388861, "learning_rate": 7.457520832752456e-06, "loss": 0.0438, "step": 7681 }, { "epoch": 0.35697026022304834, "grad_norm": 1.2530826835778048, "learning_rate": 7.456865459707309e-06, "loss": 0.0857, "step": 7682 }, { "epoch": 0.3570167286245353, "grad_norm": 0.7639768423569865, "learning_rate": 7.456210031011573e-06, "loss": 0.0439, "step": 7683 }, { "epoch": 0.3570631970260223, "grad_norm": 2.965284114920511, "learning_rate": 7.4555545466800975e-06, "loss": 0.1113, "step": 7684 }, { "epoch": 0.35710966542750927, "grad_norm": 0.6512780045961711, "learning_rate": 7.4548990067277245e-06, "loss": 0.0536, "step": 7685 }, { "epoch": 0.3571561338289963, "grad_norm": 2.7029261854155466, "learning_rate": 7.4542434111693055e-06, "loss": 0.0603, "step": 7686 }, { "epoch": 0.3572026022304833, "grad_norm": 0.8989418020968986, "learning_rate": 7.453587760019691e-06, "loss": 0.0585, "step": 7687 }, { "epoch": 0.35724907063197026, "grad_norm": 1.3548972573385407, "learning_rate": 7.45293205329373e-06, "loss": 0.0912, "step": 7688 }, { "epoch": 0.35729553903345723, "grad_norm": 0.5303205891336079, "learning_rate": 7.452276291006278e-06, "loss": 0.0412, "step": 7689 }, { "epoch": 0.35734200743494426, "grad_norm": 1.0688573531208234, "learning_rate": 7.4516204731721865e-06, "loss": 0.0504, "step": 7690 }, { "epoch": 0.35738847583643124, "grad_norm": 1.0841862305017893, "learning_rate": 7.45096459980631e-06, "loss": 0.0908, "step": 7691 }, { "epoch": 0.3574349442379182, "grad_norm": 1.649310022630214, "learning_rate": 7.4503086709235064e-06, "loss": 0.1035, "step": 7692 }, { "epoch": 0.3574814126394052, "grad_norm": 0.7046310135234044, "learning_rate": 7.449652686538633e-06, "loss": 0.0488, "step": 7693 }, { "epoch": 0.35752788104089217, "grad_norm": 1.1071355673447199, "learning_rate": 7.448996646666546e-06, "loss": 0.0415, "step": 7694 }, { "epoch": 0.3575743494423792, "grad_norm": 0.8214825694018566, "learning_rate": 7.448340551322109e-06, "loss": 0.0416, "step": 7695 }, { "epoch": 0.3576208178438662, "grad_norm": 0.9620501641021056, "learning_rate": 7.447684400520183e-06, "loss": 0.0433, "step": 7696 }, { "epoch": 0.35766728624535316, "grad_norm": 1.1416954187301445, "learning_rate": 7.4470281942756275e-06, "loss": 0.0919, "step": 7697 }, { "epoch": 0.35771375464684013, "grad_norm": 1.6513205768226311, "learning_rate": 7.446371932603308e-06, "loss": 0.0439, "step": 7698 }, { "epoch": 0.35776022304832716, "grad_norm": 1.15594381679748, "learning_rate": 7.4457156155180895e-06, "loss": 0.0656, "step": 7699 }, { "epoch": 0.35780669144981414, "grad_norm": 1.723744414903121, "learning_rate": 7.44505924303484e-06, "loss": 0.0496, "step": 7700 }, { "epoch": 0.3578531598513011, "grad_norm": 0.6035981991903856, "learning_rate": 7.444402815168422e-06, "loss": 0.0629, "step": 7701 }, { "epoch": 0.3578996282527881, "grad_norm": 1.3622146264214399, "learning_rate": 7.443746331933711e-06, "loss": 0.0756, "step": 7702 }, { "epoch": 0.35794609665427507, "grad_norm": 1.5344279042442295, "learning_rate": 7.443089793345571e-06, "loss": 0.0815, "step": 7703 }, { "epoch": 0.3579925650557621, "grad_norm": 1.8213212206416887, "learning_rate": 7.442433199418877e-06, "loss": 0.0644, "step": 7704 }, { "epoch": 0.3580390334572491, "grad_norm": 2.0146318375218595, "learning_rate": 7.441776550168501e-06, "loss": 0.074, "step": 7705 }, { "epoch": 0.35808550185873605, "grad_norm": 1.9893498444900677, "learning_rate": 7.4411198456093145e-06, "loss": 0.0719, "step": 7706 }, { "epoch": 0.35813197026022303, "grad_norm": 4.596329832479251, "learning_rate": 7.440463085756196e-06, "loss": 0.0851, "step": 7707 }, { "epoch": 0.35817843866171006, "grad_norm": 1.4847467158488825, "learning_rate": 7.439806270624019e-06, "loss": 0.0754, "step": 7708 }, { "epoch": 0.35822490706319704, "grad_norm": 0.9186584261085771, "learning_rate": 7.439149400227663e-06, "loss": 0.0665, "step": 7709 }, { "epoch": 0.358271375464684, "grad_norm": 1.3289705214552416, "learning_rate": 7.438492474582005e-06, "loss": 0.0682, "step": 7710 }, { "epoch": 0.358317843866171, "grad_norm": 1.4667253197265124, "learning_rate": 7.437835493701926e-06, "loss": 0.086, "step": 7711 }, { "epoch": 0.35836431226765797, "grad_norm": 0.6416868214250077, "learning_rate": 7.437178457602308e-06, "loss": 0.0534, "step": 7712 }, { "epoch": 0.358410780669145, "grad_norm": 1.2385041117876974, "learning_rate": 7.436521366298033e-06, "loss": 0.0741, "step": 7713 }, { "epoch": 0.358457249070632, "grad_norm": 4.928843620230104, "learning_rate": 7.4358642198039835e-06, "loss": 0.1038, "step": 7714 }, { "epoch": 0.35850371747211895, "grad_norm": 0.8375137153607647, "learning_rate": 7.435207018135047e-06, "loss": 0.0546, "step": 7715 }, { "epoch": 0.35855018587360593, "grad_norm": 0.9204368760498598, "learning_rate": 7.434549761306107e-06, "loss": 0.0731, "step": 7716 }, { "epoch": 0.35859665427509296, "grad_norm": 1.7096798870770764, "learning_rate": 7.433892449332054e-06, "loss": 0.0407, "step": 7717 }, { "epoch": 0.35864312267657994, "grad_norm": 0.832942595128429, "learning_rate": 7.433235082227774e-06, "loss": 0.0538, "step": 7718 }, { "epoch": 0.3586895910780669, "grad_norm": 1.202230661377104, "learning_rate": 7.43257766000816e-06, "loss": 0.061, "step": 7719 }, { "epoch": 0.3587360594795539, "grad_norm": 0.9315803527388105, "learning_rate": 7.4319201826881004e-06, "loss": 0.0669, "step": 7720 }, { "epoch": 0.35878252788104087, "grad_norm": 2.3300401155130195, "learning_rate": 7.43126265028249e-06, "loss": 0.0809, "step": 7721 }, { "epoch": 0.3588289962825279, "grad_norm": 0.6980315390873489, "learning_rate": 7.4306050628062225e-06, "loss": 0.0428, "step": 7722 }, { "epoch": 0.3588754646840149, "grad_norm": 1.0279029058974025, "learning_rate": 7.429947420274191e-06, "loss": 0.0518, "step": 7723 }, { "epoch": 0.35892193308550185, "grad_norm": 1.5061181438077107, "learning_rate": 7.429289722701292e-06, "loss": 0.0733, "step": 7724 }, { "epoch": 0.35896840148698883, "grad_norm": 0.761689128407033, "learning_rate": 7.428631970102426e-06, "loss": 0.0592, "step": 7725 }, { "epoch": 0.35901486988847586, "grad_norm": 0.8897023845153851, "learning_rate": 7.427974162492489e-06, "loss": 0.0439, "step": 7726 }, { "epoch": 0.35906133828996284, "grad_norm": 1.4906207160268523, "learning_rate": 7.427316299886381e-06, "loss": 0.0811, "step": 7727 }, { "epoch": 0.3591078066914498, "grad_norm": 0.8650126596317761, "learning_rate": 7.426658382299006e-06, "loss": 0.053, "step": 7728 }, { "epoch": 0.3591542750929368, "grad_norm": 1.4112605085825707, "learning_rate": 7.4260004097452635e-06, "loss": 0.0764, "step": 7729 }, { "epoch": 0.35920074349442377, "grad_norm": 0.9314131817363407, "learning_rate": 7.425342382240057e-06, "loss": 0.0768, "step": 7730 }, { "epoch": 0.3592472118959108, "grad_norm": 0.9642555725145096, "learning_rate": 7.424684299798295e-06, "loss": 0.0738, "step": 7731 }, { "epoch": 0.3592936802973978, "grad_norm": 1.3396878418922211, "learning_rate": 7.4240261624348805e-06, "loss": 0.1057, "step": 7732 }, { "epoch": 0.35934014869888475, "grad_norm": 1.0065229225316707, "learning_rate": 7.423367970164723e-06, "loss": 0.0724, "step": 7733 }, { "epoch": 0.35938661710037173, "grad_norm": 0.5083714080064543, "learning_rate": 7.42270972300273e-06, "loss": 0.0375, "step": 7734 }, { "epoch": 0.35943308550185876, "grad_norm": 2.099591175189857, "learning_rate": 7.422051420963812e-06, "loss": 0.0964, "step": 7735 }, { "epoch": 0.35947955390334574, "grad_norm": 1.2857362162998147, "learning_rate": 7.42139306406288e-06, "loss": 0.0635, "step": 7736 }, { "epoch": 0.3595260223048327, "grad_norm": 0.5335916376172003, "learning_rate": 7.420734652314847e-06, "loss": 0.0462, "step": 7737 }, { "epoch": 0.3595724907063197, "grad_norm": 0.9930740775824236, "learning_rate": 7.4200761857346256e-06, "loss": 0.0909, "step": 7738 }, { "epoch": 0.35961895910780667, "grad_norm": 1.144513183425848, "learning_rate": 7.419417664337132e-06, "loss": 0.0769, "step": 7739 }, { "epoch": 0.3596654275092937, "grad_norm": 3.524836318299841, "learning_rate": 7.418759088137282e-06, "loss": 0.0791, "step": 7740 }, { "epoch": 0.3597118959107807, "grad_norm": 0.7740991750635827, "learning_rate": 7.418100457149994e-06, "loss": 0.0494, "step": 7741 }, { "epoch": 0.35975836431226765, "grad_norm": 0.9601561743735816, "learning_rate": 7.4174417713901835e-06, "loss": 0.085, "step": 7742 }, { "epoch": 0.35980483271375463, "grad_norm": 0.9551464108872212, "learning_rate": 7.416783030872774e-06, "loss": 0.0629, "step": 7743 }, { "epoch": 0.35985130111524166, "grad_norm": 0.6546462624800349, "learning_rate": 7.416124235612685e-06, "loss": 0.0517, "step": 7744 }, { "epoch": 0.35989776951672864, "grad_norm": 0.952571056897926, "learning_rate": 7.415465385624838e-06, "loss": 0.0639, "step": 7745 }, { "epoch": 0.3599442379182156, "grad_norm": 0.9186777140342607, "learning_rate": 7.414806480924159e-06, "loss": 0.0656, "step": 7746 }, { "epoch": 0.3599907063197026, "grad_norm": 0.7763275164917482, "learning_rate": 7.414147521525572e-06, "loss": 0.0609, "step": 7747 }, { "epoch": 0.36003717472118957, "grad_norm": 0.6970106331306727, "learning_rate": 7.413488507444002e-06, "loss": 0.0464, "step": 7748 }, { "epoch": 0.3600836431226766, "grad_norm": 0.7732123639351787, "learning_rate": 7.412829438694378e-06, "loss": 0.0689, "step": 7749 }, { "epoch": 0.3601301115241636, "grad_norm": 0.8012358093389071, "learning_rate": 7.412170315291627e-06, "loss": 0.057, "step": 7750 }, { "epoch": 0.36017657992565055, "grad_norm": 0.574075395670735, "learning_rate": 7.411511137250679e-06, "loss": 0.0396, "step": 7751 }, { "epoch": 0.36022304832713753, "grad_norm": 1.0165329523734523, "learning_rate": 7.410851904586467e-06, "loss": 0.0537, "step": 7752 }, { "epoch": 0.36026951672862456, "grad_norm": 1.2826991596466066, "learning_rate": 7.410192617313922e-06, "loss": 0.076, "step": 7753 }, { "epoch": 0.36031598513011154, "grad_norm": 0.4188796116080734, "learning_rate": 7.409533275447975e-06, "loss": 0.0432, "step": 7754 }, { "epoch": 0.3603624535315985, "grad_norm": 0.7581185965670421, "learning_rate": 7.408873879003566e-06, "loss": 0.0615, "step": 7755 }, { "epoch": 0.3604089219330855, "grad_norm": 1.1495454302244061, "learning_rate": 7.408214427995628e-06, "loss": 0.0609, "step": 7756 }, { "epoch": 0.36045539033457247, "grad_norm": 1.5595091677815507, "learning_rate": 7.407554922439098e-06, "loss": 0.0731, "step": 7757 }, { "epoch": 0.3605018587360595, "grad_norm": 1.52413217249854, "learning_rate": 7.406895362348916e-06, "loss": 0.0737, "step": 7758 }, { "epoch": 0.3605483271375465, "grad_norm": 1.3926447279860226, "learning_rate": 7.406235747740019e-06, "loss": 0.0893, "step": 7759 }, { "epoch": 0.36059479553903345, "grad_norm": 0.7951108902377746, "learning_rate": 7.405576078627351e-06, "loss": 0.0561, "step": 7760 }, { "epoch": 0.36064126394052043, "grad_norm": 1.1774086375451531, "learning_rate": 7.404916355025853e-06, "loss": 0.0819, "step": 7761 }, { "epoch": 0.36068773234200746, "grad_norm": 0.9728415647959394, "learning_rate": 7.40425657695047e-06, "loss": 0.0604, "step": 7762 }, { "epoch": 0.36073420074349444, "grad_norm": 0.6748924938715012, "learning_rate": 7.403596744416142e-06, "loss": 0.0513, "step": 7763 }, { "epoch": 0.3607806691449814, "grad_norm": 1.3423420152877912, "learning_rate": 7.402936857437819e-06, "loss": 0.0692, "step": 7764 }, { "epoch": 0.3608271375464684, "grad_norm": 0.667250202175188, "learning_rate": 7.402276916030448e-06, "loss": 0.0591, "step": 7765 }, { "epoch": 0.36087360594795537, "grad_norm": 0.9339137854586204, "learning_rate": 7.401616920208976e-06, "loss": 0.059, "step": 7766 }, { "epoch": 0.3609200743494424, "grad_norm": 0.6933851257255796, "learning_rate": 7.400956869988354e-06, "loss": 0.033, "step": 7767 }, { "epoch": 0.3609665427509294, "grad_norm": 0.997521352536689, "learning_rate": 7.40029676538353e-06, "loss": 0.0439, "step": 7768 }, { "epoch": 0.36101301115241635, "grad_norm": 0.5407596620298242, "learning_rate": 7.399636606409459e-06, "loss": 0.0495, "step": 7769 }, { "epoch": 0.36105947955390333, "grad_norm": 0.7109351816094746, "learning_rate": 7.398976393081094e-06, "loss": 0.0681, "step": 7770 }, { "epoch": 0.36110594795539036, "grad_norm": 0.8341053641284786, "learning_rate": 7.398316125413388e-06, "loss": 0.0641, "step": 7771 }, { "epoch": 0.36115241635687734, "grad_norm": 2.6330422004739007, "learning_rate": 7.397655803421298e-06, "loss": 0.0669, "step": 7772 }, { "epoch": 0.3611988847583643, "grad_norm": 0.8958537479910966, "learning_rate": 7.396995427119781e-06, "loss": 0.0549, "step": 7773 }, { "epoch": 0.3612453531598513, "grad_norm": 1.4071524563519269, "learning_rate": 7.396334996523793e-06, "loss": 0.0673, "step": 7774 }, { "epoch": 0.36129182156133827, "grad_norm": 1.9467410734057442, "learning_rate": 7.395674511648298e-06, "loss": 0.0948, "step": 7775 }, { "epoch": 0.3613382899628253, "grad_norm": 0.7876880228998342, "learning_rate": 7.395013972508252e-06, "loss": 0.0501, "step": 7776 }, { "epoch": 0.3613847583643123, "grad_norm": 0.507030614757999, "learning_rate": 7.394353379118619e-06, "loss": 0.0394, "step": 7777 }, { "epoch": 0.36143122676579925, "grad_norm": 0.7030300701185896, "learning_rate": 7.3936927314943615e-06, "loss": 0.0477, "step": 7778 }, { "epoch": 0.36147769516728623, "grad_norm": 1.3075180776974613, "learning_rate": 7.3930320296504445e-06, "loss": 0.0531, "step": 7779 }, { "epoch": 0.36152416356877326, "grad_norm": 0.7649879950663043, "learning_rate": 7.392371273601834e-06, "loss": 0.053, "step": 7780 }, { "epoch": 0.36157063197026024, "grad_norm": 0.6552243551840631, "learning_rate": 7.391710463363496e-06, "loss": 0.0522, "step": 7781 }, { "epoch": 0.3616171003717472, "grad_norm": 1.4059993429352782, "learning_rate": 7.3910495989503995e-06, "loss": 0.0926, "step": 7782 }, { "epoch": 0.3616635687732342, "grad_norm": 6.081615500859389, "learning_rate": 7.390388680377512e-06, "loss": 0.0996, "step": 7783 }, { "epoch": 0.36171003717472117, "grad_norm": 2.3813002324911086, "learning_rate": 7.389727707659806e-06, "loss": 0.0548, "step": 7784 }, { "epoch": 0.3617565055762082, "grad_norm": 0.4967306718575635, "learning_rate": 7.389066680812252e-06, "loss": 0.0532, "step": 7785 }, { "epoch": 0.3618029739776952, "grad_norm": 0.8372442866880245, "learning_rate": 7.388405599849821e-06, "loss": 0.0672, "step": 7786 }, { "epoch": 0.36184944237918215, "grad_norm": 0.932919044591645, "learning_rate": 7.387744464787493e-06, "loss": 0.0508, "step": 7787 }, { "epoch": 0.36189591078066913, "grad_norm": 1.6532581405878157, "learning_rate": 7.387083275640237e-06, "loss": 0.0803, "step": 7788 }, { "epoch": 0.36194237918215616, "grad_norm": 1.1657662948278986, "learning_rate": 7.386422032423034e-06, "loss": 0.0857, "step": 7789 }, { "epoch": 0.36198884758364314, "grad_norm": 0.5685226643461754, "learning_rate": 7.38576073515086e-06, "loss": 0.0458, "step": 7790 }, { "epoch": 0.3620353159851301, "grad_norm": 1.4651134277193507, "learning_rate": 7.3850993838386946e-06, "loss": 0.1009, "step": 7791 }, { "epoch": 0.3620817843866171, "grad_norm": 1.2730420443970267, "learning_rate": 7.3844379785015175e-06, "loss": 0.0924, "step": 7792 }, { "epoch": 0.36212825278810407, "grad_norm": 1.2320416466813984, "learning_rate": 7.383776519154311e-06, "loss": 0.0757, "step": 7793 }, { "epoch": 0.3621747211895911, "grad_norm": 0.8876949373650692, "learning_rate": 7.383115005812056e-06, "loss": 0.0631, "step": 7794 }, { "epoch": 0.3622211895910781, "grad_norm": 1.1533735376876217, "learning_rate": 7.382453438489738e-06, "loss": 0.0664, "step": 7795 }, { "epoch": 0.36226765799256505, "grad_norm": 0.8091765946925047, "learning_rate": 7.381791817202344e-06, "loss": 0.073, "step": 7796 }, { "epoch": 0.362314126394052, "grad_norm": 0.5841954958509034, "learning_rate": 7.381130141964858e-06, "loss": 0.064, "step": 7797 }, { "epoch": 0.36236059479553906, "grad_norm": 1.2514331489711348, "learning_rate": 7.380468412792267e-06, "loss": 0.0551, "step": 7798 }, { "epoch": 0.36240706319702604, "grad_norm": 0.5384157026674615, "learning_rate": 7.379806629699564e-06, "loss": 0.0431, "step": 7799 }, { "epoch": 0.362453531598513, "grad_norm": 1.0728685158906448, "learning_rate": 7.379144792701734e-06, "loss": 0.0764, "step": 7800 }, { "epoch": 0.3625, "grad_norm": 1.0945784035107162, "learning_rate": 7.37848290181377e-06, "loss": 0.0546, "step": 7801 }, { "epoch": 0.36254646840148697, "grad_norm": 1.6874946607528734, "learning_rate": 7.377820957050666e-06, "loss": 0.1103, "step": 7802 }, { "epoch": 0.362592936802974, "grad_norm": 1.0184397575028334, "learning_rate": 7.377158958427413e-06, "loss": 0.0846, "step": 7803 }, { "epoch": 0.362639405204461, "grad_norm": 3.288922856153945, "learning_rate": 7.3764969059590096e-06, "loss": 0.0917, "step": 7804 }, { "epoch": 0.36268587360594795, "grad_norm": 0.7626621152892943, "learning_rate": 7.37583479966045e-06, "loss": 0.0706, "step": 7805 }, { "epoch": 0.3627323420074349, "grad_norm": 0.7651632653306203, "learning_rate": 7.375172639546731e-06, "loss": 0.0666, "step": 7806 }, { "epoch": 0.36277881040892196, "grad_norm": 0.5513086658724283, "learning_rate": 7.37451042563285e-06, "loss": 0.0508, "step": 7807 }, { "epoch": 0.36282527881040894, "grad_norm": 1.2973427088082812, "learning_rate": 7.373848157933812e-06, "loss": 0.0709, "step": 7808 }, { "epoch": 0.3628717472118959, "grad_norm": 1.4357580089727162, "learning_rate": 7.3731858364646135e-06, "loss": 0.0607, "step": 7809 }, { "epoch": 0.3629182156133829, "grad_norm": 0.8326275014589867, "learning_rate": 7.372523461240257e-06, "loss": 0.0754, "step": 7810 }, { "epoch": 0.36296468401486986, "grad_norm": 0.8009197505359077, "learning_rate": 7.37186103227575e-06, "loss": 0.0685, "step": 7811 }, { "epoch": 0.3630111524163569, "grad_norm": 0.814681964277903, "learning_rate": 7.371198549586091e-06, "loss": 0.0608, "step": 7812 }, { "epoch": 0.3630576208178439, "grad_norm": 1.4506402479931237, "learning_rate": 7.370536013186291e-06, "loss": 0.0715, "step": 7813 }, { "epoch": 0.36310408921933085, "grad_norm": 1.4239127503935705, "learning_rate": 7.369873423091355e-06, "loss": 0.0814, "step": 7814 }, { "epoch": 0.3631505576208178, "grad_norm": 0.8283823700180063, "learning_rate": 7.369210779316291e-06, "loss": 0.0546, "step": 7815 }, { "epoch": 0.36319702602230486, "grad_norm": 6.386173010376646, "learning_rate": 7.36854808187611e-06, "loss": 0.0824, "step": 7816 }, { "epoch": 0.36324349442379184, "grad_norm": 1.8349524057038895, "learning_rate": 7.367885330785822e-06, "loss": 0.0702, "step": 7817 }, { "epoch": 0.3632899628252788, "grad_norm": 0.7809174243465551, "learning_rate": 7.36722252606044e-06, "loss": 0.0645, "step": 7818 }, { "epoch": 0.3633364312267658, "grad_norm": 0.6820640355065798, "learning_rate": 7.366559667714974e-06, "loss": 0.0601, "step": 7819 }, { "epoch": 0.36338289962825276, "grad_norm": 0.9369465714620304, "learning_rate": 7.365896755764443e-06, "loss": 0.0574, "step": 7820 }, { "epoch": 0.3634293680297398, "grad_norm": 0.6634523991134108, "learning_rate": 7.36523379022386e-06, "loss": 0.0455, "step": 7821 }, { "epoch": 0.3634758364312268, "grad_norm": 1.6278678892366338, "learning_rate": 7.364570771108241e-06, "loss": 0.0947, "step": 7822 }, { "epoch": 0.36352230483271375, "grad_norm": 1.1125615763036656, "learning_rate": 7.363907698432608e-06, "loss": 0.0658, "step": 7823 }, { "epoch": 0.3635687732342007, "grad_norm": 0.9207310882564792, "learning_rate": 7.363244572211975e-06, "loss": 0.0673, "step": 7824 }, { "epoch": 0.36361524163568776, "grad_norm": 0.9881606341358602, "learning_rate": 7.362581392461367e-06, "loss": 0.0611, "step": 7825 }, { "epoch": 0.36366171003717473, "grad_norm": 2.2065261038908925, "learning_rate": 7.361918159195804e-06, "loss": 0.059, "step": 7826 }, { "epoch": 0.3637081784386617, "grad_norm": 1.0293855031137058, "learning_rate": 7.361254872430305e-06, "loss": 0.0825, "step": 7827 }, { "epoch": 0.3637546468401487, "grad_norm": 1.1662744649640633, "learning_rate": 7.3605915321799035e-06, "loss": 0.0698, "step": 7828 }, { "epoch": 0.36380111524163566, "grad_norm": 2.5196998408259583, "learning_rate": 7.359928138459615e-06, "loss": 0.0728, "step": 7829 }, { "epoch": 0.3638475836431227, "grad_norm": 0.9319184699656397, "learning_rate": 7.359264691284473e-06, "loss": 0.0684, "step": 7830 }, { "epoch": 0.3638940520446097, "grad_norm": 1.4729176058186269, "learning_rate": 7.358601190669502e-06, "loss": 0.0885, "step": 7831 }, { "epoch": 0.36394052044609665, "grad_norm": 0.5039355606267618, "learning_rate": 7.357937636629731e-06, "loss": 0.0472, "step": 7832 }, { "epoch": 0.3639869888475836, "grad_norm": 0.9442719857379896, "learning_rate": 7.357274029180191e-06, "loss": 0.0561, "step": 7833 }, { "epoch": 0.36403345724907066, "grad_norm": 0.867679217274607, "learning_rate": 7.356610368335913e-06, "loss": 0.0371, "step": 7834 }, { "epoch": 0.36407992565055763, "grad_norm": 0.8086050640220366, "learning_rate": 7.355946654111931e-06, "loss": 0.0782, "step": 7835 }, { "epoch": 0.3641263940520446, "grad_norm": 3.9625784459179925, "learning_rate": 7.355282886523276e-06, "loss": 0.0578, "step": 7836 }, { "epoch": 0.3641728624535316, "grad_norm": 1.1181216281149349, "learning_rate": 7.354619065584987e-06, "loss": 0.0455, "step": 7837 }, { "epoch": 0.36421933085501856, "grad_norm": 2.243894720604082, "learning_rate": 7.353955191312097e-06, "loss": 0.0609, "step": 7838 }, { "epoch": 0.3642657992565056, "grad_norm": 1.065101182610927, "learning_rate": 7.353291263719643e-06, "loss": 0.0708, "step": 7839 }, { "epoch": 0.3643122676579926, "grad_norm": 1.248270469960717, "learning_rate": 7.352627282822667e-06, "loss": 0.0733, "step": 7840 }, { "epoch": 0.36435873605947955, "grad_norm": 1.9024651176977883, "learning_rate": 7.351963248636206e-06, "loss": 0.0849, "step": 7841 }, { "epoch": 0.3644052044609665, "grad_norm": 1.26724480427962, "learning_rate": 7.351299161175303e-06, "loss": 0.1158, "step": 7842 }, { "epoch": 0.36445167286245356, "grad_norm": 1.5764204398280197, "learning_rate": 7.350635020454998e-06, "loss": 0.0733, "step": 7843 }, { "epoch": 0.36449814126394053, "grad_norm": 0.5808077243385382, "learning_rate": 7.349970826490336e-06, "loss": 0.0402, "step": 7844 }, { "epoch": 0.3645446096654275, "grad_norm": 1.7552832493702557, "learning_rate": 7.349306579296361e-06, "loss": 0.0713, "step": 7845 }, { "epoch": 0.3645910780669145, "grad_norm": 1.3441130511136086, "learning_rate": 7.3486422788881205e-06, "loss": 0.0616, "step": 7846 }, { "epoch": 0.36463754646840146, "grad_norm": 2.0093155715146036, "learning_rate": 7.34797792528066e-06, "loss": 0.0834, "step": 7847 }, { "epoch": 0.3646840148698885, "grad_norm": 0.5156319180634239, "learning_rate": 7.347313518489029e-06, "loss": 0.0449, "step": 7848 }, { "epoch": 0.36473048327137547, "grad_norm": 1.328222616855111, "learning_rate": 7.346649058528276e-06, "loss": 0.0792, "step": 7849 }, { "epoch": 0.36477695167286245, "grad_norm": 6.364991470366813, "learning_rate": 7.345984545413452e-06, "loss": 0.071, "step": 7850 }, { "epoch": 0.3648234200743494, "grad_norm": 1.3266100217328527, "learning_rate": 7.345319979159608e-06, "loss": 0.0706, "step": 7851 }, { "epoch": 0.36486988847583646, "grad_norm": 1.2534608130703238, "learning_rate": 7.344655359781799e-06, "loss": 0.0521, "step": 7852 }, { "epoch": 0.36491635687732343, "grad_norm": 1.4491586791875728, "learning_rate": 7.343990687295078e-06, "loss": 0.1012, "step": 7853 }, { "epoch": 0.3649628252788104, "grad_norm": 0.9471228187238673, "learning_rate": 7.343325961714501e-06, "loss": 0.0882, "step": 7854 }, { "epoch": 0.3650092936802974, "grad_norm": 10.79903558046613, "learning_rate": 7.342661183055125e-06, "loss": 0.0877, "step": 7855 }, { "epoch": 0.36505576208178436, "grad_norm": 0.8858940119824015, "learning_rate": 7.341996351332007e-06, "loss": 0.05, "step": 7856 }, { "epoch": 0.3651022304832714, "grad_norm": 0.8104811566641689, "learning_rate": 7.341331466560206e-06, "loss": 0.0813, "step": 7857 }, { "epoch": 0.36514869888475837, "grad_norm": 1.4318131310488413, "learning_rate": 7.3406665287547856e-06, "loss": 0.0714, "step": 7858 }, { "epoch": 0.36519516728624535, "grad_norm": 0.9174786242032176, "learning_rate": 7.340001537930803e-06, "loss": 0.0764, "step": 7859 }, { "epoch": 0.3652416356877323, "grad_norm": 0.6700302919487816, "learning_rate": 7.3393364941033226e-06, "loss": 0.0449, "step": 7860 }, { "epoch": 0.36528810408921936, "grad_norm": 0.9361511591350421, "learning_rate": 7.338671397287409e-06, "loss": 0.0798, "step": 7861 }, { "epoch": 0.36533457249070633, "grad_norm": 1.8074018952102686, "learning_rate": 7.3380062474981276e-06, "loss": 0.0452, "step": 7862 }, { "epoch": 0.3653810408921933, "grad_norm": 2.111973848067847, "learning_rate": 7.337341044750543e-06, "loss": 0.0888, "step": 7863 }, { "epoch": 0.3654275092936803, "grad_norm": 0.5518947753078294, "learning_rate": 7.336675789059725e-06, "loss": 0.047, "step": 7864 }, { "epoch": 0.36547397769516726, "grad_norm": 0.8110472504661872, "learning_rate": 7.33601048044074e-06, "loss": 0.04, "step": 7865 }, { "epoch": 0.3655204460966543, "grad_norm": 2.3733721272315296, "learning_rate": 7.335345118908659e-06, "loss": 0.1343, "step": 7866 }, { "epoch": 0.36556691449814127, "grad_norm": 1.0036777412998252, "learning_rate": 7.334679704478554e-06, "loss": 0.0763, "step": 7867 }, { "epoch": 0.36561338289962825, "grad_norm": 0.937462135624652, "learning_rate": 7.3340142371654944e-06, "loss": 0.0699, "step": 7868 }, { "epoch": 0.3656598513011152, "grad_norm": 0.750137599755091, "learning_rate": 7.333348716984556e-06, "loss": 0.0651, "step": 7869 }, { "epoch": 0.36570631970260226, "grad_norm": 1.533302417933758, "learning_rate": 7.332683143950816e-06, "loss": 0.0795, "step": 7870 }, { "epoch": 0.36575278810408923, "grad_norm": 1.5232683496069555, "learning_rate": 7.332017518079346e-06, "loss": 0.0726, "step": 7871 }, { "epoch": 0.3657992565055762, "grad_norm": 0.6634578212980293, "learning_rate": 7.331351839385225e-06, "loss": 0.0693, "step": 7872 }, { "epoch": 0.3658457249070632, "grad_norm": 1.4319005880905828, "learning_rate": 7.330686107883531e-06, "loss": 0.0537, "step": 7873 }, { "epoch": 0.36589219330855016, "grad_norm": 1.1462848346876036, "learning_rate": 7.330020323589344e-06, "loss": 0.0582, "step": 7874 }, { "epoch": 0.3659386617100372, "grad_norm": 4.478002840640814, "learning_rate": 7.329354486517744e-06, "loss": 0.1028, "step": 7875 }, { "epoch": 0.36598513011152417, "grad_norm": 1.9073168481684724, "learning_rate": 7.328688596683814e-06, "loss": 0.0725, "step": 7876 }, { "epoch": 0.36603159851301115, "grad_norm": 2.1474767833355903, "learning_rate": 7.328022654102634e-06, "loss": 0.09, "step": 7877 }, { "epoch": 0.3660780669144981, "grad_norm": 2.542041184452495, "learning_rate": 7.327356658789293e-06, "loss": 0.097, "step": 7878 }, { "epoch": 0.36612453531598516, "grad_norm": 0.814471159799148, "learning_rate": 7.326690610758874e-06, "loss": 0.0625, "step": 7879 }, { "epoch": 0.36617100371747213, "grad_norm": 0.9929454970879984, "learning_rate": 7.326024510026464e-06, "loss": 0.088, "step": 7880 }, { "epoch": 0.3662174721189591, "grad_norm": 0.7932636420983729, "learning_rate": 7.325358356607149e-06, "loss": 0.0727, "step": 7881 }, { "epoch": 0.3662639405204461, "grad_norm": 1.5324094789906784, "learning_rate": 7.324692150516022e-06, "loss": 0.0868, "step": 7882 }, { "epoch": 0.36631040892193306, "grad_norm": 1.6356235296035424, "learning_rate": 7.3240258917681705e-06, "loss": 0.0914, "step": 7883 }, { "epoch": 0.3663568773234201, "grad_norm": 1.5071033811120695, "learning_rate": 7.323359580378686e-06, "loss": 0.1, "step": 7884 }, { "epoch": 0.36640334572490707, "grad_norm": 0.7621114407768549, "learning_rate": 7.3226932163626616e-06, "loss": 0.087, "step": 7885 }, { "epoch": 0.36644981412639405, "grad_norm": 2.2213197460331426, "learning_rate": 7.322026799735192e-06, "loss": 0.0783, "step": 7886 }, { "epoch": 0.366496282527881, "grad_norm": 0.7799726479206581, "learning_rate": 7.321360330511371e-06, "loss": 0.0514, "step": 7887 }, { "epoch": 0.36654275092936806, "grad_norm": 0.5593658122412942, "learning_rate": 7.320693808706296e-06, "loss": 0.0613, "step": 7888 }, { "epoch": 0.36658921933085503, "grad_norm": 1.3248333045527994, "learning_rate": 7.320027234335063e-06, "loss": 0.0659, "step": 7889 }, { "epoch": 0.366635687732342, "grad_norm": 3.453023019346029, "learning_rate": 7.319360607412771e-06, "loss": 0.0655, "step": 7890 }, { "epoch": 0.366682156133829, "grad_norm": 3.7622196764126046, "learning_rate": 7.318693927954521e-06, "loss": 0.0908, "step": 7891 }, { "epoch": 0.36672862453531596, "grad_norm": 0.9406569507235176, "learning_rate": 7.318027195975412e-06, "loss": 0.0577, "step": 7892 }, { "epoch": 0.366775092936803, "grad_norm": 1.0048567962048787, "learning_rate": 7.317360411490551e-06, "loss": 0.0716, "step": 7893 }, { "epoch": 0.36682156133828997, "grad_norm": 1.3746815231917768, "learning_rate": 7.316693574515034e-06, "loss": 0.0853, "step": 7894 }, { "epoch": 0.36686802973977695, "grad_norm": 2.4202006857812224, "learning_rate": 7.31602668506397e-06, "loss": 0.0746, "step": 7895 }, { "epoch": 0.3669144981412639, "grad_norm": 1.289393978704529, "learning_rate": 7.315359743152464e-06, "loss": 0.0947, "step": 7896 }, { "epoch": 0.36696096654275095, "grad_norm": 1.1267988470815318, "learning_rate": 7.314692748795622e-06, "loss": 0.0567, "step": 7897 }, { "epoch": 0.36700743494423793, "grad_norm": 3.1579987491156545, "learning_rate": 7.314025702008555e-06, "loss": 0.0672, "step": 7898 }, { "epoch": 0.3670539033457249, "grad_norm": 1.0838021287948436, "learning_rate": 7.31335860280637e-06, "loss": 0.058, "step": 7899 }, { "epoch": 0.3671003717472119, "grad_norm": 0.6623259670281482, "learning_rate": 7.312691451204178e-06, "loss": 0.0351, "step": 7900 }, { "epoch": 0.36714684014869886, "grad_norm": 1.9776432062638922, "learning_rate": 7.312024247217089e-06, "loss": 0.1259, "step": 7901 }, { "epoch": 0.3671933085501859, "grad_norm": 1.3629356844910716, "learning_rate": 7.311356990860219e-06, "loss": 0.0729, "step": 7902 }, { "epoch": 0.36723977695167287, "grad_norm": 2.4834286576747195, "learning_rate": 7.3106896821486795e-06, "loss": 0.0653, "step": 7903 }, { "epoch": 0.36728624535315985, "grad_norm": 0.8808493953026381, "learning_rate": 7.310022321097587e-06, "loss": 0.0635, "step": 7904 }, { "epoch": 0.3673327137546468, "grad_norm": 0.7833151232285198, "learning_rate": 7.309354907722059e-06, "loss": 0.0688, "step": 7905 }, { "epoch": 0.36737918215613385, "grad_norm": 1.846252962853001, "learning_rate": 7.30868744203721e-06, "loss": 0.087, "step": 7906 }, { "epoch": 0.36742565055762083, "grad_norm": 0.7724708648755673, "learning_rate": 7.3080199240581615e-06, "loss": 0.0427, "step": 7907 }, { "epoch": 0.3674721189591078, "grad_norm": 0.6041765092416245, "learning_rate": 7.307352353800033e-06, "loss": 0.0419, "step": 7908 }, { "epoch": 0.3675185873605948, "grad_norm": 1.2049493826584816, "learning_rate": 7.306684731277943e-06, "loss": 0.0638, "step": 7909 }, { "epoch": 0.36756505576208176, "grad_norm": 0.9079814732705238, "learning_rate": 7.306017056507018e-06, "loss": 0.0546, "step": 7910 }, { "epoch": 0.3676115241635688, "grad_norm": 0.5195913759703328, "learning_rate": 7.30534932950238e-06, "loss": 0.0384, "step": 7911 }, { "epoch": 0.36765799256505577, "grad_norm": 10.015314910935922, "learning_rate": 7.304681550279152e-06, "loss": 0.0736, "step": 7912 }, { "epoch": 0.36770446096654275, "grad_norm": 1.408982903826484, "learning_rate": 7.3040137188524614e-06, "loss": 0.0635, "step": 7913 }, { "epoch": 0.3677509293680297, "grad_norm": 0.42542354515739345, "learning_rate": 7.303345835237436e-06, "loss": 0.0314, "step": 7914 }, { "epoch": 0.36779739776951675, "grad_norm": 0.5744073045389011, "learning_rate": 7.3026778994492036e-06, "loss": 0.0746, "step": 7915 }, { "epoch": 0.36784386617100373, "grad_norm": 0.5717371356162225, "learning_rate": 7.302009911502892e-06, "loss": 0.0665, "step": 7916 }, { "epoch": 0.3678903345724907, "grad_norm": 0.8785946966520971, "learning_rate": 7.301341871413634e-06, "loss": 0.0581, "step": 7917 }, { "epoch": 0.3679368029739777, "grad_norm": 2.640089106058859, "learning_rate": 7.30067377919656e-06, "loss": 0.0762, "step": 7918 }, { "epoch": 0.36798327137546466, "grad_norm": 0.6095989515671976, "learning_rate": 7.300005634866804e-06, "loss": 0.0477, "step": 7919 }, { "epoch": 0.3680297397769517, "grad_norm": 0.7202962370459521, "learning_rate": 7.2993374384394985e-06, "loss": 0.0711, "step": 7920 }, { "epoch": 0.36807620817843867, "grad_norm": 0.8549177614788557, "learning_rate": 7.2986691899297805e-06, "loss": 0.0545, "step": 7921 }, { "epoch": 0.36812267657992565, "grad_norm": 1.7442239174401784, "learning_rate": 7.298000889352786e-06, "loss": 0.0501, "step": 7922 }, { "epoch": 0.3681691449814126, "grad_norm": 1.0441552477893292, "learning_rate": 7.297332536723654e-06, "loss": 0.0631, "step": 7923 }, { "epoch": 0.36821561338289965, "grad_norm": 1.437493013780939, "learning_rate": 7.296664132057522e-06, "loss": 0.062, "step": 7924 }, { "epoch": 0.36826208178438663, "grad_norm": 0.9658828546834743, "learning_rate": 7.295995675369528e-06, "loss": 0.0583, "step": 7925 }, { "epoch": 0.3683085501858736, "grad_norm": 0.5631237110951113, "learning_rate": 7.2953271666748175e-06, "loss": 0.0526, "step": 7926 }, { "epoch": 0.3683550185873606, "grad_norm": 2.93056077035809, "learning_rate": 7.294658605988529e-06, "loss": 0.0928, "step": 7927 }, { "epoch": 0.36840148698884756, "grad_norm": 1.1167468180342752, "learning_rate": 7.29398999332581e-06, "loss": 0.07, "step": 7928 }, { "epoch": 0.3684479553903346, "grad_norm": 0.7654778144196258, "learning_rate": 7.293321328701804e-06, "loss": 0.0558, "step": 7929 }, { "epoch": 0.36849442379182157, "grad_norm": 1.1641760615553876, "learning_rate": 7.292652612131654e-06, "loss": 0.0565, "step": 7930 }, { "epoch": 0.36854089219330854, "grad_norm": 2.04924756059579, "learning_rate": 7.29198384363051e-06, "loss": 0.0647, "step": 7931 }, { "epoch": 0.3685873605947955, "grad_norm": 1.1489707133279765, "learning_rate": 7.29131502321352e-06, "loss": 0.0812, "step": 7932 }, { "epoch": 0.36863382899628255, "grad_norm": 1.5320162903531926, "learning_rate": 7.290646150895831e-06, "loss": 0.0591, "step": 7933 }, { "epoch": 0.36868029739776953, "grad_norm": 0.6125860517771927, "learning_rate": 7.289977226692598e-06, "loss": 0.0366, "step": 7934 }, { "epoch": 0.3687267657992565, "grad_norm": 1.0333019432706032, "learning_rate": 7.289308250618972e-06, "loss": 0.0722, "step": 7935 }, { "epoch": 0.3687732342007435, "grad_norm": 1.0759035852551548, "learning_rate": 7.2886392226901025e-06, "loss": 0.0585, "step": 7936 }, { "epoch": 0.36881970260223046, "grad_norm": 1.851278265047135, "learning_rate": 7.287970142921146e-06, "loss": 0.0885, "step": 7937 }, { "epoch": 0.3688661710037175, "grad_norm": 1.9112623435964728, "learning_rate": 7.287301011327258e-06, "loss": 0.0699, "step": 7938 }, { "epoch": 0.36891263940520447, "grad_norm": 3.361695744142473, "learning_rate": 7.286631827923596e-06, "loss": 0.0615, "step": 7939 }, { "epoch": 0.36895910780669144, "grad_norm": 1.4912732169526264, "learning_rate": 7.285962592725314e-06, "loss": 0.0588, "step": 7940 }, { "epoch": 0.3690055762081784, "grad_norm": 0.8006268934803672, "learning_rate": 7.285293305747576e-06, "loss": 0.0526, "step": 7941 }, { "epoch": 0.36905204460966545, "grad_norm": 0.6975989874186396, "learning_rate": 7.284623967005538e-06, "loss": 0.0395, "step": 7942 }, { "epoch": 0.36909851301115243, "grad_norm": 1.0775960442837718, "learning_rate": 7.283954576514363e-06, "loss": 0.0869, "step": 7943 }, { "epoch": 0.3691449814126394, "grad_norm": 1.3462471435425112, "learning_rate": 7.283285134289212e-06, "loss": 0.0902, "step": 7944 }, { "epoch": 0.3691914498141264, "grad_norm": 3.203873209643313, "learning_rate": 7.282615640345249e-06, "loss": 0.0622, "step": 7945 }, { "epoch": 0.36923791821561336, "grad_norm": 2.22647874745033, "learning_rate": 7.281946094697643e-06, "loss": 0.0977, "step": 7946 }, { "epoch": 0.3692843866171004, "grad_norm": 1.054124105932914, "learning_rate": 7.281276497361553e-06, "loss": 0.0561, "step": 7947 }, { "epoch": 0.36933085501858737, "grad_norm": 0.5634545077572355, "learning_rate": 7.280606848352151e-06, "loss": 0.0518, "step": 7948 }, { "epoch": 0.36937732342007434, "grad_norm": 1.1121310637125912, "learning_rate": 7.279937147684602e-06, "loss": 0.0796, "step": 7949 }, { "epoch": 0.3694237918215613, "grad_norm": 1.7530442299102225, "learning_rate": 7.279267395374078e-06, "loss": 0.0868, "step": 7950 }, { "epoch": 0.36947026022304835, "grad_norm": 2.1171391996706324, "learning_rate": 7.278597591435749e-06, "loss": 0.0706, "step": 7951 }, { "epoch": 0.36951672862453533, "grad_norm": 2.5568986897796067, "learning_rate": 7.277927735884786e-06, "loss": 0.0894, "step": 7952 }, { "epoch": 0.3695631970260223, "grad_norm": 0.7593246773941073, "learning_rate": 7.277257828736362e-06, "loss": 0.0497, "step": 7953 }, { "epoch": 0.3696096654275093, "grad_norm": 1.062308499106384, "learning_rate": 7.276587870005652e-06, "loss": 0.0668, "step": 7954 }, { "epoch": 0.36965613382899626, "grad_norm": 0.8490727110776968, "learning_rate": 7.27591785970783e-06, "loss": 0.066, "step": 7955 }, { "epoch": 0.3697026022304833, "grad_norm": 4.081306672460941, "learning_rate": 7.275247797858073e-06, "loss": 0.0957, "step": 7956 }, { "epoch": 0.36974907063197027, "grad_norm": 1.7836055278455343, "learning_rate": 7.274577684471557e-06, "loss": 0.0963, "step": 7957 }, { "epoch": 0.36979553903345724, "grad_norm": 1.5388587471569264, "learning_rate": 7.2739075195634664e-06, "loss": 0.0894, "step": 7958 }, { "epoch": 0.3698420074349442, "grad_norm": 1.1759785633991173, "learning_rate": 7.273237303148976e-06, "loss": 0.0779, "step": 7959 }, { "epoch": 0.36988847583643125, "grad_norm": 0.8279396754375415, "learning_rate": 7.2725670352432674e-06, "loss": 0.0782, "step": 7960 }, { "epoch": 0.36993494423791823, "grad_norm": 1.0862972695754753, "learning_rate": 7.271896715861524e-06, "loss": 0.0734, "step": 7961 }, { "epoch": 0.3699814126394052, "grad_norm": 1.3547198851099227, "learning_rate": 7.271226345018928e-06, "loss": 0.0884, "step": 7962 }, { "epoch": 0.3700278810408922, "grad_norm": 1.1322781502008565, "learning_rate": 7.270555922730665e-06, "loss": 0.0567, "step": 7963 }, { "epoch": 0.37007434944237916, "grad_norm": 1.0009010889261836, "learning_rate": 7.269885449011922e-06, "loss": 0.0462, "step": 7964 }, { "epoch": 0.3701208178438662, "grad_norm": 0.9709344074756348, "learning_rate": 7.269214923877883e-06, "loss": 0.09, "step": 7965 }, { "epoch": 0.37016728624535317, "grad_norm": 0.8535119125724251, "learning_rate": 7.268544347343738e-06, "loss": 0.0741, "step": 7966 }, { "epoch": 0.37021375464684014, "grad_norm": 0.9679932347266675, "learning_rate": 7.2678737194246765e-06, "loss": 0.073, "step": 7967 }, { "epoch": 0.3702602230483271, "grad_norm": 2.0349471174690015, "learning_rate": 7.267203040135888e-06, "loss": 0.0628, "step": 7968 }, { "epoch": 0.37030669144981415, "grad_norm": 1.122377108884548, "learning_rate": 7.266532309492564e-06, "loss": 0.0639, "step": 7969 }, { "epoch": 0.37035315985130113, "grad_norm": 0.8389587360758759, "learning_rate": 7.265861527509901e-06, "loss": 0.077, "step": 7970 }, { "epoch": 0.3703996282527881, "grad_norm": 1.8932567359076087, "learning_rate": 7.265190694203086e-06, "loss": 0.0597, "step": 7971 }, { "epoch": 0.3704460966542751, "grad_norm": 1.0025828974785522, "learning_rate": 7.264519809587318e-06, "loss": 0.0763, "step": 7972 }, { "epoch": 0.37049256505576206, "grad_norm": 0.7078244957141994, "learning_rate": 7.2638488736777945e-06, "loss": 0.0391, "step": 7973 }, { "epoch": 0.3705390334572491, "grad_norm": 0.9150157954900294, "learning_rate": 7.263177886489709e-06, "loss": 0.0588, "step": 7974 }, { "epoch": 0.37058550185873607, "grad_norm": 0.5827414783958019, "learning_rate": 7.2625068480382645e-06, "loss": 0.0458, "step": 7975 }, { "epoch": 0.37063197026022304, "grad_norm": 0.9039346183020254, "learning_rate": 7.261835758338658e-06, "loss": 0.0614, "step": 7976 }, { "epoch": 0.37067843866171, "grad_norm": 0.5836211346771948, "learning_rate": 7.261164617406091e-06, "loss": 0.0397, "step": 7977 }, { "epoch": 0.37072490706319705, "grad_norm": 0.7010673957159294, "learning_rate": 7.260493425255765e-06, "loss": 0.0446, "step": 7978 }, { "epoch": 0.370771375464684, "grad_norm": 0.7777282021395245, "learning_rate": 7.259822181902885e-06, "loss": 0.0586, "step": 7979 }, { "epoch": 0.370817843866171, "grad_norm": 1.0037941652279307, "learning_rate": 7.259150887362654e-06, "loss": 0.0592, "step": 7980 }, { "epoch": 0.370864312267658, "grad_norm": 1.8998178377480908, "learning_rate": 7.258479541650276e-06, "loss": 0.0697, "step": 7981 }, { "epoch": 0.37091078066914496, "grad_norm": 1.1997025186368309, "learning_rate": 7.257808144780962e-06, "loss": 0.0847, "step": 7982 }, { "epoch": 0.370957249070632, "grad_norm": 0.36307538222566965, "learning_rate": 7.257136696769915e-06, "loss": 0.0219, "step": 7983 }, { "epoch": 0.37100371747211897, "grad_norm": 1.4577730000022793, "learning_rate": 7.256465197632347e-06, "loss": 0.0772, "step": 7984 }, { "epoch": 0.37105018587360594, "grad_norm": 2.6458392089506857, "learning_rate": 7.255793647383469e-06, "loss": 0.0708, "step": 7985 }, { "epoch": 0.3710966542750929, "grad_norm": 2.3758239438495234, "learning_rate": 7.255122046038488e-06, "loss": 0.084, "step": 7986 }, { "epoch": 0.37114312267657995, "grad_norm": 0.6853864992893877, "learning_rate": 7.254450393612621e-06, "loss": 0.0603, "step": 7987 }, { "epoch": 0.3711895910780669, "grad_norm": 0.7270059657722647, "learning_rate": 7.2537786901210794e-06, "loss": 0.0657, "step": 7988 }, { "epoch": 0.3712360594795539, "grad_norm": 0.9654736860988223, "learning_rate": 7.253106935579079e-06, "loss": 0.0537, "step": 7989 }, { "epoch": 0.3712825278810409, "grad_norm": 0.8041422921308881, "learning_rate": 7.252435130001833e-06, "loss": 0.0613, "step": 7990 }, { "epoch": 0.37132899628252786, "grad_norm": 0.5632589801288697, "learning_rate": 7.251763273404561e-06, "loss": 0.0369, "step": 7991 }, { "epoch": 0.3713754646840149, "grad_norm": 0.8446542935195369, "learning_rate": 7.251091365802483e-06, "loss": 0.0507, "step": 7992 }, { "epoch": 0.37142193308550187, "grad_norm": 0.9058957913107858, "learning_rate": 7.2504194072108145e-06, "loss": 0.0824, "step": 7993 }, { "epoch": 0.37146840148698884, "grad_norm": 0.6455006018524209, "learning_rate": 7.249747397644781e-06, "loss": 0.0449, "step": 7994 }, { "epoch": 0.3715148698884758, "grad_norm": 1.6027033499942942, "learning_rate": 7.249075337119598e-06, "loss": 0.0841, "step": 7995 }, { "epoch": 0.37156133828996285, "grad_norm": 1.1130352286615346, "learning_rate": 7.248403225650491e-06, "loss": 0.0693, "step": 7996 }, { "epoch": 0.3716078066914498, "grad_norm": 0.968697698862703, "learning_rate": 7.2477310632526875e-06, "loss": 0.1062, "step": 7997 }, { "epoch": 0.3716542750929368, "grad_norm": 2.056386924187772, "learning_rate": 7.247058849941407e-06, "loss": 0.0501, "step": 7998 }, { "epoch": 0.3717007434944238, "grad_norm": 1.0297461077296073, "learning_rate": 7.246386585731879e-06, "loss": 0.039, "step": 7999 }, { "epoch": 0.37174721189591076, "grad_norm": 1.0746939671084474, "learning_rate": 7.245714270639332e-06, "loss": 0.0882, "step": 8000 }, { "epoch": 0.3717936802973978, "grad_norm": 0.8394212683551953, "learning_rate": 7.245041904678993e-06, "loss": 0.0416, "step": 8001 }, { "epoch": 0.37184014869888476, "grad_norm": 0.6712466987222697, "learning_rate": 7.244369487866092e-06, "loss": 0.0566, "step": 8002 }, { "epoch": 0.37188661710037174, "grad_norm": 0.8639192578484965, "learning_rate": 7.243697020215858e-06, "loss": 0.0626, "step": 8003 }, { "epoch": 0.3719330855018587, "grad_norm": 0.49867143971366823, "learning_rate": 7.243024501743528e-06, "loss": 0.0525, "step": 8004 }, { "epoch": 0.37197955390334575, "grad_norm": 0.8095277421569691, "learning_rate": 7.24235193246433e-06, "loss": 0.0652, "step": 8005 }, { "epoch": 0.3720260223048327, "grad_norm": 1.0605045931248762, "learning_rate": 7.241679312393502e-06, "loss": 0.0664, "step": 8006 }, { "epoch": 0.3720724907063197, "grad_norm": 0.5824838900567676, "learning_rate": 7.241006641546277e-06, "loss": 0.0364, "step": 8007 }, { "epoch": 0.3721189591078067, "grad_norm": 0.5360999749446004, "learning_rate": 7.240333919937893e-06, "loss": 0.0458, "step": 8008 }, { "epoch": 0.37216542750929366, "grad_norm": 0.8077493658020979, "learning_rate": 7.239661147583589e-06, "loss": 0.0678, "step": 8009 }, { "epoch": 0.3722118959107807, "grad_norm": 1.0946497107189672, "learning_rate": 7.238988324498602e-06, "loss": 0.0362, "step": 8010 }, { "epoch": 0.37225836431226766, "grad_norm": 1.3107906561173508, "learning_rate": 7.238315450698173e-06, "loss": 0.0551, "step": 8011 }, { "epoch": 0.37230483271375464, "grad_norm": 0.784963659611171, "learning_rate": 7.237642526197544e-06, "loss": 0.0624, "step": 8012 }, { "epoch": 0.3723513011152416, "grad_norm": 1.5692259970599434, "learning_rate": 7.236969551011954e-06, "loss": 0.0508, "step": 8013 }, { "epoch": 0.37239776951672865, "grad_norm": 0.73888565990824, "learning_rate": 7.236296525156653e-06, "loss": 0.0461, "step": 8014 }, { "epoch": 0.3724442379182156, "grad_norm": 1.1816688283285859, "learning_rate": 7.23562344864688e-06, "loss": 0.0668, "step": 8015 }, { "epoch": 0.3724907063197026, "grad_norm": 1.100992804557909, "learning_rate": 7.234950321497883e-06, "loss": 0.0928, "step": 8016 }, { "epoch": 0.3725371747211896, "grad_norm": 0.9539914783352231, "learning_rate": 7.23427714372491e-06, "loss": 0.0611, "step": 8017 }, { "epoch": 0.37258364312267656, "grad_norm": 2.239540328969514, "learning_rate": 7.233603915343207e-06, "loss": 0.0549, "step": 8018 }, { "epoch": 0.3726301115241636, "grad_norm": 0.8536352528354424, "learning_rate": 7.232930636368025e-06, "loss": 0.0506, "step": 8019 }, { "epoch": 0.37267657992565056, "grad_norm": 1.7955309595871929, "learning_rate": 7.232257306814614e-06, "loss": 0.0934, "step": 8020 }, { "epoch": 0.37272304832713754, "grad_norm": 0.9324075630343786, "learning_rate": 7.231583926698227e-06, "loss": 0.0378, "step": 8021 }, { "epoch": 0.3727695167286245, "grad_norm": 0.9506551087190557, "learning_rate": 7.230910496034112e-06, "loss": 0.0613, "step": 8022 }, { "epoch": 0.37281598513011155, "grad_norm": 1.907449173500792, "learning_rate": 7.23023701483753e-06, "loss": 0.0857, "step": 8023 }, { "epoch": 0.3728624535315985, "grad_norm": 0.6835949897453126, "learning_rate": 7.2295634831237295e-06, "loss": 0.0654, "step": 8024 }, { "epoch": 0.3729089219330855, "grad_norm": 0.6804886334079129, "learning_rate": 7.228889900907971e-06, "loss": 0.0528, "step": 8025 }, { "epoch": 0.3729553903345725, "grad_norm": 1.1010584536593297, "learning_rate": 7.22821626820551e-06, "loss": 0.0651, "step": 8026 }, { "epoch": 0.37300185873605946, "grad_norm": 0.9848517282368191, "learning_rate": 7.227542585031605e-06, "loss": 0.0684, "step": 8027 }, { "epoch": 0.3730483271375465, "grad_norm": 0.6677657861743428, "learning_rate": 7.2268688514015165e-06, "loss": 0.0586, "step": 8028 }, { "epoch": 0.37309479553903346, "grad_norm": 1.0982334655549717, "learning_rate": 7.226195067330505e-06, "loss": 0.0633, "step": 8029 }, { "epoch": 0.37314126394052044, "grad_norm": 0.6450391744351138, "learning_rate": 7.225521232833831e-06, "loss": 0.0617, "step": 8030 }, { "epoch": 0.3731877323420074, "grad_norm": 0.5609355198291611, "learning_rate": 7.22484734792676e-06, "loss": 0.0701, "step": 8031 }, { "epoch": 0.37323420074349445, "grad_norm": 0.6832388613926936, "learning_rate": 7.2241734126245546e-06, "loss": 0.0496, "step": 8032 }, { "epoch": 0.3732806691449814, "grad_norm": 2.1110479299050877, "learning_rate": 7.223499426942481e-06, "loss": 0.075, "step": 8033 }, { "epoch": 0.3733271375464684, "grad_norm": 1.0725636203797957, "learning_rate": 7.222825390895803e-06, "loss": 0.0962, "step": 8034 }, { "epoch": 0.3733736059479554, "grad_norm": 2.1659315228979015, "learning_rate": 7.222151304499792e-06, "loss": 0.0795, "step": 8035 }, { "epoch": 0.37342007434944235, "grad_norm": 1.4604849249189977, "learning_rate": 7.221477167769716e-06, "loss": 0.0727, "step": 8036 }, { "epoch": 0.3734665427509294, "grad_norm": 1.0900185870159789, "learning_rate": 7.220802980720843e-06, "loss": 0.0454, "step": 8037 }, { "epoch": 0.37351301115241636, "grad_norm": 0.6078784563841937, "learning_rate": 7.220128743368446e-06, "loss": 0.0569, "step": 8038 }, { "epoch": 0.37355947955390334, "grad_norm": 0.5515208161664346, "learning_rate": 7.219454455727796e-06, "loss": 0.0549, "step": 8039 }, { "epoch": 0.3736059479553903, "grad_norm": 0.7456387756248759, "learning_rate": 7.2187801178141655e-06, "loss": 0.053, "step": 8040 }, { "epoch": 0.37365241635687735, "grad_norm": 0.6336135760781667, "learning_rate": 7.218105729642832e-06, "loss": 0.0615, "step": 8041 }, { "epoch": 0.3736988847583643, "grad_norm": 2.7095307733997394, "learning_rate": 7.217431291229068e-06, "loss": 0.1003, "step": 8042 }, { "epoch": 0.3737453531598513, "grad_norm": 0.5239052129233902, "learning_rate": 7.216756802588152e-06, "loss": 0.0457, "step": 8043 }, { "epoch": 0.3737918215613383, "grad_norm": 2.024054299398697, "learning_rate": 7.21608226373536e-06, "loss": 0.0905, "step": 8044 }, { "epoch": 0.37383828996282525, "grad_norm": 0.48805672418479323, "learning_rate": 7.215407674685975e-06, "loss": 0.0315, "step": 8045 }, { "epoch": 0.3738847583643123, "grad_norm": 0.6256677400987988, "learning_rate": 7.214733035455271e-06, "loss": 0.0395, "step": 8046 }, { "epoch": 0.37393122676579926, "grad_norm": 0.7750181556803951, "learning_rate": 7.2140583460585355e-06, "loss": 0.065, "step": 8047 }, { "epoch": 0.37397769516728624, "grad_norm": 0.6197771807459111, "learning_rate": 7.2133836065110475e-06, "loss": 0.0686, "step": 8048 }, { "epoch": 0.3740241635687732, "grad_norm": 1.0967729224461038, "learning_rate": 7.21270881682809e-06, "loss": 0.0675, "step": 8049 }, { "epoch": 0.37407063197026025, "grad_norm": 1.0454220478848157, "learning_rate": 7.2120339770249515e-06, "loss": 0.0549, "step": 8050 }, { "epoch": 0.3741171003717472, "grad_norm": 1.2125139361781148, "learning_rate": 7.211359087116914e-06, "loss": 0.0686, "step": 8051 }, { "epoch": 0.3741635687732342, "grad_norm": 2.2762815555186826, "learning_rate": 7.210684147119264e-06, "loss": 0.09, "step": 8052 }, { "epoch": 0.3742100371747212, "grad_norm": 0.6309884261406679, "learning_rate": 7.210009157047294e-06, "loss": 0.0524, "step": 8053 }, { "epoch": 0.37425650557620815, "grad_norm": 1.6195898516217624, "learning_rate": 7.209334116916289e-06, "loss": 0.093, "step": 8054 }, { "epoch": 0.3743029739776952, "grad_norm": 0.5523010956663197, "learning_rate": 7.208659026741542e-06, "loss": 0.0474, "step": 8055 }, { "epoch": 0.37434944237918216, "grad_norm": 0.6126189043135337, "learning_rate": 7.207983886538343e-06, "loss": 0.0561, "step": 8056 }, { "epoch": 0.37439591078066914, "grad_norm": 1.0297508984120107, "learning_rate": 7.207308696321985e-06, "loss": 0.0465, "step": 8057 }, { "epoch": 0.3744423791821561, "grad_norm": 0.9030266843404928, "learning_rate": 7.206633456107761e-06, "loss": 0.0599, "step": 8058 }, { "epoch": 0.37448884758364315, "grad_norm": 1.329065162314329, "learning_rate": 7.2059581659109686e-06, "loss": 0.0924, "step": 8059 }, { "epoch": 0.3745353159851301, "grad_norm": 1.0859059360169065, "learning_rate": 7.205282825746902e-06, "loss": 0.0776, "step": 8060 }, { "epoch": 0.3745817843866171, "grad_norm": 2.00717337250031, "learning_rate": 7.204607435630857e-06, "loss": 0.0627, "step": 8061 }, { "epoch": 0.3746282527881041, "grad_norm": 1.5827139502009477, "learning_rate": 7.2039319955781354e-06, "loss": 0.0665, "step": 8062 }, { "epoch": 0.37467472118959105, "grad_norm": 1.013682966112354, "learning_rate": 7.203256505604031e-06, "loss": 0.0451, "step": 8063 }, { "epoch": 0.3747211895910781, "grad_norm": 1.1577044739457287, "learning_rate": 7.202580965723851e-06, "loss": 0.0642, "step": 8064 }, { "epoch": 0.37476765799256506, "grad_norm": 1.9165076852803826, "learning_rate": 7.201905375952892e-06, "loss": 0.0583, "step": 8065 }, { "epoch": 0.37481412639405204, "grad_norm": 2.8078480898775493, "learning_rate": 7.201229736306459e-06, "loss": 0.0996, "step": 8066 }, { "epoch": 0.374860594795539, "grad_norm": 0.7578984072980237, "learning_rate": 7.200554046799858e-06, "loss": 0.0707, "step": 8067 }, { "epoch": 0.37490706319702605, "grad_norm": 0.8626266868027443, "learning_rate": 7.199878307448389e-06, "loss": 0.0431, "step": 8068 }, { "epoch": 0.374953531598513, "grad_norm": 1.1362299387978727, "learning_rate": 7.199202518267363e-06, "loss": 0.1004, "step": 8069 }, { "epoch": 0.375, "grad_norm": 1.3525345677818235, "learning_rate": 7.1985266792720844e-06, "loss": 0.078, "step": 8070 }, { "epoch": 0.375046468401487, "grad_norm": 1.0189171640108408, "learning_rate": 7.197850790477862e-06, "loss": 0.0676, "step": 8071 }, { "epoch": 0.37509293680297395, "grad_norm": 0.4473533782399249, "learning_rate": 7.197174851900006e-06, "loss": 0.0388, "step": 8072 }, { "epoch": 0.375139405204461, "grad_norm": 0.5778509016683919, "learning_rate": 7.196498863553829e-06, "loss": 0.0594, "step": 8073 }, { "epoch": 0.37518587360594796, "grad_norm": 1.0365238113321857, "learning_rate": 7.195822825454639e-06, "loss": 0.062, "step": 8074 }, { "epoch": 0.37523234200743494, "grad_norm": 0.5283758231732533, "learning_rate": 7.195146737617751e-06, "loss": 0.0312, "step": 8075 }, { "epoch": 0.3752788104089219, "grad_norm": 0.5161539515313689, "learning_rate": 7.19447060005848e-06, "loss": 0.0513, "step": 8076 }, { "epoch": 0.37532527881040895, "grad_norm": 1.0686199694577854, "learning_rate": 7.193794412792141e-06, "loss": 0.0626, "step": 8077 }, { "epoch": 0.3753717472118959, "grad_norm": 1.140354445910034, "learning_rate": 7.193118175834047e-06, "loss": 0.0741, "step": 8078 }, { "epoch": 0.3754182156133829, "grad_norm": 0.8007179675826174, "learning_rate": 7.192441889199522e-06, "loss": 0.0684, "step": 8079 }, { "epoch": 0.3754646840148699, "grad_norm": 0.39142244364167433, "learning_rate": 7.191765552903878e-06, "loss": 0.0289, "step": 8080 }, { "epoch": 0.37551115241635685, "grad_norm": 1.2347589283149665, "learning_rate": 7.191089166962437e-06, "loss": 0.0733, "step": 8081 }, { "epoch": 0.3755576208178439, "grad_norm": 0.5321347888882199, "learning_rate": 7.1904127313905215e-06, "loss": 0.0474, "step": 8082 }, { "epoch": 0.37560408921933086, "grad_norm": 0.6598043330023298, "learning_rate": 7.189736246203452e-06, "loss": 0.0484, "step": 8083 }, { "epoch": 0.37565055762081784, "grad_norm": 1.0952474049265195, "learning_rate": 7.189059711416551e-06, "loss": 0.0754, "step": 8084 }, { "epoch": 0.3756970260223048, "grad_norm": 0.8065428501218871, "learning_rate": 7.188383127045144e-06, "loss": 0.054, "step": 8085 }, { "epoch": 0.37574349442379185, "grad_norm": 0.5410128736695059, "learning_rate": 7.187706493104557e-06, "loss": 0.0364, "step": 8086 }, { "epoch": 0.3757899628252788, "grad_norm": 1.5846210259582902, "learning_rate": 7.1870298096101144e-06, "loss": 0.0609, "step": 8087 }, { "epoch": 0.3758364312267658, "grad_norm": 0.8255881379828972, "learning_rate": 7.186353076577145e-06, "loss": 0.055, "step": 8088 }, { "epoch": 0.3758828996282528, "grad_norm": 1.0178027256627835, "learning_rate": 7.185676294020978e-06, "loss": 0.064, "step": 8089 }, { "epoch": 0.37592936802973975, "grad_norm": 0.8247987622287588, "learning_rate": 7.184999461956941e-06, "loss": 0.0398, "step": 8090 }, { "epoch": 0.3759758364312268, "grad_norm": 1.530440667261645, "learning_rate": 7.184322580400369e-06, "loss": 0.073, "step": 8091 }, { "epoch": 0.37602230483271376, "grad_norm": 1.0737683307763228, "learning_rate": 7.18364564936659e-06, "loss": 0.0708, "step": 8092 }, { "epoch": 0.37606877323420074, "grad_norm": 0.5558411031464182, "learning_rate": 7.182968668870939e-06, "loss": 0.0442, "step": 8093 }, { "epoch": 0.3761152416356877, "grad_norm": 0.5911366029269731, "learning_rate": 7.18229163892875e-06, "loss": 0.0638, "step": 8094 }, { "epoch": 0.37616171003717475, "grad_norm": 0.8193377609136123, "learning_rate": 7.181614559555358e-06, "loss": 0.0282, "step": 8095 }, { "epoch": 0.3762081784386617, "grad_norm": 0.9529073161030039, "learning_rate": 7.1809374307661e-06, "loss": 0.0438, "step": 8096 }, { "epoch": 0.3762546468401487, "grad_norm": 0.5892766196532875, "learning_rate": 7.180260252576315e-06, "loss": 0.0638, "step": 8097 }, { "epoch": 0.3763011152416357, "grad_norm": 2.7714399909824006, "learning_rate": 7.1795830250013396e-06, "loss": 0.0686, "step": 8098 }, { "epoch": 0.37634758364312265, "grad_norm": 0.7481941454972476, "learning_rate": 7.178905748056515e-06, "loss": 0.0371, "step": 8099 }, { "epoch": 0.3763940520446097, "grad_norm": 0.9571960118006073, "learning_rate": 7.178228421757183e-06, "loss": 0.0709, "step": 8100 }, { "epoch": 0.37644052044609666, "grad_norm": 0.6270163087313808, "learning_rate": 7.177551046118684e-06, "loss": 0.0509, "step": 8101 }, { "epoch": 0.37648698884758364, "grad_norm": 2.2730201753409887, "learning_rate": 7.176873621156362e-06, "loss": 0.0761, "step": 8102 }, { "epoch": 0.3765334572490706, "grad_norm": 1.4261391022671552, "learning_rate": 7.176196146885561e-06, "loss": 0.0588, "step": 8103 }, { "epoch": 0.37657992565055765, "grad_norm": 0.9873719652341894, "learning_rate": 7.1755186233216265e-06, "loss": 0.0527, "step": 8104 }, { "epoch": 0.3766263940520446, "grad_norm": 1.057679574881415, "learning_rate": 7.174841050479906e-06, "loss": 0.1088, "step": 8105 }, { "epoch": 0.3766728624535316, "grad_norm": 0.9793943863757361, "learning_rate": 7.174163428375748e-06, "loss": 0.0714, "step": 8106 }, { "epoch": 0.3767193308550186, "grad_norm": 2.5653388676355062, "learning_rate": 7.173485757024498e-06, "loss": 0.0784, "step": 8107 }, { "epoch": 0.37676579925650555, "grad_norm": 1.0738011304897064, "learning_rate": 7.172808036441508e-06, "loss": 0.0803, "step": 8108 }, { "epoch": 0.3768122676579926, "grad_norm": 0.9942747043690916, "learning_rate": 7.172130266642131e-06, "loss": 0.0739, "step": 8109 }, { "epoch": 0.37685873605947956, "grad_norm": 1.2427307562060754, "learning_rate": 7.171452447641717e-06, "loss": 0.0577, "step": 8110 }, { "epoch": 0.37690520446096654, "grad_norm": 0.39951302384596726, "learning_rate": 7.170774579455618e-06, "loss": 0.0331, "step": 8111 }, { "epoch": 0.3769516728624535, "grad_norm": 0.6152365900289412, "learning_rate": 7.170096662099191e-06, "loss": 0.0458, "step": 8112 }, { "epoch": 0.37699814126394054, "grad_norm": 0.8696686543483371, "learning_rate": 7.169418695587791e-06, "loss": 0.0495, "step": 8113 }, { "epoch": 0.3770446096654275, "grad_norm": 1.0297603757808345, "learning_rate": 7.168740679936774e-06, "loss": 0.0642, "step": 8114 }, { "epoch": 0.3770910780669145, "grad_norm": 1.124634330678965, "learning_rate": 7.1680626151615e-06, "loss": 0.0629, "step": 8115 }, { "epoch": 0.3771375464684015, "grad_norm": 0.7018327343044538, "learning_rate": 7.167384501277323e-06, "loss": 0.0419, "step": 8116 }, { "epoch": 0.37718401486988845, "grad_norm": 0.6873847253680465, "learning_rate": 7.166706338299606e-06, "loss": 0.0697, "step": 8117 }, { "epoch": 0.3772304832713755, "grad_norm": 0.67932084358041, "learning_rate": 7.166028126243711e-06, "loss": 0.0528, "step": 8118 }, { "epoch": 0.37727695167286246, "grad_norm": 1.4938664746763908, "learning_rate": 7.165349865124999e-06, "loss": 0.1027, "step": 8119 }, { "epoch": 0.37732342007434944, "grad_norm": 0.6180736432417827, "learning_rate": 7.1646715549588355e-06, "loss": 0.0628, "step": 8120 }, { "epoch": 0.3773698884758364, "grad_norm": 1.050244728716604, "learning_rate": 7.163993195760579e-06, "loss": 0.071, "step": 8121 }, { "epoch": 0.37741635687732344, "grad_norm": 3.3380848324397143, "learning_rate": 7.163314787545602e-06, "loss": 0.1185, "step": 8122 }, { "epoch": 0.3774628252788104, "grad_norm": 0.9271232309767276, "learning_rate": 7.162636330329268e-06, "loss": 0.0902, "step": 8123 }, { "epoch": 0.3775092936802974, "grad_norm": 0.5449581332474698, "learning_rate": 7.161957824126944e-06, "loss": 0.0414, "step": 8124 }, { "epoch": 0.3775557620817844, "grad_norm": 1.2256760600075363, "learning_rate": 7.161279268954e-06, "loss": 0.0492, "step": 8125 }, { "epoch": 0.37760223048327135, "grad_norm": 0.6030608231290808, "learning_rate": 7.160600664825807e-06, "loss": 0.0683, "step": 8126 }, { "epoch": 0.3776486988847584, "grad_norm": 1.0852347089235883, "learning_rate": 7.159922011757734e-06, "loss": 0.1352, "step": 8127 }, { "epoch": 0.37769516728624536, "grad_norm": 0.9616668763851298, "learning_rate": 7.159243309765154e-06, "loss": 0.0824, "step": 8128 }, { "epoch": 0.37774163568773234, "grad_norm": 1.3636037881185776, "learning_rate": 7.158564558863441e-06, "loss": 0.1117, "step": 8129 }, { "epoch": 0.3777881040892193, "grad_norm": 0.9542452464312066, "learning_rate": 7.157885759067969e-06, "loss": 0.0434, "step": 8130 }, { "epoch": 0.37783457249070634, "grad_norm": 1.2544025586316798, "learning_rate": 7.157206910394112e-06, "loss": 0.0753, "step": 8131 }, { "epoch": 0.3778810408921933, "grad_norm": 0.7951047841036954, "learning_rate": 7.15652801285725e-06, "loss": 0.0561, "step": 8132 }, { "epoch": 0.3779275092936803, "grad_norm": 0.7553668667761763, "learning_rate": 7.155849066472758e-06, "loss": 0.0542, "step": 8133 }, { "epoch": 0.3779739776951673, "grad_norm": 1.285130241846257, "learning_rate": 7.155170071256015e-06, "loss": 0.0874, "step": 8134 }, { "epoch": 0.37802044609665425, "grad_norm": 0.9372919201842388, "learning_rate": 7.154491027222403e-06, "loss": 0.0638, "step": 8135 }, { "epoch": 0.3780669144981413, "grad_norm": 0.48014090103189727, "learning_rate": 7.1538119343873e-06, "loss": 0.0411, "step": 8136 }, { "epoch": 0.37811338289962826, "grad_norm": 0.8219645767715346, "learning_rate": 7.153132792766091e-06, "loss": 0.07, "step": 8137 }, { "epoch": 0.37815985130111524, "grad_norm": 0.43528032523283766, "learning_rate": 7.152453602374158e-06, "loss": 0.0392, "step": 8138 }, { "epoch": 0.3782063197026022, "grad_norm": 0.9342456610447232, "learning_rate": 7.1517743632268855e-06, "loss": 0.0872, "step": 8139 }, { "epoch": 0.37825278810408924, "grad_norm": 0.6684770785897439, "learning_rate": 7.151095075339658e-06, "loss": 0.0472, "step": 8140 }, { "epoch": 0.3782992565055762, "grad_norm": 0.589145372800912, "learning_rate": 7.150415738727864e-06, "loss": 0.0591, "step": 8141 }, { "epoch": 0.3783457249070632, "grad_norm": 0.699448227218869, "learning_rate": 7.1497363534068905e-06, "loss": 0.0454, "step": 8142 }, { "epoch": 0.3783921933085502, "grad_norm": 1.5039278936965437, "learning_rate": 7.1490569193921246e-06, "loss": 0.0867, "step": 8143 }, { "epoch": 0.37843866171003715, "grad_norm": 0.9255948257089055, "learning_rate": 7.14837743669896e-06, "loss": 0.0579, "step": 8144 }, { "epoch": 0.3784851301115242, "grad_norm": 0.9167799306923406, "learning_rate": 7.147697905342783e-06, "loss": 0.0481, "step": 8145 }, { "epoch": 0.37853159851301116, "grad_norm": 1.2084949850294675, "learning_rate": 7.14701832533899e-06, "loss": 0.091, "step": 8146 }, { "epoch": 0.37857806691449813, "grad_norm": 0.5126139695643666, "learning_rate": 7.1463386967029715e-06, "loss": 0.0473, "step": 8147 }, { "epoch": 0.3786245353159851, "grad_norm": 1.2590204341056748, "learning_rate": 7.1456590194501215e-06, "loss": 0.0979, "step": 8148 }, { "epoch": 0.37867100371747214, "grad_norm": 0.7621763542384583, "learning_rate": 7.144979293595837e-06, "loss": 0.0392, "step": 8149 }, { "epoch": 0.3787174721189591, "grad_norm": 1.569591069940051, "learning_rate": 7.144299519155515e-06, "loss": 0.1023, "step": 8150 }, { "epoch": 0.3787639405204461, "grad_norm": 0.6977940467227162, "learning_rate": 7.143619696144551e-06, "loss": 0.0719, "step": 8151 }, { "epoch": 0.3788104089219331, "grad_norm": 0.7803940734546192, "learning_rate": 7.142939824578344e-06, "loss": 0.0534, "step": 8152 }, { "epoch": 0.37885687732342005, "grad_norm": 0.8361262197695152, "learning_rate": 7.142259904472296e-06, "loss": 0.0455, "step": 8153 }, { "epoch": 0.3789033457249071, "grad_norm": 1.3211807602764454, "learning_rate": 7.141579935841807e-06, "loss": 0.0845, "step": 8154 }, { "epoch": 0.37894981412639406, "grad_norm": 0.7782515206871322, "learning_rate": 7.140899918702276e-06, "loss": 0.0802, "step": 8155 }, { "epoch": 0.37899628252788103, "grad_norm": 13.288381267879958, "learning_rate": 7.140219853069112e-06, "loss": 0.0672, "step": 8156 }, { "epoch": 0.379042750929368, "grad_norm": 1.0783512899758376, "learning_rate": 7.139539738957713e-06, "loss": 0.0555, "step": 8157 }, { "epoch": 0.37908921933085504, "grad_norm": 0.7409135304655982, "learning_rate": 7.138859576383488e-06, "loss": 0.0424, "step": 8158 }, { "epoch": 0.379135687732342, "grad_norm": 1.24900235939988, "learning_rate": 7.138179365361843e-06, "loss": 0.0492, "step": 8159 }, { "epoch": 0.379182156133829, "grad_norm": 1.7002112699075176, "learning_rate": 7.137499105908183e-06, "loss": 0.0902, "step": 8160 }, { "epoch": 0.379228624535316, "grad_norm": 1.892219602352682, "learning_rate": 7.1368187980379215e-06, "loss": 0.0506, "step": 8161 }, { "epoch": 0.37927509293680295, "grad_norm": 1.795541339979292, "learning_rate": 7.136138441766463e-06, "loss": 0.0605, "step": 8162 }, { "epoch": 0.37932156133829, "grad_norm": 0.7729810152161185, "learning_rate": 7.135458037109222e-06, "loss": 0.0459, "step": 8163 }, { "epoch": 0.37936802973977696, "grad_norm": 3.2039939599694196, "learning_rate": 7.134777584081608e-06, "loss": 0.0686, "step": 8164 }, { "epoch": 0.37941449814126393, "grad_norm": 6.576258760410607, "learning_rate": 7.134097082699035e-06, "loss": 0.0725, "step": 8165 }, { "epoch": 0.3794609665427509, "grad_norm": 1.5798315582647333, "learning_rate": 7.133416532976917e-06, "loss": 0.1232, "step": 8166 }, { "epoch": 0.37950743494423794, "grad_norm": 2.3245455742322423, "learning_rate": 7.1327359349306694e-06, "loss": 0.0715, "step": 8167 }, { "epoch": 0.3795539033457249, "grad_norm": 1.3509374379416255, "learning_rate": 7.132055288575709e-06, "loss": 0.0583, "step": 8168 }, { "epoch": 0.3796003717472119, "grad_norm": 0.8404961425821483, "learning_rate": 7.131374593927452e-06, "loss": 0.0606, "step": 8169 }, { "epoch": 0.37964684014869887, "grad_norm": 1.3572018827135934, "learning_rate": 7.1306938510013155e-06, "loss": 0.0457, "step": 8170 }, { "epoch": 0.37969330855018585, "grad_norm": 0.6759058658626172, "learning_rate": 7.1300130598127226e-06, "loss": 0.0648, "step": 8171 }, { "epoch": 0.3797397769516729, "grad_norm": 0.7263259991398898, "learning_rate": 7.129332220377091e-06, "loss": 0.0561, "step": 8172 }, { "epoch": 0.37978624535315986, "grad_norm": 1.1405508766942296, "learning_rate": 7.128651332709844e-06, "loss": 0.0804, "step": 8173 }, { "epoch": 0.37983271375464683, "grad_norm": 1.941408866946241, "learning_rate": 7.127970396826404e-06, "loss": 0.0676, "step": 8174 }, { "epoch": 0.3798791821561338, "grad_norm": 1.1719332445237964, "learning_rate": 7.127289412742195e-06, "loss": 0.0691, "step": 8175 }, { "epoch": 0.37992565055762084, "grad_norm": 0.8544864656980848, "learning_rate": 7.126608380472642e-06, "loss": 0.0574, "step": 8176 }, { "epoch": 0.3799721189591078, "grad_norm": 1.3694501110287138, "learning_rate": 7.125927300033171e-06, "loss": 0.0764, "step": 8177 }, { "epoch": 0.3800185873605948, "grad_norm": 1.5560239791188624, "learning_rate": 7.1252461714392085e-06, "loss": 0.096, "step": 8178 }, { "epoch": 0.38006505576208177, "grad_norm": 2.9054970709541013, "learning_rate": 7.1245649947061834e-06, "loss": 0.1115, "step": 8179 }, { "epoch": 0.38011152416356875, "grad_norm": 0.7379046802888446, "learning_rate": 7.123883769849527e-06, "loss": 0.057, "step": 8180 }, { "epoch": 0.3801579925650558, "grad_norm": 1.8790077801871594, "learning_rate": 7.1232024968846645e-06, "loss": 0.0444, "step": 8181 }, { "epoch": 0.38020446096654276, "grad_norm": 1.3950973791352586, "learning_rate": 7.122521175827033e-06, "loss": 0.064, "step": 8182 }, { "epoch": 0.38025092936802973, "grad_norm": 1.0138914803348216, "learning_rate": 7.121839806692062e-06, "loss": 0.078, "step": 8183 }, { "epoch": 0.3802973977695167, "grad_norm": 1.3367026323357851, "learning_rate": 7.121158389495187e-06, "loss": 0.06, "step": 8184 }, { "epoch": 0.38034386617100374, "grad_norm": 0.5215623179434635, "learning_rate": 7.120476924251842e-06, "loss": 0.0703, "step": 8185 }, { "epoch": 0.3803903345724907, "grad_norm": 1.1812749341767366, "learning_rate": 7.1197954109774625e-06, "loss": 0.0749, "step": 8186 }, { "epoch": 0.3804368029739777, "grad_norm": 1.1988803736260643, "learning_rate": 7.1191138496874866e-06, "loss": 0.0543, "step": 8187 }, { "epoch": 0.38048327137546467, "grad_norm": 1.2976722948832478, "learning_rate": 7.118432240397352e-06, "loss": 0.0494, "step": 8188 }, { "epoch": 0.38052973977695165, "grad_norm": 1.2768717533085545, "learning_rate": 7.117750583122496e-06, "loss": 0.0475, "step": 8189 }, { "epoch": 0.3805762081784387, "grad_norm": 0.9466091101191768, "learning_rate": 7.11706887787836e-06, "loss": 0.0686, "step": 8190 }, { "epoch": 0.38062267657992566, "grad_norm": 1.037534085797479, "learning_rate": 7.116387124680388e-06, "loss": 0.0507, "step": 8191 }, { "epoch": 0.38066914498141263, "grad_norm": 2.917930393312559, "learning_rate": 7.115705323544019e-06, "loss": 0.1628, "step": 8192 }, { "epoch": 0.3807156133828996, "grad_norm": 1.9657097639510481, "learning_rate": 7.115023474484697e-06, "loss": 0.0751, "step": 8193 }, { "epoch": 0.38076208178438664, "grad_norm": 1.5564080521550987, "learning_rate": 7.1143415775178685e-06, "loss": 0.0808, "step": 8194 }, { "epoch": 0.3808085501858736, "grad_norm": 0.8254116687196965, "learning_rate": 7.113659632658978e-06, "loss": 0.0686, "step": 8195 }, { "epoch": 0.3808550185873606, "grad_norm": 0.7659338186811876, "learning_rate": 7.11297763992347e-06, "loss": 0.0661, "step": 8196 }, { "epoch": 0.38090148698884757, "grad_norm": 0.7920210358911242, "learning_rate": 7.1122955993267975e-06, "loss": 0.0583, "step": 8197 }, { "epoch": 0.38094795539033455, "grad_norm": 2.167173271597294, "learning_rate": 7.111613510884406e-06, "loss": 0.0794, "step": 8198 }, { "epoch": 0.3809944237918216, "grad_norm": 0.8111486736562166, "learning_rate": 7.110931374611745e-06, "loss": 0.0565, "step": 8199 }, { "epoch": 0.38104089219330856, "grad_norm": 1.950480319731683, "learning_rate": 7.110249190524267e-06, "loss": 0.0593, "step": 8200 }, { "epoch": 0.38108736059479553, "grad_norm": 1.151665574038753, "learning_rate": 7.109566958637423e-06, "loss": 0.0896, "step": 8201 }, { "epoch": 0.3811338289962825, "grad_norm": 0.7298140684360216, "learning_rate": 7.108884678966667e-06, "loss": 0.0548, "step": 8202 }, { "epoch": 0.38118029739776954, "grad_norm": 1.5820784228333722, "learning_rate": 7.108202351527456e-06, "loss": 0.1272, "step": 8203 }, { "epoch": 0.3812267657992565, "grad_norm": 1.0033417367513515, "learning_rate": 7.107519976335241e-06, "loss": 0.0744, "step": 8204 }, { "epoch": 0.3812732342007435, "grad_norm": 4.268286161503843, "learning_rate": 7.10683755340548e-06, "loss": 0.0827, "step": 8205 }, { "epoch": 0.38131970260223047, "grad_norm": 1.415913966925894, "learning_rate": 7.106155082753632e-06, "loss": 0.0906, "step": 8206 }, { "epoch": 0.38136617100371745, "grad_norm": 0.8165832721460056, "learning_rate": 7.105472564395155e-06, "loss": 0.0725, "step": 8207 }, { "epoch": 0.3814126394052045, "grad_norm": 0.6680419125325311, "learning_rate": 7.104789998345507e-06, "loss": 0.0461, "step": 8208 }, { "epoch": 0.38145910780669146, "grad_norm": 0.7957697322581156, "learning_rate": 7.104107384620153e-06, "loss": 0.0416, "step": 8209 }, { "epoch": 0.38150557620817843, "grad_norm": 1.1972951190923895, "learning_rate": 7.103424723234549e-06, "loss": 0.0636, "step": 8210 }, { "epoch": 0.3815520446096654, "grad_norm": 1.4409991390498265, "learning_rate": 7.102742014204163e-06, "loss": 0.0692, "step": 8211 }, { "epoch": 0.38159851301115244, "grad_norm": 0.9414038283385613, "learning_rate": 7.102059257544457e-06, "loss": 0.0536, "step": 8212 }, { "epoch": 0.3816449814126394, "grad_norm": 1.8658950270396888, "learning_rate": 7.101376453270895e-06, "loss": 0.0778, "step": 8213 }, { "epoch": 0.3816914498141264, "grad_norm": 1.261357666418384, "learning_rate": 7.100693601398946e-06, "loss": 0.0831, "step": 8214 }, { "epoch": 0.38173791821561337, "grad_norm": 0.9614071249083402, "learning_rate": 7.100010701944077e-06, "loss": 0.0635, "step": 8215 }, { "epoch": 0.38178438661710035, "grad_norm": 0.6884464586869944, "learning_rate": 7.099327754921754e-06, "loss": 0.0487, "step": 8216 }, { "epoch": 0.3818308550185874, "grad_norm": 1.6404702602131491, "learning_rate": 7.098644760347448e-06, "loss": 0.0512, "step": 8217 }, { "epoch": 0.38187732342007435, "grad_norm": 2.4265957659041204, "learning_rate": 7.097961718236629e-06, "loss": 0.0977, "step": 8218 }, { "epoch": 0.38192379182156133, "grad_norm": 0.9632723623984809, "learning_rate": 7.09727862860477e-06, "loss": 0.0544, "step": 8219 }, { "epoch": 0.3819702602230483, "grad_norm": 2.310860245111427, "learning_rate": 7.096595491467342e-06, "loss": 0.073, "step": 8220 }, { "epoch": 0.38201672862453534, "grad_norm": 0.9124553698495477, "learning_rate": 7.095912306839821e-06, "loss": 0.0566, "step": 8221 }, { "epoch": 0.3820631970260223, "grad_norm": 1.5256936784842487, "learning_rate": 7.095229074737677e-06, "loss": 0.0882, "step": 8222 }, { "epoch": 0.3821096654275093, "grad_norm": 2.5636398221626773, "learning_rate": 7.094545795176393e-06, "loss": 0.0632, "step": 8223 }, { "epoch": 0.38215613382899627, "grad_norm": 0.556519129320099, "learning_rate": 7.093862468171439e-06, "loss": 0.0375, "step": 8224 }, { "epoch": 0.38220260223048325, "grad_norm": 1.180300849375746, "learning_rate": 7.093179093738298e-06, "loss": 0.0678, "step": 8225 }, { "epoch": 0.3822490706319703, "grad_norm": 2.1489452355837977, "learning_rate": 7.092495671892449e-06, "loss": 0.1199, "step": 8226 }, { "epoch": 0.38229553903345725, "grad_norm": 1.260689327444132, "learning_rate": 7.091812202649369e-06, "loss": 0.0844, "step": 8227 }, { "epoch": 0.38234200743494423, "grad_norm": 1.2848508964550482, "learning_rate": 7.091128686024541e-06, "loss": 0.0843, "step": 8228 }, { "epoch": 0.3823884758364312, "grad_norm": 0.565993510766881, "learning_rate": 7.090445122033447e-06, "loss": 0.0454, "step": 8229 }, { "epoch": 0.38243494423791824, "grad_norm": 1.5942692993025904, "learning_rate": 7.089761510691572e-06, "loss": 0.0676, "step": 8230 }, { "epoch": 0.3824814126394052, "grad_norm": 0.7789520071640543, "learning_rate": 7.089077852014399e-06, "loss": 0.0734, "step": 8231 }, { "epoch": 0.3825278810408922, "grad_norm": 2.4149830102150758, "learning_rate": 7.088394146017412e-06, "loss": 0.0897, "step": 8232 }, { "epoch": 0.38257434944237917, "grad_norm": 1.368396916200572, "learning_rate": 7.087710392716103e-06, "loss": 0.1216, "step": 8233 }, { "epoch": 0.38262081784386615, "grad_norm": 0.6211991322448771, "learning_rate": 7.087026592125954e-06, "loss": 0.0374, "step": 8234 }, { "epoch": 0.3826672862453532, "grad_norm": 0.9702728384740984, "learning_rate": 7.086342744262457e-06, "loss": 0.0767, "step": 8235 }, { "epoch": 0.38271375464684015, "grad_norm": 0.9517999021629548, "learning_rate": 7.085658849141101e-06, "loss": 0.0912, "step": 8236 }, { "epoch": 0.38276022304832713, "grad_norm": 1.6965712217090103, "learning_rate": 7.0849749067773765e-06, "loss": 0.1183, "step": 8237 }, { "epoch": 0.3828066914498141, "grad_norm": 0.668243828375158, "learning_rate": 7.084290917186778e-06, "loss": 0.0711, "step": 8238 }, { "epoch": 0.38285315985130114, "grad_norm": 1.1255668988806076, "learning_rate": 7.083606880384793e-06, "loss": 0.0664, "step": 8239 }, { "epoch": 0.3828996282527881, "grad_norm": 0.8108497565928855, "learning_rate": 7.082922796386922e-06, "loss": 0.0943, "step": 8240 }, { "epoch": 0.3829460966542751, "grad_norm": 0.8391814324016161, "learning_rate": 7.082238665208656e-06, "loss": 0.0887, "step": 8241 }, { "epoch": 0.38299256505576207, "grad_norm": 1.029267059839825, "learning_rate": 7.081554486865493e-06, "loss": 0.0999, "step": 8242 }, { "epoch": 0.38303903345724905, "grad_norm": 1.2114896897641674, "learning_rate": 7.08087026137293e-06, "loss": 0.0544, "step": 8243 }, { "epoch": 0.3830855018587361, "grad_norm": 0.5214275243500478, "learning_rate": 7.080185988746466e-06, "loss": 0.0573, "step": 8244 }, { "epoch": 0.38313197026022305, "grad_norm": 0.6925713885142947, "learning_rate": 7.079501669001601e-06, "loss": 0.0832, "step": 8245 }, { "epoch": 0.38317843866171003, "grad_norm": 0.7799060121685243, "learning_rate": 7.078817302153831e-06, "loss": 0.0686, "step": 8246 }, { "epoch": 0.383224907063197, "grad_norm": 0.8938832950486814, "learning_rate": 7.078132888218664e-06, "loss": 0.0571, "step": 8247 }, { "epoch": 0.38327137546468404, "grad_norm": 0.8482806789265737, "learning_rate": 7.0774484272116e-06, "loss": 0.0672, "step": 8248 }, { "epoch": 0.383317843866171, "grad_norm": 0.99497194486015, "learning_rate": 7.076763919148141e-06, "loss": 0.0537, "step": 8249 }, { "epoch": 0.383364312267658, "grad_norm": 1.1226053170270682, "learning_rate": 7.076079364043795e-06, "loss": 0.0472, "step": 8250 }, { "epoch": 0.38341078066914497, "grad_norm": 0.8521616105027093, "learning_rate": 7.0753947619140654e-06, "loss": 0.0436, "step": 8251 }, { "epoch": 0.38345724907063194, "grad_norm": 0.5598969760902633, "learning_rate": 7.07471011277446e-06, "loss": 0.0698, "step": 8252 }, { "epoch": 0.383503717472119, "grad_norm": 1.2082327656118472, "learning_rate": 7.0740254166404884e-06, "loss": 0.0692, "step": 8253 }, { "epoch": 0.38355018587360595, "grad_norm": 2.9929895886437974, "learning_rate": 7.073340673527657e-06, "loss": 0.1002, "step": 8254 }, { "epoch": 0.38359665427509293, "grad_norm": 3.961317821729724, "learning_rate": 7.072655883451478e-06, "loss": 0.0686, "step": 8255 }, { "epoch": 0.3836431226765799, "grad_norm": 0.8544651045359024, "learning_rate": 7.0719710464274615e-06, "loss": 0.0664, "step": 8256 }, { "epoch": 0.38368959107806694, "grad_norm": 1.7873235544269854, "learning_rate": 7.07128616247112e-06, "loss": 0.0815, "step": 8257 }, { "epoch": 0.3837360594795539, "grad_norm": 0.810274033536985, "learning_rate": 7.070601231597967e-06, "loss": 0.0579, "step": 8258 }, { "epoch": 0.3837825278810409, "grad_norm": 0.7777951230176974, "learning_rate": 7.069916253823517e-06, "loss": 0.0427, "step": 8259 }, { "epoch": 0.38382899628252787, "grad_norm": 1.7565774592329209, "learning_rate": 7.069231229163285e-06, "loss": 0.0652, "step": 8260 }, { "epoch": 0.38387546468401484, "grad_norm": 0.45505433246589655, "learning_rate": 7.068546157632786e-06, "loss": 0.0553, "step": 8261 }, { "epoch": 0.3839219330855019, "grad_norm": 3.8244026412406407, "learning_rate": 7.067861039247542e-06, "loss": 0.0781, "step": 8262 }, { "epoch": 0.38396840148698885, "grad_norm": 1.6124076430564405, "learning_rate": 7.067175874023069e-06, "loss": 0.0791, "step": 8263 }, { "epoch": 0.38401486988847583, "grad_norm": 1.709970698286148, "learning_rate": 7.066490661974885e-06, "loss": 0.0893, "step": 8264 }, { "epoch": 0.3840613382899628, "grad_norm": 0.6657052614842418, "learning_rate": 7.0658054031185156e-06, "loss": 0.0394, "step": 8265 }, { "epoch": 0.38410780669144984, "grad_norm": 1.8885466158287154, "learning_rate": 7.0651200974694755e-06, "loss": 0.0737, "step": 8266 }, { "epoch": 0.3841542750929368, "grad_norm": 1.114715119751256, "learning_rate": 7.064434745043295e-06, "loss": 0.0526, "step": 8267 }, { "epoch": 0.3842007434944238, "grad_norm": 0.8352981975865394, "learning_rate": 7.063749345855494e-06, "loss": 0.05, "step": 8268 }, { "epoch": 0.38424721189591077, "grad_norm": 0.6408879388258091, "learning_rate": 7.063063899921597e-06, "loss": 0.0515, "step": 8269 }, { "epoch": 0.38429368029739774, "grad_norm": 3.073226873248716, "learning_rate": 7.0623784072571326e-06, "loss": 0.0509, "step": 8270 }, { "epoch": 0.3843401486988848, "grad_norm": 0.47592843048275574, "learning_rate": 7.061692867877626e-06, "loss": 0.0554, "step": 8271 }, { "epoch": 0.38438661710037175, "grad_norm": 2.8089996028704887, "learning_rate": 7.061007281798607e-06, "loss": 0.0811, "step": 8272 }, { "epoch": 0.38443308550185873, "grad_norm": 1.2496250883843498, "learning_rate": 7.060321649035602e-06, "loss": 0.0474, "step": 8273 }, { "epoch": 0.3844795539033457, "grad_norm": 1.0630632501265123, "learning_rate": 7.059635969604145e-06, "loss": 0.0783, "step": 8274 }, { "epoch": 0.38452602230483274, "grad_norm": 0.6288376450548202, "learning_rate": 7.058950243519765e-06, "loss": 0.0418, "step": 8275 }, { "epoch": 0.3845724907063197, "grad_norm": 1.0509813825523535, "learning_rate": 7.058264470797994e-06, "loss": 0.0868, "step": 8276 }, { "epoch": 0.3846189591078067, "grad_norm": 1.434285543260762, "learning_rate": 7.0575786514543685e-06, "loss": 0.0729, "step": 8277 }, { "epoch": 0.38466542750929367, "grad_norm": 0.7661122187124554, "learning_rate": 7.05689278550442e-06, "loss": 0.0567, "step": 8278 }, { "epoch": 0.38471189591078064, "grad_norm": 1.027926858937649, "learning_rate": 7.0562068729636845e-06, "loss": 0.0617, "step": 8279 }, { "epoch": 0.3847583643122677, "grad_norm": 1.0534639592643738, "learning_rate": 7.0555209138477e-06, "loss": 0.076, "step": 8280 }, { "epoch": 0.38480483271375465, "grad_norm": 0.9674744739155671, "learning_rate": 7.0548349081720025e-06, "loss": 0.0767, "step": 8281 }, { "epoch": 0.38485130111524163, "grad_norm": 0.5641683121872108, "learning_rate": 7.0541488559521335e-06, "loss": 0.0646, "step": 8282 }, { "epoch": 0.3848977695167286, "grad_norm": 0.9381163308873158, "learning_rate": 7.05346275720363e-06, "loss": 0.0508, "step": 8283 }, { "epoch": 0.38494423791821564, "grad_norm": 0.5822618281920788, "learning_rate": 7.052776611942032e-06, "loss": 0.0405, "step": 8284 }, { "epoch": 0.3849907063197026, "grad_norm": 0.915536925753291, "learning_rate": 7.052090420182886e-06, "loss": 0.0703, "step": 8285 }, { "epoch": 0.3850371747211896, "grad_norm": 3.405894793801354, "learning_rate": 7.05140418194173e-06, "loss": 0.0699, "step": 8286 }, { "epoch": 0.38508364312267657, "grad_norm": 1.0949989548630326, "learning_rate": 7.0507178972341116e-06, "loss": 0.0955, "step": 8287 }, { "epoch": 0.38513011152416354, "grad_norm": 0.6126923336059896, "learning_rate": 7.050031566075574e-06, "loss": 0.06, "step": 8288 }, { "epoch": 0.3851765799256506, "grad_norm": 1.0890045920846325, "learning_rate": 7.049345188481664e-06, "loss": 0.0607, "step": 8289 }, { "epoch": 0.38522304832713755, "grad_norm": 0.7211403539903151, "learning_rate": 7.048658764467928e-06, "loss": 0.066, "step": 8290 }, { "epoch": 0.38526951672862453, "grad_norm": 1.0054581283584343, "learning_rate": 7.047972294049916e-06, "loss": 0.0743, "step": 8291 }, { "epoch": 0.3853159851301115, "grad_norm": 0.9814967525544693, "learning_rate": 7.047285777243175e-06, "loss": 0.0791, "step": 8292 }, { "epoch": 0.38536245353159854, "grad_norm": 0.5766690089234656, "learning_rate": 7.046599214063257e-06, "loss": 0.0554, "step": 8293 }, { "epoch": 0.3854089219330855, "grad_norm": 0.6226003058172621, "learning_rate": 7.045912604525712e-06, "loss": 0.0519, "step": 8294 }, { "epoch": 0.3854553903345725, "grad_norm": 1.0182061983662356, "learning_rate": 7.045225948646094e-06, "loss": 0.0795, "step": 8295 }, { "epoch": 0.38550185873605947, "grad_norm": 0.45605573407600747, "learning_rate": 7.044539246439957e-06, "loss": 0.0401, "step": 8296 }, { "epoch": 0.38554832713754644, "grad_norm": 0.9416863594518542, "learning_rate": 7.043852497922852e-06, "loss": 0.0662, "step": 8297 }, { "epoch": 0.3855947955390335, "grad_norm": 0.7430788412055431, "learning_rate": 7.043165703110337e-06, "loss": 0.0505, "step": 8298 }, { "epoch": 0.38564126394052045, "grad_norm": 0.9981856317925347, "learning_rate": 7.042478862017969e-06, "loss": 0.0419, "step": 8299 }, { "epoch": 0.3856877323420074, "grad_norm": 1.1345677914032044, "learning_rate": 7.041791974661305e-06, "loss": 0.0522, "step": 8300 }, { "epoch": 0.3857342007434944, "grad_norm": 0.9155109883115508, "learning_rate": 7.041105041055904e-06, "loss": 0.0708, "step": 8301 }, { "epoch": 0.38578066914498144, "grad_norm": 1.472684195374552, "learning_rate": 7.040418061217325e-06, "loss": 0.0735, "step": 8302 }, { "epoch": 0.3858271375464684, "grad_norm": 1.6939953490958761, "learning_rate": 7.03973103516113e-06, "loss": 0.0757, "step": 8303 }, { "epoch": 0.3858736059479554, "grad_norm": 0.7046000166980985, "learning_rate": 7.03904396290288e-06, "loss": 0.0607, "step": 8304 }, { "epoch": 0.38592007434944237, "grad_norm": 0.8983918309121194, "learning_rate": 7.038356844458139e-06, "loss": 0.073, "step": 8305 }, { "epoch": 0.38596654275092934, "grad_norm": 0.8638789921588313, "learning_rate": 7.037669679842469e-06, "loss": 0.0588, "step": 8306 }, { "epoch": 0.3860130111524164, "grad_norm": 0.5456917137454175, "learning_rate": 7.036982469071436e-06, "loss": 0.0446, "step": 8307 }, { "epoch": 0.38605947955390335, "grad_norm": 2.223210292178113, "learning_rate": 7.0362952121606075e-06, "loss": 0.0488, "step": 8308 }, { "epoch": 0.3861059479553903, "grad_norm": 0.9694840206186174, "learning_rate": 7.035607909125549e-06, "loss": 0.0568, "step": 8309 }, { "epoch": 0.3861524163568773, "grad_norm": 1.3893813911535446, "learning_rate": 7.034920559981828e-06, "loss": 0.0799, "step": 8310 }, { "epoch": 0.38619888475836434, "grad_norm": 1.0325877142462097, "learning_rate": 7.034233164745015e-06, "loss": 0.048, "step": 8311 }, { "epoch": 0.3862453531598513, "grad_norm": 0.8328325927033522, "learning_rate": 7.03354572343068e-06, "loss": 0.0715, "step": 8312 }, { "epoch": 0.3862918215613383, "grad_norm": 1.587429920987984, "learning_rate": 7.032858236054395e-06, "loss": 0.0633, "step": 8313 }, { "epoch": 0.38633828996282527, "grad_norm": 1.0431146229091355, "learning_rate": 7.03217070263173e-06, "loss": 0.0538, "step": 8314 }, { "epoch": 0.38638475836431224, "grad_norm": 0.660210482229861, "learning_rate": 7.03148312317826e-06, "loss": 0.0432, "step": 8315 }, { "epoch": 0.3864312267657993, "grad_norm": 0.6257563576826927, "learning_rate": 7.030795497709559e-06, "loss": 0.0541, "step": 8316 }, { "epoch": 0.38647769516728625, "grad_norm": 1.3606736530045112, "learning_rate": 7.030107826241203e-06, "loss": 0.076, "step": 8317 }, { "epoch": 0.3865241635687732, "grad_norm": 2.434906626797401, "learning_rate": 7.0294201087887694e-06, "loss": 0.0883, "step": 8318 }, { "epoch": 0.3865706319702602, "grad_norm": 0.7044937445602782, "learning_rate": 7.028732345367833e-06, "loss": 0.0439, "step": 8319 }, { "epoch": 0.38661710037174724, "grad_norm": 1.2447620132690966, "learning_rate": 7.0280445359939734e-06, "loss": 0.0598, "step": 8320 }, { "epoch": 0.3866635687732342, "grad_norm": 1.275897139091529, "learning_rate": 7.027356680682771e-06, "loss": 0.0656, "step": 8321 }, { "epoch": 0.3867100371747212, "grad_norm": 0.6538052838068754, "learning_rate": 7.026668779449805e-06, "loss": 0.0428, "step": 8322 }, { "epoch": 0.38675650557620816, "grad_norm": 1.8074475447552671, "learning_rate": 7.02598083231066e-06, "loss": 0.0653, "step": 8323 }, { "epoch": 0.38680297397769514, "grad_norm": 0.5737765692680943, "learning_rate": 7.025292839280916e-06, "loss": 0.072, "step": 8324 }, { "epoch": 0.3868494423791822, "grad_norm": 1.6571507930265652, "learning_rate": 7.024604800376158e-06, "loss": 0.0702, "step": 8325 }, { "epoch": 0.38689591078066915, "grad_norm": 0.9343748872379102, "learning_rate": 7.023916715611969e-06, "loss": 0.0683, "step": 8326 }, { "epoch": 0.3869423791821561, "grad_norm": 1.4693066448734782, "learning_rate": 7.0232285850039385e-06, "loss": 0.0977, "step": 8327 }, { "epoch": 0.3869888475836431, "grad_norm": 1.1043045331466037, "learning_rate": 7.0225404085676494e-06, "loss": 0.0791, "step": 8328 }, { "epoch": 0.38703531598513014, "grad_norm": 0.8511722940795576, "learning_rate": 7.0218521863186915e-06, "loss": 0.0671, "step": 8329 }, { "epoch": 0.3870817843866171, "grad_norm": 0.7824106040444939, "learning_rate": 7.021163918272655e-06, "loss": 0.0631, "step": 8330 }, { "epoch": 0.3871282527881041, "grad_norm": 0.8138467559124497, "learning_rate": 7.020475604445127e-06, "loss": 0.063, "step": 8331 }, { "epoch": 0.38717472118959106, "grad_norm": 0.7178301889111098, "learning_rate": 7.019787244851701e-06, "loss": 0.0556, "step": 8332 }, { "epoch": 0.38722118959107804, "grad_norm": 3.1835260162367436, "learning_rate": 7.019098839507967e-06, "loss": 0.0569, "step": 8333 }, { "epoch": 0.3872676579925651, "grad_norm": 0.7200670045756536, "learning_rate": 7.01841038842952e-06, "loss": 0.0565, "step": 8334 }, { "epoch": 0.38731412639405205, "grad_norm": 1.657813902918602, "learning_rate": 7.017721891631953e-06, "loss": 0.0733, "step": 8335 }, { "epoch": 0.387360594795539, "grad_norm": 0.5920499555896083, "learning_rate": 7.017033349130863e-06, "loss": 0.0461, "step": 8336 }, { "epoch": 0.387407063197026, "grad_norm": 0.6276232674026689, "learning_rate": 7.016344760941844e-06, "loss": 0.0562, "step": 8337 }, { "epoch": 0.38745353159851303, "grad_norm": 1.0038834011382196, "learning_rate": 7.015656127080492e-06, "loss": 0.0613, "step": 8338 }, { "epoch": 0.3875, "grad_norm": 1.3888621111800954, "learning_rate": 7.0149674475624095e-06, "loss": 0.0672, "step": 8339 }, { "epoch": 0.387546468401487, "grad_norm": 0.8017159634582912, "learning_rate": 7.014278722403194e-06, "loss": 0.0394, "step": 8340 }, { "epoch": 0.38759293680297396, "grad_norm": 1.4914411928220686, "learning_rate": 7.013589951618442e-06, "loss": 0.0866, "step": 8341 }, { "epoch": 0.38763940520446094, "grad_norm": 1.1905640718793125, "learning_rate": 7.012901135223763e-06, "loss": 0.0428, "step": 8342 }, { "epoch": 0.387685873605948, "grad_norm": 0.6954360417085327, "learning_rate": 7.012212273234752e-06, "loss": 0.04, "step": 8343 }, { "epoch": 0.38773234200743495, "grad_norm": 0.9770242184694644, "learning_rate": 7.011523365667015e-06, "loss": 0.0738, "step": 8344 }, { "epoch": 0.3877788104089219, "grad_norm": 0.8606452973502529, "learning_rate": 7.010834412536158e-06, "loss": 0.0734, "step": 8345 }, { "epoch": 0.3878252788104089, "grad_norm": 0.9027502435200185, "learning_rate": 7.010145413857784e-06, "loss": 0.0549, "step": 8346 }, { "epoch": 0.38787174721189593, "grad_norm": 1.966716617758119, "learning_rate": 7.009456369647502e-06, "loss": 0.0505, "step": 8347 }, { "epoch": 0.3879182156133829, "grad_norm": 0.6379844694558693, "learning_rate": 7.008767279920917e-06, "loss": 0.0508, "step": 8348 }, { "epoch": 0.3879646840148699, "grad_norm": 1.2042311158246515, "learning_rate": 7.00807814469364e-06, "loss": 0.0631, "step": 8349 }, { "epoch": 0.38801115241635686, "grad_norm": 0.6510615641152941, "learning_rate": 7.007388963981278e-06, "loss": 0.0524, "step": 8350 }, { "epoch": 0.38805762081784384, "grad_norm": 2.5515978869007787, "learning_rate": 7.006699737799445e-06, "loss": 0.0734, "step": 8351 }, { "epoch": 0.38810408921933087, "grad_norm": 1.372537271184845, "learning_rate": 7.006010466163751e-06, "loss": 0.0783, "step": 8352 }, { "epoch": 0.38815055762081785, "grad_norm": 0.8048415777623388, "learning_rate": 7.005321149089807e-06, "loss": 0.0857, "step": 8353 }, { "epoch": 0.3881970260223048, "grad_norm": 0.8713846236941816, "learning_rate": 7.00463178659323e-06, "loss": 0.0469, "step": 8354 }, { "epoch": 0.3882434944237918, "grad_norm": 0.5999434282757536, "learning_rate": 7.003942378689632e-06, "loss": 0.0562, "step": 8355 }, { "epoch": 0.38828996282527883, "grad_norm": 1.6867855421241926, "learning_rate": 7.00325292539463e-06, "loss": 0.083, "step": 8356 }, { "epoch": 0.3883364312267658, "grad_norm": 2.9527872313840255, "learning_rate": 7.002563426723843e-06, "loss": 0.0794, "step": 8357 }, { "epoch": 0.3883828996282528, "grad_norm": 1.48670606872545, "learning_rate": 7.001873882692884e-06, "loss": 0.0872, "step": 8358 }, { "epoch": 0.38842936802973976, "grad_norm": 0.9805815211875492, "learning_rate": 7.001184293317377e-06, "loss": 0.1215, "step": 8359 }, { "epoch": 0.38847583643122674, "grad_norm": 0.9044782842149247, "learning_rate": 7.000494658612938e-06, "loss": 0.0702, "step": 8360 }, { "epoch": 0.38852230483271377, "grad_norm": 0.7895786574895574, "learning_rate": 6.999804978595189e-06, "loss": 0.0571, "step": 8361 }, { "epoch": 0.38856877323420075, "grad_norm": 0.8343046631286227, "learning_rate": 6.999115253279754e-06, "loss": 0.0612, "step": 8362 }, { "epoch": 0.3886152416356877, "grad_norm": 0.9652742292699025, "learning_rate": 6.9984254826822525e-06, "loss": 0.0536, "step": 8363 }, { "epoch": 0.3886617100371747, "grad_norm": 0.7997729999179505, "learning_rate": 6.997735666818313e-06, "loss": 0.056, "step": 8364 }, { "epoch": 0.38870817843866173, "grad_norm": 0.6994168120582086, "learning_rate": 6.997045805703556e-06, "loss": 0.0475, "step": 8365 }, { "epoch": 0.3887546468401487, "grad_norm": 1.5129027767913954, "learning_rate": 6.996355899353611e-06, "loss": 0.0835, "step": 8366 }, { "epoch": 0.3888011152416357, "grad_norm": 1.3645187000085828, "learning_rate": 6.995665947784103e-06, "loss": 0.0557, "step": 8367 }, { "epoch": 0.38884758364312266, "grad_norm": 0.9402347433475307, "learning_rate": 6.99497595101066e-06, "loss": 0.0921, "step": 8368 }, { "epoch": 0.38889405204460964, "grad_norm": 0.9147267883600037, "learning_rate": 6.994285909048914e-06, "loss": 0.0586, "step": 8369 }, { "epoch": 0.38894052044609667, "grad_norm": 0.7350860016852763, "learning_rate": 6.993595821914492e-06, "loss": 0.0437, "step": 8370 }, { "epoch": 0.38898698884758365, "grad_norm": 0.46881312690840327, "learning_rate": 6.992905689623028e-06, "loss": 0.0259, "step": 8371 }, { "epoch": 0.3890334572490706, "grad_norm": 6.290180349014111, "learning_rate": 6.9922155121901505e-06, "loss": 0.0504, "step": 8372 }, { "epoch": 0.3890799256505576, "grad_norm": 0.9418724000036139, "learning_rate": 6.991525289631496e-06, "loss": 0.0903, "step": 8373 }, { "epoch": 0.38912639405204463, "grad_norm": 1.042875189958684, "learning_rate": 6.990835021962698e-06, "loss": 0.0552, "step": 8374 }, { "epoch": 0.3891728624535316, "grad_norm": 1.4893852166059594, "learning_rate": 6.990144709199389e-06, "loss": 0.039, "step": 8375 }, { "epoch": 0.3892193308550186, "grad_norm": 1.2099247959106119, "learning_rate": 6.98945435135721e-06, "loss": 0.0733, "step": 8376 }, { "epoch": 0.38926579925650556, "grad_norm": 0.8955780475297158, "learning_rate": 6.988763948451796e-06, "loss": 0.0535, "step": 8377 }, { "epoch": 0.38931226765799254, "grad_norm": 1.0494109963412115, "learning_rate": 6.988073500498786e-06, "loss": 0.0527, "step": 8378 }, { "epoch": 0.38935873605947957, "grad_norm": 0.5597251169124658, "learning_rate": 6.987383007513817e-06, "loss": 0.0374, "step": 8379 }, { "epoch": 0.38940520446096655, "grad_norm": 1.9694237452934165, "learning_rate": 6.986692469512533e-06, "loss": 0.0744, "step": 8380 }, { "epoch": 0.3894516728624535, "grad_norm": 2.9427601820551685, "learning_rate": 6.986001886510572e-06, "loss": 0.0959, "step": 8381 }, { "epoch": 0.3894981412639405, "grad_norm": 2.155993965303105, "learning_rate": 6.985311258523578e-06, "loss": 0.0615, "step": 8382 }, { "epoch": 0.38954460966542753, "grad_norm": 1.432092607241042, "learning_rate": 6.984620585567197e-06, "loss": 0.0875, "step": 8383 }, { "epoch": 0.3895910780669145, "grad_norm": 1.0101784173484023, "learning_rate": 6.983929867657069e-06, "loss": 0.0475, "step": 8384 }, { "epoch": 0.3896375464684015, "grad_norm": 1.012598496497942, "learning_rate": 6.983239104808841e-06, "loss": 0.078, "step": 8385 }, { "epoch": 0.38968401486988846, "grad_norm": 0.7571612033437195, "learning_rate": 6.982548297038161e-06, "loss": 0.0793, "step": 8386 }, { "epoch": 0.38973048327137544, "grad_norm": 0.6654711509869031, "learning_rate": 6.9818574443606745e-06, "loss": 0.0438, "step": 8387 }, { "epoch": 0.38977695167286247, "grad_norm": 0.845098298936222, "learning_rate": 6.981166546792032e-06, "loss": 0.0593, "step": 8388 }, { "epoch": 0.38982342007434945, "grad_norm": 0.9899588581039098, "learning_rate": 6.980475604347882e-06, "loss": 0.0452, "step": 8389 }, { "epoch": 0.3898698884758364, "grad_norm": 0.7165766929613545, "learning_rate": 6.979784617043875e-06, "loss": 0.0481, "step": 8390 }, { "epoch": 0.3899163568773234, "grad_norm": 1.2098961267967079, "learning_rate": 6.979093584895662e-06, "loss": 0.0632, "step": 8391 }, { "epoch": 0.38996282527881043, "grad_norm": 0.7493144017564267, "learning_rate": 6.978402507918897e-06, "loss": 0.0705, "step": 8392 }, { "epoch": 0.3900092936802974, "grad_norm": 0.714644356719886, "learning_rate": 6.977711386129233e-06, "loss": 0.0441, "step": 8393 }, { "epoch": 0.3900557620817844, "grad_norm": 0.9114492239914568, "learning_rate": 6.977020219542323e-06, "loss": 0.0588, "step": 8394 }, { "epoch": 0.39010223048327136, "grad_norm": 1.169398929507371, "learning_rate": 6.976329008173828e-06, "loss": 0.0823, "step": 8395 }, { "epoch": 0.39014869888475834, "grad_norm": 0.8753360470280547, "learning_rate": 6.975637752039397e-06, "loss": 0.0643, "step": 8396 }, { "epoch": 0.39019516728624537, "grad_norm": 0.9709984007041577, "learning_rate": 6.974946451154694e-06, "loss": 0.0716, "step": 8397 }, { "epoch": 0.39024163568773235, "grad_norm": 1.133979554559598, "learning_rate": 6.974255105535374e-06, "loss": 0.0703, "step": 8398 }, { "epoch": 0.3902881040892193, "grad_norm": 1.2781960679728257, "learning_rate": 6.973563715197098e-06, "loss": 0.0425, "step": 8399 }, { "epoch": 0.3903345724907063, "grad_norm": 0.7880164219330922, "learning_rate": 6.972872280155528e-06, "loss": 0.0439, "step": 8400 }, { "epoch": 0.39038104089219333, "grad_norm": 2.7390040506680475, "learning_rate": 6.972180800426323e-06, "loss": 0.1441, "step": 8401 }, { "epoch": 0.3904275092936803, "grad_norm": 1.5528935121106984, "learning_rate": 6.97148927602515e-06, "loss": 0.0607, "step": 8402 }, { "epoch": 0.3904739776951673, "grad_norm": 0.9483317800060124, "learning_rate": 6.970797706967666e-06, "loss": 0.058, "step": 8403 }, { "epoch": 0.39052044609665426, "grad_norm": 0.6188759552869348, "learning_rate": 6.9701060932695415e-06, "loss": 0.0637, "step": 8404 }, { "epoch": 0.39056691449814124, "grad_norm": 1.4844852586219601, "learning_rate": 6.969414434946441e-06, "loss": 0.0609, "step": 8405 }, { "epoch": 0.39061338289962827, "grad_norm": 1.3676459689094882, "learning_rate": 6.968722732014029e-06, "loss": 0.0585, "step": 8406 }, { "epoch": 0.39065985130111525, "grad_norm": 0.7201200300922015, "learning_rate": 6.968030984487979e-06, "loss": 0.0521, "step": 8407 }, { "epoch": 0.3907063197026022, "grad_norm": 0.9023278640938031, "learning_rate": 6.967339192383952e-06, "loss": 0.0485, "step": 8408 }, { "epoch": 0.3907527881040892, "grad_norm": 0.5466966941333067, "learning_rate": 6.966647355717623e-06, "loss": 0.0365, "step": 8409 }, { "epoch": 0.39079925650557623, "grad_norm": 1.314841267281522, "learning_rate": 6.965955474504662e-06, "loss": 0.0606, "step": 8410 }, { "epoch": 0.3908457249070632, "grad_norm": 0.7903197341604189, "learning_rate": 6.965263548760739e-06, "loss": 0.0501, "step": 8411 }, { "epoch": 0.3908921933085502, "grad_norm": 2.2557997280874025, "learning_rate": 6.964571578501529e-06, "loss": 0.0836, "step": 8412 }, { "epoch": 0.39093866171003716, "grad_norm": 0.5847090069367371, "learning_rate": 6.963879563742706e-06, "loss": 0.0361, "step": 8413 }, { "epoch": 0.39098513011152414, "grad_norm": 0.8170169639205159, "learning_rate": 6.9631875044999435e-06, "loss": 0.0628, "step": 8414 }, { "epoch": 0.39103159851301117, "grad_norm": 0.8280639770285578, "learning_rate": 6.962495400788918e-06, "loss": 0.0485, "step": 8415 }, { "epoch": 0.39107806691449815, "grad_norm": 1.9964727559787594, "learning_rate": 6.961803252625305e-06, "loss": 0.0515, "step": 8416 }, { "epoch": 0.3911245353159851, "grad_norm": 1.1584992996248473, "learning_rate": 6.961111060024783e-06, "loss": 0.0883, "step": 8417 }, { "epoch": 0.3911710037174721, "grad_norm": 1.377252193981543, "learning_rate": 6.960418823003034e-06, "loss": 0.0756, "step": 8418 }, { "epoch": 0.39121747211895913, "grad_norm": 0.72748627912572, "learning_rate": 6.9597265415757335e-06, "loss": 0.0653, "step": 8419 }, { "epoch": 0.3912639405204461, "grad_norm": 0.8375938694991755, "learning_rate": 6.9590342157585645e-06, "loss": 0.069, "step": 8420 }, { "epoch": 0.3913104089219331, "grad_norm": 1.225818250731661, "learning_rate": 6.958341845567209e-06, "loss": 0.049, "step": 8421 }, { "epoch": 0.39135687732342006, "grad_norm": 1.252808723830609, "learning_rate": 6.957649431017349e-06, "loss": 0.0723, "step": 8422 }, { "epoch": 0.39140334572490704, "grad_norm": 1.4086937149428962, "learning_rate": 6.956956972124669e-06, "loss": 0.0506, "step": 8423 }, { "epoch": 0.39144981412639407, "grad_norm": 2.395258780595299, "learning_rate": 6.956264468904855e-06, "loss": 0.0468, "step": 8424 }, { "epoch": 0.39149628252788105, "grad_norm": 0.7030141233769395, "learning_rate": 6.9555719213735916e-06, "loss": 0.0488, "step": 8425 }, { "epoch": 0.391542750929368, "grad_norm": 1.7171506508617282, "learning_rate": 6.954879329546566e-06, "loss": 0.0589, "step": 8426 }, { "epoch": 0.391589219330855, "grad_norm": 0.578993533217227, "learning_rate": 6.9541866934394654e-06, "loss": 0.0241, "step": 8427 }, { "epoch": 0.39163568773234203, "grad_norm": 0.4571043299134485, "learning_rate": 6.953494013067978e-06, "loss": 0.0344, "step": 8428 }, { "epoch": 0.391682156133829, "grad_norm": 0.5030490093287271, "learning_rate": 6.952801288447798e-06, "loss": 0.0259, "step": 8429 }, { "epoch": 0.391728624535316, "grad_norm": 1.4220134010520107, "learning_rate": 6.952108519594614e-06, "loss": 0.0351, "step": 8430 }, { "epoch": 0.39177509293680296, "grad_norm": 0.8986495886481146, "learning_rate": 6.951415706524115e-06, "loss": 0.0671, "step": 8431 }, { "epoch": 0.39182156133828994, "grad_norm": 1.5152413642820655, "learning_rate": 6.950722849251996e-06, "loss": 0.0595, "step": 8432 }, { "epoch": 0.39186802973977697, "grad_norm": 0.604437434176627, "learning_rate": 6.9500299477939535e-06, "loss": 0.0423, "step": 8433 }, { "epoch": 0.39191449814126395, "grad_norm": 1.0687981063203753, "learning_rate": 6.94933700216568e-06, "loss": 0.0454, "step": 8434 }, { "epoch": 0.3919609665427509, "grad_norm": 0.9576434001433856, "learning_rate": 6.948644012382871e-06, "loss": 0.0888, "step": 8435 }, { "epoch": 0.3920074349442379, "grad_norm": 0.7848273967596043, "learning_rate": 6.947950978461226e-06, "loss": 0.0509, "step": 8436 }, { "epoch": 0.39205390334572493, "grad_norm": 0.646357124934445, "learning_rate": 6.947257900416438e-06, "loss": 0.0537, "step": 8437 }, { "epoch": 0.3921003717472119, "grad_norm": 1.3553443907570757, "learning_rate": 6.946564778264211e-06, "loss": 0.0484, "step": 8438 }, { "epoch": 0.3921468401486989, "grad_norm": 1.0910464781386249, "learning_rate": 6.945871612020244e-06, "loss": 0.0529, "step": 8439 }, { "epoch": 0.39219330855018586, "grad_norm": 2.109974368802258, "learning_rate": 6.945178401700234e-06, "loss": 0.0843, "step": 8440 }, { "epoch": 0.39223977695167284, "grad_norm": 0.8139765781385674, "learning_rate": 6.9444851473198885e-06, "loss": 0.0629, "step": 8441 }, { "epoch": 0.39228624535315987, "grad_norm": 1.2069723043756904, "learning_rate": 6.943791848894908e-06, "loss": 0.0886, "step": 8442 }, { "epoch": 0.39233271375464684, "grad_norm": 1.164969648007531, "learning_rate": 6.943098506440995e-06, "loss": 0.063, "step": 8443 }, { "epoch": 0.3923791821561338, "grad_norm": 0.5424066776234596, "learning_rate": 6.942405119973856e-06, "loss": 0.0514, "step": 8444 }, { "epoch": 0.3924256505576208, "grad_norm": 1.5641928457379464, "learning_rate": 6.9417116895091976e-06, "loss": 0.0584, "step": 8445 }, { "epoch": 0.39247211895910783, "grad_norm": 3.2423848191346782, "learning_rate": 6.941018215062725e-06, "loss": 0.1198, "step": 8446 }, { "epoch": 0.3925185873605948, "grad_norm": 0.8837321255275075, "learning_rate": 6.940324696650147e-06, "loss": 0.0749, "step": 8447 }, { "epoch": 0.3925650557620818, "grad_norm": 1.2477316399227305, "learning_rate": 6.939631134287172e-06, "loss": 0.0684, "step": 8448 }, { "epoch": 0.39261152416356876, "grad_norm": 0.788388172169938, "learning_rate": 6.938937527989511e-06, "loss": 0.0731, "step": 8449 }, { "epoch": 0.39265799256505574, "grad_norm": 1.1941243449446965, "learning_rate": 6.938243877772874e-06, "loss": 0.0545, "step": 8450 }, { "epoch": 0.39270446096654277, "grad_norm": 2.664006081763452, "learning_rate": 6.937550183652975e-06, "loss": 0.0741, "step": 8451 }, { "epoch": 0.39275092936802974, "grad_norm": 0.906521021525428, "learning_rate": 6.936856445645522e-06, "loss": 0.0648, "step": 8452 }, { "epoch": 0.3927973977695167, "grad_norm": 1.6268495073390854, "learning_rate": 6.936162663766234e-06, "loss": 0.0704, "step": 8453 }, { "epoch": 0.3928438661710037, "grad_norm": 1.5021415437453292, "learning_rate": 6.935468838030823e-06, "loss": 0.0602, "step": 8454 }, { "epoch": 0.39289033457249073, "grad_norm": 1.0872415432017313, "learning_rate": 6.934774968455007e-06, "loss": 0.0691, "step": 8455 }, { "epoch": 0.3929368029739777, "grad_norm": 5.001623327584806, "learning_rate": 6.9340810550545004e-06, "loss": 0.0621, "step": 8456 }, { "epoch": 0.3929832713754647, "grad_norm": 0.5927323479735465, "learning_rate": 6.9333870978450234e-06, "loss": 0.0438, "step": 8457 }, { "epoch": 0.39302973977695166, "grad_norm": 2.230559980853245, "learning_rate": 6.932693096842293e-06, "loss": 0.077, "step": 8458 }, { "epoch": 0.39307620817843864, "grad_norm": 0.6573763608189834, "learning_rate": 6.931999052062029e-06, "loss": 0.0673, "step": 8459 }, { "epoch": 0.39312267657992567, "grad_norm": 1.1204700128058016, "learning_rate": 6.931304963519956e-06, "loss": 0.0437, "step": 8460 }, { "epoch": 0.39316914498141264, "grad_norm": 2.4695473322195283, "learning_rate": 6.93061083123179e-06, "loss": 0.087, "step": 8461 }, { "epoch": 0.3932156133828996, "grad_norm": 1.0506207758861041, "learning_rate": 6.929916655213258e-06, "loss": 0.0473, "step": 8462 }, { "epoch": 0.3932620817843866, "grad_norm": 3.4130418777793654, "learning_rate": 6.929222435480083e-06, "loss": 0.1195, "step": 8463 }, { "epoch": 0.39330855018587363, "grad_norm": 1.0680621029028479, "learning_rate": 6.928528172047987e-06, "loss": 0.0512, "step": 8464 }, { "epoch": 0.3933550185873606, "grad_norm": 1.261063259170928, "learning_rate": 6.9278338649327e-06, "loss": 0.0499, "step": 8465 }, { "epoch": 0.3934014869888476, "grad_norm": 1.025141258967618, "learning_rate": 6.927139514149948e-06, "loss": 0.0385, "step": 8466 }, { "epoch": 0.39344795539033456, "grad_norm": 0.8081632188433907, "learning_rate": 6.926445119715456e-06, "loss": 0.0692, "step": 8467 }, { "epoch": 0.39349442379182153, "grad_norm": 0.8041055490495838, "learning_rate": 6.925750681644954e-06, "loss": 0.0966, "step": 8468 }, { "epoch": 0.39354089219330857, "grad_norm": 1.1388437983193664, "learning_rate": 6.925056199954174e-06, "loss": 0.0799, "step": 8469 }, { "epoch": 0.39358736059479554, "grad_norm": 0.7553503971104688, "learning_rate": 6.924361674658844e-06, "loss": 0.0767, "step": 8470 }, { "epoch": 0.3936338289962825, "grad_norm": 1.1655590115962051, "learning_rate": 6.923667105774696e-06, "loss": 0.0625, "step": 8471 }, { "epoch": 0.3936802973977695, "grad_norm": 1.2191183199529525, "learning_rate": 6.922972493317466e-06, "loss": 0.0518, "step": 8472 }, { "epoch": 0.39372676579925653, "grad_norm": 1.5648966701741363, "learning_rate": 6.9222778373028835e-06, "loss": 0.0905, "step": 8473 }, { "epoch": 0.3937732342007435, "grad_norm": 0.9629085400899722, "learning_rate": 6.921583137746685e-06, "loss": 0.0561, "step": 8474 }, { "epoch": 0.3938197026022305, "grad_norm": 0.9011069493483407, "learning_rate": 6.920888394664606e-06, "loss": 0.0736, "step": 8475 }, { "epoch": 0.39386617100371746, "grad_norm": 1.2669088648080555, "learning_rate": 6.920193608072382e-06, "loss": 0.0669, "step": 8476 }, { "epoch": 0.39391263940520443, "grad_norm": 1.025875333691221, "learning_rate": 6.919498777985754e-06, "loss": 0.0802, "step": 8477 }, { "epoch": 0.39395910780669147, "grad_norm": 1.0787518388542252, "learning_rate": 6.918803904420457e-06, "loss": 0.0432, "step": 8478 }, { "epoch": 0.39400557620817844, "grad_norm": 1.3920045352987342, "learning_rate": 6.918108987392232e-06, "loss": 0.0596, "step": 8479 }, { "epoch": 0.3940520446096654, "grad_norm": 2.3025365351011433, "learning_rate": 6.91741402691682e-06, "loss": 0.1085, "step": 8480 }, { "epoch": 0.3940985130111524, "grad_norm": 0.8634299190542335, "learning_rate": 6.916719023009961e-06, "loss": 0.069, "step": 8481 }, { "epoch": 0.39414498141263943, "grad_norm": 0.46795044912801714, "learning_rate": 6.9160239756874e-06, "loss": 0.0446, "step": 8482 }, { "epoch": 0.3941914498141264, "grad_norm": 3.2696342448102076, "learning_rate": 6.9153288849648805e-06, "loss": 0.0445, "step": 8483 }, { "epoch": 0.3942379182156134, "grad_norm": 3.012115808263889, "learning_rate": 6.914633750858145e-06, "loss": 0.07, "step": 8484 }, { "epoch": 0.39428438661710036, "grad_norm": 0.9765507006836303, "learning_rate": 6.9139385733829386e-06, "loss": 0.0711, "step": 8485 }, { "epoch": 0.39433085501858733, "grad_norm": 0.626968642203199, "learning_rate": 6.913243352555011e-06, "loss": 0.0545, "step": 8486 }, { "epoch": 0.39437732342007437, "grad_norm": 0.499188883211229, "learning_rate": 6.912548088390107e-06, "loss": 0.0393, "step": 8487 }, { "epoch": 0.39442379182156134, "grad_norm": 0.6544682460997518, "learning_rate": 6.911852780903976e-06, "loss": 0.0506, "step": 8488 }, { "epoch": 0.3944702602230483, "grad_norm": 3.4079470501363303, "learning_rate": 6.911157430112368e-06, "loss": 0.0912, "step": 8489 }, { "epoch": 0.3945167286245353, "grad_norm": 0.723810294098093, "learning_rate": 6.910462036031031e-06, "loss": 0.0657, "step": 8490 }, { "epoch": 0.3945631970260223, "grad_norm": 0.6717573210476258, "learning_rate": 6.909766598675719e-06, "loss": 0.0507, "step": 8491 }, { "epoch": 0.3946096654275093, "grad_norm": 0.6863598735041484, "learning_rate": 6.909071118062185e-06, "loss": 0.0623, "step": 8492 }, { "epoch": 0.3946561338289963, "grad_norm": 0.6722221572046798, "learning_rate": 6.908375594206178e-06, "loss": 0.0595, "step": 8493 }, { "epoch": 0.39470260223048326, "grad_norm": 0.6524094471149648, "learning_rate": 6.907680027123458e-06, "loss": 0.0372, "step": 8494 }, { "epoch": 0.39474907063197023, "grad_norm": 1.7541727521079822, "learning_rate": 6.906984416829776e-06, "loss": 0.067, "step": 8495 }, { "epoch": 0.39479553903345727, "grad_norm": 0.7571116304966543, "learning_rate": 6.90628876334089e-06, "loss": 0.0411, "step": 8496 }, { "epoch": 0.39484200743494424, "grad_norm": 1.9392334463081473, "learning_rate": 6.905593066672558e-06, "loss": 0.1063, "step": 8497 }, { "epoch": 0.3948884758364312, "grad_norm": 2.6321888987198987, "learning_rate": 6.9048973268405375e-06, "loss": 0.0711, "step": 8498 }, { "epoch": 0.3949349442379182, "grad_norm": 0.575567530098761, "learning_rate": 6.904201543860587e-06, "loss": 0.0539, "step": 8499 }, { "epoch": 0.3949814126394052, "grad_norm": 5.035487393077095, "learning_rate": 6.903505717748467e-06, "loss": 0.0747, "step": 8500 }, { "epoch": 0.3950278810408922, "grad_norm": 0.7645726107506563, "learning_rate": 6.9028098485199405e-06, "loss": 0.0505, "step": 8501 }, { "epoch": 0.3950743494423792, "grad_norm": 0.9284988340692538, "learning_rate": 6.902113936190767e-06, "loss": 0.0857, "step": 8502 }, { "epoch": 0.39512081784386616, "grad_norm": 0.7389167678563925, "learning_rate": 6.9014179807767115e-06, "loss": 0.0426, "step": 8503 }, { "epoch": 0.39516728624535313, "grad_norm": 1.1065215617820507, "learning_rate": 6.900721982293538e-06, "loss": 0.0661, "step": 8504 }, { "epoch": 0.39521375464684017, "grad_norm": 0.7653362079205849, "learning_rate": 6.900025940757009e-06, "loss": 0.0778, "step": 8505 }, { "epoch": 0.39526022304832714, "grad_norm": 0.7384463526139355, "learning_rate": 6.899329856182895e-06, "loss": 0.0631, "step": 8506 }, { "epoch": 0.3953066914498141, "grad_norm": 1.2047426810655255, "learning_rate": 6.898633728586961e-06, "loss": 0.0533, "step": 8507 }, { "epoch": 0.3953531598513011, "grad_norm": 0.5490120993985195, "learning_rate": 6.8979375579849725e-06, "loss": 0.0595, "step": 8508 }, { "epoch": 0.3953996282527881, "grad_norm": 0.5736614644419762, "learning_rate": 6.897241344392702e-06, "loss": 0.0535, "step": 8509 }, { "epoch": 0.3954460966542751, "grad_norm": 1.2191132174534993, "learning_rate": 6.896545087825919e-06, "loss": 0.0847, "step": 8510 }, { "epoch": 0.3954925650557621, "grad_norm": 1.2196668629583327, "learning_rate": 6.895848788300393e-06, "loss": 0.1358, "step": 8511 }, { "epoch": 0.39553903345724906, "grad_norm": 0.7827970154676926, "learning_rate": 6.895152445831896e-06, "loss": 0.0668, "step": 8512 }, { "epoch": 0.39558550185873603, "grad_norm": 0.8967593139114451, "learning_rate": 6.894456060436202e-06, "loss": 0.0738, "step": 8513 }, { "epoch": 0.39563197026022306, "grad_norm": 1.762520238680625, "learning_rate": 6.8937596321290845e-06, "loss": 0.0579, "step": 8514 }, { "epoch": 0.39567843866171004, "grad_norm": 1.5708901772940635, "learning_rate": 6.893063160926317e-06, "loss": 0.0627, "step": 8515 }, { "epoch": 0.395724907063197, "grad_norm": 1.039689290090262, "learning_rate": 6.892366646843679e-06, "loss": 0.1002, "step": 8516 }, { "epoch": 0.395771375464684, "grad_norm": 0.9170257546051992, "learning_rate": 6.8916700898969425e-06, "loss": 0.0543, "step": 8517 }, { "epoch": 0.395817843866171, "grad_norm": 2.0023722573911416, "learning_rate": 6.890973490101889e-06, "loss": 0.0777, "step": 8518 }, { "epoch": 0.395864312267658, "grad_norm": 0.5143294928894148, "learning_rate": 6.8902768474742956e-06, "loss": 0.0458, "step": 8519 }, { "epoch": 0.395910780669145, "grad_norm": 0.605708895564103, "learning_rate": 6.889580162029941e-06, "loss": 0.0448, "step": 8520 }, { "epoch": 0.39595724907063196, "grad_norm": 0.5747972402444776, "learning_rate": 6.888883433784608e-06, "loss": 0.0644, "step": 8521 }, { "epoch": 0.39600371747211893, "grad_norm": 0.6032562208703784, "learning_rate": 6.888186662754076e-06, "loss": 0.0427, "step": 8522 }, { "epoch": 0.39605018587360596, "grad_norm": 1.2878888228436642, "learning_rate": 6.887489848954129e-06, "loss": 0.0861, "step": 8523 }, { "epoch": 0.39609665427509294, "grad_norm": 1.1307308855936995, "learning_rate": 6.8867929924005516e-06, "loss": 0.0704, "step": 8524 }, { "epoch": 0.3961431226765799, "grad_norm": 3.3660538824167094, "learning_rate": 6.886096093109127e-06, "loss": 0.0656, "step": 8525 }, { "epoch": 0.3961895910780669, "grad_norm": 0.568700654872652, "learning_rate": 6.8853991510956395e-06, "loss": 0.0547, "step": 8526 }, { "epoch": 0.3962360594795539, "grad_norm": 1.7903692948474728, "learning_rate": 6.884702166375878e-06, "loss": 0.0687, "step": 8527 }, { "epoch": 0.3962825278810409, "grad_norm": 1.5995680034103614, "learning_rate": 6.884005138965629e-06, "loss": 0.0865, "step": 8528 }, { "epoch": 0.3963289962825279, "grad_norm": 0.5074776067229323, "learning_rate": 6.8833080688806795e-06, "loss": 0.0577, "step": 8529 }, { "epoch": 0.39637546468401486, "grad_norm": 0.5823583628402667, "learning_rate": 6.8826109561368216e-06, "loss": 0.0456, "step": 8530 }, { "epoch": 0.39642193308550183, "grad_norm": 0.5258553700695964, "learning_rate": 6.881913800749843e-06, "loss": 0.0574, "step": 8531 }, { "epoch": 0.39646840148698886, "grad_norm": 1.16843601839683, "learning_rate": 6.8812166027355365e-06, "loss": 0.0877, "step": 8532 }, { "epoch": 0.39651486988847584, "grad_norm": 0.9658622371841699, "learning_rate": 6.880519362109694e-06, "loss": 0.0599, "step": 8533 }, { "epoch": 0.3965613382899628, "grad_norm": 1.133123719706151, "learning_rate": 6.8798220788881086e-06, "loss": 0.0719, "step": 8534 }, { "epoch": 0.3966078066914498, "grad_norm": 1.0977033564282386, "learning_rate": 6.879124753086575e-06, "loss": 0.047, "step": 8535 }, { "epoch": 0.3966542750929368, "grad_norm": 0.692195302223141, "learning_rate": 6.878427384720888e-06, "loss": 0.0438, "step": 8536 }, { "epoch": 0.3967007434944238, "grad_norm": 1.1143054733767705, "learning_rate": 6.8777299738068425e-06, "loss": 0.0564, "step": 8537 }, { "epoch": 0.3967472118959108, "grad_norm": 0.7842009704405614, "learning_rate": 6.877032520360239e-06, "loss": 0.0729, "step": 8538 }, { "epoch": 0.39679368029739776, "grad_norm": 0.4792783742348212, "learning_rate": 6.876335024396872e-06, "loss": 0.0541, "step": 8539 }, { "epoch": 0.39684014869888473, "grad_norm": 0.9241021424166302, "learning_rate": 6.875637485932543e-06, "loss": 0.0504, "step": 8540 }, { "epoch": 0.39688661710037176, "grad_norm": 1.3958303831692362, "learning_rate": 6.874939904983048e-06, "loss": 0.0735, "step": 8541 }, { "epoch": 0.39693308550185874, "grad_norm": 0.9223009284987228, "learning_rate": 6.874242281564193e-06, "loss": 0.0612, "step": 8542 }, { "epoch": 0.3969795539033457, "grad_norm": 0.7152402379090887, "learning_rate": 6.873544615691779e-06, "loss": 0.0701, "step": 8543 }, { "epoch": 0.3970260223048327, "grad_norm": 0.7757642633133127, "learning_rate": 6.872846907381604e-06, "loss": 0.0531, "step": 8544 }, { "epoch": 0.3970724907063197, "grad_norm": 1.9811696929796174, "learning_rate": 6.872149156649479e-06, "loss": 0.0877, "step": 8545 }, { "epoch": 0.3971189591078067, "grad_norm": 0.5559978707844864, "learning_rate": 6.871451363511202e-06, "loss": 0.0729, "step": 8546 }, { "epoch": 0.3971654275092937, "grad_norm": 1.1680372623530089, "learning_rate": 6.8707535279825845e-06, "loss": 0.0628, "step": 8547 }, { "epoch": 0.39721189591078065, "grad_norm": 0.5068251488436577, "learning_rate": 6.87005565007943e-06, "loss": 0.0447, "step": 8548 }, { "epoch": 0.39725836431226763, "grad_norm": 0.7643383544891409, "learning_rate": 6.869357729817546e-06, "loss": 0.0555, "step": 8549 }, { "epoch": 0.39730483271375466, "grad_norm": 2.346625702159312, "learning_rate": 6.868659767212743e-06, "loss": 0.0759, "step": 8550 }, { "epoch": 0.39735130111524164, "grad_norm": 0.753601933660479, "learning_rate": 6.867961762280829e-06, "loss": 0.0462, "step": 8551 }, { "epoch": 0.3973977695167286, "grad_norm": 0.5451558982341742, "learning_rate": 6.867263715037615e-06, "loss": 0.0424, "step": 8552 }, { "epoch": 0.3974442379182156, "grad_norm": 0.9834291864148025, "learning_rate": 6.866565625498912e-06, "loss": 0.0403, "step": 8553 }, { "epoch": 0.3974907063197026, "grad_norm": 0.8777621740664072, "learning_rate": 6.8658674936805346e-06, "loss": 0.0496, "step": 8554 }, { "epoch": 0.3975371747211896, "grad_norm": 0.5680774771250279, "learning_rate": 6.865169319598293e-06, "loss": 0.0559, "step": 8555 }, { "epoch": 0.3975836431226766, "grad_norm": 1.4927684293087933, "learning_rate": 6.864471103268005e-06, "loss": 0.0507, "step": 8556 }, { "epoch": 0.39763011152416355, "grad_norm": 0.7740645661714878, "learning_rate": 6.863772844705484e-06, "loss": 0.0781, "step": 8557 }, { "epoch": 0.39767657992565053, "grad_norm": 1.5124345885351334, "learning_rate": 6.863074543926546e-06, "loss": 0.0667, "step": 8558 }, { "epoch": 0.39772304832713756, "grad_norm": 1.0733789522308264, "learning_rate": 6.862376200947007e-06, "loss": 0.0737, "step": 8559 }, { "epoch": 0.39776951672862454, "grad_norm": 1.8850890749431215, "learning_rate": 6.86167781578269e-06, "loss": 0.0858, "step": 8560 }, { "epoch": 0.3978159851301115, "grad_norm": 1.0552146869841152, "learning_rate": 6.860979388449409e-06, "loss": 0.0696, "step": 8561 }, { "epoch": 0.3978624535315985, "grad_norm": 1.5977946133378593, "learning_rate": 6.860280918962987e-06, "loss": 0.0515, "step": 8562 }, { "epoch": 0.3979089219330855, "grad_norm": 0.5965169070139558, "learning_rate": 6.859582407339245e-06, "loss": 0.047, "step": 8563 }, { "epoch": 0.3979553903345725, "grad_norm": 1.5551469568225968, "learning_rate": 6.8588838535940025e-06, "loss": 0.0367, "step": 8564 }, { "epoch": 0.3980018587360595, "grad_norm": 0.8924500680054076, "learning_rate": 6.858185257743084e-06, "loss": 0.0598, "step": 8565 }, { "epoch": 0.39804832713754645, "grad_norm": 1.8062259841197459, "learning_rate": 6.857486619802316e-06, "loss": 0.0764, "step": 8566 }, { "epoch": 0.39809479553903343, "grad_norm": 0.998835990503977, "learning_rate": 6.85678793978752e-06, "loss": 0.0719, "step": 8567 }, { "epoch": 0.39814126394052046, "grad_norm": 0.5083288769225145, "learning_rate": 6.856089217714522e-06, "loss": 0.0689, "step": 8568 }, { "epoch": 0.39818773234200744, "grad_norm": 0.9301202246233834, "learning_rate": 6.855390453599153e-06, "loss": 0.0733, "step": 8569 }, { "epoch": 0.3982342007434944, "grad_norm": 0.7056074093586522, "learning_rate": 6.854691647457233e-06, "loss": 0.062, "step": 8570 }, { "epoch": 0.3982806691449814, "grad_norm": 1.2976028157837498, "learning_rate": 6.853992799304597e-06, "loss": 0.0726, "step": 8571 }, { "epoch": 0.3983271375464684, "grad_norm": 1.9783468515135985, "learning_rate": 6.853293909157074e-06, "loss": 0.0817, "step": 8572 }, { "epoch": 0.3983736059479554, "grad_norm": 0.8434866819955605, "learning_rate": 6.852594977030491e-06, "loss": 0.0558, "step": 8573 }, { "epoch": 0.3984200743494424, "grad_norm": 1.54135375169486, "learning_rate": 6.851896002940684e-06, "loss": 0.0676, "step": 8574 }, { "epoch": 0.39846654275092935, "grad_norm": 1.2269995316779505, "learning_rate": 6.851196986903483e-06, "loss": 0.0755, "step": 8575 }, { "epoch": 0.39851301115241633, "grad_norm": 1.3710253201662268, "learning_rate": 6.850497928934723e-06, "loss": 0.0716, "step": 8576 }, { "epoch": 0.39855947955390336, "grad_norm": 2.4050658339589015, "learning_rate": 6.849798829050235e-06, "loss": 0.1015, "step": 8577 }, { "epoch": 0.39860594795539034, "grad_norm": 0.7992549403187106, "learning_rate": 6.849099687265858e-06, "loss": 0.0501, "step": 8578 }, { "epoch": 0.3986524163568773, "grad_norm": 0.9026090988474678, "learning_rate": 6.848400503597427e-06, "loss": 0.0764, "step": 8579 }, { "epoch": 0.3986988847583643, "grad_norm": 0.7516921379527286, "learning_rate": 6.8477012780607774e-06, "loss": 0.0678, "step": 8580 }, { "epoch": 0.3987453531598513, "grad_norm": 0.5591325272186966, "learning_rate": 6.8470020106717515e-06, "loss": 0.0214, "step": 8581 }, { "epoch": 0.3987918215613383, "grad_norm": 1.4259564470143797, "learning_rate": 6.846302701446184e-06, "loss": 0.0514, "step": 8582 }, { "epoch": 0.3988382899628253, "grad_norm": 1.0909858656624027, "learning_rate": 6.84560335039992e-06, "loss": 0.0687, "step": 8583 }, { "epoch": 0.39888475836431225, "grad_norm": 0.7966627860758594, "learning_rate": 6.844903957548795e-06, "loss": 0.065, "step": 8584 }, { "epoch": 0.39893122676579923, "grad_norm": 0.6185739295664046, "learning_rate": 6.844204522908655e-06, "loss": 0.0644, "step": 8585 }, { "epoch": 0.39897769516728626, "grad_norm": 0.8693744024838529, "learning_rate": 6.843505046495341e-06, "loss": 0.0481, "step": 8586 }, { "epoch": 0.39902416356877324, "grad_norm": 0.21938237117931517, "learning_rate": 6.842805528324698e-06, "loss": 0.0544, "step": 8587 }, { "epoch": 0.3990706319702602, "grad_norm": 0.8715628977707265, "learning_rate": 6.84210596841257e-06, "loss": 0.0459, "step": 8588 }, { "epoch": 0.3991171003717472, "grad_norm": 1.4932110688714342, "learning_rate": 6.841406366774803e-06, "loss": 0.0736, "step": 8589 }, { "epoch": 0.3991635687732342, "grad_norm": 0.7124582655423985, "learning_rate": 6.840706723427243e-06, "loss": 0.0509, "step": 8590 }, { "epoch": 0.3992100371747212, "grad_norm": 0.6112306368648002, "learning_rate": 6.840007038385739e-06, "loss": 0.0318, "step": 8591 }, { "epoch": 0.3992565055762082, "grad_norm": 1.8399179676409623, "learning_rate": 6.839307311666139e-06, "loss": 0.0746, "step": 8592 }, { "epoch": 0.39930297397769515, "grad_norm": 1.1311490178155015, "learning_rate": 6.838607543284294e-06, "loss": 0.0817, "step": 8593 }, { "epoch": 0.39934944237918213, "grad_norm": 0.5988811019408681, "learning_rate": 6.837907733256051e-06, "loss": 0.0591, "step": 8594 }, { "epoch": 0.39939591078066916, "grad_norm": 0.649502828575068, "learning_rate": 6.837207881597263e-06, "loss": 0.0674, "step": 8595 }, { "epoch": 0.39944237918215614, "grad_norm": 0.9338557967280517, "learning_rate": 6.836507988323785e-06, "loss": 0.0597, "step": 8596 }, { "epoch": 0.3994888475836431, "grad_norm": 1.4328227551532584, "learning_rate": 6.835808053451467e-06, "loss": 0.0834, "step": 8597 }, { "epoch": 0.3995353159851301, "grad_norm": 0.7057544365290775, "learning_rate": 6.835108076996165e-06, "loss": 0.0586, "step": 8598 }, { "epoch": 0.3995817843866171, "grad_norm": 1.2345773001521336, "learning_rate": 6.8344080589737325e-06, "loss": 0.0623, "step": 8599 }, { "epoch": 0.3996282527881041, "grad_norm": 4.851776393673563, "learning_rate": 6.833707999400028e-06, "loss": 0.0717, "step": 8600 }, { "epoch": 0.3996747211895911, "grad_norm": 1.6899448842726255, "learning_rate": 6.833007898290908e-06, "loss": 0.0745, "step": 8601 }, { "epoch": 0.39972118959107805, "grad_norm": 1.7296993212372425, "learning_rate": 6.832307755662227e-06, "loss": 0.0675, "step": 8602 }, { "epoch": 0.39976765799256503, "grad_norm": 0.6996254646825559, "learning_rate": 6.831607571529849e-06, "loss": 0.0588, "step": 8603 }, { "epoch": 0.39981412639405206, "grad_norm": 0.662619572530385, "learning_rate": 6.830907345909632e-06, "loss": 0.0448, "step": 8604 }, { "epoch": 0.39986059479553904, "grad_norm": 0.3755083970857076, "learning_rate": 6.830207078817437e-06, "loss": 0.0312, "step": 8605 }, { "epoch": 0.399907063197026, "grad_norm": 0.5054083958626171, "learning_rate": 6.829506770269123e-06, "loss": 0.0527, "step": 8606 }, { "epoch": 0.399953531598513, "grad_norm": 0.8524711250765572, "learning_rate": 6.828806420280557e-06, "loss": 0.0612, "step": 8607 }, { "epoch": 0.4, "grad_norm": 1.0950136590267456, "learning_rate": 6.828106028867601e-06, "loss": 0.0837, "step": 8608 }, { "epoch": 0.400046468401487, "grad_norm": 1.0319662458590375, "learning_rate": 6.827405596046117e-06, "loss": 0.1067, "step": 8609 }, { "epoch": 0.400092936802974, "grad_norm": 1.246291232873265, "learning_rate": 6.8267051218319766e-06, "loss": 0.0615, "step": 8610 }, { "epoch": 0.40013940520446095, "grad_norm": 0.5710606851515038, "learning_rate": 6.8260046062410405e-06, "loss": 0.0507, "step": 8611 }, { "epoch": 0.40018587360594793, "grad_norm": 0.8565388837556117, "learning_rate": 6.825304049289179e-06, "loss": 0.0645, "step": 8612 }, { "epoch": 0.40023234200743496, "grad_norm": 0.9187513400700831, "learning_rate": 6.824603450992259e-06, "loss": 0.0638, "step": 8613 }, { "epoch": 0.40027881040892194, "grad_norm": 1.7647799853134665, "learning_rate": 6.82390281136615e-06, "loss": 0.0792, "step": 8614 }, { "epoch": 0.4003252788104089, "grad_norm": 1.239387218870013, "learning_rate": 6.8232021304267225e-06, "loss": 0.0685, "step": 8615 }, { "epoch": 0.4003717472118959, "grad_norm": 0.8998429646620575, "learning_rate": 6.82250140818985e-06, "loss": 0.0625, "step": 8616 }, { "epoch": 0.4004182156133829, "grad_norm": 1.0496081789118448, "learning_rate": 6.821800644671401e-06, "loss": 0.0678, "step": 8617 }, { "epoch": 0.4004646840148699, "grad_norm": 2.5309959544845695, "learning_rate": 6.82109983988725e-06, "loss": 0.0542, "step": 8618 }, { "epoch": 0.4005111524163569, "grad_norm": 1.2751264701155711, "learning_rate": 6.820398993853271e-06, "loss": 0.0518, "step": 8619 }, { "epoch": 0.40055762081784385, "grad_norm": 2.001339376509122, "learning_rate": 6.819698106585339e-06, "loss": 0.0999, "step": 8620 }, { "epoch": 0.40060408921933083, "grad_norm": 0.6136943972673164, "learning_rate": 6.818997178099329e-06, "loss": 0.038, "step": 8621 }, { "epoch": 0.40065055762081786, "grad_norm": 0.9525829846106737, "learning_rate": 6.81829620841112e-06, "loss": 0.0785, "step": 8622 }, { "epoch": 0.40069702602230484, "grad_norm": 1.0543679493668145, "learning_rate": 6.817595197536586e-06, "loss": 0.0883, "step": 8623 }, { "epoch": 0.4007434944237918, "grad_norm": 0.685982006084088, "learning_rate": 6.8168941454916105e-06, "loss": 0.042, "step": 8624 }, { "epoch": 0.4007899628252788, "grad_norm": 2.1859799050849573, "learning_rate": 6.816193052292069e-06, "loss": 0.0653, "step": 8625 }, { "epoch": 0.4008364312267658, "grad_norm": 0.5453768974217459, "learning_rate": 6.8154919179538425e-06, "loss": 0.0401, "step": 8626 }, { "epoch": 0.4008828996282528, "grad_norm": 0.7654943335644651, "learning_rate": 6.814790742492814e-06, "loss": 0.0508, "step": 8627 }, { "epoch": 0.4009293680297398, "grad_norm": 0.9916861130108501, "learning_rate": 6.814089525924865e-06, "loss": 0.0774, "step": 8628 }, { "epoch": 0.40097583643122675, "grad_norm": 0.5156449946744012, "learning_rate": 6.81338826826588e-06, "loss": 0.0596, "step": 8629 }, { "epoch": 0.4010223048327137, "grad_norm": 1.8254759506193983, "learning_rate": 6.812686969531741e-06, "loss": 0.0695, "step": 8630 }, { "epoch": 0.40106877323420076, "grad_norm": 1.2141125655534148, "learning_rate": 6.8119856297383345e-06, "loss": 0.084, "step": 8631 }, { "epoch": 0.40111524163568774, "grad_norm": 0.7785630904626196, "learning_rate": 6.811284248901546e-06, "loss": 0.0594, "step": 8632 }, { "epoch": 0.4011617100371747, "grad_norm": 1.5119508621224496, "learning_rate": 6.810582827037263e-06, "loss": 0.0496, "step": 8633 }, { "epoch": 0.4012081784386617, "grad_norm": 1.0952766796045363, "learning_rate": 6.809881364161376e-06, "loss": 0.0597, "step": 8634 }, { "epoch": 0.4012546468401487, "grad_norm": 0.6954610690434047, "learning_rate": 6.809179860289768e-06, "loss": 0.0519, "step": 8635 }, { "epoch": 0.4013011152416357, "grad_norm": 0.5751240345602647, "learning_rate": 6.808478315438334e-06, "loss": 0.0478, "step": 8636 }, { "epoch": 0.4013475836431227, "grad_norm": 1.1567392172358506, "learning_rate": 6.8077767296229614e-06, "loss": 0.1031, "step": 8637 }, { "epoch": 0.40139405204460965, "grad_norm": 1.1428739533272465, "learning_rate": 6.807075102859543e-06, "loss": 0.0572, "step": 8638 }, { "epoch": 0.4014405204460966, "grad_norm": 0.6421756285539577, "learning_rate": 6.806373435163972e-06, "loss": 0.0303, "step": 8639 }, { "epoch": 0.40148698884758366, "grad_norm": 0.76367211971061, "learning_rate": 6.8056717265521436e-06, "loss": 0.081, "step": 8640 }, { "epoch": 0.40153345724907064, "grad_norm": 0.849290315336898, "learning_rate": 6.804969977039948e-06, "loss": 0.0627, "step": 8641 }, { "epoch": 0.4015799256505576, "grad_norm": 0.9090183186419792, "learning_rate": 6.8042681866432815e-06, "loss": 0.0686, "step": 8642 }, { "epoch": 0.4016263940520446, "grad_norm": 1.662603134600298, "learning_rate": 6.803566355378044e-06, "loss": 0.076, "step": 8643 }, { "epoch": 0.4016728624535316, "grad_norm": 1.133403218425339, "learning_rate": 6.8028644832601295e-06, "loss": 0.055, "step": 8644 }, { "epoch": 0.4017193308550186, "grad_norm": 1.861324884600313, "learning_rate": 6.802162570305436e-06, "loss": 0.0835, "step": 8645 }, { "epoch": 0.4017657992565056, "grad_norm": 0.7743451305571406, "learning_rate": 6.801460616529867e-06, "loss": 0.0703, "step": 8646 }, { "epoch": 0.40181226765799255, "grad_norm": 0.960355897949613, "learning_rate": 6.800758621949315e-06, "loss": 0.0697, "step": 8647 }, { "epoch": 0.4018587360594795, "grad_norm": 0.7438629024187741, "learning_rate": 6.800056586579687e-06, "loss": 0.0421, "step": 8648 }, { "epoch": 0.40190520446096656, "grad_norm": 0.9994742853426533, "learning_rate": 6.799354510436883e-06, "loss": 0.0708, "step": 8649 }, { "epoch": 0.40195167286245354, "grad_norm": 2.0978140985616998, "learning_rate": 6.798652393536804e-06, "loss": 0.0419, "step": 8650 }, { "epoch": 0.4019981412639405, "grad_norm": 0.8317358156833855, "learning_rate": 6.797950235895358e-06, "loss": 0.0605, "step": 8651 }, { "epoch": 0.4020446096654275, "grad_norm": 0.5511054536402158, "learning_rate": 6.797248037528443e-06, "loss": 0.0588, "step": 8652 }, { "epoch": 0.4020910780669145, "grad_norm": 0.8744732698169692, "learning_rate": 6.796545798451971e-06, "loss": 0.0687, "step": 8653 }, { "epoch": 0.4021375464684015, "grad_norm": 1.4750417289072193, "learning_rate": 6.795843518681846e-06, "loss": 0.0543, "step": 8654 }, { "epoch": 0.4021840148698885, "grad_norm": 0.7125579264806486, "learning_rate": 6.7951411982339735e-06, "loss": 0.0512, "step": 8655 }, { "epoch": 0.40223048327137545, "grad_norm": 0.720961206647958, "learning_rate": 6.794438837124264e-06, "loss": 0.0608, "step": 8656 }, { "epoch": 0.4022769516728624, "grad_norm": 1.565176256459706, "learning_rate": 6.793736435368628e-06, "loss": 0.1249, "step": 8657 }, { "epoch": 0.40232342007434946, "grad_norm": 1.3858692871298486, "learning_rate": 6.793033992982971e-06, "loss": 0.0659, "step": 8658 }, { "epoch": 0.40236988847583643, "grad_norm": 0.5609625413631282, "learning_rate": 6.792331509983209e-06, "loss": 0.0555, "step": 8659 }, { "epoch": 0.4024163568773234, "grad_norm": 0.862937008647731, "learning_rate": 6.791628986385251e-06, "loss": 0.0797, "step": 8660 }, { "epoch": 0.4024628252788104, "grad_norm": 0.8152317375609518, "learning_rate": 6.790926422205012e-06, "loss": 0.0598, "step": 8661 }, { "epoch": 0.4025092936802974, "grad_norm": 0.769074883197717, "learning_rate": 6.7902238174584025e-06, "loss": 0.0555, "step": 8662 }, { "epoch": 0.4025557620817844, "grad_norm": 0.5731058090948503, "learning_rate": 6.789521172161342e-06, "loss": 0.045, "step": 8663 }, { "epoch": 0.4026022304832714, "grad_norm": 1.1749607296852258, "learning_rate": 6.788818486329741e-06, "loss": 0.054, "step": 8664 }, { "epoch": 0.40264869888475835, "grad_norm": 0.6043130637633518, "learning_rate": 6.78811575997952e-06, "loss": 0.041, "step": 8665 }, { "epoch": 0.4026951672862453, "grad_norm": 1.2200398182833834, "learning_rate": 6.7874129931265955e-06, "loss": 0.0719, "step": 8666 }, { "epoch": 0.40274163568773236, "grad_norm": 0.9313264868451233, "learning_rate": 6.786710185786884e-06, "loss": 0.0432, "step": 8667 }, { "epoch": 0.40278810408921933, "grad_norm": 1.1216484404873317, "learning_rate": 6.786007337976307e-06, "loss": 0.08, "step": 8668 }, { "epoch": 0.4028345724907063, "grad_norm": 2.18437163146897, "learning_rate": 6.785304449710785e-06, "loss": 0.0733, "step": 8669 }, { "epoch": 0.4028810408921933, "grad_norm": 0.6496710159860537, "learning_rate": 6.784601521006238e-06, "loss": 0.0495, "step": 8670 }, { "epoch": 0.4029275092936803, "grad_norm": 0.7830987988767407, "learning_rate": 6.783898551878586e-06, "loss": 0.0588, "step": 8671 }, { "epoch": 0.4029739776951673, "grad_norm": 1.0726688686461991, "learning_rate": 6.783195542343757e-06, "loss": 0.0422, "step": 8672 }, { "epoch": 0.4030204460966543, "grad_norm": 1.0014466847059376, "learning_rate": 6.78249249241767e-06, "loss": 0.0683, "step": 8673 }, { "epoch": 0.40306691449814125, "grad_norm": 0.7762927508297175, "learning_rate": 6.781789402116253e-06, "loss": 0.0553, "step": 8674 }, { "epoch": 0.4031133828996282, "grad_norm": 0.51655525654287, "learning_rate": 6.781086271455433e-06, "loss": 0.0383, "step": 8675 }, { "epoch": 0.40315985130111526, "grad_norm": 0.9056247855049232, "learning_rate": 6.780383100451131e-06, "loss": 0.0737, "step": 8676 }, { "epoch": 0.40320631970260223, "grad_norm": 0.6273321450698314, "learning_rate": 6.779679889119279e-06, "loss": 0.0755, "step": 8677 }, { "epoch": 0.4032527881040892, "grad_norm": 2.3006458869658153, "learning_rate": 6.778976637475805e-06, "loss": 0.0711, "step": 8678 }, { "epoch": 0.4032992565055762, "grad_norm": 1.6669900202069712, "learning_rate": 6.778273345536637e-06, "loss": 0.0698, "step": 8679 }, { "epoch": 0.4033457249070632, "grad_norm": 0.487203575351218, "learning_rate": 6.777570013317707e-06, "loss": 0.0466, "step": 8680 }, { "epoch": 0.4033921933085502, "grad_norm": 0.6474067065382525, "learning_rate": 6.7768666408349445e-06, "loss": 0.0382, "step": 8681 }, { "epoch": 0.40343866171003717, "grad_norm": 0.665636077753171, "learning_rate": 6.776163228104283e-06, "loss": 0.0524, "step": 8682 }, { "epoch": 0.40348513011152415, "grad_norm": 1.025339826212401, "learning_rate": 6.775459775141655e-06, "loss": 0.0889, "step": 8683 }, { "epoch": 0.4035315985130111, "grad_norm": 6.998460194509693, "learning_rate": 6.774756281962994e-06, "loss": 0.0698, "step": 8684 }, { "epoch": 0.40357806691449816, "grad_norm": 0.6626334473198915, "learning_rate": 6.774052748584237e-06, "loss": 0.0672, "step": 8685 }, { "epoch": 0.40362453531598513, "grad_norm": 0.5451789144564658, "learning_rate": 6.773349175021316e-06, "loss": 0.0605, "step": 8686 }, { "epoch": 0.4036710037174721, "grad_norm": 1.0542927480758542, "learning_rate": 6.772645561290171e-06, "loss": 0.0728, "step": 8687 }, { "epoch": 0.4037174721189591, "grad_norm": 0.8962063702814793, "learning_rate": 6.771941907406738e-06, "loss": 0.0517, "step": 8688 }, { "epoch": 0.4037639405204461, "grad_norm": 1.1141682502154098, "learning_rate": 6.771238213386954e-06, "loss": 0.0487, "step": 8689 }, { "epoch": 0.4038104089219331, "grad_norm": 0.8972463963290448, "learning_rate": 6.7705344792467645e-06, "loss": 0.0551, "step": 8690 }, { "epoch": 0.40385687732342007, "grad_norm": 1.4541203855639002, "learning_rate": 6.769830705002101e-06, "loss": 0.0896, "step": 8691 }, { "epoch": 0.40390334572490705, "grad_norm": 0.6057941662827012, "learning_rate": 6.769126890668912e-06, "loss": 0.0467, "step": 8692 }, { "epoch": 0.403949814126394, "grad_norm": 0.6863286080393245, "learning_rate": 6.768423036263136e-06, "loss": 0.0483, "step": 8693 }, { "epoch": 0.40399628252788106, "grad_norm": 0.5050226269337803, "learning_rate": 6.767719141800718e-06, "loss": 0.0382, "step": 8694 }, { "epoch": 0.40404275092936803, "grad_norm": 0.866601818440805, "learning_rate": 6.7670152072976e-06, "loss": 0.0612, "step": 8695 }, { "epoch": 0.404089219330855, "grad_norm": 0.5353860373313364, "learning_rate": 6.7663112327697285e-06, "loss": 0.0451, "step": 8696 }, { "epoch": 0.404135687732342, "grad_norm": 0.6265760659581756, "learning_rate": 6.765607218233048e-06, "loss": 0.0548, "step": 8697 }, { "epoch": 0.404182156133829, "grad_norm": 0.5518101565269812, "learning_rate": 6.764903163703506e-06, "loss": 0.0461, "step": 8698 }, { "epoch": 0.404228624535316, "grad_norm": 0.5420478591030853, "learning_rate": 6.764199069197051e-06, "loss": 0.0537, "step": 8699 }, { "epoch": 0.40427509293680297, "grad_norm": 0.9499964414762022, "learning_rate": 6.7634949347296284e-06, "loss": 0.0707, "step": 8700 }, { "epoch": 0.40432156133828995, "grad_norm": 0.5922181324684057, "learning_rate": 6.76279076031719e-06, "loss": 0.0278, "step": 8701 }, { "epoch": 0.4043680297397769, "grad_norm": 1.1968198576618363, "learning_rate": 6.762086545975686e-06, "loss": 0.0445, "step": 8702 }, { "epoch": 0.40441449814126396, "grad_norm": 1.097227952257313, "learning_rate": 6.761382291721066e-06, "loss": 0.1096, "step": 8703 }, { "epoch": 0.40446096654275093, "grad_norm": 1.1322585072111933, "learning_rate": 6.760677997569285e-06, "loss": 0.0597, "step": 8704 }, { "epoch": 0.4045074349442379, "grad_norm": 1.067399418995206, "learning_rate": 6.759973663536293e-06, "loss": 0.0516, "step": 8705 }, { "epoch": 0.4045539033457249, "grad_norm": 0.6990598127577474, "learning_rate": 6.759269289638046e-06, "loss": 0.0483, "step": 8706 }, { "epoch": 0.4046003717472119, "grad_norm": 0.9054127525409291, "learning_rate": 6.758564875890499e-06, "loss": 0.0616, "step": 8707 }, { "epoch": 0.4046468401486989, "grad_norm": 0.7747244272912934, "learning_rate": 6.757860422309603e-06, "loss": 0.0604, "step": 8708 }, { "epoch": 0.40469330855018587, "grad_norm": 3.828242529533146, "learning_rate": 6.757155928911322e-06, "loss": 0.0558, "step": 8709 }, { "epoch": 0.40473977695167285, "grad_norm": 0.6770315317843412, "learning_rate": 6.756451395711608e-06, "loss": 0.0513, "step": 8710 }, { "epoch": 0.4047862453531598, "grad_norm": 0.6896421597514693, "learning_rate": 6.7557468227264224e-06, "loss": 0.0397, "step": 8711 }, { "epoch": 0.40483271375464686, "grad_norm": 0.7720929849965387, "learning_rate": 6.755042209971722e-06, "loss": 0.0592, "step": 8712 }, { "epoch": 0.40487918215613383, "grad_norm": 0.7008413134420481, "learning_rate": 6.75433755746347e-06, "loss": 0.054, "step": 8713 }, { "epoch": 0.4049256505576208, "grad_norm": 0.7664119420385029, "learning_rate": 6.753632865217625e-06, "loss": 0.0519, "step": 8714 }, { "epoch": 0.4049721189591078, "grad_norm": 2.1722806613867176, "learning_rate": 6.75292813325015e-06, "loss": 0.0772, "step": 8715 }, { "epoch": 0.4050185873605948, "grad_norm": 1.167598002216562, "learning_rate": 6.752223361577008e-06, "loss": 0.0695, "step": 8716 }, { "epoch": 0.4050650557620818, "grad_norm": 1.4703614481328553, "learning_rate": 6.751518550214162e-06, "loss": 0.0628, "step": 8717 }, { "epoch": 0.40511152416356877, "grad_norm": 0.8654537731043412, "learning_rate": 6.750813699177577e-06, "loss": 0.078, "step": 8718 }, { "epoch": 0.40515799256505575, "grad_norm": 0.6245399593100253, "learning_rate": 6.750108808483221e-06, "loss": 0.061, "step": 8719 }, { "epoch": 0.4052044609665427, "grad_norm": 2.0365061335117742, "learning_rate": 6.749403878147056e-06, "loss": 0.0822, "step": 8720 }, { "epoch": 0.40525092936802976, "grad_norm": 1.9683576975755128, "learning_rate": 6.748698908185054e-06, "loss": 0.0983, "step": 8721 }, { "epoch": 0.40529739776951673, "grad_norm": 1.909073878015544, "learning_rate": 6.747993898613181e-06, "loss": 0.0872, "step": 8722 }, { "epoch": 0.4053438661710037, "grad_norm": 0.651879557397825, "learning_rate": 6.7472888494474065e-06, "loss": 0.061, "step": 8723 }, { "epoch": 0.4053903345724907, "grad_norm": 0.5770892923565167, "learning_rate": 6.7465837607037e-06, "loss": 0.047, "step": 8724 }, { "epoch": 0.4054368029739777, "grad_norm": 0.9600589986546273, "learning_rate": 6.745878632398033e-06, "loss": 0.0483, "step": 8725 }, { "epoch": 0.4054832713754647, "grad_norm": 0.9921841153128084, "learning_rate": 6.745173464546378e-06, "loss": 0.0402, "step": 8726 }, { "epoch": 0.40552973977695167, "grad_norm": 1.5605493942127382, "learning_rate": 6.744468257164707e-06, "loss": 0.1227, "step": 8727 }, { "epoch": 0.40557620817843865, "grad_norm": 1.3979569349683025, "learning_rate": 6.743763010268995e-06, "loss": 0.0868, "step": 8728 }, { "epoch": 0.4056226765799256, "grad_norm": 1.0567805242782513, "learning_rate": 6.743057723875214e-06, "loss": 0.0741, "step": 8729 }, { "epoch": 0.40566914498141265, "grad_norm": 1.097071138587029, "learning_rate": 6.74235239799934e-06, "loss": 0.0718, "step": 8730 }, { "epoch": 0.40571561338289963, "grad_norm": 1.7654866655244494, "learning_rate": 6.741647032657354e-06, "loss": 0.0907, "step": 8731 }, { "epoch": 0.4057620817843866, "grad_norm": 1.0951510864247953, "learning_rate": 6.740941627865226e-06, "loss": 0.0613, "step": 8732 }, { "epoch": 0.4058085501858736, "grad_norm": 0.9132224053043584, "learning_rate": 6.74023618363894e-06, "loss": 0.0434, "step": 8733 }, { "epoch": 0.4058550185873606, "grad_norm": 1.078623852845318, "learning_rate": 6.739530699994472e-06, "loss": 0.0791, "step": 8734 }, { "epoch": 0.4059014869888476, "grad_norm": 1.7884367838397695, "learning_rate": 6.738825176947803e-06, "loss": 0.0716, "step": 8735 }, { "epoch": 0.40594795539033457, "grad_norm": 1.3490622456494257, "learning_rate": 6.738119614514913e-06, "loss": 0.0771, "step": 8736 }, { "epoch": 0.40599442379182155, "grad_norm": 2.6536686944050785, "learning_rate": 6.737414012711786e-06, "loss": 0.0603, "step": 8737 }, { "epoch": 0.4060408921933085, "grad_norm": 0.8644230334397198, "learning_rate": 6.736708371554403e-06, "loss": 0.0441, "step": 8738 }, { "epoch": 0.40608736059479555, "grad_norm": 0.6548143552310286, "learning_rate": 6.736002691058745e-06, "loss": 0.0406, "step": 8739 }, { "epoch": 0.40613382899628253, "grad_norm": 0.714230176533582, "learning_rate": 6.735296971240801e-06, "loss": 0.0715, "step": 8740 }, { "epoch": 0.4061802973977695, "grad_norm": 0.4833063953583637, "learning_rate": 6.734591212116555e-06, "loss": 0.0486, "step": 8741 }, { "epoch": 0.4062267657992565, "grad_norm": 0.760209932077742, "learning_rate": 6.733885413701991e-06, "loss": 0.0405, "step": 8742 }, { "epoch": 0.4062732342007435, "grad_norm": 1.1559644698948057, "learning_rate": 6.733179576013098e-06, "loss": 0.0612, "step": 8743 }, { "epoch": 0.4063197026022305, "grad_norm": 0.8462800214631255, "learning_rate": 6.732473699065863e-06, "loss": 0.0446, "step": 8744 }, { "epoch": 0.40636617100371747, "grad_norm": 0.739572726734749, "learning_rate": 6.731767782876275e-06, "loss": 0.0872, "step": 8745 }, { "epoch": 0.40641263940520445, "grad_norm": 1.1542895893389467, "learning_rate": 6.7310618274603255e-06, "loss": 0.0824, "step": 8746 }, { "epoch": 0.4064591078066915, "grad_norm": 0.7235002426572813, "learning_rate": 6.730355832834001e-06, "loss": 0.0491, "step": 8747 }, { "epoch": 0.40650557620817845, "grad_norm": 0.7889041198012131, "learning_rate": 6.729649799013298e-06, "loss": 0.0487, "step": 8748 }, { "epoch": 0.40655204460966543, "grad_norm": 0.7251757873341598, "learning_rate": 6.728943726014208e-06, "loss": 0.0722, "step": 8749 }, { "epoch": 0.4065985130111524, "grad_norm": 0.5351076683335115, "learning_rate": 6.728237613852721e-06, "loss": 0.0475, "step": 8750 }, { "epoch": 0.4066449814126394, "grad_norm": 0.7113380666604434, "learning_rate": 6.7275314625448315e-06, "loss": 0.0303, "step": 8751 }, { "epoch": 0.4066914498141264, "grad_norm": 0.5085267830657599, "learning_rate": 6.726825272106539e-06, "loss": 0.0425, "step": 8752 }, { "epoch": 0.4067379182156134, "grad_norm": 0.9816136165100701, "learning_rate": 6.726119042553836e-06, "loss": 0.0495, "step": 8753 }, { "epoch": 0.40678438661710037, "grad_norm": 1.119050137243449, "learning_rate": 6.725412773902717e-06, "loss": 0.0441, "step": 8754 }, { "epoch": 0.40683085501858735, "grad_norm": 0.46016424400669886, "learning_rate": 6.724706466169188e-06, "loss": 0.0447, "step": 8755 }, { "epoch": 0.4068773234200744, "grad_norm": 1.390802220802975, "learning_rate": 6.724000119369238e-06, "loss": 0.0681, "step": 8756 }, { "epoch": 0.40692379182156135, "grad_norm": 0.7280203626014548, "learning_rate": 6.7232937335188734e-06, "loss": 0.0613, "step": 8757 }, { "epoch": 0.40697026022304833, "grad_norm": 1.0879714459046657, "learning_rate": 6.722587308634091e-06, "loss": 0.0584, "step": 8758 }, { "epoch": 0.4070167286245353, "grad_norm": 1.3156009641150934, "learning_rate": 6.721880844730892e-06, "loss": 0.0589, "step": 8759 }, { "epoch": 0.4070631970260223, "grad_norm": 1.1291825236747661, "learning_rate": 6.72117434182528e-06, "loss": 0.0864, "step": 8760 }, { "epoch": 0.4071096654275093, "grad_norm": 1.2843622499963419, "learning_rate": 6.720467799933259e-06, "loss": 0.0707, "step": 8761 }, { "epoch": 0.4071561338289963, "grad_norm": 1.1845447967924478, "learning_rate": 6.719761219070832e-06, "loss": 0.0521, "step": 8762 }, { "epoch": 0.40720260223048327, "grad_norm": 1.1605388818527236, "learning_rate": 6.719054599254001e-06, "loss": 0.0504, "step": 8763 }, { "epoch": 0.40724907063197024, "grad_norm": 1.151682251428587, "learning_rate": 6.718347940498776e-06, "loss": 0.0942, "step": 8764 }, { "epoch": 0.4072955390334573, "grad_norm": 1.7705197170665306, "learning_rate": 6.717641242821161e-06, "loss": 0.0678, "step": 8765 }, { "epoch": 0.40734200743494425, "grad_norm": 0.9303469214215782, "learning_rate": 6.716934506237165e-06, "loss": 0.0543, "step": 8766 }, { "epoch": 0.40738847583643123, "grad_norm": 1.4221730931885523, "learning_rate": 6.7162277307627944e-06, "loss": 0.0729, "step": 8767 }, { "epoch": 0.4074349442379182, "grad_norm": 1.3551270972333134, "learning_rate": 6.715520916414058e-06, "loss": 0.0812, "step": 8768 }, { "epoch": 0.4074814126394052, "grad_norm": 0.9344377648060355, "learning_rate": 6.7148140632069695e-06, "loss": 0.0505, "step": 8769 }, { "epoch": 0.4075278810408922, "grad_norm": 1.4870158206126614, "learning_rate": 6.714107171157537e-06, "loss": 0.0875, "step": 8770 }, { "epoch": 0.4075743494423792, "grad_norm": 1.066901808240758, "learning_rate": 6.713400240281772e-06, "loss": 0.0611, "step": 8771 }, { "epoch": 0.40762081784386617, "grad_norm": 0.9109326012591941, "learning_rate": 6.712693270595691e-06, "loss": 0.0686, "step": 8772 }, { "epoch": 0.40766728624535314, "grad_norm": 0.6600176555876727, "learning_rate": 6.711986262115301e-06, "loss": 0.1048, "step": 8773 }, { "epoch": 0.4077137546468402, "grad_norm": 1.0152471995170855, "learning_rate": 6.711279214856623e-06, "loss": 0.0766, "step": 8774 }, { "epoch": 0.40776022304832715, "grad_norm": 0.9415806653849689, "learning_rate": 6.710572128835669e-06, "loss": 0.0718, "step": 8775 }, { "epoch": 0.40780669144981413, "grad_norm": 0.7962499748580018, "learning_rate": 6.709865004068455e-06, "loss": 0.0609, "step": 8776 }, { "epoch": 0.4078531598513011, "grad_norm": 0.33212271223366957, "learning_rate": 6.709157840571e-06, "loss": 0.0472, "step": 8777 }, { "epoch": 0.4078996282527881, "grad_norm": 1.0564470783206803, "learning_rate": 6.708450638359322e-06, "loss": 0.056, "step": 8778 }, { "epoch": 0.4079460966542751, "grad_norm": 0.7114694711534354, "learning_rate": 6.707743397449437e-06, "loss": 0.05, "step": 8779 }, { "epoch": 0.4079925650557621, "grad_norm": 2.4028147172486864, "learning_rate": 6.7070361178573675e-06, "loss": 0.0724, "step": 8780 }, { "epoch": 0.40803903345724907, "grad_norm": 1.0639256037368343, "learning_rate": 6.706328799599133e-06, "loss": 0.0624, "step": 8781 }, { "epoch": 0.40808550185873604, "grad_norm": 0.5278666497364675, "learning_rate": 6.705621442690757e-06, "loss": 0.0527, "step": 8782 }, { "epoch": 0.4081319702602231, "grad_norm": 1.5979157408989846, "learning_rate": 6.704914047148258e-06, "loss": 0.076, "step": 8783 }, { "epoch": 0.40817843866171005, "grad_norm": 0.5245217894343724, "learning_rate": 6.704206612987664e-06, "loss": 0.0449, "step": 8784 }, { "epoch": 0.40822490706319703, "grad_norm": 0.7203091539000062, "learning_rate": 6.703499140224994e-06, "loss": 0.03, "step": 8785 }, { "epoch": 0.408271375464684, "grad_norm": 0.8252470762586572, "learning_rate": 6.702791628876276e-06, "loss": 0.0662, "step": 8786 }, { "epoch": 0.408317843866171, "grad_norm": 0.6631884950890742, "learning_rate": 6.702084078957537e-06, "loss": 0.0545, "step": 8787 }, { "epoch": 0.408364312267658, "grad_norm": 2.4329040076056194, "learning_rate": 6.7013764904848e-06, "loss": 0.0976, "step": 8788 }, { "epoch": 0.408410780669145, "grad_norm": 0.6874629305304713, "learning_rate": 6.700668863474096e-06, "loss": 0.0604, "step": 8789 }, { "epoch": 0.40845724907063197, "grad_norm": 1.1497578904427055, "learning_rate": 6.699961197941453e-06, "loss": 0.0783, "step": 8790 }, { "epoch": 0.40850371747211894, "grad_norm": 0.792437433119869, "learning_rate": 6.6992534939029e-06, "loss": 0.0895, "step": 8791 }, { "epoch": 0.408550185873606, "grad_norm": 0.9851356304557745, "learning_rate": 6.698545751374465e-06, "loss": 0.061, "step": 8792 }, { "epoch": 0.40859665427509295, "grad_norm": 1.6544897245168737, "learning_rate": 6.697837970372182e-06, "loss": 0.0898, "step": 8793 }, { "epoch": 0.40864312267657993, "grad_norm": 0.7769805450740213, "learning_rate": 6.697130150912081e-06, "loss": 0.056, "step": 8794 }, { "epoch": 0.4086895910780669, "grad_norm": 0.6798531811870374, "learning_rate": 6.696422293010196e-06, "loss": 0.081, "step": 8795 }, { "epoch": 0.4087360594795539, "grad_norm": 0.5209631800601089, "learning_rate": 6.695714396682565e-06, "loss": 0.0352, "step": 8796 }, { "epoch": 0.4087825278810409, "grad_norm": 0.42794961649647284, "learning_rate": 6.6950064619452124e-06, "loss": 0.0329, "step": 8797 }, { "epoch": 0.4088289962825279, "grad_norm": 0.6856503766430914, "learning_rate": 6.6942984888141825e-06, "loss": 0.0464, "step": 8798 }, { "epoch": 0.40887546468401487, "grad_norm": 0.6937051218363792, "learning_rate": 6.6935904773055085e-06, "loss": 0.0802, "step": 8799 }, { "epoch": 0.40892193308550184, "grad_norm": 0.8293883493732933, "learning_rate": 6.692882427435226e-06, "loss": 0.0617, "step": 8800 }, { "epoch": 0.4089684014869889, "grad_norm": 1.07514364987995, "learning_rate": 6.692174339219376e-06, "loss": 0.0448, "step": 8801 }, { "epoch": 0.40901486988847585, "grad_norm": 1.0041989523840742, "learning_rate": 6.691466212673996e-06, "loss": 0.0764, "step": 8802 }, { "epoch": 0.40906133828996283, "grad_norm": 0.8002083918661876, "learning_rate": 6.690758047815127e-06, "loss": 0.0642, "step": 8803 }, { "epoch": 0.4091078066914498, "grad_norm": 1.9544146040928756, "learning_rate": 6.690049844658807e-06, "loss": 0.0733, "step": 8804 }, { "epoch": 0.4091542750929368, "grad_norm": 1.0848509749647712, "learning_rate": 6.68934160322108e-06, "loss": 0.0731, "step": 8805 }, { "epoch": 0.4092007434944238, "grad_norm": 0.7162180640618225, "learning_rate": 6.688633323517988e-06, "loss": 0.0737, "step": 8806 }, { "epoch": 0.4092472118959108, "grad_norm": 0.7157065413240199, "learning_rate": 6.687925005565573e-06, "loss": 0.0555, "step": 8807 }, { "epoch": 0.40929368029739777, "grad_norm": 0.7756518183850343, "learning_rate": 6.687216649379883e-06, "loss": 0.0602, "step": 8808 }, { "epoch": 0.40934014869888474, "grad_norm": 1.4287043044526002, "learning_rate": 6.6865082549769576e-06, "loss": 0.0659, "step": 8809 }, { "epoch": 0.4093866171003718, "grad_norm": 1.110607449972657, "learning_rate": 6.685799822372845e-06, "loss": 0.0705, "step": 8810 }, { "epoch": 0.40943308550185875, "grad_norm": 1.0989665464178986, "learning_rate": 6.685091351583593e-06, "loss": 0.0804, "step": 8811 }, { "epoch": 0.4094795539033457, "grad_norm": 1.2702952688801896, "learning_rate": 6.684382842625248e-06, "loss": 0.075, "step": 8812 }, { "epoch": 0.4095260223048327, "grad_norm": 1.1341453500749483, "learning_rate": 6.683674295513859e-06, "loss": 0.1, "step": 8813 }, { "epoch": 0.4095724907063197, "grad_norm": 0.6466580325085853, "learning_rate": 6.682965710265475e-06, "loss": 0.0496, "step": 8814 }, { "epoch": 0.4096189591078067, "grad_norm": 0.5773711087549023, "learning_rate": 6.682257086896148e-06, "loss": 0.073, "step": 8815 }, { "epoch": 0.4096654275092937, "grad_norm": 0.9078110128895619, "learning_rate": 6.681548425421924e-06, "loss": 0.0718, "step": 8816 }, { "epoch": 0.40971189591078067, "grad_norm": 0.959992842050317, "learning_rate": 6.6808397258588605e-06, "loss": 0.0786, "step": 8817 }, { "epoch": 0.40975836431226764, "grad_norm": 0.6984426913038106, "learning_rate": 6.680130988223009e-06, "loss": 0.04, "step": 8818 }, { "epoch": 0.4098048327137547, "grad_norm": 1.2894590826318453, "learning_rate": 6.679422212530421e-06, "loss": 0.0836, "step": 8819 }, { "epoch": 0.40985130111524165, "grad_norm": 0.6746191873142281, "learning_rate": 6.678713398797154e-06, "loss": 0.0639, "step": 8820 }, { "epoch": 0.4098977695167286, "grad_norm": 1.4783402121184568, "learning_rate": 6.67800454703926e-06, "loss": 0.0424, "step": 8821 }, { "epoch": 0.4099442379182156, "grad_norm": 0.5748668740478052, "learning_rate": 6.677295657272797e-06, "loss": 0.0522, "step": 8822 }, { "epoch": 0.4099907063197026, "grad_norm": 3.0588888697664256, "learning_rate": 6.676586729513823e-06, "loss": 0.0995, "step": 8823 }, { "epoch": 0.4100371747211896, "grad_norm": 1.1336740673990604, "learning_rate": 6.6758777637783934e-06, "loss": 0.0601, "step": 8824 }, { "epoch": 0.4100836431226766, "grad_norm": 1.1317450511583156, "learning_rate": 6.675168760082571e-06, "loss": 0.0683, "step": 8825 }, { "epoch": 0.41013011152416357, "grad_norm": 0.7450658371354655, "learning_rate": 6.674459718442411e-06, "loss": 0.0603, "step": 8826 }, { "epoch": 0.41017657992565054, "grad_norm": 1.0581315379333571, "learning_rate": 6.673750638873977e-06, "loss": 0.0604, "step": 8827 }, { "epoch": 0.4102230483271376, "grad_norm": 0.9089874778436601, "learning_rate": 6.673041521393329e-06, "loss": 0.0571, "step": 8828 }, { "epoch": 0.41026951672862455, "grad_norm": 1.478738993583951, "learning_rate": 6.672332366016529e-06, "loss": 0.0484, "step": 8829 }, { "epoch": 0.4103159851301115, "grad_norm": 0.750228528325516, "learning_rate": 6.6716231727596424e-06, "loss": 0.0397, "step": 8830 }, { "epoch": 0.4103624535315985, "grad_norm": 0.8475948902154599, "learning_rate": 6.67091394163873e-06, "loss": 0.0965, "step": 8831 }, { "epoch": 0.4104089219330855, "grad_norm": 0.647382278407686, "learning_rate": 6.67020467266986e-06, "loss": 0.0574, "step": 8832 }, { "epoch": 0.4104553903345725, "grad_norm": 0.9136121674292287, "learning_rate": 6.669495365869094e-06, "loss": 0.0564, "step": 8833 }, { "epoch": 0.4105018587360595, "grad_norm": 0.9692138677105013, "learning_rate": 6.6687860212525035e-06, "loss": 0.0775, "step": 8834 }, { "epoch": 0.41054832713754646, "grad_norm": 0.7326391848189738, "learning_rate": 6.668076638836152e-06, "loss": 0.063, "step": 8835 }, { "epoch": 0.41059479553903344, "grad_norm": 1.2787282238385616, "learning_rate": 6.667367218636108e-06, "loss": 0.0614, "step": 8836 }, { "epoch": 0.4106412639405205, "grad_norm": 0.6044423938590484, "learning_rate": 6.666657760668445e-06, "loss": 0.0337, "step": 8837 }, { "epoch": 0.41068773234200745, "grad_norm": 0.6513760250526303, "learning_rate": 6.6659482649492255e-06, "loss": 0.0666, "step": 8838 }, { "epoch": 0.4107342007434944, "grad_norm": 0.5763066639096214, "learning_rate": 6.665238731494527e-06, "loss": 0.0446, "step": 8839 }, { "epoch": 0.4107806691449814, "grad_norm": 1.6461272036065167, "learning_rate": 6.6645291603204175e-06, "loss": 0.0666, "step": 8840 }, { "epoch": 0.4108271375464684, "grad_norm": 0.5240953542641198, "learning_rate": 6.663819551442969e-06, "loss": 0.0509, "step": 8841 }, { "epoch": 0.4108736059479554, "grad_norm": 1.072013613524113, "learning_rate": 6.66310990487826e-06, "loss": 0.0657, "step": 8842 }, { "epoch": 0.4109200743494424, "grad_norm": 0.42216460031240954, "learning_rate": 6.66240022064236e-06, "loss": 0.0405, "step": 8843 }, { "epoch": 0.41096654275092936, "grad_norm": 1.3364711745336735, "learning_rate": 6.6616904987513455e-06, "loss": 0.0699, "step": 8844 }, { "epoch": 0.41101301115241634, "grad_norm": 2.4593417186441595, "learning_rate": 6.66098073922129e-06, "loss": 0.0814, "step": 8845 }, { "epoch": 0.4110594795539034, "grad_norm": 1.1585427044738656, "learning_rate": 6.660270942068275e-06, "loss": 0.0577, "step": 8846 }, { "epoch": 0.41110594795539035, "grad_norm": 2.4490256716462278, "learning_rate": 6.659561107308375e-06, "loss": 0.0463, "step": 8847 }, { "epoch": 0.4111524163568773, "grad_norm": 1.0277174975651802, "learning_rate": 6.65885123495767e-06, "loss": 0.0437, "step": 8848 }, { "epoch": 0.4111988847583643, "grad_norm": 0.5265915317201211, "learning_rate": 6.6581413250322395e-06, "loss": 0.0467, "step": 8849 }, { "epoch": 0.4112453531598513, "grad_norm": 0.8984610019301114, "learning_rate": 6.6574313775481605e-06, "loss": 0.0601, "step": 8850 }, { "epoch": 0.4112918215613383, "grad_norm": 0.5712079976052211, "learning_rate": 6.656721392521517e-06, "loss": 0.0414, "step": 8851 }, { "epoch": 0.4113382899628253, "grad_norm": 1.5012774489757965, "learning_rate": 6.65601136996839e-06, "loss": 0.0542, "step": 8852 }, { "epoch": 0.41138475836431226, "grad_norm": 1.3163386025139099, "learning_rate": 6.6553013099048625e-06, "loss": 0.0853, "step": 8853 }, { "epoch": 0.41143122676579924, "grad_norm": 0.8872239548208449, "learning_rate": 6.654591212347019e-06, "loss": 0.0555, "step": 8854 }, { "epoch": 0.4114776951672863, "grad_norm": 0.9402335805343845, "learning_rate": 6.653881077310944e-06, "loss": 0.073, "step": 8855 }, { "epoch": 0.41152416356877325, "grad_norm": 1.3385320144132897, "learning_rate": 6.65317090481272e-06, "loss": 0.0708, "step": 8856 }, { "epoch": 0.4115706319702602, "grad_norm": 0.6738735207681791, "learning_rate": 6.652460694868436e-06, "loss": 0.0556, "step": 8857 }, { "epoch": 0.4116171003717472, "grad_norm": 0.7564285365920346, "learning_rate": 6.651750447494178e-06, "loss": 0.0703, "step": 8858 }, { "epoch": 0.4116635687732342, "grad_norm": 0.6893841625054566, "learning_rate": 6.651040162706034e-06, "loss": 0.0545, "step": 8859 }, { "epoch": 0.4117100371747212, "grad_norm": 1.0445341681896083, "learning_rate": 6.650329840520091e-06, "loss": 0.068, "step": 8860 }, { "epoch": 0.4117565055762082, "grad_norm": 1.8183580796788472, "learning_rate": 6.649619480952444e-06, "loss": 0.0737, "step": 8861 }, { "epoch": 0.41180297397769516, "grad_norm": 0.853569958854042, "learning_rate": 6.648909084019176e-06, "loss": 0.0675, "step": 8862 }, { "epoch": 0.41184944237918214, "grad_norm": 0.8925630758082376, "learning_rate": 6.648198649736384e-06, "loss": 0.0652, "step": 8863 }, { "epoch": 0.41189591078066917, "grad_norm": 0.8925630758082376, "learning_rate": 6.648198649736384e-06, "loss": 0.0656, "step": 8864 }, { "epoch": 0.41194237918215615, "grad_norm": 0.6214579109376234, "learning_rate": 6.647488178120156e-06, "loss": 0.0612, "step": 8865 }, { "epoch": 0.4119888475836431, "grad_norm": 0.7491326998483229, "learning_rate": 6.6467776691865885e-06, "loss": 0.0603, "step": 8866 }, { "epoch": 0.4120353159851301, "grad_norm": 1.7848864566118587, "learning_rate": 6.6460671229517725e-06, "loss": 0.078, "step": 8867 }, { "epoch": 0.4120817843866171, "grad_norm": 1.1034536674448168, "learning_rate": 6.645356539431805e-06, "loss": 0.0673, "step": 8868 }, { "epoch": 0.4121282527881041, "grad_norm": 0.9358565943760448, "learning_rate": 6.644645918642778e-06, "loss": 0.0725, "step": 8869 }, { "epoch": 0.4121747211895911, "grad_norm": 0.45244203631866475, "learning_rate": 6.643935260600792e-06, "loss": 0.0428, "step": 8870 }, { "epoch": 0.41222118959107806, "grad_norm": 0.4778556882328261, "learning_rate": 6.6432245653219405e-06, "loss": 0.0352, "step": 8871 }, { "epoch": 0.41226765799256504, "grad_norm": 0.8661187080114091, "learning_rate": 6.642513832822324e-06, "loss": 0.0696, "step": 8872 }, { "epoch": 0.41231412639405207, "grad_norm": 0.6159705871596381, "learning_rate": 6.641803063118039e-06, "loss": 0.0497, "step": 8873 }, { "epoch": 0.41236059479553905, "grad_norm": 0.6426038166814818, "learning_rate": 6.64109225622519e-06, "loss": 0.0671, "step": 8874 }, { "epoch": 0.412407063197026, "grad_norm": 0.947161067312943, "learning_rate": 6.640381412159872e-06, "loss": 0.0764, "step": 8875 }, { "epoch": 0.412453531598513, "grad_norm": 0.5856612735852331, "learning_rate": 6.6396705309381895e-06, "loss": 0.071, "step": 8876 }, { "epoch": 0.4125, "grad_norm": 1.2851045098440692, "learning_rate": 6.638959612576243e-06, "loss": 0.0652, "step": 8877 }, { "epoch": 0.412546468401487, "grad_norm": 0.6076926883549709, "learning_rate": 6.638248657090136e-06, "loss": 0.0511, "step": 8878 }, { "epoch": 0.412592936802974, "grad_norm": 1.0171833459283721, "learning_rate": 6.637537664495973e-06, "loss": 0.0646, "step": 8879 }, { "epoch": 0.41263940520446096, "grad_norm": 0.8132168890876477, "learning_rate": 6.636826634809859e-06, "loss": 0.0656, "step": 8880 }, { "epoch": 0.41268587360594794, "grad_norm": 0.694200040309597, "learning_rate": 6.6361155680478985e-06, "loss": 0.0499, "step": 8881 }, { "epoch": 0.41273234200743497, "grad_norm": 1.2141209159399398, "learning_rate": 6.635404464226197e-06, "loss": 0.056, "step": 8882 }, { "epoch": 0.41277881040892195, "grad_norm": 0.8512387832657015, "learning_rate": 6.6346933233608655e-06, "loss": 0.0563, "step": 8883 }, { "epoch": 0.4128252788104089, "grad_norm": 1.1074737639056254, "learning_rate": 6.633982145468009e-06, "loss": 0.0688, "step": 8884 }, { "epoch": 0.4128717472118959, "grad_norm": 1.0129182159962842, "learning_rate": 6.6332709305637376e-06, "loss": 0.0496, "step": 8885 }, { "epoch": 0.4129182156133829, "grad_norm": 2.439840382048396, "learning_rate": 6.63255967866416e-06, "loss": 0.0317, "step": 8886 }, { "epoch": 0.4129646840148699, "grad_norm": 0.5961836364829782, "learning_rate": 6.631848389785387e-06, "loss": 0.0497, "step": 8887 }, { "epoch": 0.4130111524163569, "grad_norm": 0.46403301283899984, "learning_rate": 6.631137063943531e-06, "loss": 0.0454, "step": 8888 }, { "epoch": 0.41305762081784386, "grad_norm": 0.995815494506309, "learning_rate": 6.630425701154704e-06, "loss": 0.0778, "step": 8889 }, { "epoch": 0.41310408921933084, "grad_norm": 2.1512833404430314, "learning_rate": 6.629714301435017e-06, "loss": 0.0764, "step": 8890 }, { "epoch": 0.41315055762081787, "grad_norm": 0.47502294603804657, "learning_rate": 6.629002864800589e-06, "loss": 0.0497, "step": 8891 }, { "epoch": 0.41319702602230485, "grad_norm": 1.0186746732022767, "learning_rate": 6.62829139126753e-06, "loss": 0.041, "step": 8892 }, { "epoch": 0.4132434944237918, "grad_norm": 0.8952608309603136, "learning_rate": 6.6275798808519565e-06, "loss": 0.0754, "step": 8893 }, { "epoch": 0.4132899628252788, "grad_norm": 0.9846876019706687, "learning_rate": 6.6268683335699865e-06, "loss": 0.057, "step": 8894 }, { "epoch": 0.4133364312267658, "grad_norm": 1.502124839043808, "learning_rate": 6.626156749437736e-06, "loss": 0.0537, "step": 8895 }, { "epoch": 0.4133828996282528, "grad_norm": 3.089147238623831, "learning_rate": 6.625445128471324e-06, "loss": 0.11, "step": 8896 }, { "epoch": 0.4134293680297398, "grad_norm": 1.0483699173555434, "learning_rate": 6.624733470686871e-06, "loss": 0.0356, "step": 8897 }, { "epoch": 0.41347583643122676, "grad_norm": 2.108144344592648, "learning_rate": 6.624021776100492e-06, "loss": 0.1002, "step": 8898 }, { "epoch": 0.41352230483271374, "grad_norm": 0.9812921911515655, "learning_rate": 6.6233100447283105e-06, "loss": 0.0449, "step": 8899 }, { "epoch": 0.41356877323420077, "grad_norm": 0.7836427947700362, "learning_rate": 6.622598276586449e-06, "loss": 0.0768, "step": 8900 }, { "epoch": 0.41361524163568775, "grad_norm": 1.2389026468227646, "learning_rate": 6.6218864716910305e-06, "loss": 0.0489, "step": 8901 }, { "epoch": 0.4136617100371747, "grad_norm": 0.6206726298735977, "learning_rate": 6.621174630058173e-06, "loss": 0.0501, "step": 8902 }, { "epoch": 0.4137081784386617, "grad_norm": 0.8180686829000346, "learning_rate": 6.620462751704008e-06, "loss": 0.0682, "step": 8903 }, { "epoch": 0.4137546468401487, "grad_norm": 0.7880660210858456, "learning_rate": 6.619750836644653e-06, "loss": 0.0515, "step": 8904 }, { "epoch": 0.4138011152416357, "grad_norm": 0.5645752964040803, "learning_rate": 6.619038884896238e-06, "loss": 0.0433, "step": 8905 }, { "epoch": 0.4138475836431227, "grad_norm": 1.135729146631436, "learning_rate": 6.618326896474887e-06, "loss": 0.0654, "step": 8906 }, { "epoch": 0.41389405204460966, "grad_norm": 0.5146165584782804, "learning_rate": 6.617614871396729e-06, "loss": 0.0622, "step": 8907 }, { "epoch": 0.41394052044609664, "grad_norm": 0.764673990509553, "learning_rate": 6.616902809677892e-06, "loss": 0.0457, "step": 8908 }, { "epoch": 0.41398698884758367, "grad_norm": 0.6346433396259427, "learning_rate": 6.6161907113345045e-06, "loss": 0.0378, "step": 8909 }, { "epoch": 0.41403345724907065, "grad_norm": 1.0868530502254097, "learning_rate": 6.615478576382698e-06, "loss": 0.0729, "step": 8910 }, { "epoch": 0.4140799256505576, "grad_norm": 1.614292380703436, "learning_rate": 6.6147664048385985e-06, "loss": 0.067, "step": 8911 }, { "epoch": 0.4141263940520446, "grad_norm": 1.737653551701298, "learning_rate": 6.614054196718343e-06, "loss": 0.0724, "step": 8912 }, { "epoch": 0.4141728624535316, "grad_norm": 0.8196202833867993, "learning_rate": 6.613341952038059e-06, "loss": 0.0596, "step": 8913 }, { "epoch": 0.4142193308550186, "grad_norm": 0.7758116012232977, "learning_rate": 6.612629670813883e-06, "loss": 0.0644, "step": 8914 }, { "epoch": 0.4142657992565056, "grad_norm": 0.6377192785181097, "learning_rate": 6.6119173530619485e-06, "loss": 0.0818, "step": 8915 }, { "epoch": 0.41431226765799256, "grad_norm": 0.5814352096219334, "learning_rate": 6.611204998798389e-06, "loss": 0.058, "step": 8916 }, { "epoch": 0.41435873605947954, "grad_norm": 0.9186870818653076, "learning_rate": 6.6104926080393395e-06, "loss": 0.0703, "step": 8917 }, { "epoch": 0.41440520446096657, "grad_norm": 0.8111028869269459, "learning_rate": 6.609780180800938e-06, "loss": 0.0718, "step": 8918 }, { "epoch": 0.41445167286245355, "grad_norm": 0.6652729319774102, "learning_rate": 6.60906771709932e-06, "loss": 0.0215, "step": 8919 }, { "epoch": 0.4144981412639405, "grad_norm": 0.7662289564974025, "learning_rate": 6.608355216950626e-06, "loss": 0.0585, "step": 8920 }, { "epoch": 0.4145446096654275, "grad_norm": 1.4474264147427345, "learning_rate": 6.6076426803709935e-06, "loss": 0.089, "step": 8921 }, { "epoch": 0.4145910780669145, "grad_norm": 10.825381478018436, "learning_rate": 6.606930107376562e-06, "loss": 0.0491, "step": 8922 }, { "epoch": 0.4146375464684015, "grad_norm": 0.6987356576213846, "learning_rate": 6.606217497983473e-06, "loss": 0.0707, "step": 8923 }, { "epoch": 0.4146840148698885, "grad_norm": 1.7970055535448022, "learning_rate": 6.605504852207866e-06, "loss": 0.0762, "step": 8924 }, { "epoch": 0.41473048327137546, "grad_norm": 1.0245641690753056, "learning_rate": 6.604792170065886e-06, "loss": 0.0687, "step": 8925 }, { "epoch": 0.41477695167286244, "grad_norm": 1.2656417074887025, "learning_rate": 6.604079451573672e-06, "loss": 0.0802, "step": 8926 }, { "epoch": 0.41482342007434947, "grad_norm": 0.4869853718001164, "learning_rate": 6.603366696747373e-06, "loss": 0.0378, "step": 8927 }, { "epoch": 0.41486988847583645, "grad_norm": 2.393740316293032, "learning_rate": 6.602653905603128e-06, "loss": 0.0844, "step": 8928 }, { "epoch": 0.4149163568773234, "grad_norm": 0.8895936137985672, "learning_rate": 6.601941078157087e-06, "loss": 0.0624, "step": 8929 }, { "epoch": 0.4149628252788104, "grad_norm": 0.7705805183657327, "learning_rate": 6.601228214425394e-06, "loss": 0.0702, "step": 8930 }, { "epoch": 0.4150092936802974, "grad_norm": 1.519720856481837, "learning_rate": 6.600515314424197e-06, "loss": 0.0549, "step": 8931 }, { "epoch": 0.4150557620817844, "grad_norm": 0.6333242848267948, "learning_rate": 6.599802378169643e-06, "loss": 0.0571, "step": 8932 }, { "epoch": 0.4151022304832714, "grad_norm": 1.073295264240329, "learning_rate": 6.599089405677881e-06, "loss": 0.069, "step": 8933 }, { "epoch": 0.41514869888475836, "grad_norm": 1.4161582932044299, "learning_rate": 6.5983763969650625e-06, "loss": 0.0975, "step": 8934 }, { "epoch": 0.41519516728624534, "grad_norm": 0.9876742743283399, "learning_rate": 6.597663352047334e-06, "loss": 0.0622, "step": 8935 }, { "epoch": 0.41524163568773237, "grad_norm": 0.5513453537072158, "learning_rate": 6.596950270940852e-06, "loss": 0.0677, "step": 8936 }, { "epoch": 0.41528810408921935, "grad_norm": 1.4494424851218088, "learning_rate": 6.596237153661764e-06, "loss": 0.1039, "step": 8937 }, { "epoch": 0.4153345724907063, "grad_norm": 0.48226402168511645, "learning_rate": 6.5955240002262235e-06, "loss": 0.0416, "step": 8938 }, { "epoch": 0.4153810408921933, "grad_norm": 0.6460623514856373, "learning_rate": 6.594810810650388e-06, "loss": 0.0657, "step": 8939 }, { "epoch": 0.4154275092936803, "grad_norm": 0.6484812483553936, "learning_rate": 6.594097584950406e-06, "loss": 0.0578, "step": 8940 }, { "epoch": 0.4154739776951673, "grad_norm": 0.9456442327824433, "learning_rate": 6.593384323142437e-06, "loss": 0.0591, "step": 8941 }, { "epoch": 0.4155204460966543, "grad_norm": 1.066174098860619, "learning_rate": 6.592671025242638e-06, "loss": 0.0756, "step": 8942 }, { "epoch": 0.41556691449814126, "grad_norm": 0.8283145900428756, "learning_rate": 6.5919576912671615e-06, "loss": 0.0635, "step": 8943 }, { "epoch": 0.41561338289962824, "grad_norm": 1.184981184161275, "learning_rate": 6.591244321232171e-06, "loss": 0.076, "step": 8944 }, { "epoch": 0.41565985130111527, "grad_norm": 1.072692016774812, "learning_rate": 6.590530915153818e-06, "loss": 0.0973, "step": 8945 }, { "epoch": 0.41570631970260224, "grad_norm": 1.1026895673636903, "learning_rate": 6.5898174730482675e-06, "loss": 0.0637, "step": 8946 }, { "epoch": 0.4157527881040892, "grad_norm": 0.39836980897362245, "learning_rate": 6.589103994931678e-06, "loss": 0.0557, "step": 8947 }, { "epoch": 0.4157992565055762, "grad_norm": 1.3743961027890659, "learning_rate": 6.5883904808202105e-06, "loss": 0.0551, "step": 8948 }, { "epoch": 0.4158457249070632, "grad_norm": 0.6341718019943813, "learning_rate": 6.587676930730027e-06, "loss": 0.0543, "step": 8949 }, { "epoch": 0.4158921933085502, "grad_norm": 1.004186897485305, "learning_rate": 6.58696334467729e-06, "loss": 0.0758, "step": 8950 }, { "epoch": 0.4159386617100372, "grad_norm": 0.8518794434722925, "learning_rate": 6.586249722678162e-06, "loss": 0.0923, "step": 8951 }, { "epoch": 0.41598513011152416, "grad_norm": 0.6218891954687765, "learning_rate": 6.585536064748809e-06, "loss": 0.0589, "step": 8952 }, { "epoch": 0.41603159851301114, "grad_norm": 0.9265676606394369, "learning_rate": 6.584822370905395e-06, "loss": 0.0505, "step": 8953 }, { "epoch": 0.41607806691449817, "grad_norm": 0.597776495764822, "learning_rate": 6.584108641164088e-06, "loss": 0.0413, "step": 8954 }, { "epoch": 0.41612453531598514, "grad_norm": 1.1164718036203507, "learning_rate": 6.5833948755410504e-06, "loss": 0.0667, "step": 8955 }, { "epoch": 0.4161710037174721, "grad_norm": 1.3032626253831752, "learning_rate": 6.582681074052455e-06, "loss": 0.0795, "step": 8956 }, { "epoch": 0.4162174721189591, "grad_norm": 0.6260699178798038, "learning_rate": 6.581967236714466e-06, "loss": 0.0683, "step": 8957 }, { "epoch": 0.4162639405204461, "grad_norm": 0.6539584707789441, "learning_rate": 6.5812533635432525e-06, "loss": 0.0561, "step": 8958 }, { "epoch": 0.4163104089219331, "grad_norm": 0.7944638146257805, "learning_rate": 6.5805394545549905e-06, "loss": 0.0761, "step": 8959 }, { "epoch": 0.4163568773234201, "grad_norm": 1.5908764556209394, "learning_rate": 6.5798255097658425e-06, "loss": 0.0815, "step": 8960 }, { "epoch": 0.41640334572490706, "grad_norm": 0.8931331783705345, "learning_rate": 6.579111529191986e-06, "loss": 0.0602, "step": 8961 }, { "epoch": 0.41644981412639404, "grad_norm": 0.35354107599018414, "learning_rate": 6.57839751284959e-06, "loss": 0.0222, "step": 8962 }, { "epoch": 0.41649628252788107, "grad_norm": 0.6813182167725121, "learning_rate": 6.57768346075483e-06, "loss": 0.057, "step": 8963 }, { "epoch": 0.41654275092936804, "grad_norm": 1.4843110649997382, "learning_rate": 6.5769693729238785e-06, "loss": 0.0672, "step": 8964 }, { "epoch": 0.416589219330855, "grad_norm": 0.7231746627894263, "learning_rate": 6.576255249372911e-06, "loss": 0.0678, "step": 8965 }, { "epoch": 0.416635687732342, "grad_norm": 0.7164326227732309, "learning_rate": 6.575541090118105e-06, "loss": 0.0553, "step": 8966 }, { "epoch": 0.416682156133829, "grad_norm": 1.9884281095820153, "learning_rate": 6.574826895175632e-06, "loss": 0.1051, "step": 8967 }, { "epoch": 0.416728624535316, "grad_norm": 2.1048495629536896, "learning_rate": 6.574112664561676e-06, "loss": 0.0815, "step": 8968 }, { "epoch": 0.416775092936803, "grad_norm": 0.9769122210435128, "learning_rate": 6.5733983982924076e-06, "loss": 0.046, "step": 8969 }, { "epoch": 0.41682156133828996, "grad_norm": 0.5321142963182539, "learning_rate": 6.572684096384011e-06, "loss": 0.0504, "step": 8970 }, { "epoch": 0.41686802973977694, "grad_norm": 1.102626847300804, "learning_rate": 6.571969758852665e-06, "loss": 0.0397, "step": 8971 }, { "epoch": 0.41691449814126397, "grad_norm": 0.8149068959260943, "learning_rate": 6.571255385714549e-06, "loss": 0.0625, "step": 8972 }, { "epoch": 0.41696096654275094, "grad_norm": 0.531746525608324, "learning_rate": 6.570540976985845e-06, "loss": 0.0481, "step": 8973 }, { "epoch": 0.4170074349442379, "grad_norm": 0.752711610993642, "learning_rate": 6.569826532682736e-06, "loss": 0.0308, "step": 8974 }, { "epoch": 0.4170539033457249, "grad_norm": 1.3739329876439836, "learning_rate": 6.569112052821401e-06, "loss": 0.0868, "step": 8975 }, { "epoch": 0.4171003717472119, "grad_norm": 1.0876563041999232, "learning_rate": 6.568397537418029e-06, "loss": 0.08, "step": 8976 }, { "epoch": 0.4171468401486989, "grad_norm": 0.9070943427437114, "learning_rate": 6.567682986488801e-06, "loss": 0.0547, "step": 8977 }, { "epoch": 0.4171933085501859, "grad_norm": 0.6467962078501399, "learning_rate": 6.566968400049904e-06, "loss": 0.0509, "step": 8978 }, { "epoch": 0.41723977695167286, "grad_norm": 1.2654891707847902, "learning_rate": 6.566253778117523e-06, "loss": 0.0722, "step": 8979 }, { "epoch": 0.41728624535315983, "grad_norm": 0.9956351984907517, "learning_rate": 6.565539120707846e-06, "loss": 0.0657, "step": 8980 }, { "epoch": 0.41733271375464687, "grad_norm": 0.6954473438463586, "learning_rate": 6.56482442783706e-06, "loss": 0.0534, "step": 8981 }, { "epoch": 0.41737918215613384, "grad_norm": 0.6879491733111905, "learning_rate": 6.564109699521352e-06, "loss": 0.0828, "step": 8982 }, { "epoch": 0.4174256505576208, "grad_norm": 1.2614643242376264, "learning_rate": 6.563394935776917e-06, "loss": 0.0752, "step": 8983 }, { "epoch": 0.4174721189591078, "grad_norm": 0.7209965039865814, "learning_rate": 6.562680136619939e-06, "loss": 0.0813, "step": 8984 }, { "epoch": 0.4175185873605948, "grad_norm": 0.7021332891956471, "learning_rate": 6.56196530206661e-06, "loss": 0.0486, "step": 8985 }, { "epoch": 0.4175650557620818, "grad_norm": 0.6725259734639523, "learning_rate": 6.561250432133125e-06, "loss": 0.0685, "step": 8986 }, { "epoch": 0.4176115241635688, "grad_norm": 0.7031168798017735, "learning_rate": 6.560535526835673e-06, "loss": 0.0632, "step": 8987 }, { "epoch": 0.41765799256505576, "grad_norm": 0.6596056524077041, "learning_rate": 6.5598205861904505e-06, "loss": 0.0496, "step": 8988 }, { "epoch": 0.41770446096654273, "grad_norm": 0.601267069700117, "learning_rate": 6.55910561021365e-06, "loss": 0.0601, "step": 8989 }, { "epoch": 0.41775092936802977, "grad_norm": 1.304400164730587, "learning_rate": 6.558390598921465e-06, "loss": 0.0869, "step": 8990 }, { "epoch": 0.41779739776951674, "grad_norm": 0.9390971922464549, "learning_rate": 6.557675552330093e-06, "loss": 0.05, "step": 8991 }, { "epoch": 0.4178438661710037, "grad_norm": 0.8440536644213065, "learning_rate": 6.556960470455731e-06, "loss": 0.0361, "step": 8992 }, { "epoch": 0.4178903345724907, "grad_norm": 0.8430137786623683, "learning_rate": 6.556245353314575e-06, "loss": 0.0489, "step": 8993 }, { "epoch": 0.4179368029739777, "grad_norm": 0.577919448464202, "learning_rate": 6.555530200922824e-06, "loss": 0.0641, "step": 8994 }, { "epoch": 0.4179832713754647, "grad_norm": 0.8026992299612341, "learning_rate": 6.554815013296678e-06, "loss": 0.0582, "step": 8995 }, { "epoch": 0.4180297397769517, "grad_norm": 1.0620265642842506, "learning_rate": 6.554099790452334e-06, "loss": 0.0648, "step": 8996 }, { "epoch": 0.41807620817843866, "grad_norm": 0.617401499681276, "learning_rate": 6.553384532405996e-06, "loss": 0.0476, "step": 8997 }, { "epoch": 0.41812267657992563, "grad_norm": 0.9009001835183845, "learning_rate": 6.552669239173861e-06, "loss": 0.0699, "step": 8998 }, { "epoch": 0.41816914498141267, "grad_norm": 0.5827943843513825, "learning_rate": 6.551953910772134e-06, "loss": 0.0579, "step": 8999 }, { "epoch": 0.41821561338289964, "grad_norm": 0.6387650386362043, "learning_rate": 6.5512385472170185e-06, "loss": 0.0526, "step": 9000 }, { "epoch": 0.4182620817843866, "grad_norm": 0.6943210842091443, "learning_rate": 6.5505231485247165e-06, "loss": 0.065, "step": 9001 }, { "epoch": 0.4183085501858736, "grad_norm": 0.9556122793050904, "learning_rate": 6.549807714711435e-06, "loss": 0.0653, "step": 9002 }, { "epoch": 0.41835501858736057, "grad_norm": 0.6433194839292142, "learning_rate": 6.549092245793376e-06, "loss": 0.0558, "step": 9003 }, { "epoch": 0.4184014869888476, "grad_norm": 0.6215198193390807, "learning_rate": 6.548376741786747e-06, "loss": 0.0659, "step": 9004 }, { "epoch": 0.4184479553903346, "grad_norm": 0.5454188612793843, "learning_rate": 6.547661202707756e-06, "loss": 0.0719, "step": 9005 }, { "epoch": 0.41849442379182156, "grad_norm": 0.9956308537959718, "learning_rate": 6.546945628572609e-06, "loss": 0.0637, "step": 9006 }, { "epoch": 0.41854089219330853, "grad_norm": 1.0209612693864445, "learning_rate": 6.546230019397517e-06, "loss": 0.048, "step": 9007 }, { "epoch": 0.41858736059479557, "grad_norm": 0.8775488946126195, "learning_rate": 6.545514375198686e-06, "loss": 0.0599, "step": 9008 }, { "epoch": 0.41863382899628254, "grad_norm": 0.5291733691034957, "learning_rate": 6.544798695992329e-06, "loss": 0.0371, "step": 9009 }, { "epoch": 0.4186802973977695, "grad_norm": 4.027866883287285, "learning_rate": 6.544082981794655e-06, "loss": 0.0814, "step": 9010 }, { "epoch": 0.4187267657992565, "grad_norm": 0.6653091356171249, "learning_rate": 6.543367232621876e-06, "loss": 0.0494, "step": 9011 }, { "epoch": 0.41877323420074347, "grad_norm": 0.43176401613997034, "learning_rate": 6.542651448490207e-06, "loss": 0.0424, "step": 9012 }, { "epoch": 0.4188197026022305, "grad_norm": 1.4773932439751352, "learning_rate": 6.541935629415856e-06, "loss": 0.0678, "step": 9013 }, { "epoch": 0.4188661710037175, "grad_norm": 0.679340759061882, "learning_rate": 6.541219775415042e-06, "loss": 0.039, "step": 9014 }, { "epoch": 0.41891263940520446, "grad_norm": 0.9149877459294022, "learning_rate": 6.540503886503977e-06, "loss": 0.06, "step": 9015 }, { "epoch": 0.41895910780669143, "grad_norm": 1.860768228649243, "learning_rate": 6.539787962698878e-06, "loss": 0.0781, "step": 9016 }, { "epoch": 0.41900557620817847, "grad_norm": 0.7737374727796879, "learning_rate": 6.539072004015962e-06, "loss": 0.0642, "step": 9017 }, { "epoch": 0.41905204460966544, "grad_norm": 0.8086198029767393, "learning_rate": 6.538356010471445e-06, "loss": 0.064, "step": 9018 }, { "epoch": 0.4190985130111524, "grad_norm": 1.3654234698381944, "learning_rate": 6.537639982081544e-06, "loss": 0.0674, "step": 9019 }, { "epoch": 0.4191449814126394, "grad_norm": 0.6012665143195199, "learning_rate": 6.536923918862479e-06, "loss": 0.0642, "step": 9020 }, { "epoch": 0.41919144981412637, "grad_norm": 1.0738947039612616, "learning_rate": 6.536207820830471e-06, "loss": 0.0749, "step": 9021 }, { "epoch": 0.4192379182156134, "grad_norm": 1.0738947039612616, "learning_rate": 6.536207820830471e-06, "loss": 0.0447, "step": 9022 }, { "epoch": 0.4192843866171004, "grad_norm": 1.1000332049106458, "learning_rate": 6.535491688001738e-06, "loss": 0.0474, "step": 9023 }, { "epoch": 0.41933085501858736, "grad_norm": 0.6370600534662743, "learning_rate": 6.534775520392502e-06, "loss": 0.0737, "step": 9024 }, { "epoch": 0.41937732342007433, "grad_norm": 1.0405287248225463, "learning_rate": 6.534059318018988e-06, "loss": 0.06, "step": 9025 }, { "epoch": 0.41942379182156136, "grad_norm": 2.3602564414693776, "learning_rate": 6.533343080897412e-06, "loss": 0.0643, "step": 9026 }, { "epoch": 0.41947026022304834, "grad_norm": 1.0925767803388022, "learning_rate": 6.5326268090440035e-06, "loss": 0.0601, "step": 9027 }, { "epoch": 0.4195167286245353, "grad_norm": 0.5235874611071517, "learning_rate": 6.531910502474985e-06, "loss": 0.0725, "step": 9028 }, { "epoch": 0.4195631970260223, "grad_norm": 2.475266267611144, "learning_rate": 6.53119416120658e-06, "loss": 0.0753, "step": 9029 }, { "epoch": 0.41960966542750927, "grad_norm": 0.5899692703965868, "learning_rate": 6.530477785255016e-06, "loss": 0.0435, "step": 9030 }, { "epoch": 0.4196561338289963, "grad_norm": 1.3646572064650766, "learning_rate": 6.529761374636521e-06, "loss": 0.0857, "step": 9031 }, { "epoch": 0.4197026022304833, "grad_norm": 0.7423566748551156, "learning_rate": 6.529044929367319e-06, "loss": 0.0932, "step": 9032 }, { "epoch": 0.41974907063197026, "grad_norm": 1.357282980170288, "learning_rate": 6.528328449463639e-06, "loss": 0.0813, "step": 9033 }, { "epoch": 0.41979553903345723, "grad_norm": 1.032603720583424, "learning_rate": 6.527611934941713e-06, "loss": 0.0699, "step": 9034 }, { "epoch": 0.41984200743494426, "grad_norm": 1.2690179948014761, "learning_rate": 6.5268953858177695e-06, "loss": 0.0956, "step": 9035 }, { "epoch": 0.41988847583643124, "grad_norm": 0.565341177695476, "learning_rate": 6.526178802108036e-06, "loss": 0.05, "step": 9036 }, { "epoch": 0.4199349442379182, "grad_norm": 1.2996115074450814, "learning_rate": 6.52546218382875e-06, "loss": 0.0789, "step": 9037 }, { "epoch": 0.4199814126394052, "grad_norm": 0.825060438563881, "learning_rate": 6.524745530996137e-06, "loss": 0.0817, "step": 9038 }, { "epoch": 0.42002788104089217, "grad_norm": 1.325523265001425, "learning_rate": 6.524028843626433e-06, "loss": 0.0837, "step": 9039 }, { "epoch": 0.4200743494423792, "grad_norm": 0.6210679628239854, "learning_rate": 6.523312121735874e-06, "loss": 0.0528, "step": 9040 }, { "epoch": 0.4201208178438662, "grad_norm": 1.2575789569582378, "learning_rate": 6.522595365340689e-06, "loss": 0.0828, "step": 9041 }, { "epoch": 0.42016728624535316, "grad_norm": 1.6056462046199094, "learning_rate": 6.521878574457119e-06, "loss": 0.0527, "step": 9042 }, { "epoch": 0.42021375464684013, "grad_norm": 0.7778745183123026, "learning_rate": 6.521161749101397e-06, "loss": 0.0627, "step": 9043 }, { "epoch": 0.42026022304832716, "grad_norm": 0.8990633901110028, "learning_rate": 6.520444889289761e-06, "loss": 0.0627, "step": 9044 }, { "epoch": 0.42030669144981414, "grad_norm": 1.349379670998768, "learning_rate": 6.519727995038445e-06, "loss": 0.0719, "step": 9045 }, { "epoch": 0.4203531598513011, "grad_norm": 0.8601178610401357, "learning_rate": 6.519011066363694e-06, "loss": 0.0471, "step": 9046 }, { "epoch": 0.4203996282527881, "grad_norm": 0.6949905983713309, "learning_rate": 6.518294103281741e-06, "loss": 0.0804, "step": 9047 }, { "epoch": 0.42044609665427507, "grad_norm": 0.7387646184955362, "learning_rate": 6.5175771058088296e-06, "loss": 0.0545, "step": 9048 }, { "epoch": 0.4204925650557621, "grad_norm": 0.7009890148346452, "learning_rate": 6.516860073961201e-06, "loss": 0.0556, "step": 9049 }, { "epoch": 0.4205390334572491, "grad_norm": 0.6488231222786023, "learning_rate": 6.5161430077550935e-06, "loss": 0.0456, "step": 9050 }, { "epoch": 0.42058550185873605, "grad_norm": 1.8771822284991384, "learning_rate": 6.515425907206752e-06, "loss": 0.079, "step": 9051 }, { "epoch": 0.42063197026022303, "grad_norm": 1.4457370267118315, "learning_rate": 6.514708772332419e-06, "loss": 0.0529, "step": 9052 }, { "epoch": 0.42067843866171006, "grad_norm": 0.9866607406004789, "learning_rate": 6.513991603148337e-06, "loss": 0.0637, "step": 9053 }, { "epoch": 0.42072490706319704, "grad_norm": 1.6056724985999693, "learning_rate": 6.513274399670752e-06, "loss": 0.0537, "step": 9054 }, { "epoch": 0.420771375464684, "grad_norm": 0.5879182146863332, "learning_rate": 6.512557161915909e-06, "loss": 0.0384, "step": 9055 }, { "epoch": 0.420817843866171, "grad_norm": 0.5794250779147846, "learning_rate": 6.511839889900055e-06, "loss": 0.029, "step": 9056 }, { "epoch": 0.42086431226765797, "grad_norm": 0.5761168023859831, "learning_rate": 6.511122583639435e-06, "loss": 0.0375, "step": 9057 }, { "epoch": 0.420910780669145, "grad_norm": 3.8535822114041722, "learning_rate": 6.510405243150299e-06, "loss": 0.0731, "step": 9058 }, { "epoch": 0.420957249070632, "grad_norm": 2.0732739971816705, "learning_rate": 6.5096878684488954e-06, "loss": 0.0723, "step": 9059 }, { "epoch": 0.42100371747211895, "grad_norm": 1.5436516291724394, "learning_rate": 6.508970459551471e-06, "loss": 0.0682, "step": 9060 }, { "epoch": 0.42105018587360593, "grad_norm": 0.9612915161760364, "learning_rate": 6.50825301647428e-06, "loss": 0.0604, "step": 9061 }, { "epoch": 0.42109665427509296, "grad_norm": 0.672200726862781, "learning_rate": 6.507535539233568e-06, "loss": 0.0463, "step": 9062 }, { "epoch": 0.42114312267657994, "grad_norm": 0.8827386097163852, "learning_rate": 6.506818027845592e-06, "loss": 0.0662, "step": 9063 }, { "epoch": 0.4211895910780669, "grad_norm": 1.8236208580993434, "learning_rate": 6.506100482326599e-06, "loss": 0.0827, "step": 9064 }, { "epoch": 0.4212360594795539, "grad_norm": 0.9703287594344104, "learning_rate": 6.505382902692846e-06, "loss": 0.0612, "step": 9065 }, { "epoch": 0.42128252788104087, "grad_norm": 0.5765725190252424, "learning_rate": 6.504665288960588e-06, "loss": 0.0472, "step": 9066 }, { "epoch": 0.4213289962825279, "grad_norm": 0.989015326253294, "learning_rate": 6.503947641146074e-06, "loss": 0.06, "step": 9067 }, { "epoch": 0.4213754646840149, "grad_norm": 0.9637826858529353, "learning_rate": 6.503229959265566e-06, "loss": 0.0744, "step": 9068 }, { "epoch": 0.42142193308550185, "grad_norm": 1.7157202729859757, "learning_rate": 6.502512243335316e-06, "loss": 0.0435, "step": 9069 }, { "epoch": 0.42146840148698883, "grad_norm": 1.2168804913164262, "learning_rate": 6.501794493371582e-06, "loss": 0.0591, "step": 9070 }, { "epoch": 0.42151486988847586, "grad_norm": 2.2498641206367944, "learning_rate": 6.501076709390622e-06, "loss": 0.0824, "step": 9071 }, { "epoch": 0.42156133828996284, "grad_norm": 1.0131821470453342, "learning_rate": 6.500358891408696e-06, "loss": 0.0558, "step": 9072 }, { "epoch": 0.4216078066914498, "grad_norm": 1.0816025754213276, "learning_rate": 6.499641039442061e-06, "loss": 0.0803, "step": 9073 }, { "epoch": 0.4216542750929368, "grad_norm": 0.8877906726272113, "learning_rate": 6.498923153506977e-06, "loss": 0.0629, "step": 9074 }, { "epoch": 0.42170074349442377, "grad_norm": 0.6545537255138312, "learning_rate": 6.498205233619708e-06, "loss": 0.054, "step": 9075 }, { "epoch": 0.4217472118959108, "grad_norm": 0.6736763833706035, "learning_rate": 6.497487279796513e-06, "loss": 0.0418, "step": 9076 }, { "epoch": 0.4217936802973978, "grad_norm": 0.7563575364635065, "learning_rate": 6.4967692920536555e-06, "loss": 0.0614, "step": 9077 }, { "epoch": 0.42184014869888475, "grad_norm": 0.8221257547454502, "learning_rate": 6.4960512704073974e-06, "loss": 0.0966, "step": 9078 }, { "epoch": 0.42188661710037173, "grad_norm": 0.493777210954584, "learning_rate": 6.495333214874003e-06, "loss": 0.0554, "step": 9079 }, { "epoch": 0.42193308550185876, "grad_norm": 1.1999417816599223, "learning_rate": 6.494615125469738e-06, "loss": 0.0655, "step": 9080 }, { "epoch": 0.42197955390334574, "grad_norm": 2.4057869032438997, "learning_rate": 6.493897002210869e-06, "loss": 0.0685, "step": 9081 }, { "epoch": 0.4220260223048327, "grad_norm": 0.6529460881706513, "learning_rate": 6.493178845113658e-06, "loss": 0.0507, "step": 9082 }, { "epoch": 0.4220724907063197, "grad_norm": 0.9649167262323491, "learning_rate": 6.492460654194377e-06, "loss": 0.0518, "step": 9083 }, { "epoch": 0.42211895910780667, "grad_norm": 0.49703963440149823, "learning_rate": 6.49174242946929e-06, "loss": 0.051, "step": 9084 }, { "epoch": 0.4221654275092937, "grad_norm": 1.6007604439922223, "learning_rate": 6.4910241709546675e-06, "loss": 0.0856, "step": 9085 }, { "epoch": 0.4222118959107807, "grad_norm": 0.8101584542960845, "learning_rate": 6.490305878666776e-06, "loss": 0.0601, "step": 9086 }, { "epoch": 0.42225836431226765, "grad_norm": 1.1021089956130212, "learning_rate": 6.48958755262189e-06, "loss": 0.0654, "step": 9087 }, { "epoch": 0.42230483271375463, "grad_norm": 0.9128741364305528, "learning_rate": 6.488869192836279e-06, "loss": 0.0619, "step": 9088 }, { "epoch": 0.42235130111524166, "grad_norm": 1.3608677473219075, "learning_rate": 6.488150799326212e-06, "loss": 0.0671, "step": 9089 }, { "epoch": 0.42239776951672864, "grad_norm": 4.486031838367028, "learning_rate": 6.4874323721079646e-06, "loss": 0.0635, "step": 9090 }, { "epoch": 0.4224442379182156, "grad_norm": 0.5271356726360785, "learning_rate": 6.486713911197807e-06, "loss": 0.0455, "step": 9091 }, { "epoch": 0.4224907063197026, "grad_norm": 0.7659483646863807, "learning_rate": 6.485995416612016e-06, "loss": 0.0647, "step": 9092 }, { "epoch": 0.42253717472118957, "grad_norm": 1.0498563021750054, "learning_rate": 6.485276888366864e-06, "loss": 0.0743, "step": 9093 }, { "epoch": 0.4225836431226766, "grad_norm": 3.1748984029609804, "learning_rate": 6.484558326478626e-06, "loss": 0.0727, "step": 9094 }, { "epoch": 0.4226301115241636, "grad_norm": 0.5953078956902766, "learning_rate": 6.483839730963582e-06, "loss": 0.0467, "step": 9095 }, { "epoch": 0.42267657992565055, "grad_norm": 2.9910044791737422, "learning_rate": 6.4831211018380044e-06, "loss": 0.0874, "step": 9096 }, { "epoch": 0.42272304832713753, "grad_norm": 1.1077934576638202, "learning_rate": 6.482402439118174e-06, "loss": 0.0743, "step": 9097 }, { "epoch": 0.42276951672862456, "grad_norm": 0.7644834091249441, "learning_rate": 6.481683742820367e-06, "loss": 0.0701, "step": 9098 }, { "epoch": 0.42281598513011154, "grad_norm": 0.739479387183616, "learning_rate": 6.480965012960865e-06, "loss": 0.0451, "step": 9099 }, { "epoch": 0.4228624535315985, "grad_norm": 1.0975278275920393, "learning_rate": 6.480246249555946e-06, "loss": 0.0974, "step": 9100 }, { "epoch": 0.4229089219330855, "grad_norm": 2.8698325863255514, "learning_rate": 6.479527452621892e-06, "loss": 0.0726, "step": 9101 }, { "epoch": 0.42295539033457247, "grad_norm": 1.9146502651422719, "learning_rate": 6.478808622174985e-06, "loss": 0.0664, "step": 9102 }, { "epoch": 0.4230018587360595, "grad_norm": 0.7887425702268915, "learning_rate": 6.478089758231504e-06, "loss": 0.0323, "step": 9103 }, { "epoch": 0.4230483271375465, "grad_norm": 0.4323743725894142, "learning_rate": 6.4773708608077355e-06, "loss": 0.0375, "step": 9104 }, { "epoch": 0.42309479553903345, "grad_norm": 1.1735681954386858, "learning_rate": 6.476651929919962e-06, "loss": 0.0754, "step": 9105 }, { "epoch": 0.42314126394052043, "grad_norm": 1.6672855597372003, "learning_rate": 6.475932965584468e-06, "loss": 0.0535, "step": 9106 }, { "epoch": 0.42318773234200746, "grad_norm": 0.48167822182027176, "learning_rate": 6.47521396781754e-06, "loss": 0.0277, "step": 9107 }, { "epoch": 0.42323420074349444, "grad_norm": 0.47262307152412425, "learning_rate": 6.474494936635463e-06, "loss": 0.0375, "step": 9108 }, { "epoch": 0.4232806691449814, "grad_norm": 1.551449399514131, "learning_rate": 6.473775872054522e-06, "loss": 0.0751, "step": 9109 }, { "epoch": 0.4233271375464684, "grad_norm": 0.6563583690866869, "learning_rate": 6.473056774091007e-06, "loss": 0.0617, "step": 9110 }, { "epoch": 0.42337360594795537, "grad_norm": 2.9513746136685906, "learning_rate": 6.472337642761206e-06, "loss": 0.0726, "step": 9111 }, { "epoch": 0.4234200743494424, "grad_norm": 0.880258828235116, "learning_rate": 6.471618478081409e-06, "loss": 0.0655, "step": 9112 }, { "epoch": 0.4234665427509294, "grad_norm": 0.880258828235116, "learning_rate": 6.471618478081409e-06, "loss": 0.0572, "step": 9113 }, { "epoch": 0.42351301115241635, "grad_norm": 2.3528752928310195, "learning_rate": 6.470899280067902e-06, "loss": 0.0784, "step": 9114 }, { "epoch": 0.42355947955390333, "grad_norm": 0.7315869782340193, "learning_rate": 6.4701800487369804e-06, "loss": 0.0731, "step": 9115 }, { "epoch": 0.42360594795539036, "grad_norm": 0.4567160748697737, "learning_rate": 6.469460784104932e-06, "loss": 0.0344, "step": 9116 }, { "epoch": 0.42365241635687734, "grad_norm": 1.1881070281949626, "learning_rate": 6.468741486188052e-06, "loss": 0.1033, "step": 9117 }, { "epoch": 0.4236988847583643, "grad_norm": 1.4185942820511221, "learning_rate": 6.46802215500263e-06, "loss": 0.0759, "step": 9118 }, { "epoch": 0.4237453531598513, "grad_norm": 1.474498821709028, "learning_rate": 6.467302790564961e-06, "loss": 0.0503, "step": 9119 }, { "epoch": 0.42379182156133827, "grad_norm": 0.719215996588324, "learning_rate": 6.46658339289134e-06, "loss": 0.0703, "step": 9120 }, { "epoch": 0.4238382899628253, "grad_norm": 0.7421737684619792, "learning_rate": 6.4658639619980626e-06, "loss": 0.0718, "step": 9121 }, { "epoch": 0.4238847583643123, "grad_norm": 1.6268438598918706, "learning_rate": 6.465144497901422e-06, "loss": 0.0749, "step": 9122 }, { "epoch": 0.42393122676579925, "grad_norm": 1.488326380629625, "learning_rate": 6.464425000617718e-06, "loss": 0.0803, "step": 9123 }, { "epoch": 0.42397769516728623, "grad_norm": 0.8386452537466388, "learning_rate": 6.463705470163244e-06, "loss": 0.0585, "step": 9124 }, { "epoch": 0.42402416356877326, "grad_norm": 1.2094343846126252, "learning_rate": 6.462985906554302e-06, "loss": 0.0915, "step": 9125 }, { "epoch": 0.42407063197026024, "grad_norm": 0.7750282681065397, "learning_rate": 6.462266309807191e-06, "loss": 0.0585, "step": 9126 }, { "epoch": 0.4241171003717472, "grad_norm": 0.7011395140672788, "learning_rate": 6.461546679938208e-06, "loss": 0.0472, "step": 9127 }, { "epoch": 0.4241635687732342, "grad_norm": 1.0077858680052936, "learning_rate": 6.460827016963653e-06, "loss": 0.0899, "step": 9128 }, { "epoch": 0.42421003717472117, "grad_norm": 1.0920357999786936, "learning_rate": 6.46010732089983e-06, "loss": 0.0818, "step": 9129 }, { "epoch": 0.4242565055762082, "grad_norm": 1.1819768377276403, "learning_rate": 6.45938759176304e-06, "loss": 0.0567, "step": 9130 }, { "epoch": 0.4243029739776952, "grad_norm": 1.0247375626290307, "learning_rate": 6.458667829569583e-06, "loss": 0.0464, "step": 9131 }, { "epoch": 0.42434944237918215, "grad_norm": 1.291306834199853, "learning_rate": 6.457948034335767e-06, "loss": 0.0661, "step": 9132 }, { "epoch": 0.42439591078066913, "grad_norm": 0.9981287014736044, "learning_rate": 6.457228206077891e-06, "loss": 0.0975, "step": 9133 }, { "epoch": 0.42444237918215616, "grad_norm": 0.9534990663151754, "learning_rate": 6.4565083448122645e-06, "loss": 0.0558, "step": 9134 }, { "epoch": 0.42448884758364314, "grad_norm": 0.8102688978163171, "learning_rate": 6.455788450555191e-06, "loss": 0.0636, "step": 9135 }, { "epoch": 0.4245353159851301, "grad_norm": 0.44325017498225977, "learning_rate": 6.455068523322975e-06, "loss": 0.0451, "step": 9136 }, { "epoch": 0.4245817843866171, "grad_norm": 0.7153625052117025, "learning_rate": 6.4543485631319265e-06, "loss": 0.0429, "step": 9137 }, { "epoch": 0.42462825278810407, "grad_norm": 0.7808102204958107, "learning_rate": 6.453628569998352e-06, "loss": 0.0428, "step": 9138 }, { "epoch": 0.4246747211895911, "grad_norm": 0.6480880321197648, "learning_rate": 6.452908543938562e-06, "loss": 0.0586, "step": 9139 }, { "epoch": 0.4247211895910781, "grad_norm": 1.430498679817211, "learning_rate": 6.452188484968862e-06, "loss": 0.0838, "step": 9140 }, { "epoch": 0.42476765799256505, "grad_norm": 0.774153155596902, "learning_rate": 6.4514683931055666e-06, "loss": 0.0504, "step": 9141 }, { "epoch": 0.424814126394052, "grad_norm": 0.5579391124474022, "learning_rate": 6.450748268364983e-06, "loss": 0.0365, "step": 9142 }, { "epoch": 0.42486059479553906, "grad_norm": 0.5237026697579356, "learning_rate": 6.450028110763425e-06, "loss": 0.0657, "step": 9143 }, { "epoch": 0.42490706319702604, "grad_norm": 0.7830005937538451, "learning_rate": 6.449307920317205e-06, "loss": 0.0525, "step": 9144 }, { "epoch": 0.424953531598513, "grad_norm": 0.6833913985205341, "learning_rate": 6.448587697042633e-06, "loss": 0.072, "step": 9145 }, { "epoch": 0.425, "grad_norm": 0.30568761939518263, "learning_rate": 6.447867440956027e-06, "loss": 0.0491, "step": 9146 }, { "epoch": 0.42504646840148697, "grad_norm": 0.9427065034539973, "learning_rate": 6.447147152073699e-06, "loss": 0.0696, "step": 9147 }, { "epoch": 0.425092936802974, "grad_norm": 1.2112153685656748, "learning_rate": 6.446426830411964e-06, "loss": 0.0527, "step": 9148 }, { "epoch": 0.425139405204461, "grad_norm": 0.700158736755508, "learning_rate": 6.4457064759871405e-06, "loss": 0.0587, "step": 9149 }, { "epoch": 0.42518587360594795, "grad_norm": 0.5301516484961522, "learning_rate": 6.444986088815543e-06, "loss": 0.0528, "step": 9150 }, { "epoch": 0.4252323420074349, "grad_norm": 1.8889808741997827, "learning_rate": 6.444265668913489e-06, "loss": 0.0585, "step": 9151 }, { "epoch": 0.42527881040892196, "grad_norm": 1.0487115718862356, "learning_rate": 6.4435452162972975e-06, "loss": 0.0547, "step": 9152 }, { "epoch": 0.42532527881040894, "grad_norm": 1.2853347588003403, "learning_rate": 6.442824730983289e-06, "loss": 0.0648, "step": 9153 }, { "epoch": 0.4253717472118959, "grad_norm": 1.490071475479314, "learning_rate": 6.442104212987782e-06, "loss": 0.0612, "step": 9154 }, { "epoch": 0.4254182156133829, "grad_norm": 1.0359926630343315, "learning_rate": 6.441383662327095e-06, "loss": 0.0756, "step": 9155 }, { "epoch": 0.42546468401486986, "grad_norm": 1.1095367191556356, "learning_rate": 6.440663079017553e-06, "loss": 0.0471, "step": 9156 }, { "epoch": 0.4255111524163569, "grad_norm": 0.8782595091789083, "learning_rate": 6.439942463075474e-06, "loss": 0.0525, "step": 9157 }, { "epoch": 0.4255576208178439, "grad_norm": 0.885248800759811, "learning_rate": 6.439221814517185e-06, "loss": 0.0478, "step": 9158 }, { "epoch": 0.42560408921933085, "grad_norm": 0.4951172862758984, "learning_rate": 6.438501133359006e-06, "loss": 0.0543, "step": 9159 }, { "epoch": 0.4256505576208178, "grad_norm": 0.5014420657737101, "learning_rate": 6.437780419617261e-06, "loss": 0.0475, "step": 9160 }, { "epoch": 0.42569702602230486, "grad_norm": 4.48303184320378, "learning_rate": 6.437059673308277e-06, "loss": 0.0785, "step": 9161 }, { "epoch": 0.42574349442379184, "grad_norm": 0.517900680107964, "learning_rate": 6.43633889444838e-06, "loss": 0.0361, "step": 9162 }, { "epoch": 0.4257899628252788, "grad_norm": 0.6510078691079452, "learning_rate": 6.435618083053895e-06, "loss": 0.065, "step": 9163 }, { "epoch": 0.4258364312267658, "grad_norm": 1.359418426503814, "learning_rate": 6.434897239141148e-06, "loss": 0.0649, "step": 9164 }, { "epoch": 0.42588289962825276, "grad_norm": 0.8546237264133003, "learning_rate": 6.434176362726469e-06, "loss": 0.0687, "step": 9165 }, { "epoch": 0.4259293680297398, "grad_norm": 1.2269024915151745, "learning_rate": 6.433455453826187e-06, "loss": 0.08, "step": 9166 }, { "epoch": 0.4259758364312268, "grad_norm": 0.7982558123733765, "learning_rate": 6.432734512456629e-06, "loss": 0.0648, "step": 9167 }, { "epoch": 0.42602230483271375, "grad_norm": 0.7780476839311786, "learning_rate": 6.432013538634129e-06, "loss": 0.0381, "step": 9168 }, { "epoch": 0.4260687732342007, "grad_norm": 0.9138268603764396, "learning_rate": 6.431292532375012e-06, "loss": 0.0528, "step": 9169 }, { "epoch": 0.42611524163568776, "grad_norm": 0.9099756554478067, "learning_rate": 6.4305714936956145e-06, "loss": 0.0767, "step": 9170 }, { "epoch": 0.42616171003717473, "grad_norm": 1.7897989521485778, "learning_rate": 6.429850422612266e-06, "loss": 0.0789, "step": 9171 }, { "epoch": 0.4262081784386617, "grad_norm": 0.9949054894114439, "learning_rate": 6.4291293191413005e-06, "loss": 0.0733, "step": 9172 }, { "epoch": 0.4262546468401487, "grad_norm": 0.5431956114775864, "learning_rate": 6.428408183299052e-06, "loss": 0.0483, "step": 9173 }, { "epoch": 0.42630111524163566, "grad_norm": 1.3458876307065135, "learning_rate": 6.427687015101855e-06, "loss": 0.0615, "step": 9174 }, { "epoch": 0.4263475836431227, "grad_norm": 0.5698280745164529, "learning_rate": 6.426965814566045e-06, "loss": 0.0654, "step": 9175 }, { "epoch": 0.4263940520446097, "grad_norm": 1.622429980145001, "learning_rate": 6.426244581707956e-06, "loss": 0.0743, "step": 9176 }, { "epoch": 0.42644052044609665, "grad_norm": 0.5630769993330413, "learning_rate": 6.425523316543928e-06, "loss": 0.0552, "step": 9177 }, { "epoch": 0.4264869888475836, "grad_norm": 1.045083913549704, "learning_rate": 6.424802019090296e-06, "loss": 0.0507, "step": 9178 }, { "epoch": 0.42653345724907066, "grad_norm": 0.6503047790691744, "learning_rate": 6.424080689363396e-06, "loss": 0.0489, "step": 9179 }, { "epoch": 0.42657992565055763, "grad_norm": 1.5947944110463181, "learning_rate": 6.423359327379575e-06, "loss": 0.0498, "step": 9180 }, { "epoch": 0.4266263940520446, "grad_norm": 0.5697050857709487, "learning_rate": 6.4226379331551625e-06, "loss": 0.0584, "step": 9181 }, { "epoch": 0.4266728624535316, "grad_norm": 0.4628973719521978, "learning_rate": 6.421916506706506e-06, "loss": 0.0532, "step": 9182 }, { "epoch": 0.42671933085501856, "grad_norm": 0.5981115535504853, "learning_rate": 6.421195048049943e-06, "loss": 0.058, "step": 9183 }, { "epoch": 0.4267657992565056, "grad_norm": 0.4786058916442086, "learning_rate": 6.420473557201816e-06, "loss": 0.0497, "step": 9184 }, { "epoch": 0.4268122676579926, "grad_norm": 1.2430153095068301, "learning_rate": 6.41975203417847e-06, "loss": 0.0646, "step": 9185 }, { "epoch": 0.42685873605947955, "grad_norm": 0.9735744125061679, "learning_rate": 6.419030478996243e-06, "loss": 0.0487, "step": 9186 }, { "epoch": 0.4269052044609665, "grad_norm": 1.0410623938139372, "learning_rate": 6.418308891671484e-06, "loss": 0.0596, "step": 9187 }, { "epoch": 0.42695167286245356, "grad_norm": 3.7116510763386765, "learning_rate": 6.417587272220537e-06, "loss": 0.1062, "step": 9188 }, { "epoch": 0.42699814126394053, "grad_norm": 0.8314076149524067, "learning_rate": 6.4168656206597435e-06, "loss": 0.0576, "step": 9189 }, { "epoch": 0.4270446096654275, "grad_norm": 0.6946139406021423, "learning_rate": 6.416143937005455e-06, "loss": 0.0749, "step": 9190 }, { "epoch": 0.4270910780669145, "grad_norm": 1.142242806475438, "learning_rate": 6.415422221274015e-06, "loss": 0.0489, "step": 9191 }, { "epoch": 0.42713754646840146, "grad_norm": 0.8161765837774297, "learning_rate": 6.414700473481772e-06, "loss": 0.0549, "step": 9192 }, { "epoch": 0.4271840148698885, "grad_norm": 0.8769780795749156, "learning_rate": 6.413978693645073e-06, "loss": 0.0625, "step": 9193 }, { "epoch": 0.42723048327137547, "grad_norm": 0.7290641951966264, "learning_rate": 6.41325688178027e-06, "loss": 0.0706, "step": 9194 }, { "epoch": 0.42727695167286245, "grad_norm": 0.3757356063129789, "learning_rate": 6.412535037903711e-06, "loss": 0.0348, "step": 9195 }, { "epoch": 0.4273234200743494, "grad_norm": 1.1260419064866962, "learning_rate": 6.4118131620317445e-06, "loss": 0.092, "step": 9196 }, { "epoch": 0.42736988847583646, "grad_norm": 1.3717199754977212, "learning_rate": 6.411091254180728e-06, "loss": 0.0889, "step": 9197 }, { "epoch": 0.42741635687732343, "grad_norm": 1.8393475541930087, "learning_rate": 6.410369314367007e-06, "loss": 0.0499, "step": 9198 }, { "epoch": 0.4274628252788104, "grad_norm": 0.7857902719549265, "learning_rate": 6.409647342606935e-06, "loss": 0.0485, "step": 9199 }, { "epoch": 0.4275092936802974, "grad_norm": 0.5623026894897274, "learning_rate": 6.408925338916871e-06, "loss": 0.0509, "step": 9200 }, { "epoch": 0.42755576208178436, "grad_norm": 0.6888420858002942, "learning_rate": 6.408203303313162e-06, "loss": 0.057, "step": 9201 }, { "epoch": 0.4276022304832714, "grad_norm": 1.2184378117958117, "learning_rate": 6.407481235812167e-06, "loss": 0.0605, "step": 9202 }, { "epoch": 0.42764869888475837, "grad_norm": 1.3777605337470067, "learning_rate": 6.406759136430242e-06, "loss": 0.0905, "step": 9203 }, { "epoch": 0.42769516728624535, "grad_norm": 1.520944965691005, "learning_rate": 6.4060370051837405e-06, "loss": 0.0599, "step": 9204 }, { "epoch": 0.4277416356877323, "grad_norm": 0.5658045116952651, "learning_rate": 6.405314842089021e-06, "loss": 0.0337, "step": 9205 }, { "epoch": 0.42778810408921936, "grad_norm": 0.7329260297170925, "learning_rate": 6.404592647162442e-06, "loss": 0.0622, "step": 9206 }, { "epoch": 0.42783457249070633, "grad_norm": 1.597855737309906, "learning_rate": 6.403870420420362e-06, "loss": 0.0752, "step": 9207 }, { "epoch": 0.4278810408921933, "grad_norm": 2.1114284514703003, "learning_rate": 6.403148161879137e-06, "loss": 0.1034, "step": 9208 }, { "epoch": 0.4279275092936803, "grad_norm": 0.9652998983965797, "learning_rate": 6.402425871555132e-06, "loss": 0.0526, "step": 9209 }, { "epoch": 0.42797397769516726, "grad_norm": 1.3664365095522322, "learning_rate": 6.401703549464703e-06, "loss": 0.0763, "step": 9210 }, { "epoch": 0.4280204460966543, "grad_norm": 0.7427454624402385, "learning_rate": 6.400981195624214e-06, "loss": 0.0501, "step": 9211 }, { "epoch": 0.42806691449814127, "grad_norm": 2.895546746745509, "learning_rate": 6.400258810050027e-06, "loss": 0.0619, "step": 9212 }, { "epoch": 0.42811338289962825, "grad_norm": 0.7755190263890152, "learning_rate": 6.399536392758503e-06, "loss": 0.0511, "step": 9213 }, { "epoch": 0.4281598513011152, "grad_norm": 0.8204544325986888, "learning_rate": 6.398813943766008e-06, "loss": 0.0645, "step": 9214 }, { "epoch": 0.42820631970260226, "grad_norm": 1.0707034230832877, "learning_rate": 6.398091463088904e-06, "loss": 0.071, "step": 9215 }, { "epoch": 0.42825278810408923, "grad_norm": 0.7367745345108553, "learning_rate": 6.397368950743556e-06, "loss": 0.0544, "step": 9216 }, { "epoch": 0.4282992565055762, "grad_norm": 1.2306640895877445, "learning_rate": 6.396646406746333e-06, "loss": 0.0868, "step": 9217 }, { "epoch": 0.4283457249070632, "grad_norm": 0.7030449457035667, "learning_rate": 6.395923831113597e-06, "loss": 0.054, "step": 9218 }, { "epoch": 0.42839219330855016, "grad_norm": 0.8413996086438689, "learning_rate": 6.395201223861719e-06, "loss": 0.0636, "step": 9219 }, { "epoch": 0.4284386617100372, "grad_norm": 1.1690911136637863, "learning_rate": 6.394478585007063e-06, "loss": 0.0659, "step": 9220 }, { "epoch": 0.42848513011152417, "grad_norm": 0.7199876442966017, "learning_rate": 6.393755914566e-06, "loss": 0.0516, "step": 9221 }, { "epoch": 0.42853159851301115, "grad_norm": 0.7443981810389811, "learning_rate": 6.393033212554899e-06, "loss": 0.0428, "step": 9222 }, { "epoch": 0.4285780669144981, "grad_norm": 0.4271123849354512, "learning_rate": 6.392310478990129e-06, "loss": 0.0414, "step": 9223 }, { "epoch": 0.42862453531598516, "grad_norm": 1.591958150140148, "learning_rate": 6.391587713888062e-06, "loss": 0.0808, "step": 9224 }, { "epoch": 0.42867100371747213, "grad_norm": 0.7805338528476412, "learning_rate": 6.390864917265067e-06, "loss": 0.0558, "step": 9225 }, { "epoch": 0.4287174721189591, "grad_norm": 1.6254254667336838, "learning_rate": 6.39014208913752e-06, "loss": 0.0866, "step": 9226 }, { "epoch": 0.4287639405204461, "grad_norm": 1.946778229532561, "learning_rate": 6.3894192295217905e-06, "loss": 0.0952, "step": 9227 }, { "epoch": 0.42881040892193306, "grad_norm": 0.5858296839432576, "learning_rate": 6.388696338434253e-06, "loss": 0.0515, "step": 9228 }, { "epoch": 0.4288568773234201, "grad_norm": 0.6526083173243609, "learning_rate": 6.387973415891282e-06, "loss": 0.0432, "step": 9229 }, { "epoch": 0.42890334572490707, "grad_norm": 6.415501123791065, "learning_rate": 6.387250461909252e-06, "loss": 0.0527, "step": 9230 }, { "epoch": 0.42894981412639405, "grad_norm": 1.0185660470071802, "learning_rate": 6.38652747650454e-06, "loss": 0.0683, "step": 9231 }, { "epoch": 0.428996282527881, "grad_norm": 0.7142262830818399, "learning_rate": 6.38580445969352e-06, "loss": 0.0654, "step": 9232 }, { "epoch": 0.42904275092936806, "grad_norm": 0.5112579425824456, "learning_rate": 6.3850814114925705e-06, "loss": 0.0481, "step": 9233 }, { "epoch": 0.42908921933085503, "grad_norm": 0.7438183541643131, "learning_rate": 6.384358331918071e-06, "loss": 0.0808, "step": 9234 }, { "epoch": 0.429135687732342, "grad_norm": 1.493963805500888, "learning_rate": 6.383635220986396e-06, "loss": 0.0524, "step": 9235 }, { "epoch": 0.429182156133829, "grad_norm": 0.918013821243957, "learning_rate": 6.38291207871393e-06, "loss": 0.0734, "step": 9236 }, { "epoch": 0.42922862453531596, "grad_norm": 1.0600445075390252, "learning_rate": 6.382188905117047e-06, "loss": 0.0645, "step": 9237 }, { "epoch": 0.429275092936803, "grad_norm": 1.7270857406779527, "learning_rate": 6.381465700212131e-06, "loss": 0.0812, "step": 9238 }, { "epoch": 0.42932156133828997, "grad_norm": 1.2670866752944923, "learning_rate": 6.380742464015565e-06, "loss": 0.0516, "step": 9239 }, { "epoch": 0.42936802973977695, "grad_norm": 0.9505790662037159, "learning_rate": 6.380019196543726e-06, "loss": 0.078, "step": 9240 }, { "epoch": 0.4294144981412639, "grad_norm": 0.7543126289901133, "learning_rate": 6.3792958978130024e-06, "loss": 0.0547, "step": 9241 }, { "epoch": 0.42946096654275095, "grad_norm": 0.7133645298559042, "learning_rate": 6.378572567839773e-06, "loss": 0.0683, "step": 9242 }, { "epoch": 0.42950743494423793, "grad_norm": 0.4397022733436571, "learning_rate": 6.3778492066404265e-06, "loss": 0.028, "step": 9243 }, { "epoch": 0.4295539033457249, "grad_norm": 0.7089570756309139, "learning_rate": 6.377125814231344e-06, "loss": 0.0656, "step": 9244 }, { "epoch": 0.4296003717472119, "grad_norm": 0.5440398311371067, "learning_rate": 6.376402390628911e-06, "loss": 0.0428, "step": 9245 }, { "epoch": 0.42964684014869886, "grad_norm": 1.0033111189461932, "learning_rate": 6.375678935849517e-06, "loss": 0.0681, "step": 9246 }, { "epoch": 0.4296933085501859, "grad_norm": 0.762372627818366, "learning_rate": 6.374955449909547e-06, "loss": 0.0458, "step": 9247 }, { "epoch": 0.42973977695167287, "grad_norm": 1.0399681242504286, "learning_rate": 6.374231932825389e-06, "loss": 0.0766, "step": 9248 }, { "epoch": 0.42978624535315985, "grad_norm": 0.6530652333102578, "learning_rate": 6.37350838461343e-06, "loss": 0.0591, "step": 9249 }, { "epoch": 0.4298327137546468, "grad_norm": 1.1936757451404096, "learning_rate": 6.3727848052900605e-06, "loss": 0.0868, "step": 9250 }, { "epoch": 0.42987918215613385, "grad_norm": 1.0855067598726469, "learning_rate": 6.372061194871672e-06, "loss": 0.0612, "step": 9251 }, { "epoch": 0.42992565055762083, "grad_norm": 0.6727551636170376, "learning_rate": 6.3713375533746525e-06, "loss": 0.0445, "step": 9252 }, { "epoch": 0.4299721189591078, "grad_norm": 0.8807601353766576, "learning_rate": 6.370613880815395e-06, "loss": 0.0552, "step": 9253 }, { "epoch": 0.4300185873605948, "grad_norm": 0.5192708051902265, "learning_rate": 6.3698901772102895e-06, "loss": 0.0384, "step": 9254 }, { "epoch": 0.43006505576208176, "grad_norm": 0.5315663258961759, "learning_rate": 6.369166442575729e-06, "loss": 0.0418, "step": 9255 }, { "epoch": 0.4301115241635688, "grad_norm": 1.392558285587344, "learning_rate": 6.36844267692811e-06, "loss": 0.0731, "step": 9256 }, { "epoch": 0.43015799256505577, "grad_norm": 0.5882051474725322, "learning_rate": 6.367718880283822e-06, "loss": 0.0413, "step": 9257 }, { "epoch": 0.43020446096654275, "grad_norm": 0.9887725565415654, "learning_rate": 6.366995052659263e-06, "loss": 0.0602, "step": 9258 }, { "epoch": 0.4302509293680297, "grad_norm": 0.5867607676519793, "learning_rate": 6.366271194070826e-06, "loss": 0.041, "step": 9259 }, { "epoch": 0.43029739776951675, "grad_norm": 0.7133952347857373, "learning_rate": 6.3655473045349106e-06, "loss": 0.0485, "step": 9260 }, { "epoch": 0.43034386617100373, "grad_norm": 0.7547652070060553, "learning_rate": 6.364823384067909e-06, "loss": 0.049, "step": 9261 }, { "epoch": 0.4303903345724907, "grad_norm": 0.6591144710568322, "learning_rate": 6.364099432686222e-06, "loss": 0.0306, "step": 9262 }, { "epoch": 0.4304368029739777, "grad_norm": 0.940460345532396, "learning_rate": 6.3633754504062485e-06, "loss": 0.065, "step": 9263 }, { "epoch": 0.43048327137546466, "grad_norm": 0.9067137438573797, "learning_rate": 6.362651437244385e-06, "loss": 0.081, "step": 9264 }, { "epoch": 0.4305297397769517, "grad_norm": 1.4308942017678172, "learning_rate": 6.361927393217033e-06, "loss": 0.0528, "step": 9265 }, { "epoch": 0.43057620817843867, "grad_norm": 0.8914828018732188, "learning_rate": 6.361203318340592e-06, "loss": 0.0746, "step": 9266 }, { "epoch": 0.43062267657992565, "grad_norm": 1.4128994129130086, "learning_rate": 6.360479212631463e-06, "loss": 0.1254, "step": 9267 }, { "epoch": 0.4306691449814126, "grad_norm": 0.8879063698965286, "learning_rate": 6.359755076106049e-06, "loss": 0.0595, "step": 9268 }, { "epoch": 0.43071561338289965, "grad_norm": 1.076176643666773, "learning_rate": 6.359030908780749e-06, "loss": 0.0581, "step": 9269 }, { "epoch": 0.43076208178438663, "grad_norm": 0.7237832531986825, "learning_rate": 6.35830671067197e-06, "loss": 0.0478, "step": 9270 }, { "epoch": 0.4308085501858736, "grad_norm": 0.7271749116343306, "learning_rate": 6.357582481796114e-06, "loss": 0.0632, "step": 9271 }, { "epoch": 0.4308550185873606, "grad_norm": 0.8965069404542824, "learning_rate": 6.356858222169587e-06, "loss": 0.0667, "step": 9272 }, { "epoch": 0.43090148698884756, "grad_norm": 0.8782173700618978, "learning_rate": 6.356133931808792e-06, "loss": 0.0498, "step": 9273 }, { "epoch": 0.4309479553903346, "grad_norm": 9.52140380396493, "learning_rate": 6.3554096107301366e-06, "loss": 0.0797, "step": 9274 }, { "epoch": 0.43099442379182157, "grad_norm": 0.7913338283356572, "learning_rate": 6.354685258950027e-06, "loss": 0.0732, "step": 9275 }, { "epoch": 0.43104089219330854, "grad_norm": 1.1074178945479343, "learning_rate": 6.35396087648487e-06, "loss": 0.0661, "step": 9276 }, { "epoch": 0.4310873605947955, "grad_norm": 0.8858012668649271, "learning_rate": 6.353236463351076e-06, "loss": 0.0504, "step": 9277 }, { "epoch": 0.43113382899628255, "grad_norm": 0.7698066721585279, "learning_rate": 6.352512019565049e-06, "loss": 0.06, "step": 9278 }, { "epoch": 0.43118029739776953, "grad_norm": 0.6045671485086912, "learning_rate": 6.3517875451432036e-06, "loss": 0.0599, "step": 9279 }, { "epoch": 0.4312267657992565, "grad_norm": 1.350919502125976, "learning_rate": 6.351063040101947e-06, "loss": 0.0615, "step": 9280 }, { "epoch": 0.4312732342007435, "grad_norm": 1.8798142036185523, "learning_rate": 6.350338504457689e-06, "loss": 0.0735, "step": 9281 }, { "epoch": 0.43131970260223046, "grad_norm": 0.8258404170414716, "learning_rate": 6.3496139382268444e-06, "loss": 0.072, "step": 9282 }, { "epoch": 0.4313661710037175, "grad_norm": 0.6979502289187921, "learning_rate": 6.3488893414258235e-06, "loss": 0.0554, "step": 9283 }, { "epoch": 0.43141263940520447, "grad_norm": 0.7421460453392932, "learning_rate": 6.348164714071039e-06, "loss": 0.0731, "step": 9284 }, { "epoch": 0.43145910780669144, "grad_norm": 1.4152196756217865, "learning_rate": 6.347440056178904e-06, "loss": 0.0553, "step": 9285 }, { "epoch": 0.4315055762081784, "grad_norm": 0.6359702996490172, "learning_rate": 6.346715367765836e-06, "loss": 0.0485, "step": 9286 }, { "epoch": 0.43155204460966545, "grad_norm": 1.8525028360598113, "learning_rate": 6.345990648848246e-06, "loss": 0.0799, "step": 9287 }, { "epoch": 0.43159851301115243, "grad_norm": 0.4575465370001414, "learning_rate": 6.345265899442549e-06, "loss": 0.0133, "step": 9288 }, { "epoch": 0.4316449814126394, "grad_norm": 1.348552252473654, "learning_rate": 6.344541119565167e-06, "loss": 0.0773, "step": 9289 }, { "epoch": 0.4316914498141264, "grad_norm": 0.6807717942468717, "learning_rate": 6.343816309232512e-06, "loss": 0.0523, "step": 9290 }, { "epoch": 0.43173791821561336, "grad_norm": 0.8985134509421416, "learning_rate": 6.343091468461003e-06, "loss": 0.0568, "step": 9291 }, { "epoch": 0.4317843866171004, "grad_norm": 0.7864872194062844, "learning_rate": 6.34236659726706e-06, "loss": 0.0658, "step": 9292 }, { "epoch": 0.43183085501858737, "grad_norm": 0.9573262211674354, "learning_rate": 6.341641695667099e-06, "loss": 0.0477, "step": 9293 }, { "epoch": 0.43187732342007434, "grad_norm": 0.7359931716924697, "learning_rate": 6.340916763677543e-06, "loss": 0.0708, "step": 9294 }, { "epoch": 0.4319237918215613, "grad_norm": 1.025703308971092, "learning_rate": 6.34019180131481e-06, "loss": 0.0666, "step": 9295 }, { "epoch": 0.43197026022304835, "grad_norm": 0.5653328388049945, "learning_rate": 6.339466808595323e-06, "loss": 0.0589, "step": 9296 }, { "epoch": 0.43201672862453533, "grad_norm": 1.1317919757572201, "learning_rate": 6.338741785535503e-06, "loss": 0.0467, "step": 9297 }, { "epoch": 0.4320631970260223, "grad_norm": 0.8382775889875332, "learning_rate": 6.338016732151772e-06, "loss": 0.0483, "step": 9298 }, { "epoch": 0.4321096654275093, "grad_norm": 1.7377534085607793, "learning_rate": 6.337291648460554e-06, "loss": 0.0742, "step": 9299 }, { "epoch": 0.43215613382899626, "grad_norm": 1.1611472758905705, "learning_rate": 6.3365665344782725e-06, "loss": 0.0598, "step": 9300 }, { "epoch": 0.4322026022304833, "grad_norm": 0.6103875432263379, "learning_rate": 6.335841390221353e-06, "loss": 0.0522, "step": 9301 }, { "epoch": 0.43224907063197027, "grad_norm": 0.45441124707497343, "learning_rate": 6.3351162157062186e-06, "loss": 0.05, "step": 9302 }, { "epoch": 0.43229553903345724, "grad_norm": 1.873420028321845, "learning_rate": 6.334391010949298e-06, "loss": 0.1162, "step": 9303 }, { "epoch": 0.4323420074349442, "grad_norm": 2.1320663772771526, "learning_rate": 6.333665775967016e-06, "loss": 0.0554, "step": 9304 }, { "epoch": 0.43238847583643125, "grad_norm": 0.962237896807603, "learning_rate": 6.332940510775801e-06, "loss": 0.0472, "step": 9305 }, { "epoch": 0.43243494423791823, "grad_norm": 1.0960430871020694, "learning_rate": 6.3322152153920805e-06, "loss": 0.0638, "step": 9306 }, { "epoch": 0.4324814126394052, "grad_norm": 1.0262032940392, "learning_rate": 6.331489889832282e-06, "loss": 0.0703, "step": 9307 }, { "epoch": 0.4325278810408922, "grad_norm": 0.7796546381105268, "learning_rate": 6.330764534112837e-06, "loss": 0.0485, "step": 9308 }, { "epoch": 0.43257434944237916, "grad_norm": 0.5969745091043683, "learning_rate": 6.330039148250176e-06, "loss": 0.0718, "step": 9309 }, { "epoch": 0.4326208178438662, "grad_norm": 1.0808396571801602, "learning_rate": 6.329313732260726e-06, "loss": 0.0613, "step": 9310 }, { "epoch": 0.43266728624535317, "grad_norm": 1.6181730952482412, "learning_rate": 6.328588286160923e-06, "loss": 0.0717, "step": 9311 }, { "epoch": 0.43271375464684014, "grad_norm": 1.1811273542216556, "learning_rate": 6.327862809967196e-06, "loss": 0.0578, "step": 9312 }, { "epoch": 0.4327602230483271, "grad_norm": 1.614012554875005, "learning_rate": 6.327137303695979e-06, "loss": 0.0649, "step": 9313 }, { "epoch": 0.43280669144981415, "grad_norm": 0.6201935560179825, "learning_rate": 6.326411767363705e-06, "loss": 0.0606, "step": 9314 }, { "epoch": 0.43285315985130113, "grad_norm": 1.527547518216028, "learning_rate": 6.325686200986808e-06, "loss": 0.093, "step": 9315 }, { "epoch": 0.4328996282527881, "grad_norm": 1.5662238124932968, "learning_rate": 6.324960604581725e-06, "loss": 0.1217, "step": 9316 }, { "epoch": 0.4329460966542751, "grad_norm": 0.731326842058435, "learning_rate": 6.324234978164887e-06, "loss": 0.0467, "step": 9317 }, { "epoch": 0.43299256505576206, "grad_norm": 0.3092495061569661, "learning_rate": 6.323509321752737e-06, "loss": 0.0417, "step": 9318 }, { "epoch": 0.4330390334572491, "grad_norm": 0.7943285674942849, "learning_rate": 6.322783635361704e-06, "loss": 0.0683, "step": 9319 }, { "epoch": 0.43308550185873607, "grad_norm": 1.2097368778117308, "learning_rate": 6.322057919008232e-06, "loss": 0.0746, "step": 9320 }, { "epoch": 0.43313197026022304, "grad_norm": 0.7171123355835374, "learning_rate": 6.321332172708756e-06, "loss": 0.0779, "step": 9321 }, { "epoch": 0.43317843866171, "grad_norm": 7.854764050885341, "learning_rate": 6.3206063964797146e-06, "loss": 0.0853, "step": 9322 }, { "epoch": 0.43322490706319705, "grad_norm": 0.6924892057896637, "learning_rate": 6.319880590337549e-06, "loss": 0.061, "step": 9323 }, { "epoch": 0.433271375464684, "grad_norm": 0.583241962332714, "learning_rate": 6.3191547542987e-06, "loss": 0.0598, "step": 9324 }, { "epoch": 0.433317843866171, "grad_norm": 1.1437179196351452, "learning_rate": 6.318428888379606e-06, "loss": 0.0789, "step": 9325 }, { "epoch": 0.433364312267658, "grad_norm": 0.8136232551371582, "learning_rate": 6.3177029925967105e-06, "loss": 0.0668, "step": 9326 }, { "epoch": 0.43341078066914496, "grad_norm": 0.46624074788559255, "learning_rate": 6.316977066966456e-06, "loss": 0.0471, "step": 9327 }, { "epoch": 0.433457249070632, "grad_norm": 1.0349898764198415, "learning_rate": 6.316251111505285e-06, "loss": 0.057, "step": 9328 }, { "epoch": 0.43350371747211897, "grad_norm": 0.8163942563843335, "learning_rate": 6.315525126229639e-06, "loss": 0.0878, "step": 9329 }, { "epoch": 0.43355018587360594, "grad_norm": 0.728435540494832, "learning_rate": 6.314799111155967e-06, "loss": 0.0461, "step": 9330 }, { "epoch": 0.4335966542750929, "grad_norm": 0.7257089259005146, "learning_rate": 6.314073066300709e-06, "loss": 0.0501, "step": 9331 }, { "epoch": 0.43364312267657995, "grad_norm": 0.7077920932112183, "learning_rate": 6.313346991680315e-06, "loss": 0.0568, "step": 9332 }, { "epoch": 0.4336895910780669, "grad_norm": 0.5438355610378957, "learning_rate": 6.312620887311228e-06, "loss": 0.0385, "step": 9333 }, { "epoch": 0.4337360594795539, "grad_norm": 2.5442221225556834, "learning_rate": 6.311894753209896e-06, "loss": 0.0387, "step": 9334 }, { "epoch": 0.4337825278810409, "grad_norm": 1.1472916996620945, "learning_rate": 6.311168589392768e-06, "loss": 0.0851, "step": 9335 }, { "epoch": 0.43382899628252786, "grad_norm": 1.54617386425346, "learning_rate": 6.3104423958762915e-06, "loss": 0.0799, "step": 9336 }, { "epoch": 0.4338754646840149, "grad_norm": 0.5782884026250829, "learning_rate": 6.309716172676916e-06, "loss": 0.0413, "step": 9337 }, { "epoch": 0.43392193308550187, "grad_norm": 1.1961010119189743, "learning_rate": 6.308989919811088e-06, "loss": 0.0579, "step": 9338 }, { "epoch": 0.43396840148698884, "grad_norm": 3.5285240321676388, "learning_rate": 6.308263637295264e-06, "loss": 0.0739, "step": 9339 }, { "epoch": 0.4340148698884758, "grad_norm": 0.9891763812823386, "learning_rate": 6.30753732514589e-06, "loss": 0.0553, "step": 9340 }, { "epoch": 0.43406133828996285, "grad_norm": 0.7205295937851097, "learning_rate": 6.306810983379419e-06, "loss": 0.0418, "step": 9341 }, { "epoch": 0.4341078066914498, "grad_norm": 1.4763178122265737, "learning_rate": 6.306084612012307e-06, "loss": 0.0617, "step": 9342 }, { "epoch": 0.4341542750929368, "grad_norm": 0.985501136773987, "learning_rate": 6.305358211061e-06, "loss": 0.0709, "step": 9343 }, { "epoch": 0.4342007434944238, "grad_norm": 1.2721034365985113, "learning_rate": 6.304631780541957e-06, "loss": 0.0647, "step": 9344 }, { "epoch": 0.43424721189591076, "grad_norm": 1.3609192137277824, "learning_rate": 6.303905320471633e-06, "loss": 0.089, "step": 9345 }, { "epoch": 0.4342936802973978, "grad_norm": 1.3299849069197223, "learning_rate": 6.303178830866478e-06, "loss": 0.0627, "step": 9346 }, { "epoch": 0.43434014869888476, "grad_norm": 1.1375960292046112, "learning_rate": 6.302452311742953e-06, "loss": 0.0619, "step": 9347 }, { "epoch": 0.43438661710037174, "grad_norm": 0.7615500952821519, "learning_rate": 6.301725763117511e-06, "loss": 0.0656, "step": 9348 }, { "epoch": 0.4344330855018587, "grad_norm": 0.6346190214699696, "learning_rate": 6.300999185006612e-06, "loss": 0.0424, "step": 9349 }, { "epoch": 0.43447955390334575, "grad_norm": 0.821300747159603, "learning_rate": 6.30027257742671e-06, "loss": 0.0781, "step": 9350 }, { "epoch": 0.4345260223048327, "grad_norm": 0.578709443008215, "learning_rate": 6.299545940394267e-06, "loss": 0.0359, "step": 9351 }, { "epoch": 0.4345724907063197, "grad_norm": 0.8946941656966431, "learning_rate": 6.2988192739257414e-06, "loss": 0.0686, "step": 9352 }, { "epoch": 0.4346189591078067, "grad_norm": 1.5699988158839842, "learning_rate": 6.29809257803759e-06, "loss": 0.1021, "step": 9353 }, { "epoch": 0.43466542750929366, "grad_norm": 0.5282981358495031, "learning_rate": 6.297365852746279e-06, "loss": 0.0463, "step": 9354 }, { "epoch": 0.4347118959107807, "grad_norm": 0.6515005429778365, "learning_rate": 6.2966390980682625e-06, "loss": 0.0444, "step": 9355 }, { "epoch": 0.43475836431226766, "grad_norm": 0.9383868076821926, "learning_rate": 6.2959123140200075e-06, "loss": 0.0584, "step": 9356 }, { "epoch": 0.43480483271375464, "grad_norm": 0.5340559980817535, "learning_rate": 6.2951855006179756e-06, "loss": 0.0523, "step": 9357 }, { "epoch": 0.4348513011152416, "grad_norm": 0.47356140124146323, "learning_rate": 6.294458657878626e-06, "loss": 0.058, "step": 9358 }, { "epoch": 0.43489776951672865, "grad_norm": 2.444249059941744, "learning_rate": 6.293731785818429e-06, "loss": 0.0685, "step": 9359 }, { "epoch": 0.4349442379182156, "grad_norm": 1.44746050532042, "learning_rate": 6.293004884453843e-06, "loss": 0.0834, "step": 9360 }, { "epoch": 0.4349907063197026, "grad_norm": 2.855043578304061, "learning_rate": 6.292277953801335e-06, "loss": 0.075, "step": 9361 }, { "epoch": 0.4350371747211896, "grad_norm": 0.7454428036941368, "learning_rate": 6.291550993877373e-06, "loss": 0.0695, "step": 9362 }, { "epoch": 0.43508364312267656, "grad_norm": 1.191814250906689, "learning_rate": 6.29082400469842e-06, "loss": 0.077, "step": 9363 }, { "epoch": 0.4351301115241636, "grad_norm": 1.6953619850046024, "learning_rate": 6.290096986280946e-06, "loss": 0.0533, "step": 9364 }, { "epoch": 0.43517657992565056, "grad_norm": 1.4308425836504084, "learning_rate": 6.289369938641418e-06, "loss": 0.0674, "step": 9365 }, { "epoch": 0.43522304832713754, "grad_norm": 0.8323627704785659, "learning_rate": 6.288642861796302e-06, "loss": 0.0489, "step": 9366 }, { "epoch": 0.4352695167286245, "grad_norm": 0.7733590563293732, "learning_rate": 6.28791575576207e-06, "loss": 0.0694, "step": 9367 }, { "epoch": 0.43531598513011155, "grad_norm": 0.9385706129550271, "learning_rate": 6.28718862055519e-06, "loss": 0.0501, "step": 9368 }, { "epoch": 0.4353624535315985, "grad_norm": 0.9161546516656262, "learning_rate": 6.286461456192135e-06, "loss": 0.0501, "step": 9369 }, { "epoch": 0.4354089219330855, "grad_norm": 1.317901359751587, "learning_rate": 6.285734262689371e-06, "loss": 0.0952, "step": 9370 }, { "epoch": 0.4354553903345725, "grad_norm": 0.6594919134234565, "learning_rate": 6.285007040063376e-06, "loss": 0.0477, "step": 9371 }, { "epoch": 0.43550185873605946, "grad_norm": 1.004847795473009, "learning_rate": 6.2842797883306165e-06, "loss": 0.0661, "step": 9372 }, { "epoch": 0.4355483271375465, "grad_norm": 2.8987058992981716, "learning_rate": 6.2835525075075685e-06, "loss": 0.0722, "step": 9373 }, { "epoch": 0.43559479553903346, "grad_norm": 1.7517376358078864, "learning_rate": 6.2828251976107066e-06, "loss": 0.0792, "step": 9374 }, { "epoch": 0.43564126394052044, "grad_norm": 1.2677613806111072, "learning_rate": 6.282097858656501e-06, "loss": 0.0782, "step": 9375 }, { "epoch": 0.4356877323420074, "grad_norm": 0.5481113577749237, "learning_rate": 6.281370490661433e-06, "loss": 0.0559, "step": 9376 }, { "epoch": 0.43573420074349445, "grad_norm": 0.7213073677356061, "learning_rate": 6.280643093641974e-06, "loss": 0.0699, "step": 9377 }, { "epoch": 0.4357806691449814, "grad_norm": 0.7909316514492495, "learning_rate": 6.2799156676146e-06, "loss": 0.0698, "step": 9378 }, { "epoch": 0.4358271375464684, "grad_norm": 1.9139640593525376, "learning_rate": 6.2791882125957885e-06, "loss": 0.0677, "step": 9379 }, { "epoch": 0.4358736059479554, "grad_norm": 0.7504992174026807, "learning_rate": 6.278460728602019e-06, "loss": 0.0703, "step": 9380 }, { "epoch": 0.43592007434944235, "grad_norm": 0.6641217132277285, "learning_rate": 6.277733215649768e-06, "loss": 0.0543, "step": 9381 }, { "epoch": 0.4359665427509294, "grad_norm": 2.277214333005931, "learning_rate": 6.277005673755514e-06, "loss": 0.0985, "step": 9382 }, { "epoch": 0.43601301115241636, "grad_norm": 0.894815527837439, "learning_rate": 6.27627810293574e-06, "loss": 0.0598, "step": 9383 }, { "epoch": 0.43605947955390334, "grad_norm": 0.6728412337384998, "learning_rate": 6.275550503206922e-06, "loss": 0.0597, "step": 9384 }, { "epoch": 0.4361059479553903, "grad_norm": 0.728306931445586, "learning_rate": 6.274822874585542e-06, "loss": 0.0683, "step": 9385 }, { "epoch": 0.43615241635687735, "grad_norm": 1.3127241851382596, "learning_rate": 6.2740952170880835e-06, "loss": 0.0615, "step": 9386 }, { "epoch": 0.4361988847583643, "grad_norm": 1.5585623998277176, "learning_rate": 6.273367530731026e-06, "loss": 0.0645, "step": 9387 }, { "epoch": 0.4362453531598513, "grad_norm": 1.0622403343961058, "learning_rate": 6.272639815530854e-06, "loss": 0.0787, "step": 9388 }, { "epoch": 0.4362918215613383, "grad_norm": 2.028624610163614, "learning_rate": 6.271912071504052e-06, "loss": 0.0693, "step": 9389 }, { "epoch": 0.43633828996282525, "grad_norm": 0.6948645148306719, "learning_rate": 6.271184298667103e-06, "loss": 0.0627, "step": 9390 }, { "epoch": 0.4363847583643123, "grad_norm": 0.8414244704792866, "learning_rate": 6.27045649703649e-06, "loss": 0.0535, "step": 9391 }, { "epoch": 0.43643122676579926, "grad_norm": 1.2326623484433838, "learning_rate": 6.269728666628701e-06, "loss": 0.0946, "step": 9392 }, { "epoch": 0.43647769516728624, "grad_norm": 0.7643218667082626, "learning_rate": 6.2690008074602215e-06, "loss": 0.0518, "step": 9393 }, { "epoch": 0.4365241635687732, "grad_norm": 1.1889714880691882, "learning_rate": 6.268272919547537e-06, "loss": 0.1049, "step": 9394 }, { "epoch": 0.43657063197026025, "grad_norm": 0.6756454093205471, "learning_rate": 6.2675450029071385e-06, "loss": 0.0579, "step": 9395 }, { "epoch": 0.4366171003717472, "grad_norm": 0.8219819383817849, "learning_rate": 6.26681705755551e-06, "loss": 0.053, "step": 9396 }, { "epoch": 0.4366635687732342, "grad_norm": 0.8658089744468367, "learning_rate": 6.266089083509142e-06, "loss": 0.0952, "step": 9397 }, { "epoch": 0.4367100371747212, "grad_norm": 0.5319229986738045, "learning_rate": 6.265361080784523e-06, "loss": 0.0606, "step": 9398 }, { "epoch": 0.43675650557620815, "grad_norm": 1.272380907512501, "learning_rate": 6.2646330493981445e-06, "loss": 0.0461, "step": 9399 }, { "epoch": 0.4368029739776952, "grad_norm": 0.867677761881555, "learning_rate": 6.263904989366496e-06, "loss": 0.0738, "step": 9400 }, { "epoch": 0.43684944237918216, "grad_norm": 1.019394221889964, "learning_rate": 6.263176900706071e-06, "loss": 0.0629, "step": 9401 }, { "epoch": 0.43689591078066914, "grad_norm": 0.732540635002201, "learning_rate": 6.262448783433358e-06, "loss": 0.0614, "step": 9402 }, { "epoch": 0.4369423791821561, "grad_norm": 0.9280893134173365, "learning_rate": 6.261720637564851e-06, "loss": 0.0774, "step": 9403 }, { "epoch": 0.43698884758364315, "grad_norm": 1.045456750355086, "learning_rate": 6.260992463117045e-06, "loss": 0.0644, "step": 9404 }, { "epoch": 0.4370353159851301, "grad_norm": 1.078712479268607, "learning_rate": 6.260264260106433e-06, "loss": 0.067, "step": 9405 }, { "epoch": 0.4370817843866171, "grad_norm": 0.9411514461245867, "learning_rate": 6.2595360285495065e-06, "loss": 0.0518, "step": 9406 }, { "epoch": 0.4371282527881041, "grad_norm": 0.6453809009779546, "learning_rate": 6.258807768462766e-06, "loss": 0.0415, "step": 9407 }, { "epoch": 0.43717472118959105, "grad_norm": 0.9502644450959947, "learning_rate": 6.258079479862702e-06, "loss": 0.0664, "step": 9408 }, { "epoch": 0.4372211895910781, "grad_norm": 0.967428416783195, "learning_rate": 6.257351162765815e-06, "loss": 0.0478, "step": 9409 }, { "epoch": 0.43726765799256506, "grad_norm": 0.39630389240128844, "learning_rate": 6.2566228171886025e-06, "loss": 0.0377, "step": 9410 }, { "epoch": 0.43731412639405204, "grad_norm": 0.5575409687804308, "learning_rate": 6.255894443147557e-06, "loss": 0.0419, "step": 9411 }, { "epoch": 0.437360594795539, "grad_norm": 0.7479253168529535, "learning_rate": 6.255166040659184e-06, "loss": 0.0584, "step": 9412 }, { "epoch": 0.43740706319702605, "grad_norm": 0.5522740218058086, "learning_rate": 6.254437609739978e-06, "loss": 0.0496, "step": 9413 }, { "epoch": 0.437453531598513, "grad_norm": 1.2822677879275206, "learning_rate": 6.25370915040644e-06, "loss": 0.0911, "step": 9414 }, { "epoch": 0.4375, "grad_norm": 2.2242911007176955, "learning_rate": 6.252980662675069e-06, "loss": 0.1366, "step": 9415 }, { "epoch": 0.437546468401487, "grad_norm": 0.8920186909157836, "learning_rate": 6.252252146562369e-06, "loss": 0.0482, "step": 9416 }, { "epoch": 0.43759293680297395, "grad_norm": 0.9254710707370318, "learning_rate": 6.251523602084839e-06, "loss": 0.0613, "step": 9417 }, { "epoch": 0.437639405204461, "grad_norm": 0.8289250735685743, "learning_rate": 6.250795029258983e-06, "loss": 0.0641, "step": 9418 }, { "epoch": 0.43768587360594796, "grad_norm": 0.8190682327323743, "learning_rate": 6.250066428101302e-06, "loss": 0.0536, "step": 9419 }, { "epoch": 0.43773234200743494, "grad_norm": 0.8572427451392165, "learning_rate": 6.249337798628301e-06, "loss": 0.0535, "step": 9420 }, { "epoch": 0.4377788104089219, "grad_norm": 0.7292151175548827, "learning_rate": 6.2486091408564845e-06, "loss": 0.0464, "step": 9421 }, { "epoch": 0.43782527881040895, "grad_norm": 0.6563898882866105, "learning_rate": 6.247880454802357e-06, "loss": 0.0559, "step": 9422 }, { "epoch": 0.4378717472118959, "grad_norm": 0.950759617031879, "learning_rate": 6.2471517404824225e-06, "loss": 0.0661, "step": 9423 }, { "epoch": 0.4379182156133829, "grad_norm": 12.82219501324587, "learning_rate": 6.24642299791319e-06, "loss": 0.0754, "step": 9424 }, { "epoch": 0.4379646840148699, "grad_norm": 0.9901333121654894, "learning_rate": 6.2456942271111634e-06, "loss": 0.0625, "step": 9425 }, { "epoch": 0.43801115241635685, "grad_norm": 0.6879405017467733, "learning_rate": 6.244965428092852e-06, "loss": 0.0587, "step": 9426 }, { "epoch": 0.4380576208178439, "grad_norm": 1.038581685004234, "learning_rate": 6.244236600874763e-06, "loss": 0.0559, "step": 9427 }, { "epoch": 0.43810408921933086, "grad_norm": 1.1958643324929508, "learning_rate": 6.243507745473406e-06, "loss": 0.0717, "step": 9428 }, { "epoch": 0.43815055762081784, "grad_norm": 4.954752283451561, "learning_rate": 6.24277886190529e-06, "loss": 0.1657, "step": 9429 }, { "epoch": 0.4381970260223048, "grad_norm": 0.6558564179336936, "learning_rate": 6.242049950186925e-06, "loss": 0.0472, "step": 9430 }, { "epoch": 0.43824349442379185, "grad_norm": 1.016488656874362, "learning_rate": 6.241321010334821e-06, "loss": 0.0689, "step": 9431 }, { "epoch": 0.4382899628252788, "grad_norm": 1.282516359480619, "learning_rate": 6.2405920423654885e-06, "loss": 0.0301, "step": 9432 }, { "epoch": 0.4383364312267658, "grad_norm": 0.5112412127395214, "learning_rate": 6.239863046295443e-06, "loss": 0.0581, "step": 9433 }, { "epoch": 0.4383828996282528, "grad_norm": 1.2501050923509343, "learning_rate": 6.239134022141192e-06, "loss": 0.0972, "step": 9434 }, { "epoch": 0.43842936802973975, "grad_norm": 0.6163519446249097, "learning_rate": 6.238404969919252e-06, "loss": 0.0512, "step": 9435 }, { "epoch": 0.4384758364312268, "grad_norm": 1.5808799552284902, "learning_rate": 6.237675889646138e-06, "loss": 0.0664, "step": 9436 }, { "epoch": 0.43852230483271376, "grad_norm": 1.5148558503951994, "learning_rate": 6.23694678133836e-06, "loss": 0.1129, "step": 9437 }, { "epoch": 0.43856877323420074, "grad_norm": 0.8039195177638607, "learning_rate": 6.236217645012435e-06, "loss": 0.0649, "step": 9438 }, { "epoch": 0.4386152416356877, "grad_norm": 0.9642731775766482, "learning_rate": 6.2354884806848825e-06, "loss": 0.0797, "step": 9439 }, { "epoch": 0.43866171003717475, "grad_norm": 0.7397960590148133, "learning_rate": 6.234759288372213e-06, "loss": 0.0679, "step": 9440 }, { "epoch": 0.4387081784386617, "grad_norm": 2.5030647332328857, "learning_rate": 6.234030068090947e-06, "loss": 0.061, "step": 9441 }, { "epoch": 0.4387546468401487, "grad_norm": 3.1370438466704416, "learning_rate": 6.2333008198576e-06, "loss": 0.0576, "step": 9442 }, { "epoch": 0.4388011152416357, "grad_norm": 1.0590016613067628, "learning_rate": 6.232571543688692e-06, "loss": 0.071, "step": 9443 }, { "epoch": 0.43884758364312265, "grad_norm": 0.7958601764394616, "learning_rate": 6.231842239600743e-06, "loss": 0.057, "step": 9444 }, { "epoch": 0.4388940520446097, "grad_norm": 0.7086043369571104, "learning_rate": 6.231112907610269e-06, "loss": 0.0587, "step": 9445 }, { "epoch": 0.43894052044609666, "grad_norm": 0.7631575544725474, "learning_rate": 6.2303835477337925e-06, "loss": 0.0683, "step": 9446 }, { "epoch": 0.43898698884758364, "grad_norm": 0.6832798099056898, "learning_rate": 6.229654159987833e-06, "loss": 0.0612, "step": 9447 }, { "epoch": 0.4390334572490706, "grad_norm": 0.7713749603604348, "learning_rate": 6.228924744388914e-06, "loss": 0.0373, "step": 9448 }, { "epoch": 0.43907992565055765, "grad_norm": 0.9948584332618869, "learning_rate": 6.228195300953554e-06, "loss": 0.0696, "step": 9449 }, { "epoch": 0.4391263940520446, "grad_norm": 0.733571191731355, "learning_rate": 6.227465829698279e-06, "loss": 0.0615, "step": 9450 }, { "epoch": 0.4391728624535316, "grad_norm": 1.1114729613315741, "learning_rate": 6.226736330639612e-06, "loss": 0.05, "step": 9451 }, { "epoch": 0.4392193308550186, "grad_norm": 1.3840722258964788, "learning_rate": 6.226006803794074e-06, "loss": 0.0674, "step": 9452 }, { "epoch": 0.43926579925650555, "grad_norm": 0.8854956570537156, "learning_rate": 6.225277249178194e-06, "loss": 0.0636, "step": 9453 }, { "epoch": 0.4393122676579926, "grad_norm": 1.621572709800757, "learning_rate": 6.224547666808494e-06, "loss": 0.0461, "step": 9454 }, { "epoch": 0.43935873605947956, "grad_norm": 0.5603424905053149, "learning_rate": 6.223818056701498e-06, "loss": 0.0533, "step": 9455 }, { "epoch": 0.43940520446096654, "grad_norm": 1.3595812276540724, "learning_rate": 6.223088418873737e-06, "loss": 0.0612, "step": 9456 }, { "epoch": 0.4394516728624535, "grad_norm": 2.7004222050184485, "learning_rate": 6.222358753341738e-06, "loss": 0.0865, "step": 9457 }, { "epoch": 0.43949814126394054, "grad_norm": 1.1591356094683332, "learning_rate": 6.221629060122024e-06, "loss": 0.0763, "step": 9458 }, { "epoch": 0.4395446096654275, "grad_norm": 1.0351875579141, "learning_rate": 6.220899339231126e-06, "loss": 0.0629, "step": 9459 }, { "epoch": 0.4395910780669145, "grad_norm": 1.0079751364404417, "learning_rate": 6.220169590685574e-06, "loss": 0.0547, "step": 9460 }, { "epoch": 0.4396375464684015, "grad_norm": 0.4914775169349635, "learning_rate": 6.219439814501897e-06, "loss": 0.0576, "step": 9461 }, { "epoch": 0.43968401486988845, "grad_norm": 4.0246882696665836, "learning_rate": 6.218710010696623e-06, "loss": 0.0735, "step": 9462 }, { "epoch": 0.4397304832713755, "grad_norm": 1.0319982717366005, "learning_rate": 6.217980179286286e-06, "loss": 0.0645, "step": 9463 }, { "epoch": 0.43977695167286246, "grad_norm": 0.37506866695138913, "learning_rate": 6.217250320287416e-06, "loss": 0.0475, "step": 9464 }, { "epoch": 0.43982342007434944, "grad_norm": 0.7083273109316317, "learning_rate": 6.216520433716544e-06, "loss": 0.0527, "step": 9465 }, { "epoch": 0.4398698884758364, "grad_norm": 1.432818541103711, "learning_rate": 6.215790519590205e-06, "loss": 0.0408, "step": 9466 }, { "epoch": 0.43991635687732344, "grad_norm": 0.5728907068711023, "learning_rate": 6.21506057792493e-06, "loss": 0.0407, "step": 9467 }, { "epoch": 0.4399628252788104, "grad_norm": 4.304328274727478, "learning_rate": 6.214330608737255e-06, "loss": 0.0959, "step": 9468 }, { "epoch": 0.4400092936802974, "grad_norm": 0.9718089182650286, "learning_rate": 6.213600612043713e-06, "loss": 0.0888, "step": 9469 }, { "epoch": 0.4400557620817844, "grad_norm": 0.6594942809115447, "learning_rate": 6.212870587860841e-06, "loss": 0.0366, "step": 9470 }, { "epoch": 0.44010223048327135, "grad_norm": 1.3064721554605363, "learning_rate": 6.212140536205172e-06, "loss": 0.0635, "step": 9471 }, { "epoch": 0.4401486988847584, "grad_norm": 0.8317531375673558, "learning_rate": 6.211410457093245e-06, "loss": 0.0497, "step": 9472 }, { "epoch": 0.44019516728624536, "grad_norm": 1.2144303321926646, "learning_rate": 6.210680350541597e-06, "loss": 0.0699, "step": 9473 }, { "epoch": 0.44024163568773234, "grad_norm": 0.40137671979424644, "learning_rate": 6.209950216566764e-06, "loss": 0.0396, "step": 9474 }, { "epoch": 0.4402881040892193, "grad_norm": 0.6054615779469145, "learning_rate": 6.209220055185285e-06, "loss": 0.0625, "step": 9475 }, { "epoch": 0.44033457249070634, "grad_norm": 0.986081631053469, "learning_rate": 6.2084898664137e-06, "loss": 0.0769, "step": 9476 }, { "epoch": 0.4403810408921933, "grad_norm": 0.48848182332488027, "learning_rate": 6.207759650268547e-06, "loss": 0.0459, "step": 9477 }, { "epoch": 0.4404275092936803, "grad_norm": 0.6580711730773436, "learning_rate": 6.207029406766366e-06, "loss": 0.0721, "step": 9478 }, { "epoch": 0.4404739776951673, "grad_norm": 2.465414459279239, "learning_rate": 6.206299135923699e-06, "loss": 0.1023, "step": 9479 }, { "epoch": 0.44052044609665425, "grad_norm": 0.6702877042938626, "learning_rate": 6.205568837757088e-06, "loss": 0.0657, "step": 9480 }, { "epoch": 0.4405669144981413, "grad_norm": 0.7071564761252717, "learning_rate": 6.204838512283073e-06, "loss": 0.0861, "step": 9481 }, { "epoch": 0.44061338289962826, "grad_norm": 1.0655607633864523, "learning_rate": 6.204108159518197e-06, "loss": 0.0658, "step": 9482 }, { "epoch": 0.44065985130111524, "grad_norm": 1.0179010210849089, "learning_rate": 6.203377779479005e-06, "loss": 0.0441, "step": 9483 }, { "epoch": 0.4407063197026022, "grad_norm": 1.4187042636072318, "learning_rate": 6.202647372182039e-06, "loss": 0.0958, "step": 9484 }, { "epoch": 0.44075278810408924, "grad_norm": 1.3051862731333082, "learning_rate": 6.201916937643845e-06, "loss": 0.06, "step": 9485 }, { "epoch": 0.4407992565055762, "grad_norm": 0.5916571285511375, "learning_rate": 6.201186475880967e-06, "loss": 0.0463, "step": 9486 }, { "epoch": 0.4408457249070632, "grad_norm": 0.712658689660231, "learning_rate": 6.200455986909953e-06, "loss": 0.0663, "step": 9487 }, { "epoch": 0.4408921933085502, "grad_norm": 0.9804084259354207, "learning_rate": 6.199725470747344e-06, "loss": 0.066, "step": 9488 }, { "epoch": 0.44093866171003715, "grad_norm": 1.2967300218924243, "learning_rate": 6.198994927409691e-06, "loss": 0.0871, "step": 9489 }, { "epoch": 0.4409851301115242, "grad_norm": 0.713605759396454, "learning_rate": 6.198264356913544e-06, "loss": 0.0456, "step": 9490 }, { "epoch": 0.44103159851301116, "grad_norm": 2.438216190562147, "learning_rate": 6.197533759275445e-06, "loss": 0.0586, "step": 9491 }, { "epoch": 0.44107806691449813, "grad_norm": 0.6761241336717847, "learning_rate": 6.196803134511949e-06, "loss": 0.0429, "step": 9492 }, { "epoch": 0.4411245353159851, "grad_norm": 0.7396599581074031, "learning_rate": 6.1960724826395994e-06, "loss": 0.0742, "step": 9493 }, { "epoch": 0.44117100371747214, "grad_norm": 4.6725379825179125, "learning_rate": 6.1953418036749505e-06, "loss": 0.08, "step": 9494 }, { "epoch": 0.4412174721189591, "grad_norm": 1.0246765942448595, "learning_rate": 6.194611097634551e-06, "loss": 0.0592, "step": 9495 }, { "epoch": 0.4412639405204461, "grad_norm": 0.6832897115385443, "learning_rate": 6.193880364534952e-06, "loss": 0.0428, "step": 9496 }, { "epoch": 0.4413104089219331, "grad_norm": 3.407689101920403, "learning_rate": 6.193149604392709e-06, "loss": 0.0689, "step": 9497 }, { "epoch": 0.44135687732342005, "grad_norm": 0.7269319714753509, "learning_rate": 6.192418817224371e-06, "loss": 0.0335, "step": 9498 }, { "epoch": 0.4414033457249071, "grad_norm": 0.8486765483701673, "learning_rate": 6.1916880030464896e-06, "loss": 0.0493, "step": 9499 }, { "epoch": 0.44144981412639406, "grad_norm": 1.2424762365074384, "learning_rate": 6.190957161875622e-06, "loss": 0.0997, "step": 9500 }, { "epoch": 0.44149628252788103, "grad_norm": 0.6384215197987564, "learning_rate": 6.190226293728322e-06, "loss": 0.0779, "step": 9501 }, { "epoch": 0.441542750929368, "grad_norm": 2.2563611037659115, "learning_rate": 6.189495398621142e-06, "loss": 0.0862, "step": 9502 }, { "epoch": 0.44158921933085504, "grad_norm": 0.7372527164383834, "learning_rate": 6.188764476570641e-06, "loss": 0.0746, "step": 9503 }, { "epoch": 0.441635687732342, "grad_norm": 0.4530869993809524, "learning_rate": 6.188033527593374e-06, "loss": 0.0329, "step": 9504 }, { "epoch": 0.441682156133829, "grad_norm": 1.1820572608648783, "learning_rate": 6.187302551705895e-06, "loss": 0.0959, "step": 9505 }, { "epoch": 0.441728624535316, "grad_norm": 0.8618879559898474, "learning_rate": 6.186571548924764e-06, "loss": 0.0371, "step": 9506 }, { "epoch": 0.44177509293680295, "grad_norm": 0.8166770457080483, "learning_rate": 6.18584051926654e-06, "loss": 0.0461, "step": 9507 }, { "epoch": 0.44182156133829, "grad_norm": 0.6508599855909176, "learning_rate": 6.185109462747779e-06, "loss": 0.0327, "step": 9508 }, { "epoch": 0.44186802973977696, "grad_norm": 0.6445502480974727, "learning_rate": 6.184378379385042e-06, "loss": 0.0508, "step": 9509 }, { "epoch": 0.44191449814126393, "grad_norm": 1.0666574883187405, "learning_rate": 6.183647269194888e-06, "loss": 0.079, "step": 9510 }, { "epoch": 0.4419609665427509, "grad_norm": 1.2602543328908131, "learning_rate": 6.1829161321938776e-06, "loss": 0.0747, "step": 9511 }, { "epoch": 0.44200743494423794, "grad_norm": 0.8100094443648153, "learning_rate": 6.182184968398571e-06, "loss": 0.0718, "step": 9512 }, { "epoch": 0.4420539033457249, "grad_norm": 0.6973518902431427, "learning_rate": 6.181453777825532e-06, "loss": 0.0729, "step": 9513 }, { "epoch": 0.4421003717472119, "grad_norm": 0.8258439314142015, "learning_rate": 6.180722560491321e-06, "loss": 0.0651, "step": 9514 }, { "epoch": 0.44214684014869887, "grad_norm": 0.6645910791465379, "learning_rate": 6.1799913164125e-06, "loss": 0.0568, "step": 9515 }, { "epoch": 0.44219330855018585, "grad_norm": 0.5064692255569374, "learning_rate": 6.179260045605637e-06, "loss": 0.0361, "step": 9516 }, { "epoch": 0.4422397769516729, "grad_norm": 0.6711355998398251, "learning_rate": 6.17852874808729e-06, "loss": 0.0696, "step": 9517 }, { "epoch": 0.44228624535315986, "grad_norm": 0.9096932182407955, "learning_rate": 6.177797423874027e-06, "loss": 0.0543, "step": 9518 }, { "epoch": 0.44233271375464683, "grad_norm": 1.4300492318192404, "learning_rate": 6.177066072982414e-06, "loss": 0.0629, "step": 9519 }, { "epoch": 0.4423791821561338, "grad_norm": 0.5106490439400864, "learning_rate": 6.1763346954290135e-06, "loss": 0.0593, "step": 9520 }, { "epoch": 0.44242565055762084, "grad_norm": 0.8327634224919364, "learning_rate": 6.1756032912303955e-06, "loss": 0.0497, "step": 9521 }, { "epoch": 0.4424721189591078, "grad_norm": 0.7484443452973392, "learning_rate": 6.174871860403126e-06, "loss": 0.0555, "step": 9522 }, { "epoch": 0.4425185873605948, "grad_norm": 0.7936548014400564, "learning_rate": 6.174140402963771e-06, "loss": 0.0471, "step": 9523 }, { "epoch": 0.44256505576208177, "grad_norm": 0.7288555185646409, "learning_rate": 6.1734089189289e-06, "loss": 0.0602, "step": 9524 }, { "epoch": 0.44261152416356875, "grad_norm": 1.4641543787400435, "learning_rate": 6.172677408315083e-06, "loss": 0.0625, "step": 9525 }, { "epoch": 0.4426579925650558, "grad_norm": 0.5432537405232226, "learning_rate": 6.171945871138889e-06, "loss": 0.0427, "step": 9526 }, { "epoch": 0.44270446096654276, "grad_norm": 0.6064677387653973, "learning_rate": 6.171214307416885e-06, "loss": 0.0438, "step": 9527 }, { "epoch": 0.44275092936802973, "grad_norm": 1.5666710028576294, "learning_rate": 6.170482717165648e-06, "loss": 0.0635, "step": 9528 }, { "epoch": 0.4427973977695167, "grad_norm": 0.5804461640239591, "learning_rate": 6.169751100401743e-06, "loss": 0.0505, "step": 9529 }, { "epoch": 0.44284386617100374, "grad_norm": 1.2907047181530682, "learning_rate": 6.169019457141745e-06, "loss": 0.074, "step": 9530 }, { "epoch": 0.4428903345724907, "grad_norm": 0.8152507269075313, "learning_rate": 6.168287787402226e-06, "loss": 0.0531, "step": 9531 }, { "epoch": 0.4429368029739777, "grad_norm": 2.56346887455487, "learning_rate": 6.167556091199758e-06, "loss": 0.0551, "step": 9532 }, { "epoch": 0.44298327137546467, "grad_norm": 0.9005013422426282, "learning_rate": 6.166824368550917e-06, "loss": 0.0681, "step": 9533 }, { "epoch": 0.44302973977695165, "grad_norm": 0.7708349704722067, "learning_rate": 6.166092619472276e-06, "loss": 0.049, "step": 9534 }, { "epoch": 0.4430762081784387, "grad_norm": 0.9290679711342519, "learning_rate": 6.16536084398041e-06, "loss": 0.0666, "step": 9535 }, { "epoch": 0.44312267657992566, "grad_norm": 4.558120781465524, "learning_rate": 6.164629042091894e-06, "loss": 0.064, "step": 9536 }, { "epoch": 0.44316914498141263, "grad_norm": 0.9932151668358538, "learning_rate": 6.163897213823304e-06, "loss": 0.0593, "step": 9537 }, { "epoch": 0.4432156133828996, "grad_norm": 1.1337505012953562, "learning_rate": 6.163165359191218e-06, "loss": 0.069, "step": 9538 }, { "epoch": 0.44326208178438664, "grad_norm": 1.404000339570314, "learning_rate": 6.1624334782122116e-06, "loss": 0.0754, "step": 9539 }, { "epoch": 0.4433085501858736, "grad_norm": 1.417030049966965, "learning_rate": 6.161701570902864e-06, "loss": 0.0706, "step": 9540 }, { "epoch": 0.4433550185873606, "grad_norm": 1.0964231147565138, "learning_rate": 6.160969637279752e-06, "loss": 0.0806, "step": 9541 }, { "epoch": 0.44340148698884757, "grad_norm": 0.692956484413477, "learning_rate": 6.160237677359457e-06, "loss": 0.0744, "step": 9542 }, { "epoch": 0.44344795539033455, "grad_norm": 1.2681254081654285, "learning_rate": 6.159505691158558e-06, "loss": 0.0856, "step": 9543 }, { "epoch": 0.4434944237918216, "grad_norm": 1.0937506872559326, "learning_rate": 6.1587736786936326e-06, "loss": 0.0875, "step": 9544 }, { "epoch": 0.44354089219330856, "grad_norm": 1.4900931681294525, "learning_rate": 6.158041639981267e-06, "loss": 0.0326, "step": 9545 }, { "epoch": 0.44358736059479553, "grad_norm": 1.2823867721567286, "learning_rate": 6.157309575038036e-06, "loss": 0.0662, "step": 9546 }, { "epoch": 0.4436338289962825, "grad_norm": 0.4583313216031702, "learning_rate": 6.156577483880527e-06, "loss": 0.0389, "step": 9547 }, { "epoch": 0.44368029739776954, "grad_norm": 1.224278110356909, "learning_rate": 6.155845366525321e-06, "loss": 0.0928, "step": 9548 }, { "epoch": 0.4437267657992565, "grad_norm": 0.5506289262484056, "learning_rate": 6.1551132229889995e-06, "loss": 0.0335, "step": 9549 }, { "epoch": 0.4437732342007435, "grad_norm": 1.6681752719998648, "learning_rate": 6.154381053288148e-06, "loss": 0.0493, "step": 9550 }, { "epoch": 0.44381970260223047, "grad_norm": 0.5477448204642996, "learning_rate": 6.153648857439353e-06, "loss": 0.0582, "step": 9551 }, { "epoch": 0.44386617100371745, "grad_norm": 1.019762124955877, "learning_rate": 6.1529166354591955e-06, "loss": 0.0801, "step": 9552 }, { "epoch": 0.4439126394052045, "grad_norm": 1.8833510294906501, "learning_rate": 6.152184387364262e-06, "loss": 0.0941, "step": 9553 }, { "epoch": 0.44395910780669146, "grad_norm": 1.6542467833697336, "learning_rate": 6.15145211317114e-06, "loss": 0.0715, "step": 9554 }, { "epoch": 0.44400557620817843, "grad_norm": 0.9484950980693075, "learning_rate": 6.150719812896416e-06, "loss": 0.063, "step": 9555 }, { "epoch": 0.4440520446096654, "grad_norm": 1.0028911315698754, "learning_rate": 6.149987486556677e-06, "loss": 0.0766, "step": 9556 }, { "epoch": 0.44409851301115244, "grad_norm": 1.223651076067462, "learning_rate": 6.149255134168513e-06, "loss": 0.0642, "step": 9557 }, { "epoch": 0.4441449814126394, "grad_norm": 0.6981937938206636, "learning_rate": 6.148522755748507e-06, "loss": 0.0506, "step": 9558 }, { "epoch": 0.4441914498141264, "grad_norm": 0.7326422763541075, "learning_rate": 6.147790351313254e-06, "loss": 0.0756, "step": 9559 }, { "epoch": 0.44423791821561337, "grad_norm": 0.863333221809484, "learning_rate": 6.147057920879341e-06, "loss": 0.0707, "step": 9560 }, { "epoch": 0.44428438661710035, "grad_norm": 0.6477167492485413, "learning_rate": 6.146325464463358e-06, "loss": 0.0603, "step": 9561 }, { "epoch": 0.4443308550185874, "grad_norm": 0.4101059276113569, "learning_rate": 6.145592982081898e-06, "loss": 0.0493, "step": 9562 }, { "epoch": 0.44437732342007435, "grad_norm": 2.322103540477512, "learning_rate": 6.1448604737515515e-06, "loss": 0.0644, "step": 9563 }, { "epoch": 0.44442379182156133, "grad_norm": 0.8337803248265362, "learning_rate": 6.14412793948891e-06, "loss": 0.0703, "step": 9564 }, { "epoch": 0.4444702602230483, "grad_norm": 0.951938271171144, "learning_rate": 6.143395379310566e-06, "loss": 0.0653, "step": 9565 }, { "epoch": 0.44451672862453534, "grad_norm": 1.0322175861053033, "learning_rate": 6.142662793233114e-06, "loss": 0.0678, "step": 9566 }, { "epoch": 0.4445631970260223, "grad_norm": 0.7893352132237029, "learning_rate": 6.141930181273146e-06, "loss": 0.0541, "step": 9567 }, { "epoch": 0.4446096654275093, "grad_norm": 2.174482872999838, "learning_rate": 6.141197543447257e-06, "loss": 0.0588, "step": 9568 }, { "epoch": 0.44465613382899627, "grad_norm": 0.9027393068730144, "learning_rate": 6.140464879772046e-06, "loss": 0.0586, "step": 9569 }, { "epoch": 0.44470260223048325, "grad_norm": 0.7588164804532316, "learning_rate": 6.139732190264101e-06, "loss": 0.0865, "step": 9570 }, { "epoch": 0.4447490706319703, "grad_norm": 0.784909064614231, "learning_rate": 6.138999474940025e-06, "loss": 0.0732, "step": 9571 }, { "epoch": 0.44479553903345725, "grad_norm": 0.9610904180718077, "learning_rate": 6.1382667338164115e-06, "loss": 0.0599, "step": 9572 }, { "epoch": 0.44484200743494423, "grad_norm": 0.6799305362491312, "learning_rate": 6.137533966909858e-06, "loss": 0.0587, "step": 9573 }, { "epoch": 0.4448884758364312, "grad_norm": 0.5686845400561678, "learning_rate": 6.136801174236964e-06, "loss": 0.0442, "step": 9574 }, { "epoch": 0.44493494423791824, "grad_norm": 0.765264090123855, "learning_rate": 6.136068355814325e-06, "loss": 0.0745, "step": 9575 }, { "epoch": 0.4449814126394052, "grad_norm": 0.8525116076439139, "learning_rate": 6.1353355116585445e-06, "loss": 0.0567, "step": 9576 }, { "epoch": 0.4450278810408922, "grad_norm": 0.5191584098564045, "learning_rate": 6.134602641786217e-06, "loss": 0.0522, "step": 9577 }, { "epoch": 0.44507434944237917, "grad_norm": 0.6994109854846619, "learning_rate": 6.133869746213947e-06, "loss": 0.0609, "step": 9578 }, { "epoch": 0.44512081784386615, "grad_norm": 0.5430487505802373, "learning_rate": 6.133136824958334e-06, "loss": 0.0454, "step": 9579 }, { "epoch": 0.4451672862453532, "grad_norm": 0.8238251580374408, "learning_rate": 6.132403878035979e-06, "loss": 0.0638, "step": 9580 }, { "epoch": 0.44521375464684015, "grad_norm": 1.1444846863782063, "learning_rate": 6.131670905463486e-06, "loss": 0.0818, "step": 9581 }, { "epoch": 0.44526022304832713, "grad_norm": 2.463507047668214, "learning_rate": 6.1309379072574536e-06, "loss": 0.0775, "step": 9582 }, { "epoch": 0.4453066914498141, "grad_norm": 0.8950751011263617, "learning_rate": 6.130204883434488e-06, "loss": 0.0595, "step": 9583 }, { "epoch": 0.44535315985130114, "grad_norm": 0.4832449268115024, "learning_rate": 6.129471834011192e-06, "loss": 0.0409, "step": 9584 }, { "epoch": 0.4453996282527881, "grad_norm": 1.780693065500563, "learning_rate": 6.12873875900417e-06, "loss": 0.1016, "step": 9585 }, { "epoch": 0.4454460966542751, "grad_norm": 0.6292826437685284, "learning_rate": 6.128005658430029e-06, "loss": 0.0693, "step": 9586 }, { "epoch": 0.44549256505576207, "grad_norm": 0.6901000609708663, "learning_rate": 6.127272532305372e-06, "loss": 0.0482, "step": 9587 }, { "epoch": 0.44553903345724905, "grad_norm": 0.9495703940245758, "learning_rate": 6.126539380646804e-06, "loss": 0.0553, "step": 9588 }, { "epoch": 0.4455855018587361, "grad_norm": 0.6117399797796326, "learning_rate": 6.125806203470934e-06, "loss": 0.0554, "step": 9589 }, { "epoch": 0.44563197026022305, "grad_norm": 1.2905809747919035, "learning_rate": 6.125073000794369e-06, "loss": 0.0766, "step": 9590 }, { "epoch": 0.44567843866171003, "grad_norm": 0.6279285107572085, "learning_rate": 6.124339772633718e-06, "loss": 0.0473, "step": 9591 }, { "epoch": 0.445724907063197, "grad_norm": 1.4596577513621982, "learning_rate": 6.123606519005585e-06, "loss": 0.0659, "step": 9592 }, { "epoch": 0.44577137546468404, "grad_norm": 0.927665515973014, "learning_rate": 6.122873239926584e-06, "loss": 0.0498, "step": 9593 }, { "epoch": 0.445817843866171, "grad_norm": 1.3713149217662375, "learning_rate": 6.122139935413321e-06, "loss": 0.1138, "step": 9594 }, { "epoch": 0.445864312267658, "grad_norm": 1.963821053019732, "learning_rate": 6.121406605482407e-06, "loss": 0.0647, "step": 9595 }, { "epoch": 0.44591078066914497, "grad_norm": 1.1592755910745733, "learning_rate": 6.120673250150455e-06, "loss": 0.0655, "step": 9596 }, { "epoch": 0.44595724907063194, "grad_norm": 1.1312133163015665, "learning_rate": 6.119939869434073e-06, "loss": 0.0705, "step": 9597 }, { "epoch": 0.446003717472119, "grad_norm": 1.4568770315596233, "learning_rate": 6.119206463349875e-06, "loss": 0.0674, "step": 9598 }, { "epoch": 0.44605018587360595, "grad_norm": 4.780033687726117, "learning_rate": 6.118473031914471e-06, "loss": 0.0796, "step": 9599 }, { "epoch": 0.44609665427509293, "grad_norm": 0.5816100575604437, "learning_rate": 6.1177395751444775e-06, "loss": 0.0691, "step": 9600 }, { "epoch": 0.4461431226765799, "grad_norm": 0.5554275092611158, "learning_rate": 6.117006093056506e-06, "loss": 0.0393, "step": 9601 }, { "epoch": 0.44618959107806694, "grad_norm": 0.7072357783448171, "learning_rate": 6.11627258566717e-06, "loss": 0.0526, "step": 9602 }, { "epoch": 0.4462360594795539, "grad_norm": 0.9689902144521519, "learning_rate": 6.115539052993086e-06, "loss": 0.0679, "step": 9603 }, { "epoch": 0.4462825278810409, "grad_norm": 1.3633617670628573, "learning_rate": 6.114805495050867e-06, "loss": 0.061, "step": 9604 }, { "epoch": 0.44632899628252787, "grad_norm": 0.8814301297474828, "learning_rate": 6.114071911857132e-06, "loss": 0.0597, "step": 9605 }, { "epoch": 0.44637546468401484, "grad_norm": 0.664034616663783, "learning_rate": 6.113338303428495e-06, "loss": 0.0611, "step": 9606 }, { "epoch": 0.4464219330855019, "grad_norm": 0.8618390858231053, "learning_rate": 6.112604669781572e-06, "loss": 0.0674, "step": 9607 }, { "epoch": 0.44646840148698885, "grad_norm": 0.8654324257870714, "learning_rate": 6.111871010932985e-06, "loss": 0.0568, "step": 9608 }, { "epoch": 0.44651486988847583, "grad_norm": 1.076993213284953, "learning_rate": 6.1111373268993465e-06, "loss": 0.0757, "step": 9609 }, { "epoch": 0.4465613382899628, "grad_norm": 0.5901584036974201, "learning_rate": 6.110403617697282e-06, "loss": 0.0574, "step": 9610 }, { "epoch": 0.44660780669144984, "grad_norm": 0.8431082540830267, "learning_rate": 6.1096698833434035e-06, "loss": 0.0707, "step": 9611 }, { "epoch": 0.4466542750929368, "grad_norm": 1.161965237212846, "learning_rate": 6.108936123854335e-06, "loss": 0.0819, "step": 9612 }, { "epoch": 0.4467007434944238, "grad_norm": 0.7223139558937894, "learning_rate": 6.108202339246697e-06, "loss": 0.0574, "step": 9613 }, { "epoch": 0.44674721189591077, "grad_norm": 0.6629224372336194, "learning_rate": 6.107468529537109e-06, "loss": 0.0532, "step": 9614 }, { "epoch": 0.44679368029739774, "grad_norm": 0.9622896428745265, "learning_rate": 6.1067346947421925e-06, "loss": 0.0779, "step": 9615 }, { "epoch": 0.4468401486988848, "grad_norm": 1.5248459769449032, "learning_rate": 6.106000834878571e-06, "loss": 0.0767, "step": 9616 }, { "epoch": 0.44688661710037175, "grad_norm": 1.3980875057580637, "learning_rate": 6.105266949962867e-06, "loss": 0.051, "step": 9617 }, { "epoch": 0.44693308550185873, "grad_norm": 1.1790213479377165, "learning_rate": 6.104533040011703e-06, "loss": 0.0694, "step": 9618 }, { "epoch": 0.4469795539033457, "grad_norm": 0.9976959688729232, "learning_rate": 6.103799105041703e-06, "loss": 0.0535, "step": 9619 }, { "epoch": 0.44702602230483274, "grad_norm": 2.341125055482052, "learning_rate": 6.1030651450694924e-06, "loss": 0.0595, "step": 9620 }, { "epoch": 0.4470724907063197, "grad_norm": 0.423046176114467, "learning_rate": 6.102331160111693e-06, "loss": 0.0369, "step": 9621 }, { "epoch": 0.4471189591078067, "grad_norm": 0.818984496436234, "learning_rate": 6.1015971501849354e-06, "loss": 0.0461, "step": 9622 }, { "epoch": 0.44716542750929367, "grad_norm": 0.6095021356235625, "learning_rate": 6.100863115305842e-06, "loss": 0.049, "step": 9623 }, { "epoch": 0.44721189591078064, "grad_norm": 0.49850240876816565, "learning_rate": 6.10012905549104e-06, "loss": 0.0356, "step": 9624 }, { "epoch": 0.4472583643122677, "grad_norm": 1.5959305599593754, "learning_rate": 6.099394970757157e-06, "loss": 0.0559, "step": 9625 }, { "epoch": 0.44730483271375465, "grad_norm": 3.680183621333897, "learning_rate": 6.098660861120821e-06, "loss": 0.0784, "step": 9626 }, { "epoch": 0.44735130111524163, "grad_norm": 1.5058709414159557, "learning_rate": 6.0979267265986594e-06, "loss": 0.0735, "step": 9627 }, { "epoch": 0.4473977695167286, "grad_norm": 0.9169762392989569, "learning_rate": 6.097192567207304e-06, "loss": 0.0476, "step": 9628 }, { "epoch": 0.44744423791821564, "grad_norm": 3.1293891000509277, "learning_rate": 6.0964583829633806e-06, "loss": 0.1153, "step": 9629 }, { "epoch": 0.4474907063197026, "grad_norm": 1.010197994619825, "learning_rate": 6.095724173883521e-06, "loss": 0.0779, "step": 9630 }, { "epoch": 0.4475371747211896, "grad_norm": 0.5434189507817756, "learning_rate": 6.094989939984356e-06, "loss": 0.0503, "step": 9631 }, { "epoch": 0.44758364312267657, "grad_norm": 1.2749581103144494, "learning_rate": 6.094255681282516e-06, "loss": 0.0584, "step": 9632 }, { "epoch": 0.44763011152416354, "grad_norm": 0.6319924233648108, "learning_rate": 6.093521397794633e-06, "loss": 0.066, "step": 9633 }, { "epoch": 0.4476765799256506, "grad_norm": 0.6753958530656232, "learning_rate": 6.092787089537341e-06, "loss": 0.0786, "step": 9634 }, { "epoch": 0.44772304832713755, "grad_norm": 0.7267161839767645, "learning_rate": 6.092052756527269e-06, "loss": 0.076, "step": 9635 }, { "epoch": 0.44776951672862453, "grad_norm": 0.7084381914152332, "learning_rate": 6.091318398781053e-06, "loss": 0.0492, "step": 9636 }, { "epoch": 0.4478159851301115, "grad_norm": 0.6310926270471994, "learning_rate": 6.0905840163153286e-06, "loss": 0.0681, "step": 9637 }, { "epoch": 0.44786245353159854, "grad_norm": 0.4929826029270219, "learning_rate": 6.089849609146726e-06, "loss": 0.0588, "step": 9638 }, { "epoch": 0.4479089219330855, "grad_norm": 0.9937488790016049, "learning_rate": 6.089115177291884e-06, "loss": 0.0825, "step": 9639 }, { "epoch": 0.4479553903345725, "grad_norm": 2.0598503356477766, "learning_rate": 6.088380720767438e-06, "loss": 0.0681, "step": 9640 }, { "epoch": 0.44800185873605947, "grad_norm": 1.5353135863556988, "learning_rate": 6.087646239590022e-06, "loss": 0.084, "step": 9641 }, { "epoch": 0.44804832713754644, "grad_norm": 0.6729286746118779, "learning_rate": 6.086911733776273e-06, "loss": 0.0586, "step": 9642 }, { "epoch": 0.4480947955390335, "grad_norm": 0.6689828843902422, "learning_rate": 6.0861772033428305e-06, "loss": 0.0667, "step": 9643 }, { "epoch": 0.44814126394052045, "grad_norm": 0.6620077445928024, "learning_rate": 6.085442648306331e-06, "loss": 0.0668, "step": 9644 }, { "epoch": 0.4481877323420074, "grad_norm": 1.2657380311271915, "learning_rate": 6.084708068683411e-06, "loss": 0.0897, "step": 9645 }, { "epoch": 0.4482342007434944, "grad_norm": 0.9433766554955039, "learning_rate": 6.083973464490716e-06, "loss": 0.0772, "step": 9646 }, { "epoch": 0.44828066914498144, "grad_norm": 1.0324355735875617, "learning_rate": 6.083238835744878e-06, "loss": 0.0898, "step": 9647 }, { "epoch": 0.4483271375464684, "grad_norm": 2.22687226668661, "learning_rate": 6.082504182462541e-06, "loss": 0.0534, "step": 9648 }, { "epoch": 0.4483736059479554, "grad_norm": 0.48306325596053296, "learning_rate": 6.0817695046603455e-06, "loss": 0.0378, "step": 9649 }, { "epoch": 0.44842007434944237, "grad_norm": 0.5590006821932471, "learning_rate": 6.08103480235493e-06, "loss": 0.0579, "step": 9650 }, { "epoch": 0.44846654275092934, "grad_norm": 0.8669157885480945, "learning_rate": 6.080300075562942e-06, "loss": 0.07, "step": 9651 }, { "epoch": 0.4485130111524164, "grad_norm": 1.6494708267471143, "learning_rate": 6.079565324301018e-06, "loss": 0.0642, "step": 9652 }, { "epoch": 0.44855947955390335, "grad_norm": 1.9223490768719274, "learning_rate": 6.078830548585803e-06, "loss": 0.069, "step": 9653 }, { "epoch": 0.4486059479553903, "grad_norm": 0.5100379936002921, "learning_rate": 6.078095748433943e-06, "loss": 0.0441, "step": 9654 }, { "epoch": 0.4486524163568773, "grad_norm": 0.7488520774611218, "learning_rate": 6.0773609238620765e-06, "loss": 0.0545, "step": 9655 }, { "epoch": 0.44869888475836434, "grad_norm": 0.8988215804541781, "learning_rate": 6.0766260748868536e-06, "loss": 0.0736, "step": 9656 }, { "epoch": 0.4487453531598513, "grad_norm": 1.100261895737129, "learning_rate": 6.075891201524915e-06, "loss": 0.0739, "step": 9657 }, { "epoch": 0.4487918215613383, "grad_norm": 1.5024015289397588, "learning_rate": 6.0751563037929095e-06, "loss": 0.0557, "step": 9658 }, { "epoch": 0.44883828996282527, "grad_norm": 0.7392857288907749, "learning_rate": 6.074421381707482e-06, "loss": 0.0638, "step": 9659 }, { "epoch": 0.44888475836431224, "grad_norm": 0.535962086890735, "learning_rate": 6.073686435285279e-06, "loss": 0.0427, "step": 9660 }, { "epoch": 0.4489312267657993, "grad_norm": 0.6732414219917596, "learning_rate": 6.072951464542949e-06, "loss": 0.0594, "step": 9661 }, { "epoch": 0.44897769516728625, "grad_norm": 0.6792083677970303, "learning_rate": 6.072216469497138e-06, "loss": 0.0593, "step": 9662 }, { "epoch": 0.4490241635687732, "grad_norm": 1.2456246918336384, "learning_rate": 6.071481450164497e-06, "loss": 0.0686, "step": 9663 }, { "epoch": 0.4490706319702602, "grad_norm": 1.0904759547036451, "learning_rate": 6.070746406561672e-06, "loss": 0.0533, "step": 9664 }, { "epoch": 0.44911710037174724, "grad_norm": 0.7827159679059964, "learning_rate": 6.070011338705315e-06, "loss": 0.0478, "step": 9665 }, { "epoch": 0.4491635687732342, "grad_norm": 1.0850257855617267, "learning_rate": 6.0692762466120746e-06, "loss": 0.0559, "step": 9666 }, { "epoch": 0.4492100371747212, "grad_norm": 0.5965592243905472, "learning_rate": 6.068541130298601e-06, "loss": 0.0521, "step": 9667 }, { "epoch": 0.44925650557620816, "grad_norm": 0.44185665862417106, "learning_rate": 6.0678059897815465e-06, "loss": 0.0427, "step": 9668 }, { "epoch": 0.44930297397769514, "grad_norm": 0.9574792419429501, "learning_rate": 6.067070825077563e-06, "loss": 0.0542, "step": 9669 }, { "epoch": 0.4493494423791822, "grad_norm": 0.7855278390833146, "learning_rate": 6.066335636203302e-06, "loss": 0.058, "step": 9670 }, { "epoch": 0.44939591078066915, "grad_norm": 0.6809971839768527, "learning_rate": 6.065600423175416e-06, "loss": 0.0553, "step": 9671 }, { "epoch": 0.4494423791821561, "grad_norm": 2.1455997012576606, "learning_rate": 6.06486518601056e-06, "loss": 0.076, "step": 9672 }, { "epoch": 0.4494888475836431, "grad_norm": 0.6351700718667188, "learning_rate": 6.064129924725387e-06, "loss": 0.0398, "step": 9673 }, { "epoch": 0.44953531598513014, "grad_norm": 0.9922662789853531, "learning_rate": 6.06339463933655e-06, "loss": 0.0723, "step": 9674 }, { "epoch": 0.4495817843866171, "grad_norm": 0.831267347735975, "learning_rate": 6.062659329860706e-06, "loss": 0.0631, "step": 9675 }, { "epoch": 0.4496282527881041, "grad_norm": 0.664575527165418, "learning_rate": 6.061923996314509e-06, "loss": 0.0503, "step": 9676 }, { "epoch": 0.44967472118959106, "grad_norm": 1.931764633388539, "learning_rate": 6.061188638714616e-06, "loss": 0.0639, "step": 9677 }, { "epoch": 0.44972118959107804, "grad_norm": 0.5578192401988262, "learning_rate": 6.060453257077686e-06, "loss": 0.0633, "step": 9678 }, { "epoch": 0.4497676579925651, "grad_norm": 0.7893646304296175, "learning_rate": 6.059717851420369e-06, "loss": 0.0557, "step": 9679 }, { "epoch": 0.44981412639405205, "grad_norm": 0.41108512435553185, "learning_rate": 6.058982421759331e-06, "loss": 0.0373, "step": 9680 }, { "epoch": 0.449860594795539, "grad_norm": 1.371678037965828, "learning_rate": 6.058246968111226e-06, "loss": 0.0801, "step": 9681 }, { "epoch": 0.449907063197026, "grad_norm": 0.9085387909915583, "learning_rate": 6.0575114904927114e-06, "loss": 0.0582, "step": 9682 }, { "epoch": 0.44995353159851303, "grad_norm": 0.8170113178496559, "learning_rate": 6.0567759889204494e-06, "loss": 0.0638, "step": 9683 }, { "epoch": 0.45, "grad_norm": 0.7069344782801262, "learning_rate": 6.056040463411099e-06, "loss": 0.0542, "step": 9684 }, { "epoch": 0.450046468401487, "grad_norm": 0.4619846665658838, "learning_rate": 6.055304913981322e-06, "loss": 0.0394, "step": 9685 }, { "epoch": 0.45009293680297396, "grad_norm": 1.0474106925929805, "learning_rate": 6.054569340647777e-06, "loss": 0.0529, "step": 9686 }, { "epoch": 0.45013940520446094, "grad_norm": 1.191685945807408, "learning_rate": 6.053833743427126e-06, "loss": 0.0453, "step": 9687 }, { "epoch": 0.450185873605948, "grad_norm": 0.5493094918287529, "learning_rate": 6.053098122336033e-06, "loss": 0.0312, "step": 9688 }, { "epoch": 0.45023234200743495, "grad_norm": 1.723570972975321, "learning_rate": 6.052362477391157e-06, "loss": 0.0543, "step": 9689 }, { "epoch": 0.4502788104089219, "grad_norm": 0.7228974983372005, "learning_rate": 6.051626808609165e-06, "loss": 0.0494, "step": 9690 }, { "epoch": 0.4503252788104089, "grad_norm": 0.7795710519771579, "learning_rate": 6.050891116006718e-06, "loss": 0.038, "step": 9691 }, { "epoch": 0.45037174721189593, "grad_norm": 0.8451542102764508, "learning_rate": 6.0501553996004835e-06, "loss": 0.0454, "step": 9692 }, { "epoch": 0.4504182156133829, "grad_norm": 0.8324764995499171, "learning_rate": 6.049419659407122e-06, "loss": 0.0563, "step": 9693 }, { "epoch": 0.4504646840148699, "grad_norm": 0.9730524079069264, "learning_rate": 6.0486838954433005e-06, "loss": 0.0671, "step": 9694 }, { "epoch": 0.45051115241635686, "grad_norm": 0.84040358905883, "learning_rate": 6.047948107725686e-06, "loss": 0.0676, "step": 9695 }, { "epoch": 0.45055762081784384, "grad_norm": 0.6583940562547883, "learning_rate": 6.047212296270944e-06, "loss": 0.0481, "step": 9696 }, { "epoch": 0.45060408921933087, "grad_norm": 0.6846586588266346, "learning_rate": 6.046476461095741e-06, "loss": 0.0634, "step": 9697 }, { "epoch": 0.45065055762081785, "grad_norm": 0.7863653774946271, "learning_rate": 6.045740602216743e-06, "loss": 0.0642, "step": 9698 }, { "epoch": 0.4506970260223048, "grad_norm": 1.2014089079761987, "learning_rate": 6.045004719650621e-06, "loss": 0.0615, "step": 9699 }, { "epoch": 0.4507434944237918, "grad_norm": 1.1875191553941862, "learning_rate": 6.0442688134140426e-06, "loss": 0.08, "step": 9700 }, { "epoch": 0.45078996282527883, "grad_norm": 1.3213963314208412, "learning_rate": 6.043532883523676e-06, "loss": 0.0478, "step": 9701 }, { "epoch": 0.4508364312267658, "grad_norm": 1.3913351937138863, "learning_rate": 6.042796929996193e-06, "loss": 0.0878, "step": 9702 }, { "epoch": 0.4508828996282528, "grad_norm": 0.5582834075578577, "learning_rate": 6.04206095284826e-06, "loss": 0.0695, "step": 9703 }, { "epoch": 0.45092936802973976, "grad_norm": 0.7458698040978293, "learning_rate": 6.04132495209655e-06, "loss": 0.0528, "step": 9704 }, { "epoch": 0.45097583643122674, "grad_norm": 0.6982101382141347, "learning_rate": 6.040588927757735e-06, "loss": 0.0649, "step": 9705 }, { "epoch": 0.45102230483271377, "grad_norm": 0.5182404190464367, "learning_rate": 6.039852879848484e-06, "loss": 0.0296, "step": 9706 }, { "epoch": 0.45106877323420075, "grad_norm": 0.5979347776232276, "learning_rate": 6.039116808385472e-06, "loss": 0.0465, "step": 9707 }, { "epoch": 0.4511152416356877, "grad_norm": 1.1771446804362597, "learning_rate": 6.03838071338537e-06, "loss": 0.0695, "step": 9708 }, { "epoch": 0.4511617100371747, "grad_norm": 0.7692651061385405, "learning_rate": 6.037644594864853e-06, "loss": 0.0621, "step": 9709 }, { "epoch": 0.45120817843866173, "grad_norm": 0.7926517695027264, "learning_rate": 6.036908452840592e-06, "loss": 0.0703, "step": 9710 }, { "epoch": 0.4512546468401487, "grad_norm": 0.5450516495862352, "learning_rate": 6.036172287329265e-06, "loss": 0.0388, "step": 9711 }, { "epoch": 0.4513011152416357, "grad_norm": 0.708642691962167, "learning_rate": 6.0354360983475445e-06, "loss": 0.0596, "step": 9712 }, { "epoch": 0.45134758364312266, "grad_norm": 0.9080579558440903, "learning_rate": 6.034699885912107e-06, "loss": 0.0816, "step": 9713 }, { "epoch": 0.45139405204460964, "grad_norm": 0.8135599016983421, "learning_rate": 6.033963650039627e-06, "loss": 0.0661, "step": 9714 }, { "epoch": 0.45144052044609667, "grad_norm": 1.3938783945865008, "learning_rate": 6.033227390746783e-06, "loss": 0.0696, "step": 9715 }, { "epoch": 0.45148698884758365, "grad_norm": 0.3437283402578035, "learning_rate": 6.03249110805025e-06, "loss": 0.0334, "step": 9716 }, { "epoch": 0.4515334572490706, "grad_norm": 0.693439665982486, "learning_rate": 6.031754801966708e-06, "loss": 0.0636, "step": 9717 }, { "epoch": 0.4515799256505576, "grad_norm": 0.8475978407830971, "learning_rate": 6.031018472512833e-06, "loss": 0.0624, "step": 9718 }, { "epoch": 0.45162639405204463, "grad_norm": 1.2350239905601341, "learning_rate": 6.030282119705306e-06, "loss": 0.0752, "step": 9719 }, { "epoch": 0.4516728624535316, "grad_norm": 0.629813151530663, "learning_rate": 6.029545743560803e-06, "loss": 0.06, "step": 9720 }, { "epoch": 0.4517193308550186, "grad_norm": 0.5389128183023922, "learning_rate": 6.028809344096006e-06, "loss": 0.0328, "step": 9721 }, { "epoch": 0.45176579925650556, "grad_norm": 0.7790310632116308, "learning_rate": 6.0280729213275945e-06, "loss": 0.0571, "step": 9722 }, { "epoch": 0.45181226765799254, "grad_norm": 0.6399055759976224, "learning_rate": 6.027336475272248e-06, "loss": 0.054, "step": 9723 }, { "epoch": 0.45185873605947957, "grad_norm": 1.4288044870924301, "learning_rate": 6.026600005946652e-06, "loss": 0.0704, "step": 9724 }, { "epoch": 0.45190520446096655, "grad_norm": 2.569057524537409, "learning_rate": 6.025863513367484e-06, "loss": 0.054, "step": 9725 }, { "epoch": 0.4519516728624535, "grad_norm": 0.8354996805641014, "learning_rate": 6.0251269975514275e-06, "loss": 0.0646, "step": 9726 }, { "epoch": 0.4519981412639405, "grad_norm": 0.7985231170270859, "learning_rate": 6.024390458515164e-06, "loss": 0.0651, "step": 9727 }, { "epoch": 0.45204460966542753, "grad_norm": 0.6170841096853632, "learning_rate": 6.02365389627538e-06, "loss": 0.0569, "step": 9728 }, { "epoch": 0.4520910780669145, "grad_norm": 0.6563611416309096, "learning_rate": 6.022917310848758e-06, "loss": 0.0603, "step": 9729 }, { "epoch": 0.4521375464684015, "grad_norm": 1.7800627236811815, "learning_rate": 6.0221807022519815e-06, "loss": 0.0449, "step": 9730 }, { "epoch": 0.45218401486988846, "grad_norm": 0.6074127757906073, "learning_rate": 6.021444070501739e-06, "loss": 0.0599, "step": 9731 }, { "epoch": 0.45223048327137544, "grad_norm": 0.319165720637928, "learning_rate": 6.02070741561471e-06, "loss": 0.0324, "step": 9732 }, { "epoch": 0.45227695167286247, "grad_norm": 1.11144335576843, "learning_rate": 6.019970737607584e-06, "loss": 0.0707, "step": 9733 }, { "epoch": 0.45232342007434945, "grad_norm": 0.507188724482709, "learning_rate": 6.019234036497049e-06, "loss": 0.0558, "step": 9734 }, { "epoch": 0.4523698884758364, "grad_norm": 1.0813773525682187, "learning_rate": 6.018497312299789e-06, "loss": 0.0597, "step": 9735 }, { "epoch": 0.4524163568773234, "grad_norm": 0.6044506314440348, "learning_rate": 6.017760565032493e-06, "loss": 0.053, "step": 9736 }, { "epoch": 0.45246282527881043, "grad_norm": 0.4985257390021454, "learning_rate": 6.017023794711851e-06, "loss": 0.0402, "step": 9737 }, { "epoch": 0.4525092936802974, "grad_norm": 0.7631412862203159, "learning_rate": 6.016287001354549e-06, "loss": 0.0407, "step": 9738 }, { "epoch": 0.4525557620817844, "grad_norm": 0.7772512905763007, "learning_rate": 6.015550184977275e-06, "loss": 0.0617, "step": 9739 }, { "epoch": 0.45260223048327136, "grad_norm": 0.9262517205634052, "learning_rate": 6.01481334559672e-06, "loss": 0.0796, "step": 9740 }, { "epoch": 0.45264869888475834, "grad_norm": 1.0108884823582005, "learning_rate": 6.014076483229577e-06, "loss": 0.0631, "step": 9741 }, { "epoch": 0.45269516728624537, "grad_norm": 0.5701087980057876, "learning_rate": 6.013339597892532e-06, "loss": 0.0473, "step": 9742 }, { "epoch": 0.45274163568773235, "grad_norm": 0.9462258624970877, "learning_rate": 6.0126026896022825e-06, "loss": 0.0697, "step": 9743 }, { "epoch": 0.4527881040892193, "grad_norm": 0.6344613710937286, "learning_rate": 6.011865758375512e-06, "loss": 0.0511, "step": 9744 }, { "epoch": 0.4528345724907063, "grad_norm": 0.43976282622573354, "learning_rate": 6.01112880422892e-06, "loss": 0.0329, "step": 9745 }, { "epoch": 0.45288104089219333, "grad_norm": 2.6130544011670715, "learning_rate": 6.010391827179196e-06, "loss": 0.1032, "step": 9746 }, { "epoch": 0.4529275092936803, "grad_norm": 1.1365735099308647, "learning_rate": 6.009654827243032e-06, "loss": 0.0838, "step": 9747 }, { "epoch": 0.4529739776951673, "grad_norm": 1.3623388791137667, "learning_rate": 6.008917804437124e-06, "loss": 0.0569, "step": 9748 }, { "epoch": 0.45302044609665426, "grad_norm": 0.7426884602339959, "learning_rate": 6.008180758778167e-06, "loss": 0.0662, "step": 9749 }, { "epoch": 0.45306691449814124, "grad_norm": 0.5574908760975459, "learning_rate": 6.007443690282855e-06, "loss": 0.0519, "step": 9750 }, { "epoch": 0.45311338289962827, "grad_norm": 0.53612830931874, "learning_rate": 6.006706598967881e-06, "loss": 0.0532, "step": 9751 }, { "epoch": 0.45315985130111525, "grad_norm": 1.1644822930383008, "learning_rate": 6.005969484849945e-06, "loss": 0.0752, "step": 9752 }, { "epoch": 0.4532063197026022, "grad_norm": 0.5963019842540284, "learning_rate": 6.0052323479457405e-06, "loss": 0.0435, "step": 9753 }, { "epoch": 0.4532527881040892, "grad_norm": 1.0175753579421374, "learning_rate": 6.004495188271965e-06, "loss": 0.074, "step": 9754 }, { "epoch": 0.45329925650557623, "grad_norm": 0.6603476953813996, "learning_rate": 6.003758005845319e-06, "loss": 0.0378, "step": 9755 }, { "epoch": 0.4533457249070632, "grad_norm": 0.9250424622255436, "learning_rate": 6.0030208006824934e-06, "loss": 0.0731, "step": 9756 }, { "epoch": 0.4533921933085502, "grad_norm": 0.9058776189469127, "learning_rate": 6.002283572800194e-06, "loss": 0.0371, "step": 9757 }, { "epoch": 0.45343866171003716, "grad_norm": 0.5911067611231268, "learning_rate": 6.001546322215116e-06, "loss": 0.0394, "step": 9758 }, { "epoch": 0.45348513011152414, "grad_norm": 1.2912620767810925, "learning_rate": 6.000809048943959e-06, "loss": 0.0784, "step": 9759 }, { "epoch": 0.45353159851301117, "grad_norm": 0.8267894074924594, "learning_rate": 6.0000717530034236e-06, "loss": 0.0688, "step": 9760 }, { "epoch": 0.45357806691449815, "grad_norm": 0.9231162158175277, "learning_rate": 5.999334434410211e-06, "loss": 0.0491, "step": 9761 }, { "epoch": 0.4536245353159851, "grad_norm": 0.8370985476633391, "learning_rate": 5.998597093181022e-06, "loss": 0.0693, "step": 9762 }, { "epoch": 0.4536710037174721, "grad_norm": 1.106698771538843, "learning_rate": 5.997859729332558e-06, "loss": 0.0333, "step": 9763 }, { "epoch": 0.45371747211895913, "grad_norm": 0.7896360751486898, "learning_rate": 5.99712234288152e-06, "loss": 0.0584, "step": 9764 }, { "epoch": 0.4537639405204461, "grad_norm": 1.77382630192916, "learning_rate": 5.996384933844612e-06, "loss": 0.0856, "step": 9765 }, { "epoch": 0.4538104089219331, "grad_norm": 0.7276690638872875, "learning_rate": 5.995647502238536e-06, "loss": 0.0522, "step": 9766 }, { "epoch": 0.45385687732342006, "grad_norm": 0.5093123570266793, "learning_rate": 5.994910048079998e-06, "loss": 0.0646, "step": 9767 }, { "epoch": 0.45390334572490704, "grad_norm": 0.9670258910869547, "learning_rate": 5.994172571385699e-06, "loss": 0.0566, "step": 9768 }, { "epoch": 0.45394981412639407, "grad_norm": 0.8746751081175437, "learning_rate": 5.993435072172344e-06, "loss": 0.0436, "step": 9769 }, { "epoch": 0.45399628252788105, "grad_norm": 1.738756086834464, "learning_rate": 5.992697550456641e-06, "loss": 0.0616, "step": 9770 }, { "epoch": 0.454042750929368, "grad_norm": 0.6964334802863013, "learning_rate": 5.991960006255292e-06, "loss": 0.0387, "step": 9771 }, { "epoch": 0.454089219330855, "grad_norm": 1.1876225215939966, "learning_rate": 5.991222439585007e-06, "loss": 0.0723, "step": 9772 }, { "epoch": 0.45413568773234203, "grad_norm": 1.169231926491834, "learning_rate": 5.990484850462488e-06, "loss": 0.0693, "step": 9773 }, { "epoch": 0.454182156133829, "grad_norm": 1.9446660666134454, "learning_rate": 5.989747238904447e-06, "loss": 0.0568, "step": 9774 }, { "epoch": 0.454228624535316, "grad_norm": 0.7105586279762034, "learning_rate": 5.989009604927587e-06, "loss": 0.0535, "step": 9775 }, { "epoch": 0.45427509293680296, "grad_norm": 1.6890495679059165, "learning_rate": 5.988271948548619e-06, "loss": 0.075, "step": 9776 }, { "epoch": 0.45432156133828994, "grad_norm": 0.4022764300842477, "learning_rate": 5.987534269784252e-06, "loss": 0.0345, "step": 9777 }, { "epoch": 0.45436802973977697, "grad_norm": 1.3032598535249535, "learning_rate": 5.986796568651195e-06, "loss": 0.0663, "step": 9778 }, { "epoch": 0.45441449814126395, "grad_norm": 0.6504464849343559, "learning_rate": 5.986058845166156e-06, "loss": 0.0748, "step": 9779 }, { "epoch": 0.4544609665427509, "grad_norm": 0.6904345194049549, "learning_rate": 5.985321099345847e-06, "loss": 0.0351, "step": 9780 }, { "epoch": 0.4545074349442379, "grad_norm": 0.9519520957133308, "learning_rate": 5.984583331206978e-06, "loss": 0.0628, "step": 9781 }, { "epoch": 0.45455390334572493, "grad_norm": 0.7538846415670301, "learning_rate": 5.98384554076626e-06, "loss": 0.0526, "step": 9782 }, { "epoch": 0.4546003717472119, "grad_norm": 1.0117419728881496, "learning_rate": 5.983107728040404e-06, "loss": 0.0578, "step": 9783 }, { "epoch": 0.4546468401486989, "grad_norm": 0.5781492672390979, "learning_rate": 5.982369893046126e-06, "loss": 0.0485, "step": 9784 }, { "epoch": 0.45469330855018586, "grad_norm": 2.0288733668452004, "learning_rate": 5.9816320358001335e-06, "loss": 0.075, "step": 9785 }, { "epoch": 0.45473977695167284, "grad_norm": 0.9433552465380982, "learning_rate": 5.9808941563191436e-06, "loss": 0.0486, "step": 9786 }, { "epoch": 0.45478624535315987, "grad_norm": 0.6093206490157576, "learning_rate": 5.980156254619868e-06, "loss": 0.0468, "step": 9787 }, { "epoch": 0.45483271375464684, "grad_norm": 0.732884726895338, "learning_rate": 5.979418330719021e-06, "loss": 0.042, "step": 9788 }, { "epoch": 0.4548791821561338, "grad_norm": 0.446106708555973, "learning_rate": 5.97868038463332e-06, "loss": 0.0393, "step": 9789 }, { "epoch": 0.4549256505576208, "grad_norm": 0.7062085677701418, "learning_rate": 5.977942416379476e-06, "loss": 0.0632, "step": 9790 }, { "epoch": 0.45497211895910783, "grad_norm": 1.6723263758544664, "learning_rate": 5.9772044259742094e-06, "loss": 0.0549, "step": 9791 }, { "epoch": 0.4550185873605948, "grad_norm": 0.6349709531480181, "learning_rate": 5.976466413434231e-06, "loss": 0.0538, "step": 9792 }, { "epoch": 0.4550650557620818, "grad_norm": 0.556661429558675, "learning_rate": 5.975728378776263e-06, "loss": 0.0475, "step": 9793 }, { "epoch": 0.45511152416356876, "grad_norm": 0.6892203168343994, "learning_rate": 5.97499032201702e-06, "loss": 0.0576, "step": 9794 }, { "epoch": 0.45515799256505574, "grad_norm": 0.6259986776753036, "learning_rate": 5.9742522431732185e-06, "loss": 0.0396, "step": 9795 }, { "epoch": 0.45520446096654277, "grad_norm": 1.245541061317673, "learning_rate": 5.97351414226158e-06, "loss": 0.0586, "step": 9796 }, { "epoch": 0.45525092936802974, "grad_norm": 1.1089202120764299, "learning_rate": 5.9727760192988215e-06, "loss": 0.0698, "step": 9797 }, { "epoch": 0.4552973977695167, "grad_norm": 1.0885764980011077, "learning_rate": 5.972037874301662e-06, "loss": 0.0914, "step": 9798 }, { "epoch": 0.4553438661710037, "grad_norm": 1.4922416916247854, "learning_rate": 5.971299707286822e-06, "loss": 0.0758, "step": 9799 }, { "epoch": 0.45539033457249073, "grad_norm": 0.4627766434675763, "learning_rate": 5.97056151827102e-06, "loss": 0.055, "step": 9800 }, { "epoch": 0.4554368029739777, "grad_norm": 0.6458022912800202, "learning_rate": 5.9698233072709785e-06, "loss": 0.0543, "step": 9801 }, { "epoch": 0.4554832713754647, "grad_norm": 0.8726096466978406, "learning_rate": 5.96908507430342e-06, "loss": 0.0563, "step": 9802 }, { "epoch": 0.45552973977695166, "grad_norm": 0.9036737994732584, "learning_rate": 5.9683468193850635e-06, "loss": 0.0642, "step": 9803 }, { "epoch": 0.45557620817843864, "grad_norm": 0.6148377660320005, "learning_rate": 5.967608542532631e-06, "loss": 0.0499, "step": 9804 }, { "epoch": 0.45562267657992567, "grad_norm": 0.7161501011226541, "learning_rate": 5.966870243762849e-06, "loss": 0.053, "step": 9805 }, { "epoch": 0.45566914498141264, "grad_norm": 1.1746227623397403, "learning_rate": 5.966131923092438e-06, "loss": 0.0506, "step": 9806 }, { "epoch": 0.4557156133828996, "grad_norm": 0.5706180829736168, "learning_rate": 5.9653935805381206e-06, "loss": 0.0468, "step": 9807 }, { "epoch": 0.4557620817843866, "grad_norm": 1.7796365614151899, "learning_rate": 5.964655216116624e-06, "loss": 0.0614, "step": 9808 }, { "epoch": 0.45580855018587363, "grad_norm": 0.5294074088025439, "learning_rate": 5.96391682984467e-06, "loss": 0.0445, "step": 9809 }, { "epoch": 0.4558550185873606, "grad_norm": 0.44156251436840643, "learning_rate": 5.963178421738987e-06, "loss": 0.0604, "step": 9810 }, { "epoch": 0.4559014869888476, "grad_norm": 1.154189610719422, "learning_rate": 5.962439991816298e-06, "loss": 0.0529, "step": 9811 }, { "epoch": 0.45594795539033456, "grad_norm": 0.46872358241204365, "learning_rate": 5.96170154009333e-06, "loss": 0.0361, "step": 9812 }, { "epoch": 0.45599442379182153, "grad_norm": 1.5688236692561732, "learning_rate": 5.96096306658681e-06, "loss": 0.0538, "step": 9813 }, { "epoch": 0.45604089219330857, "grad_norm": 0.8627220850084099, "learning_rate": 5.960224571313465e-06, "loss": 0.0407, "step": 9814 }, { "epoch": 0.45608736059479554, "grad_norm": 1.6785884724608204, "learning_rate": 5.959486054290025e-06, "loss": 0.0841, "step": 9815 }, { "epoch": 0.4561338289962825, "grad_norm": 1.2303387931794751, "learning_rate": 5.958747515533213e-06, "loss": 0.0605, "step": 9816 }, { "epoch": 0.4561802973977695, "grad_norm": 1.298622469873709, "learning_rate": 5.958008955059762e-06, "loss": 0.0864, "step": 9817 }, { "epoch": 0.45622676579925653, "grad_norm": 0.6216808124938377, "learning_rate": 5.9572703728864016e-06, "loss": 0.0511, "step": 9818 }, { "epoch": 0.4562732342007435, "grad_norm": 0.45909218658972234, "learning_rate": 5.956531769029856e-06, "loss": 0.0481, "step": 9819 }, { "epoch": 0.4563197026022305, "grad_norm": 0.5726435676436324, "learning_rate": 5.955793143506863e-06, "loss": 0.0497, "step": 9820 }, { "epoch": 0.45636617100371746, "grad_norm": 0.7014083627474992, "learning_rate": 5.955054496334147e-06, "loss": 0.0452, "step": 9821 }, { "epoch": 0.45641263940520443, "grad_norm": 1.258537298390173, "learning_rate": 5.954315827528442e-06, "loss": 0.0726, "step": 9822 }, { "epoch": 0.45645910780669147, "grad_norm": 3.448338623981513, "learning_rate": 5.953577137106478e-06, "loss": 0.0685, "step": 9823 }, { "epoch": 0.45650557620817844, "grad_norm": 2.2771072275536217, "learning_rate": 5.95283842508499e-06, "loss": 0.1006, "step": 9824 }, { "epoch": 0.4565520446096654, "grad_norm": 0.6839571493177747, "learning_rate": 5.952099691480708e-06, "loss": 0.0578, "step": 9825 }, { "epoch": 0.4565985130111524, "grad_norm": 0.9333481836674032, "learning_rate": 5.9513609363103665e-06, "loss": 0.0594, "step": 9826 }, { "epoch": 0.45664498141263943, "grad_norm": 0.7250026272739503, "learning_rate": 5.950622159590699e-06, "loss": 0.0639, "step": 9827 }, { "epoch": 0.4566914498141264, "grad_norm": 0.4432984202407377, "learning_rate": 5.949883361338437e-06, "loss": 0.0399, "step": 9828 }, { "epoch": 0.4567379182156134, "grad_norm": 0.7460321832319072, "learning_rate": 5.949144541570318e-06, "loss": 0.0532, "step": 9829 }, { "epoch": 0.45678438661710036, "grad_norm": 0.5765798398733042, "learning_rate": 5.948405700303077e-06, "loss": 0.0605, "step": 9830 }, { "epoch": 0.45683085501858733, "grad_norm": 1.5293334142183126, "learning_rate": 5.9476668375534485e-06, "loss": 0.0698, "step": 9831 }, { "epoch": 0.45687732342007437, "grad_norm": 1.1027199788223043, "learning_rate": 5.946927953338168e-06, "loss": 0.0678, "step": 9832 }, { "epoch": 0.45692379182156134, "grad_norm": 0.4125383440960004, "learning_rate": 5.946189047673974e-06, "loss": 0.0478, "step": 9833 }, { "epoch": 0.4569702602230483, "grad_norm": 0.4107042869577898, "learning_rate": 5.945450120577601e-06, "loss": 0.0503, "step": 9834 }, { "epoch": 0.4570167286245353, "grad_norm": 0.583376219499161, "learning_rate": 5.944711172065788e-06, "loss": 0.035, "step": 9835 }, { "epoch": 0.4570631970260223, "grad_norm": 0.5549323115516059, "learning_rate": 5.943972202155273e-06, "loss": 0.0459, "step": 9836 }, { "epoch": 0.4571096654275093, "grad_norm": 0.8116341827085468, "learning_rate": 5.943233210862795e-06, "loss": 0.0733, "step": 9837 }, { "epoch": 0.4571561338289963, "grad_norm": 0.5537965839895534, "learning_rate": 5.94249419820509e-06, "loss": 0.0582, "step": 9838 }, { "epoch": 0.45720260223048326, "grad_norm": 0.7652752451932234, "learning_rate": 5.9417551641989e-06, "loss": 0.0623, "step": 9839 }, { "epoch": 0.45724907063197023, "grad_norm": 1.0911508183357295, "learning_rate": 5.941016108860965e-06, "loss": 0.0633, "step": 9840 }, { "epoch": 0.45729553903345727, "grad_norm": 0.9420065295526007, "learning_rate": 5.940277032208024e-06, "loss": 0.0579, "step": 9841 }, { "epoch": 0.45734200743494424, "grad_norm": 0.9692235809459275, "learning_rate": 5.93953793425682e-06, "loss": 0.0856, "step": 9842 }, { "epoch": 0.4573884758364312, "grad_norm": 0.4608019951129523, "learning_rate": 5.9387988150240916e-06, "loss": 0.0399, "step": 9843 }, { "epoch": 0.4574349442379182, "grad_norm": 0.45384815430299996, "learning_rate": 5.938059674526583e-06, "loss": 0.0567, "step": 9844 }, { "epoch": 0.4574814126394052, "grad_norm": 1.2630220797452913, "learning_rate": 5.937320512781033e-06, "loss": 0.0643, "step": 9845 }, { "epoch": 0.4575278810408922, "grad_norm": 0.8236628023620318, "learning_rate": 5.9365813298041896e-06, "loss": 0.0692, "step": 9846 }, { "epoch": 0.4575743494423792, "grad_norm": 0.897710590534524, "learning_rate": 5.935842125612793e-06, "loss": 0.0699, "step": 9847 }, { "epoch": 0.45762081784386616, "grad_norm": 0.7371645118464382, "learning_rate": 5.935102900223585e-06, "loss": 0.0511, "step": 9848 }, { "epoch": 0.45766728624535313, "grad_norm": 1.6766385647217747, "learning_rate": 5.934363653653315e-06, "loss": 0.0951, "step": 9849 }, { "epoch": 0.45771375464684017, "grad_norm": 0.8929693008256288, "learning_rate": 5.933624385918723e-06, "loss": 0.0555, "step": 9850 }, { "epoch": 0.45776022304832714, "grad_norm": 0.7590700672248564, "learning_rate": 5.932885097036558e-06, "loss": 0.0669, "step": 9851 }, { "epoch": 0.4578066914498141, "grad_norm": 1.0202777778332457, "learning_rate": 5.9321457870235624e-06, "loss": 0.0577, "step": 9852 }, { "epoch": 0.4578531598513011, "grad_norm": 0.708956708758167, "learning_rate": 5.931406455896483e-06, "loss": 0.077, "step": 9853 }, { "epoch": 0.4578996282527881, "grad_norm": 0.9344347000275991, "learning_rate": 5.930667103672068e-06, "loss": 0.0533, "step": 9854 }, { "epoch": 0.4579460966542751, "grad_norm": 0.7382178812275256, "learning_rate": 5.929927730367064e-06, "loss": 0.0803, "step": 9855 }, { "epoch": 0.4579925650557621, "grad_norm": 0.62873654948671, "learning_rate": 5.929188335998217e-06, "loss": 0.0521, "step": 9856 }, { "epoch": 0.45803903345724906, "grad_norm": 0.671650444225764, "learning_rate": 5.928448920582277e-06, "loss": 0.0624, "step": 9857 }, { "epoch": 0.45808550185873603, "grad_norm": 0.728820463942441, "learning_rate": 5.927709484135992e-06, "loss": 0.0761, "step": 9858 }, { "epoch": 0.45813197026022306, "grad_norm": 0.6486659076149158, "learning_rate": 5.92697002667611e-06, "loss": 0.0342, "step": 9859 }, { "epoch": 0.45817843866171004, "grad_norm": 0.692272082841243, "learning_rate": 5.92623054821938e-06, "loss": 0.0456, "step": 9860 }, { "epoch": 0.458224907063197, "grad_norm": 0.7321017842856123, "learning_rate": 5.925491048782557e-06, "loss": 0.0718, "step": 9861 }, { "epoch": 0.458271375464684, "grad_norm": 1.2154687975845695, "learning_rate": 5.924751528382385e-06, "loss": 0.0525, "step": 9862 }, { "epoch": 0.458317843866171, "grad_norm": 0.37991594212426416, "learning_rate": 5.9240119870356185e-06, "loss": 0.0596, "step": 9863 }, { "epoch": 0.458364312267658, "grad_norm": 0.776041676830256, "learning_rate": 5.923272424759008e-06, "loss": 0.0636, "step": 9864 }, { "epoch": 0.458410780669145, "grad_norm": 0.7613469319581269, "learning_rate": 5.922532841569304e-06, "loss": 0.0604, "step": 9865 }, { "epoch": 0.45845724907063196, "grad_norm": 2.445315301582964, "learning_rate": 5.921793237483262e-06, "loss": 0.0936, "step": 9866 }, { "epoch": 0.45850371747211893, "grad_norm": 0.9165155994255324, "learning_rate": 5.921053612517632e-06, "loss": 0.0912, "step": 9867 }, { "epoch": 0.45855018587360596, "grad_norm": 0.5708920077533991, "learning_rate": 5.920313966689169e-06, "loss": 0.0695, "step": 9868 }, { "epoch": 0.45859665427509294, "grad_norm": 0.7423784367894251, "learning_rate": 5.9195743000146244e-06, "loss": 0.0379, "step": 9869 }, { "epoch": 0.4586431226765799, "grad_norm": 1.1885920691175222, "learning_rate": 5.9188346125107555e-06, "loss": 0.0557, "step": 9870 }, { "epoch": 0.4586895910780669, "grad_norm": 0.6707849547071879, "learning_rate": 5.9180949041943156e-06, "loss": 0.0648, "step": 9871 }, { "epoch": 0.4587360594795539, "grad_norm": 1.1189156222115524, "learning_rate": 5.917355175082059e-06, "loss": 0.063, "step": 9872 }, { "epoch": 0.4587825278810409, "grad_norm": 0.6446642055415165, "learning_rate": 5.916615425190744e-06, "loss": 0.0572, "step": 9873 }, { "epoch": 0.4588289962825279, "grad_norm": 0.5913309284944486, "learning_rate": 5.915875654537122e-06, "loss": 0.051, "step": 9874 }, { "epoch": 0.45887546468401486, "grad_norm": 0.9819942683098208, "learning_rate": 5.915135863137955e-06, "loss": 0.0721, "step": 9875 }, { "epoch": 0.45892193308550183, "grad_norm": 1.0179180308523257, "learning_rate": 5.9143960510099965e-06, "loss": 0.0596, "step": 9876 }, { "epoch": 0.45896840148698886, "grad_norm": 0.7556418135864613, "learning_rate": 5.9136562181700055e-06, "loss": 0.0579, "step": 9877 }, { "epoch": 0.45901486988847584, "grad_norm": 0.7749229893831581, "learning_rate": 5.912916364634739e-06, "loss": 0.0463, "step": 9878 }, { "epoch": 0.4590613382899628, "grad_norm": 0.4957019035516888, "learning_rate": 5.912176490420957e-06, "loss": 0.0569, "step": 9879 }, { "epoch": 0.4591078066914498, "grad_norm": 1.3505882170354435, "learning_rate": 5.911436595545417e-06, "loss": 0.0718, "step": 9880 }, { "epoch": 0.4591542750929368, "grad_norm": 0.4464822535509126, "learning_rate": 5.910696680024876e-06, "loss": 0.0409, "step": 9881 }, { "epoch": 0.4592007434944238, "grad_norm": 0.8046889830981887, "learning_rate": 5.909956743876101e-06, "loss": 0.0597, "step": 9882 }, { "epoch": 0.4592472118959108, "grad_norm": 0.8850731619983553, "learning_rate": 5.9092167871158454e-06, "loss": 0.0596, "step": 9883 }, { "epoch": 0.45929368029739776, "grad_norm": 1.7967828165868749, "learning_rate": 5.908476809760874e-06, "loss": 0.0647, "step": 9884 }, { "epoch": 0.45934014869888473, "grad_norm": 0.7702322654839996, "learning_rate": 5.907736811827947e-06, "loss": 0.0688, "step": 9885 }, { "epoch": 0.45938661710037176, "grad_norm": 0.7428424673944525, "learning_rate": 5.906996793333824e-06, "loss": 0.0531, "step": 9886 }, { "epoch": 0.45943308550185874, "grad_norm": 0.5821921396443382, "learning_rate": 5.906256754295271e-06, "loss": 0.0562, "step": 9887 }, { "epoch": 0.4594795539033457, "grad_norm": 1.0204599718463523, "learning_rate": 5.9055166947290474e-06, "loss": 0.0686, "step": 9888 }, { "epoch": 0.4595260223048327, "grad_norm": 1.0490872384418324, "learning_rate": 5.904776614651917e-06, "loss": 0.063, "step": 9889 }, { "epoch": 0.4595724907063197, "grad_norm": 1.2639005652857116, "learning_rate": 5.904036514080647e-06, "loss": 0.0878, "step": 9890 }, { "epoch": 0.4596189591078067, "grad_norm": 0.8159785009768149, "learning_rate": 5.903296393031996e-06, "loss": 0.0795, "step": 9891 }, { "epoch": 0.4596654275092937, "grad_norm": 0.8777629527652456, "learning_rate": 5.902556251522731e-06, "loss": 0.0693, "step": 9892 }, { "epoch": 0.45971189591078065, "grad_norm": 1.0161373918756742, "learning_rate": 5.901816089569619e-06, "loss": 0.0356, "step": 9893 }, { "epoch": 0.45975836431226763, "grad_norm": 0.4926050328659872, "learning_rate": 5.9010759071894195e-06, "loss": 0.0544, "step": 9894 }, { "epoch": 0.45980483271375466, "grad_norm": 0.8600389838531377, "learning_rate": 5.900335704398906e-06, "loss": 0.0528, "step": 9895 }, { "epoch": 0.45985130111524164, "grad_norm": 0.5911361112660045, "learning_rate": 5.89959548121484e-06, "loss": 0.0619, "step": 9896 }, { "epoch": 0.4598977695167286, "grad_norm": 0.6898554004903444, "learning_rate": 5.89885523765399e-06, "loss": 0.0441, "step": 9897 }, { "epoch": 0.4599442379182156, "grad_norm": 0.43736452472146387, "learning_rate": 5.898114973733121e-06, "loss": 0.0389, "step": 9898 }, { "epoch": 0.4599907063197026, "grad_norm": 0.4891467612464529, "learning_rate": 5.897374689469004e-06, "loss": 0.0322, "step": 9899 }, { "epoch": 0.4600371747211896, "grad_norm": 0.567411676507734, "learning_rate": 5.896634384878407e-06, "loss": 0.0624, "step": 9900 }, { "epoch": 0.4600836431226766, "grad_norm": 1.4091079879302026, "learning_rate": 5.8958940599780955e-06, "loss": 0.0704, "step": 9901 }, { "epoch": 0.46013011152416355, "grad_norm": 1.0323514043524031, "learning_rate": 5.8951537147848424e-06, "loss": 0.0594, "step": 9902 }, { "epoch": 0.46017657992565053, "grad_norm": 0.9020682878053097, "learning_rate": 5.894413349315413e-06, "loss": 0.0632, "step": 9903 }, { "epoch": 0.46022304832713756, "grad_norm": 0.42862769462405254, "learning_rate": 5.893672963586581e-06, "loss": 0.0474, "step": 9904 }, { "epoch": 0.46026951672862454, "grad_norm": 2.8047442646565184, "learning_rate": 5.892932557615117e-06, "loss": 0.0727, "step": 9905 }, { "epoch": 0.4603159851301115, "grad_norm": 0.35910855267603, "learning_rate": 5.89219213141779e-06, "loss": 0.0537, "step": 9906 }, { "epoch": 0.4603624535315985, "grad_norm": 0.8062226055509487, "learning_rate": 5.8914516850113725e-06, "loss": 0.0681, "step": 9907 }, { "epoch": 0.4604089219330855, "grad_norm": 0.6115854895766045, "learning_rate": 5.8907112184126355e-06, "loss": 0.0541, "step": 9908 }, { "epoch": 0.4604553903345725, "grad_norm": 1.299420414888164, "learning_rate": 5.889970731638352e-06, "loss": 0.0398, "step": 9909 }, { "epoch": 0.4605018587360595, "grad_norm": 0.9929475358177337, "learning_rate": 5.8892302247052945e-06, "loss": 0.0461, "step": 9910 }, { "epoch": 0.46054832713754645, "grad_norm": 0.49159967139740385, "learning_rate": 5.888489697630237e-06, "loss": 0.0517, "step": 9911 }, { "epoch": 0.46059479553903343, "grad_norm": 0.7498461556214945, "learning_rate": 5.887749150429953e-06, "loss": 0.0656, "step": 9912 }, { "epoch": 0.46064126394052046, "grad_norm": 0.8554827475229388, "learning_rate": 5.887008583121216e-06, "loss": 0.0498, "step": 9913 }, { "epoch": 0.46068773234200744, "grad_norm": 0.9289569989592209, "learning_rate": 5.886267995720802e-06, "loss": 0.0471, "step": 9914 }, { "epoch": 0.4607342007434944, "grad_norm": 1.2951326071323934, "learning_rate": 5.885527388245484e-06, "loss": 0.0698, "step": 9915 }, { "epoch": 0.4607806691449814, "grad_norm": 0.4985070859436147, "learning_rate": 5.884786760712038e-06, "loss": 0.0489, "step": 9916 }, { "epoch": 0.4608271375464684, "grad_norm": 0.8902102314778869, "learning_rate": 5.884046113137243e-06, "loss": 0.0574, "step": 9917 }, { "epoch": 0.4608736059479554, "grad_norm": 0.6344795966410396, "learning_rate": 5.883305445537871e-06, "loss": 0.0418, "step": 9918 }, { "epoch": 0.4609200743494424, "grad_norm": 0.5343501200811354, "learning_rate": 5.882564757930702e-06, "loss": 0.036, "step": 9919 }, { "epoch": 0.46096654275092935, "grad_norm": 0.64258858900794, "learning_rate": 5.881824050332512e-06, "loss": 0.0407, "step": 9920 }, { "epoch": 0.46101301115241633, "grad_norm": 0.6958478905111672, "learning_rate": 5.881083322760078e-06, "loss": 0.0736, "step": 9921 }, { "epoch": 0.46105947955390336, "grad_norm": 0.8015122535203115, "learning_rate": 5.8803425752301814e-06, "loss": 0.0631, "step": 9922 }, { "epoch": 0.46110594795539034, "grad_norm": 1.0216275676137647, "learning_rate": 5.879601807759599e-06, "loss": 0.0742, "step": 9923 }, { "epoch": 0.4611524163568773, "grad_norm": 0.5684452222607124, "learning_rate": 5.878861020365109e-06, "loss": 0.0478, "step": 9924 }, { "epoch": 0.4611988847583643, "grad_norm": 0.6783220815215001, "learning_rate": 5.878120213063491e-06, "loss": 0.0393, "step": 9925 }, { "epoch": 0.4612453531598513, "grad_norm": 1.4312356629954934, "learning_rate": 5.8773793858715265e-06, "loss": 0.0775, "step": 9926 }, { "epoch": 0.4612918215613383, "grad_norm": 0.4777017394733711, "learning_rate": 5.876638538805996e-06, "loss": 0.0548, "step": 9927 }, { "epoch": 0.4613382899628253, "grad_norm": 0.6677568700842771, "learning_rate": 5.8758976718836795e-06, "loss": 0.0577, "step": 9928 }, { "epoch": 0.46138475836431225, "grad_norm": 0.8066119631745863, "learning_rate": 5.875156785121361e-06, "loss": 0.0713, "step": 9929 }, { "epoch": 0.46143122676579923, "grad_norm": 0.3848596720787096, "learning_rate": 5.874415878535817e-06, "loss": 0.0396, "step": 9930 }, { "epoch": 0.46147769516728626, "grad_norm": 0.5927556788955157, "learning_rate": 5.8736749521438355e-06, "loss": 0.0477, "step": 9931 }, { "epoch": 0.46152416356877324, "grad_norm": 0.9042476388325457, "learning_rate": 5.872934005962195e-06, "loss": 0.0703, "step": 9932 }, { "epoch": 0.4615706319702602, "grad_norm": 0.9320625025039003, "learning_rate": 5.872193040007681e-06, "loss": 0.0452, "step": 9933 }, { "epoch": 0.4616171003717472, "grad_norm": 0.6705122347052004, "learning_rate": 5.871452054297078e-06, "loss": 0.042, "step": 9934 }, { "epoch": 0.4616635687732342, "grad_norm": 0.7267932460797099, "learning_rate": 5.8707110488471674e-06, "loss": 0.0632, "step": 9935 }, { "epoch": 0.4617100371747212, "grad_norm": 1.306119362770542, "learning_rate": 5.869970023674736e-06, "loss": 0.0534, "step": 9936 }, { "epoch": 0.4617565055762082, "grad_norm": 1.6838355055865832, "learning_rate": 5.869228978796566e-06, "loss": 0.0491, "step": 9937 }, { "epoch": 0.46180297397769515, "grad_norm": 0.5383258617835603, "learning_rate": 5.8684879142294475e-06, "loss": 0.0503, "step": 9938 }, { "epoch": 0.46184944237918213, "grad_norm": 1.1077910424760051, "learning_rate": 5.867746829990162e-06, "loss": 0.0768, "step": 9939 }, { "epoch": 0.46189591078066916, "grad_norm": 1.2902915512473405, "learning_rate": 5.867005726095497e-06, "loss": 0.0586, "step": 9940 }, { "epoch": 0.46194237918215614, "grad_norm": 1.6372094266252462, "learning_rate": 5.866264602562242e-06, "loss": 0.0898, "step": 9941 }, { "epoch": 0.4619888475836431, "grad_norm": 0.7295383506548633, "learning_rate": 5.865523459407179e-06, "loss": 0.0701, "step": 9942 }, { "epoch": 0.4620353159851301, "grad_norm": 1.923890455756592, "learning_rate": 5.864782296647101e-06, "loss": 0.0904, "step": 9943 }, { "epoch": 0.4620817843866171, "grad_norm": 0.9001138497428933, "learning_rate": 5.864041114298794e-06, "loss": 0.0549, "step": 9944 }, { "epoch": 0.4621282527881041, "grad_norm": 0.8822618849611106, "learning_rate": 5.863299912379045e-06, "loss": 0.0898, "step": 9945 }, { "epoch": 0.4621747211895911, "grad_norm": 1.8214242633056494, "learning_rate": 5.862558690904645e-06, "loss": 0.0751, "step": 9946 }, { "epoch": 0.46222118959107805, "grad_norm": 0.5599771623940367, "learning_rate": 5.861817449892383e-06, "loss": 0.0626, "step": 9947 }, { "epoch": 0.46226765799256503, "grad_norm": 0.9019824081290078, "learning_rate": 5.861076189359049e-06, "loss": 0.062, "step": 9948 }, { "epoch": 0.46231412639405206, "grad_norm": 0.8002589397304113, "learning_rate": 5.860334909321432e-06, "loss": 0.0748, "step": 9949 }, { "epoch": 0.46236059479553904, "grad_norm": 0.8662786089952055, "learning_rate": 5.859593609796324e-06, "loss": 0.0576, "step": 9950 }, { "epoch": 0.462407063197026, "grad_norm": 1.3673865160959036, "learning_rate": 5.8588522908005176e-06, "loss": 0.0779, "step": 9951 }, { "epoch": 0.462453531598513, "grad_norm": 0.5161344642169405, "learning_rate": 5.858110952350802e-06, "loss": 0.0615, "step": 9952 }, { "epoch": 0.4625, "grad_norm": 0.7548470185050008, "learning_rate": 5.8573695944639696e-06, "loss": 0.0431, "step": 9953 }, { "epoch": 0.462546468401487, "grad_norm": 0.771331583626317, "learning_rate": 5.856628217156813e-06, "loss": 0.0653, "step": 9954 }, { "epoch": 0.462592936802974, "grad_norm": 0.6053018778388607, "learning_rate": 5.8558868204461265e-06, "loss": 0.0593, "step": 9955 }, { "epoch": 0.46263940520446095, "grad_norm": 1.3079560999849498, "learning_rate": 5.855145404348703e-06, "loss": 0.0981, "step": 9956 }, { "epoch": 0.46268587360594793, "grad_norm": 0.8819541468809338, "learning_rate": 5.854403968881335e-06, "loss": 0.0835, "step": 9957 }, { "epoch": 0.46273234200743496, "grad_norm": 1.006353242313599, "learning_rate": 5.853662514060821e-06, "loss": 0.0431, "step": 9958 }, { "epoch": 0.46277881040892194, "grad_norm": 0.4815138252772284, "learning_rate": 5.8529210399039486e-06, "loss": 0.0536, "step": 9959 }, { "epoch": 0.4628252788104089, "grad_norm": 0.9233137275695592, "learning_rate": 5.8521795464275185e-06, "loss": 0.0635, "step": 9960 }, { "epoch": 0.4628717472118959, "grad_norm": 0.7088695857776449, "learning_rate": 5.851438033648326e-06, "loss": 0.0629, "step": 9961 }, { "epoch": 0.4629182156133829, "grad_norm": 0.7589387532312696, "learning_rate": 5.850696501583164e-06, "loss": 0.06, "step": 9962 }, { "epoch": 0.4629646840148699, "grad_norm": 0.9560655090847171, "learning_rate": 5.849954950248831e-06, "loss": 0.067, "step": 9963 }, { "epoch": 0.4630111524163569, "grad_norm": 0.6577060983890846, "learning_rate": 5.849213379662125e-06, "loss": 0.068, "step": 9964 }, { "epoch": 0.46305762081784385, "grad_norm": 0.7427082297882811, "learning_rate": 5.8484717898398425e-06, "loss": 0.0716, "step": 9965 }, { "epoch": 0.46310408921933083, "grad_norm": 0.734141461565696, "learning_rate": 5.847730180798779e-06, "loss": 0.063, "step": 9966 }, { "epoch": 0.46315055762081786, "grad_norm": 0.7520646002569222, "learning_rate": 5.846988552555736e-06, "loss": 0.0578, "step": 9967 }, { "epoch": 0.46319702602230484, "grad_norm": 2.2959661756233065, "learning_rate": 5.846246905127511e-06, "loss": 0.0906, "step": 9968 }, { "epoch": 0.4632434944237918, "grad_norm": 4.418852170533002, "learning_rate": 5.8455052385309015e-06, "loss": 0.0806, "step": 9969 }, { "epoch": 0.4632899628252788, "grad_norm": 0.37720682964492636, "learning_rate": 5.84476355278271e-06, "loss": 0.02, "step": 9970 }, { "epoch": 0.4633364312267658, "grad_norm": 0.5272597536673446, "learning_rate": 5.844021847899735e-06, "loss": 0.0476, "step": 9971 }, { "epoch": 0.4633828996282528, "grad_norm": 1.1633949880651697, "learning_rate": 5.843280123898777e-06, "loss": 0.0672, "step": 9972 }, { "epoch": 0.4634293680297398, "grad_norm": 1.6476906468958163, "learning_rate": 5.8425383807966365e-06, "loss": 0.0781, "step": 9973 }, { "epoch": 0.46347583643122675, "grad_norm": 0.38173796566629303, "learning_rate": 5.841796618610113e-06, "loss": 0.0518, "step": 9974 }, { "epoch": 0.4635223048327137, "grad_norm": 0.5663564406887023, "learning_rate": 5.841054837356012e-06, "loss": 0.0426, "step": 9975 }, { "epoch": 0.46356877323420076, "grad_norm": 0.8957375502091823, "learning_rate": 5.840313037051133e-06, "loss": 0.0661, "step": 9976 }, { "epoch": 0.46361524163568774, "grad_norm": 0.6190186111083862, "learning_rate": 5.83957121771228e-06, "loss": 0.0643, "step": 9977 }, { "epoch": 0.4636617100371747, "grad_norm": 0.5262932359564889, "learning_rate": 5.838829379356254e-06, "loss": 0.035, "step": 9978 }, { "epoch": 0.4637081784386617, "grad_norm": 0.47805849938702827, "learning_rate": 5.838087521999861e-06, "loss": 0.0402, "step": 9979 }, { "epoch": 0.4637546468401487, "grad_norm": 0.8444340408460865, "learning_rate": 5.837345645659903e-06, "loss": 0.0646, "step": 9980 }, { "epoch": 0.4638011152416357, "grad_norm": 0.6211521760717867, "learning_rate": 5.836603750353184e-06, "loss": 0.0646, "step": 9981 }, { "epoch": 0.4638475836431227, "grad_norm": 0.5981994195867729, "learning_rate": 5.835861836096512e-06, "loss": 0.0538, "step": 9982 }, { "epoch": 0.46389405204460965, "grad_norm": 0.7513563509582254, "learning_rate": 5.835119902906688e-06, "loss": 0.0494, "step": 9983 }, { "epoch": 0.4639405204460966, "grad_norm": 0.5306096951980678, "learning_rate": 5.834377950800519e-06, "loss": 0.0483, "step": 9984 }, { "epoch": 0.46398698884758366, "grad_norm": 0.9902844001959216, "learning_rate": 5.833635979794811e-06, "loss": 0.0412, "step": 9985 }, { "epoch": 0.46403345724907064, "grad_norm": 0.7202129626505996, "learning_rate": 5.832893989906372e-06, "loss": 0.0737, "step": 9986 }, { "epoch": 0.4640799256505576, "grad_norm": 0.6693032112889721, "learning_rate": 5.832151981152007e-06, "loss": 0.0715, "step": 9987 }, { "epoch": 0.4641263940520446, "grad_norm": 0.7714740733956676, "learning_rate": 5.831409953548524e-06, "loss": 0.0672, "step": 9988 }, { "epoch": 0.4641728624535316, "grad_norm": 0.6659796043502789, "learning_rate": 5.8306679071127304e-06, "loss": 0.0541, "step": 9989 }, { "epoch": 0.4642193308550186, "grad_norm": 0.9930448944138973, "learning_rate": 5.829925841861433e-06, "loss": 0.0668, "step": 9990 }, { "epoch": 0.4642657992565056, "grad_norm": 1.18542052409938, "learning_rate": 5.829183757811443e-06, "loss": 0.0615, "step": 9991 }, { "epoch": 0.46431226765799255, "grad_norm": 0.6634807684750278, "learning_rate": 5.828441654979569e-06, "loss": 0.0397, "step": 9992 }, { "epoch": 0.4643587360594795, "grad_norm": 0.7304703570948605, "learning_rate": 5.827699533382618e-06, "loss": 0.0432, "step": 9993 }, { "epoch": 0.46440520446096656, "grad_norm": 0.4987818643318336, "learning_rate": 5.8269573930374035e-06, "loss": 0.0327, "step": 9994 }, { "epoch": 0.46445167286245354, "grad_norm": 0.7247869609943215, "learning_rate": 5.826215233960732e-06, "loss": 0.0462, "step": 9995 }, { "epoch": 0.4644981412639405, "grad_norm": 0.6701257845079989, "learning_rate": 5.825473056169416e-06, "loss": 0.0664, "step": 9996 }, { "epoch": 0.4645446096654275, "grad_norm": 0.6781252987470349, "learning_rate": 5.824730859680268e-06, "loss": 0.0597, "step": 9997 }, { "epoch": 0.4645910780669145, "grad_norm": 0.9341156418255083, "learning_rate": 5.823988644510095e-06, "loss": 0.061, "step": 9998 }, { "epoch": 0.4646375464684015, "grad_norm": 1.2731172987971981, "learning_rate": 5.823246410675714e-06, "loss": 0.0803, "step": 9999 }, { "epoch": 0.4646840148698885, "grad_norm": 0.9056249004564131, "learning_rate": 5.822504158193936e-06, "loss": 0.0591, "step": 10000 }, { "epoch": 0.46473048327137545, "grad_norm": 1.0412540711743252, "learning_rate": 5.821761887081572e-06, "loss": 0.0503, "step": 10001 }, { "epoch": 0.4647769516728624, "grad_norm": 0.5729578359812704, "learning_rate": 5.821019597355436e-06, "loss": 0.0564, "step": 10002 }, { "epoch": 0.46482342007434946, "grad_norm": 0.37979580480504566, "learning_rate": 5.820277289032343e-06, "loss": 0.045, "step": 10003 }, { "epoch": 0.46486988847583643, "grad_norm": 0.8188565814952157, "learning_rate": 5.819534962129105e-06, "loss": 0.0665, "step": 10004 }, { "epoch": 0.4649163568773234, "grad_norm": 0.7715850261036701, "learning_rate": 5.8187926166625386e-06, "loss": 0.0675, "step": 10005 }, { "epoch": 0.4649628252788104, "grad_norm": 1.3744383203448483, "learning_rate": 5.818050252649459e-06, "loss": 0.0598, "step": 10006 }, { "epoch": 0.4650092936802974, "grad_norm": 1.2821209346009326, "learning_rate": 5.817307870106676e-06, "loss": 0.0997, "step": 10007 }, { "epoch": 0.4650557620817844, "grad_norm": 0.6352124629852333, "learning_rate": 5.816565469051013e-06, "loss": 0.046, "step": 10008 }, { "epoch": 0.4651022304832714, "grad_norm": 0.7625997404025298, "learning_rate": 5.815823049499283e-06, "loss": 0.0479, "step": 10009 }, { "epoch": 0.46514869888475835, "grad_norm": 0.6141662058499575, "learning_rate": 5.8150806114683e-06, "loss": 0.0567, "step": 10010 }, { "epoch": 0.4651951672862453, "grad_norm": 1.5310761295975843, "learning_rate": 5.8143381549748855e-06, "loss": 0.0997, "step": 10011 }, { "epoch": 0.46524163568773236, "grad_norm": 1.148436881035711, "learning_rate": 5.8135956800358525e-06, "loss": 0.0551, "step": 10012 }, { "epoch": 0.46528810408921933, "grad_norm": 2.099383641882489, "learning_rate": 5.812853186668023e-06, "loss": 0.0988, "step": 10013 }, { "epoch": 0.4653345724907063, "grad_norm": 0.7252252641260446, "learning_rate": 5.8121106748882125e-06, "loss": 0.0499, "step": 10014 }, { "epoch": 0.4653810408921933, "grad_norm": 0.8224032589033553, "learning_rate": 5.8113681447132395e-06, "loss": 0.0572, "step": 10015 }, { "epoch": 0.4654275092936803, "grad_norm": 1.0989768593660139, "learning_rate": 5.810625596159926e-06, "loss": 0.0616, "step": 10016 }, { "epoch": 0.4654739776951673, "grad_norm": 3.731952826151704, "learning_rate": 5.809883029245088e-06, "loss": 0.0665, "step": 10017 }, { "epoch": 0.4655204460966543, "grad_norm": 0.5809197093408549, "learning_rate": 5.8091404439855485e-06, "loss": 0.0194, "step": 10018 }, { "epoch": 0.46556691449814125, "grad_norm": 0.8018100799752146, "learning_rate": 5.808397840398124e-06, "loss": 0.0768, "step": 10019 }, { "epoch": 0.4656133828996282, "grad_norm": 0.7442677583453615, "learning_rate": 5.80765521849964e-06, "loss": 0.0466, "step": 10020 }, { "epoch": 0.46565985130111526, "grad_norm": 0.5060779129428294, "learning_rate": 5.806912578306914e-06, "loss": 0.0485, "step": 10021 }, { "epoch": 0.46570631970260223, "grad_norm": 0.6458758947983462, "learning_rate": 5.806169919836768e-06, "loss": 0.0682, "step": 10022 }, { "epoch": 0.4657527881040892, "grad_norm": 1.0889509773826873, "learning_rate": 5.8054272431060275e-06, "loss": 0.0692, "step": 10023 }, { "epoch": 0.4657992565055762, "grad_norm": 0.9869260116787423, "learning_rate": 5.8046845481315094e-06, "loss": 0.0873, "step": 10024 }, { "epoch": 0.4658457249070632, "grad_norm": 0.8573852494463784, "learning_rate": 5.803941834930042e-06, "loss": 0.0644, "step": 10025 }, { "epoch": 0.4658921933085502, "grad_norm": 0.7342331453097526, "learning_rate": 5.803199103518444e-06, "loss": 0.0747, "step": 10026 }, { "epoch": 0.46593866171003717, "grad_norm": 1.0264540862804359, "learning_rate": 5.8024563539135414e-06, "loss": 0.0404, "step": 10027 }, { "epoch": 0.46598513011152415, "grad_norm": 0.7833129244446204, "learning_rate": 5.801713586132158e-06, "loss": 0.0571, "step": 10028 }, { "epoch": 0.4660315985130111, "grad_norm": 0.44199258609286385, "learning_rate": 5.800970800191118e-06, "loss": 0.0502, "step": 10029 }, { "epoch": 0.46607806691449816, "grad_norm": 0.9711792860421292, "learning_rate": 5.800227996107247e-06, "loss": 0.066, "step": 10030 }, { "epoch": 0.46612453531598513, "grad_norm": 0.6857328316909719, "learning_rate": 5.7994851738973675e-06, "loss": 0.0575, "step": 10031 }, { "epoch": 0.4661710037174721, "grad_norm": 1.0830671734015629, "learning_rate": 5.79874233357831e-06, "loss": 0.0609, "step": 10032 }, { "epoch": 0.4662174721189591, "grad_norm": 0.7246155267506795, "learning_rate": 5.797999475166897e-06, "loss": 0.0562, "step": 10033 }, { "epoch": 0.4662639405204461, "grad_norm": 0.6497315332249476, "learning_rate": 5.7972565986799556e-06, "loss": 0.0572, "step": 10034 }, { "epoch": 0.4663104089219331, "grad_norm": 1.032812348421132, "learning_rate": 5.796513704134314e-06, "loss": 0.0797, "step": 10035 }, { "epoch": 0.46635687732342007, "grad_norm": 0.8729673709756434, "learning_rate": 5.795770791546798e-06, "loss": 0.0377, "step": 10036 }, { "epoch": 0.46640334572490705, "grad_norm": 0.9745353373014067, "learning_rate": 5.795027860934236e-06, "loss": 0.0602, "step": 10037 }, { "epoch": 0.466449814126394, "grad_norm": 1.9688794639524863, "learning_rate": 5.794284912313457e-06, "loss": 0.0645, "step": 10038 }, { "epoch": 0.46649628252788106, "grad_norm": 0.6211190379710115, "learning_rate": 5.793541945701288e-06, "loss": 0.0664, "step": 10039 }, { "epoch": 0.46654275092936803, "grad_norm": 0.67257343536541, "learning_rate": 5.7927989611145584e-06, "loss": 0.064, "step": 10040 }, { "epoch": 0.466589219330855, "grad_norm": 0.6051182024000721, "learning_rate": 5.792055958570099e-06, "loss": 0.0536, "step": 10041 }, { "epoch": 0.466635687732342, "grad_norm": 1.0938922641821176, "learning_rate": 5.791312938084738e-06, "loss": 0.0625, "step": 10042 }, { "epoch": 0.466682156133829, "grad_norm": 0.603045770358726, "learning_rate": 5.790569899675305e-06, "loss": 0.0666, "step": 10043 }, { "epoch": 0.466728624535316, "grad_norm": 0.5406142372697477, "learning_rate": 5.789826843358634e-06, "loss": 0.0639, "step": 10044 }, { "epoch": 0.46677509293680297, "grad_norm": 1.197855791205324, "learning_rate": 5.789083769151552e-06, "loss": 0.0754, "step": 10045 }, { "epoch": 0.46682156133828995, "grad_norm": 0.9302402679795178, "learning_rate": 5.788340677070892e-06, "loss": 0.0655, "step": 10046 }, { "epoch": 0.4668680297397769, "grad_norm": 0.8298015757987387, "learning_rate": 5.787597567133487e-06, "loss": 0.0693, "step": 10047 }, { "epoch": 0.46691449814126396, "grad_norm": 0.5797376105895005, "learning_rate": 5.786854439356167e-06, "loss": 0.0635, "step": 10048 }, { "epoch": 0.46696096654275093, "grad_norm": 0.49353178941210957, "learning_rate": 5.7861112937557674e-06, "loss": 0.0391, "step": 10049 }, { "epoch": 0.4670074349442379, "grad_norm": 0.7069448490168078, "learning_rate": 5.785368130349119e-06, "loss": 0.0686, "step": 10050 }, { "epoch": 0.4670539033457249, "grad_norm": 0.7737580347676121, "learning_rate": 5.784624949153055e-06, "loss": 0.0494, "step": 10051 }, { "epoch": 0.4671003717472119, "grad_norm": 1.4393812382523046, "learning_rate": 5.78388175018441e-06, "loss": 0.0847, "step": 10052 }, { "epoch": 0.4671468401486989, "grad_norm": 0.7178641624476745, "learning_rate": 5.78313853346002e-06, "loss": 0.0519, "step": 10053 }, { "epoch": 0.46719330855018587, "grad_norm": 0.9769764764618548, "learning_rate": 5.782395298996717e-06, "loss": 0.0706, "step": 10054 }, { "epoch": 0.46723977695167285, "grad_norm": 3.93232359984513, "learning_rate": 5.781652046811337e-06, "loss": 0.0799, "step": 10055 }, { "epoch": 0.4672862453531598, "grad_norm": 0.4684852525107833, "learning_rate": 5.780908776920715e-06, "loss": 0.0487, "step": 10056 }, { "epoch": 0.46733271375464686, "grad_norm": 0.9102284832769483, "learning_rate": 5.780165489341688e-06, "loss": 0.0546, "step": 10057 }, { "epoch": 0.46737918215613383, "grad_norm": 0.6679937400412184, "learning_rate": 5.779422184091092e-06, "loss": 0.0627, "step": 10058 }, { "epoch": 0.4674256505576208, "grad_norm": 0.6706691917419623, "learning_rate": 5.7786788611857625e-06, "loss": 0.0608, "step": 10059 }, { "epoch": 0.4674721189591078, "grad_norm": 0.7409502325428352, "learning_rate": 5.777935520642537e-06, "loss": 0.0606, "step": 10060 }, { "epoch": 0.4675185873605948, "grad_norm": 1.6029690420392257, "learning_rate": 5.777192162478253e-06, "loss": 0.0573, "step": 10061 }, { "epoch": 0.4675650557620818, "grad_norm": 0.637131592348598, "learning_rate": 5.776448786709751e-06, "loss": 0.0615, "step": 10062 }, { "epoch": 0.46761152416356877, "grad_norm": 0.9881328772869364, "learning_rate": 5.775705393353863e-06, "loss": 0.0573, "step": 10063 }, { "epoch": 0.46765799256505575, "grad_norm": 0.6875038472980232, "learning_rate": 5.774961982427434e-06, "loss": 0.0312, "step": 10064 }, { "epoch": 0.4677044609665427, "grad_norm": 1.7320920307897274, "learning_rate": 5.774218553947299e-06, "loss": 0.0648, "step": 10065 }, { "epoch": 0.46775092936802976, "grad_norm": 1.1976695395035313, "learning_rate": 5.773475107930301e-06, "loss": 0.0738, "step": 10066 }, { "epoch": 0.46779739776951673, "grad_norm": 0.5908125150145378, "learning_rate": 5.772731644393277e-06, "loss": 0.0602, "step": 10067 }, { "epoch": 0.4678438661710037, "grad_norm": 0.4124686912155361, "learning_rate": 5.7719881633530665e-06, "loss": 0.0362, "step": 10068 }, { "epoch": 0.4678903345724907, "grad_norm": 1.0633880992622777, "learning_rate": 5.771244664826512e-06, "loss": 0.0513, "step": 10069 }, { "epoch": 0.4679368029739777, "grad_norm": 0.6981342032027353, "learning_rate": 5.770501148830455e-06, "loss": 0.0563, "step": 10070 }, { "epoch": 0.4679832713754647, "grad_norm": 1.4840339564704828, "learning_rate": 5.769757615381736e-06, "loss": 0.0657, "step": 10071 }, { "epoch": 0.46802973977695167, "grad_norm": 2.0770374760711, "learning_rate": 5.769014064497196e-06, "loss": 0.0754, "step": 10072 }, { "epoch": 0.46807620817843865, "grad_norm": 0.4688007682027049, "learning_rate": 5.768270496193678e-06, "loss": 0.0516, "step": 10073 }, { "epoch": 0.4681226765799256, "grad_norm": 0.6567522272996293, "learning_rate": 5.767526910488026e-06, "loss": 0.047, "step": 10074 }, { "epoch": 0.46816914498141265, "grad_norm": 0.6021074873382091, "learning_rate": 5.76678330739708e-06, "loss": 0.0674, "step": 10075 }, { "epoch": 0.46821561338289963, "grad_norm": 2.0370692377747477, "learning_rate": 5.766039686937688e-06, "loss": 0.0437, "step": 10076 }, { "epoch": 0.4682620817843866, "grad_norm": 0.5679667665166253, "learning_rate": 5.765296049126687e-06, "loss": 0.0348, "step": 10077 }, { "epoch": 0.4683085501858736, "grad_norm": 1.0631913488788807, "learning_rate": 5.764552393980928e-06, "loss": 0.0768, "step": 10078 }, { "epoch": 0.4683550185873606, "grad_norm": 0.5220766633103395, "learning_rate": 5.763808721517251e-06, "loss": 0.0409, "step": 10079 }, { "epoch": 0.4684014869888476, "grad_norm": 0.5474002002455998, "learning_rate": 5.763065031752503e-06, "loss": 0.0393, "step": 10080 }, { "epoch": 0.46844795539033457, "grad_norm": 0.7748774551350618, "learning_rate": 5.762321324703528e-06, "loss": 0.0496, "step": 10081 }, { "epoch": 0.46849442379182155, "grad_norm": 0.6660587827190945, "learning_rate": 5.761577600387175e-06, "loss": 0.0481, "step": 10082 }, { "epoch": 0.4685408921933085, "grad_norm": 0.23198748947392192, "learning_rate": 5.760833858820286e-06, "loss": 0.0225, "step": 10083 }, { "epoch": 0.46858736059479555, "grad_norm": 0.580259869822879, "learning_rate": 5.760090100019708e-06, "loss": 0.0551, "step": 10084 }, { "epoch": 0.46863382899628253, "grad_norm": 0.9660128471738719, "learning_rate": 5.759346324002293e-06, "loss": 0.0713, "step": 10085 }, { "epoch": 0.4686802973977695, "grad_norm": 1.2494987731710254, "learning_rate": 5.758602530784883e-06, "loss": 0.0604, "step": 10086 }, { "epoch": 0.4687267657992565, "grad_norm": 0.4129702379545378, "learning_rate": 5.757858720384325e-06, "loss": 0.0364, "step": 10087 }, { "epoch": 0.4687732342007435, "grad_norm": 1.2123115466359355, "learning_rate": 5.757114892817473e-06, "loss": 0.0811, "step": 10088 }, { "epoch": 0.4688197026022305, "grad_norm": 1.3112475890049846, "learning_rate": 5.756371048101169e-06, "loss": 0.0425, "step": 10089 }, { "epoch": 0.46886617100371747, "grad_norm": 0.6738164947965785, "learning_rate": 5.7556271862522665e-06, "loss": 0.0592, "step": 10090 }, { "epoch": 0.46891263940520445, "grad_norm": 0.8822448557520832, "learning_rate": 5.754883307287612e-06, "loss": 0.0797, "step": 10091 }, { "epoch": 0.4689591078066915, "grad_norm": 1.0148323980064218, "learning_rate": 5.754139411224056e-06, "loss": 0.0911, "step": 10092 }, { "epoch": 0.46900557620817845, "grad_norm": 0.9296406192510772, "learning_rate": 5.753395498078448e-06, "loss": 0.0658, "step": 10093 }, { "epoch": 0.46905204460966543, "grad_norm": 0.5569619907890578, "learning_rate": 5.75265156786764e-06, "loss": 0.0497, "step": 10094 }, { "epoch": 0.4690985130111524, "grad_norm": 0.39143583029720114, "learning_rate": 5.751907620608482e-06, "loss": 0.0294, "step": 10095 }, { "epoch": 0.4691449814126394, "grad_norm": 1.5426052375443928, "learning_rate": 5.751163656317824e-06, "loss": 0.0688, "step": 10096 }, { "epoch": 0.4691914498141264, "grad_norm": 0.45381266195292813, "learning_rate": 5.7504196750125195e-06, "loss": 0.0363, "step": 10097 }, { "epoch": 0.4692379182156134, "grad_norm": 0.4831786775386074, "learning_rate": 5.749675676709419e-06, "loss": 0.063, "step": 10098 }, { "epoch": 0.46928438661710037, "grad_norm": 0.596903671129829, "learning_rate": 5.748931661425375e-06, "loss": 0.0558, "step": 10099 }, { "epoch": 0.46933085501858735, "grad_norm": 2.806004385797564, "learning_rate": 5.748187629177244e-06, "loss": 0.0691, "step": 10100 }, { "epoch": 0.4693773234200744, "grad_norm": 0.8654889153888032, "learning_rate": 5.747443579981872e-06, "loss": 0.083, "step": 10101 }, { "epoch": 0.46942379182156135, "grad_norm": 0.6800142826819419, "learning_rate": 5.746699513856118e-06, "loss": 0.054, "step": 10102 }, { "epoch": 0.46947026022304833, "grad_norm": 0.9618472765847292, "learning_rate": 5.745955430816833e-06, "loss": 0.0504, "step": 10103 }, { "epoch": 0.4695167286245353, "grad_norm": 0.977503839817135, "learning_rate": 5.745211330880872e-06, "loss": 0.0599, "step": 10104 }, { "epoch": 0.4695631970260223, "grad_norm": 1.0955086408853794, "learning_rate": 5.744467214065092e-06, "loss": 0.0903, "step": 10105 }, { "epoch": 0.4696096654275093, "grad_norm": 0.520978671031844, "learning_rate": 5.743723080386345e-06, "loss": 0.0553, "step": 10106 }, { "epoch": 0.4696561338289963, "grad_norm": 1.1621255455975181, "learning_rate": 5.742978929861488e-06, "loss": 0.0616, "step": 10107 }, { "epoch": 0.46970260223048327, "grad_norm": 0.7997106683254526, "learning_rate": 5.742234762507375e-06, "loss": 0.0756, "step": 10108 }, { "epoch": 0.46974907063197024, "grad_norm": 1.0751739990596465, "learning_rate": 5.741490578340865e-06, "loss": 0.1214, "step": 10109 }, { "epoch": 0.4697955390334573, "grad_norm": 1.6181305570809636, "learning_rate": 5.740746377378812e-06, "loss": 0.0722, "step": 10110 }, { "epoch": 0.46984200743494425, "grad_norm": 0.8365553032129417, "learning_rate": 5.7400021596380725e-06, "loss": 0.077, "step": 10111 }, { "epoch": 0.46988847583643123, "grad_norm": 0.7575400085217081, "learning_rate": 5.739257925135508e-06, "loss": 0.0389, "step": 10112 }, { "epoch": 0.4699349442379182, "grad_norm": 0.556220991589439, "learning_rate": 5.738513673887972e-06, "loss": 0.0574, "step": 10113 }, { "epoch": 0.4699814126394052, "grad_norm": 0.6139580478411539, "learning_rate": 5.737769405912324e-06, "loss": 0.0406, "step": 10114 }, { "epoch": 0.4700278810408922, "grad_norm": 0.5112565072349403, "learning_rate": 5.737025121225424e-06, "loss": 0.0564, "step": 10115 }, { "epoch": 0.4700743494423792, "grad_norm": 0.7754428854027315, "learning_rate": 5.736280819844128e-06, "loss": 0.0433, "step": 10116 }, { "epoch": 0.47012081784386617, "grad_norm": 0.8064241075326081, "learning_rate": 5.735536501785296e-06, "loss": 0.0748, "step": 10117 }, { "epoch": 0.47016728624535314, "grad_norm": 0.4724380010794492, "learning_rate": 5.734792167065789e-06, "loss": 0.0544, "step": 10118 }, { "epoch": 0.4702137546468402, "grad_norm": 1.1129066758032005, "learning_rate": 5.734047815702466e-06, "loss": 0.0625, "step": 10119 }, { "epoch": 0.47026022304832715, "grad_norm": 0.6815285706312503, "learning_rate": 5.733303447712186e-06, "loss": 0.0561, "step": 10120 }, { "epoch": 0.47030669144981413, "grad_norm": 0.43791043280830944, "learning_rate": 5.732559063111814e-06, "loss": 0.0442, "step": 10121 }, { "epoch": 0.4703531598513011, "grad_norm": 1.070056767998338, "learning_rate": 5.731814661918206e-06, "loss": 0.0707, "step": 10122 }, { "epoch": 0.4703996282527881, "grad_norm": 0.4881174299979955, "learning_rate": 5.731070244148226e-06, "loss": 0.0545, "step": 10123 }, { "epoch": 0.4704460966542751, "grad_norm": 0.891763016328656, "learning_rate": 5.730325809818736e-06, "loss": 0.0706, "step": 10124 }, { "epoch": 0.4704925650557621, "grad_norm": 0.7013948221979285, "learning_rate": 5.729581358946596e-06, "loss": 0.055, "step": 10125 }, { "epoch": 0.47053903345724907, "grad_norm": 0.6359992936654911, "learning_rate": 5.728836891548673e-06, "loss": 0.0446, "step": 10126 }, { "epoch": 0.47058550185873604, "grad_norm": 0.453545480697894, "learning_rate": 5.728092407641827e-06, "loss": 0.0575, "step": 10127 }, { "epoch": 0.4706319702602231, "grad_norm": 0.5014840713199302, "learning_rate": 5.727347907242919e-06, "loss": 0.0434, "step": 10128 }, { "epoch": 0.47067843866171005, "grad_norm": 0.6790848658931755, "learning_rate": 5.726603390368819e-06, "loss": 0.0611, "step": 10129 }, { "epoch": 0.47072490706319703, "grad_norm": 0.6866861699417831, "learning_rate": 5.725858857036385e-06, "loss": 0.0368, "step": 10130 }, { "epoch": 0.470771375464684, "grad_norm": 0.7186254788434525, "learning_rate": 5.7251143072624845e-06, "loss": 0.0742, "step": 10131 }, { "epoch": 0.470817843866171, "grad_norm": 0.6914983651539055, "learning_rate": 5.724369741063982e-06, "loss": 0.064, "step": 10132 }, { "epoch": 0.470864312267658, "grad_norm": 0.6398616814563376, "learning_rate": 5.723625158457741e-06, "loss": 0.0694, "step": 10133 }, { "epoch": 0.470910780669145, "grad_norm": 0.49908085846484324, "learning_rate": 5.722880559460631e-06, "loss": 0.0441, "step": 10134 }, { "epoch": 0.47095724907063197, "grad_norm": 0.5394047896399474, "learning_rate": 5.722135944089515e-06, "loss": 0.0493, "step": 10135 }, { "epoch": 0.47100371747211894, "grad_norm": 1.0961821921420585, "learning_rate": 5.721391312361259e-06, "loss": 0.0651, "step": 10136 }, { "epoch": 0.471050185873606, "grad_norm": 0.7594867725347256, "learning_rate": 5.72064666429273e-06, "loss": 0.0688, "step": 10137 }, { "epoch": 0.47109665427509295, "grad_norm": 0.7053365682627745, "learning_rate": 5.719901999900797e-06, "loss": 0.0564, "step": 10138 }, { "epoch": 0.47114312267657993, "grad_norm": 1.0257142280553841, "learning_rate": 5.719157319202325e-06, "loss": 0.0852, "step": 10139 }, { "epoch": 0.4711895910780669, "grad_norm": 0.6186116501883269, "learning_rate": 5.718412622214184e-06, "loss": 0.0529, "step": 10140 }, { "epoch": 0.4712360594795539, "grad_norm": 0.7084892047381766, "learning_rate": 5.71766790895324e-06, "loss": 0.0589, "step": 10141 }, { "epoch": 0.4712825278810409, "grad_norm": 0.494593321313822, "learning_rate": 5.716923179436362e-06, "loss": 0.0437, "step": 10142 }, { "epoch": 0.4713289962825279, "grad_norm": 0.8074136896983923, "learning_rate": 5.7161784336804195e-06, "loss": 0.0776, "step": 10143 }, { "epoch": 0.47137546468401487, "grad_norm": 1.0807673994000577, "learning_rate": 5.715433671702284e-06, "loss": 0.0754, "step": 10144 }, { "epoch": 0.47142193308550184, "grad_norm": 1.20264681754429, "learning_rate": 5.71468889351882e-06, "loss": 0.0563, "step": 10145 }, { "epoch": 0.4714684014869889, "grad_norm": 0.6900922065463986, "learning_rate": 5.713944099146902e-06, "loss": 0.073, "step": 10146 }, { "epoch": 0.47151486988847585, "grad_norm": 1.0152200899098345, "learning_rate": 5.7131992886034e-06, "loss": 0.0715, "step": 10147 }, { "epoch": 0.47156133828996283, "grad_norm": 0.9153525350991892, "learning_rate": 5.7124544619051816e-06, "loss": 0.0616, "step": 10148 }, { "epoch": 0.4716078066914498, "grad_norm": 2.3636893050805607, "learning_rate": 5.711709619069119e-06, "loss": 0.0836, "step": 10149 }, { "epoch": 0.4716542750929368, "grad_norm": 0.7152162791260612, "learning_rate": 5.7109647601120854e-06, "loss": 0.0476, "step": 10150 }, { "epoch": 0.4717007434944238, "grad_norm": 0.7441303970171184, "learning_rate": 5.7102198850509524e-06, "loss": 0.0454, "step": 10151 }, { "epoch": 0.4717472118959108, "grad_norm": 0.7971982710602415, "learning_rate": 5.709474993902589e-06, "loss": 0.0673, "step": 10152 }, { "epoch": 0.47179368029739777, "grad_norm": 0.6170253786817321, "learning_rate": 5.708730086683874e-06, "loss": 0.0503, "step": 10153 }, { "epoch": 0.47184014869888474, "grad_norm": 0.3536777998619871, "learning_rate": 5.707985163411675e-06, "loss": 0.0401, "step": 10154 }, { "epoch": 0.4718866171003718, "grad_norm": 0.6581638080218415, "learning_rate": 5.707240224102866e-06, "loss": 0.0553, "step": 10155 }, { "epoch": 0.47193308550185875, "grad_norm": 0.634534624039138, "learning_rate": 5.706495268774323e-06, "loss": 0.058, "step": 10156 }, { "epoch": 0.4719795539033457, "grad_norm": 0.6313420103424381, "learning_rate": 5.705750297442917e-06, "loss": 0.0483, "step": 10157 }, { "epoch": 0.4720260223048327, "grad_norm": 0.6151479954547311, "learning_rate": 5.705005310125526e-06, "loss": 0.0534, "step": 10158 }, { "epoch": 0.4720724907063197, "grad_norm": 0.7477454812729557, "learning_rate": 5.704260306839021e-06, "loss": 0.0392, "step": 10159 }, { "epoch": 0.4721189591078067, "grad_norm": 1.481355859941932, "learning_rate": 5.7035152876002785e-06, "loss": 0.049, "step": 10160 }, { "epoch": 0.4721654275092937, "grad_norm": 0.5750322860257763, "learning_rate": 5.702770252426175e-06, "loss": 0.0415, "step": 10161 }, { "epoch": 0.47221189591078067, "grad_norm": 0.7236254374830403, "learning_rate": 5.702025201333586e-06, "loss": 0.0485, "step": 10162 }, { "epoch": 0.47225836431226764, "grad_norm": 0.500228423598167, "learning_rate": 5.7012801343393876e-06, "loss": 0.0477, "step": 10163 }, { "epoch": 0.4723048327137547, "grad_norm": 0.3947021454336698, "learning_rate": 5.700535051460455e-06, "loss": 0.0308, "step": 10164 }, { "epoch": 0.47235130111524165, "grad_norm": 0.6110827246149652, "learning_rate": 5.699789952713667e-06, "loss": 0.0641, "step": 10165 }, { "epoch": 0.4723977695167286, "grad_norm": 0.8404488753047499, "learning_rate": 5.699044838115901e-06, "loss": 0.0818, "step": 10166 }, { "epoch": 0.4724442379182156, "grad_norm": 0.7706473355331628, "learning_rate": 5.698299707684031e-06, "loss": 0.0716, "step": 10167 }, { "epoch": 0.4724907063197026, "grad_norm": 0.7512152258707696, "learning_rate": 5.6975545614349405e-06, "loss": 0.0626, "step": 10168 }, { "epoch": 0.4725371747211896, "grad_norm": 0.48825834294131165, "learning_rate": 5.6968093993855035e-06, "loss": 0.0326, "step": 10169 }, { "epoch": 0.4725836431226766, "grad_norm": 0.6128899296357383, "learning_rate": 5.696064221552601e-06, "loss": 0.0402, "step": 10170 }, { "epoch": 0.47263011152416357, "grad_norm": 1.4194041348994422, "learning_rate": 5.69531902795311e-06, "loss": 0.0767, "step": 10171 }, { "epoch": 0.47267657992565054, "grad_norm": 0.6441265749825327, "learning_rate": 5.694573818603911e-06, "loss": 0.0587, "step": 10172 }, { "epoch": 0.4727230483271376, "grad_norm": 0.6901574116595801, "learning_rate": 5.693828593521886e-06, "loss": 0.0392, "step": 10173 }, { "epoch": 0.47276951672862455, "grad_norm": 0.745386021436481, "learning_rate": 5.693083352723911e-06, "loss": 0.0624, "step": 10174 }, { "epoch": 0.4728159851301115, "grad_norm": 0.6053196692582631, "learning_rate": 5.69233809622687e-06, "loss": 0.0565, "step": 10175 }, { "epoch": 0.4728624535315985, "grad_norm": 0.49907791954980985, "learning_rate": 5.691592824047642e-06, "loss": 0.0489, "step": 10176 }, { "epoch": 0.4729089219330855, "grad_norm": 0.4876895154212635, "learning_rate": 5.690847536203109e-06, "loss": 0.0612, "step": 10177 }, { "epoch": 0.4729553903345725, "grad_norm": 0.7028978531342752, "learning_rate": 5.690102232710153e-06, "loss": 0.0383, "step": 10178 }, { "epoch": 0.4730018587360595, "grad_norm": 0.6623140805192839, "learning_rate": 5.689356913585653e-06, "loss": 0.0647, "step": 10179 }, { "epoch": 0.47304832713754646, "grad_norm": 0.8794851361617378, "learning_rate": 5.688611578846495e-06, "loss": 0.0561, "step": 10180 }, { "epoch": 0.47309479553903344, "grad_norm": 0.7434267034874313, "learning_rate": 5.687866228509559e-06, "loss": 0.0659, "step": 10181 }, { "epoch": 0.4731412639405205, "grad_norm": 0.6848667259862669, "learning_rate": 5.687120862591729e-06, "loss": 0.052, "step": 10182 }, { "epoch": 0.47318773234200745, "grad_norm": 1.2688926748601939, "learning_rate": 5.68637548110989e-06, "loss": 0.0776, "step": 10183 }, { "epoch": 0.4732342007434944, "grad_norm": 0.7366496968864109, "learning_rate": 5.685630084080922e-06, "loss": 0.0519, "step": 10184 }, { "epoch": 0.4732806691449814, "grad_norm": 0.7979893746779925, "learning_rate": 5.684884671521712e-06, "loss": 0.0572, "step": 10185 }, { "epoch": 0.4733271375464684, "grad_norm": 2.0076888190766873, "learning_rate": 5.684139243449143e-06, "loss": 0.0909, "step": 10186 }, { "epoch": 0.4733736059479554, "grad_norm": 0.805322940154131, "learning_rate": 5.6833937998801e-06, "loss": 0.0637, "step": 10187 }, { "epoch": 0.4734200743494424, "grad_norm": 1.5667092978013784, "learning_rate": 5.682648340831468e-06, "loss": 0.0606, "step": 10188 }, { "epoch": 0.47346654275092936, "grad_norm": 0.5704692236118779, "learning_rate": 5.6819028663201325e-06, "loss": 0.0425, "step": 10189 }, { "epoch": 0.47351301115241634, "grad_norm": 0.9586752113441366, "learning_rate": 5.68115737636298e-06, "loss": 0.0621, "step": 10190 }, { "epoch": 0.4735594795539034, "grad_norm": 0.6419014645900768, "learning_rate": 5.680411870976897e-06, "loss": 0.0625, "step": 10191 }, { "epoch": 0.47360594795539035, "grad_norm": 0.46382944602906206, "learning_rate": 5.679666350178767e-06, "loss": 0.0389, "step": 10192 }, { "epoch": 0.4736524163568773, "grad_norm": 0.7264335548187106, "learning_rate": 5.678920813985479e-06, "loss": 0.0494, "step": 10193 }, { "epoch": 0.4736988847583643, "grad_norm": 0.9059672938727634, "learning_rate": 5.678175262413921e-06, "loss": 0.0819, "step": 10194 }, { "epoch": 0.4737453531598513, "grad_norm": 1.3056874022537035, "learning_rate": 5.677429695480979e-06, "loss": 0.094, "step": 10195 }, { "epoch": 0.4737918215613383, "grad_norm": 0.6687542052747405, "learning_rate": 5.676684113203541e-06, "loss": 0.059, "step": 10196 }, { "epoch": 0.4738382899628253, "grad_norm": 0.6082100664922521, "learning_rate": 5.675938515598497e-06, "loss": 0.0415, "step": 10197 }, { "epoch": 0.47388475836431226, "grad_norm": 0.5930018249423719, "learning_rate": 5.6751929026827325e-06, "loss": 0.0406, "step": 10198 }, { "epoch": 0.47393122676579924, "grad_norm": 0.7070533915261005, "learning_rate": 5.6744472744731395e-06, "loss": 0.0507, "step": 10199 }, { "epoch": 0.4739776951672863, "grad_norm": 0.5801356557313957, "learning_rate": 5.673701630986606e-06, "loss": 0.0692, "step": 10200 }, { "epoch": 0.47402416356877325, "grad_norm": 0.9705403838238089, "learning_rate": 5.672955972240019e-06, "loss": 0.068, "step": 10201 }, { "epoch": 0.4740706319702602, "grad_norm": 0.5483924449022879, "learning_rate": 5.672210298250273e-06, "loss": 0.0556, "step": 10202 }, { "epoch": 0.4741171003717472, "grad_norm": 0.7326689420956264, "learning_rate": 5.671464609034257e-06, "loss": 0.0532, "step": 10203 }, { "epoch": 0.4741635687732342, "grad_norm": 0.9301531072152873, "learning_rate": 5.6707189046088606e-06, "loss": 0.1016, "step": 10204 }, { "epoch": 0.4742100371747212, "grad_norm": 0.4481934518295989, "learning_rate": 5.6699731849909735e-06, "loss": 0.04, "step": 10205 }, { "epoch": 0.4742565055762082, "grad_norm": 0.6807109553971125, "learning_rate": 5.66922745019749e-06, "loss": 0.0481, "step": 10206 }, { "epoch": 0.47430297397769516, "grad_norm": 0.9821142357528402, "learning_rate": 5.668481700245301e-06, "loss": 0.0697, "step": 10207 }, { "epoch": 0.47434944237918214, "grad_norm": 0.46783551175217586, "learning_rate": 5.667735935151295e-06, "loss": 0.0338, "step": 10208 }, { "epoch": 0.47439591078066917, "grad_norm": 2.3472091711279766, "learning_rate": 5.666990154932371e-06, "loss": 0.0557, "step": 10209 }, { "epoch": 0.47444237918215615, "grad_norm": 0.598576265959179, "learning_rate": 5.666244359605415e-06, "loss": 0.0606, "step": 10210 }, { "epoch": 0.4744888475836431, "grad_norm": 0.8332410368650844, "learning_rate": 5.665498549187325e-06, "loss": 0.0643, "step": 10211 }, { "epoch": 0.4745353159851301, "grad_norm": 0.3672051414881231, "learning_rate": 5.6647527236949915e-06, "loss": 0.0286, "step": 10212 }, { "epoch": 0.4745817843866171, "grad_norm": 0.6666617234836784, "learning_rate": 5.6640068831453085e-06, "loss": 0.0624, "step": 10213 }, { "epoch": 0.4746282527881041, "grad_norm": 0.5718506746299453, "learning_rate": 5.663261027555172e-06, "loss": 0.0747, "step": 10214 }, { "epoch": 0.4746747211895911, "grad_norm": 0.9699797769512551, "learning_rate": 5.662515156941476e-06, "loss": 0.0766, "step": 10215 }, { "epoch": 0.47472118959107806, "grad_norm": 0.9577127427253035, "learning_rate": 5.661769271321113e-06, "loss": 0.0615, "step": 10216 }, { "epoch": 0.47476765799256504, "grad_norm": 0.8092334687249343, "learning_rate": 5.661023370710979e-06, "loss": 0.0707, "step": 10217 }, { "epoch": 0.47481412639405207, "grad_norm": 0.9293093823344728, "learning_rate": 5.660277455127971e-06, "loss": 0.0701, "step": 10218 }, { "epoch": 0.47486059479553905, "grad_norm": 0.641164140611206, "learning_rate": 5.659531524588984e-06, "loss": 0.052, "step": 10219 }, { "epoch": 0.474907063197026, "grad_norm": 0.5397403626846757, "learning_rate": 5.6587855791109124e-06, "loss": 0.0641, "step": 10220 }, { "epoch": 0.474953531598513, "grad_norm": 0.8820316787383042, "learning_rate": 5.658039618710656e-06, "loss": 0.0689, "step": 10221 }, { "epoch": 0.475, "grad_norm": 0.6244172052806761, "learning_rate": 5.657293643405107e-06, "loss": 0.0393, "step": 10222 }, { "epoch": 0.475046468401487, "grad_norm": 1.737080234916235, "learning_rate": 5.656547653211168e-06, "loss": 0.0545, "step": 10223 }, { "epoch": 0.475092936802974, "grad_norm": 0.7062408545368927, "learning_rate": 5.655801648145732e-06, "loss": 0.0563, "step": 10224 }, { "epoch": 0.47513940520446096, "grad_norm": 0.7508722838612962, "learning_rate": 5.6550556282256985e-06, "loss": 0.0462, "step": 10225 }, { "epoch": 0.47518587360594794, "grad_norm": 0.5859292408481699, "learning_rate": 5.654309593467966e-06, "loss": 0.0361, "step": 10226 }, { "epoch": 0.47523234200743497, "grad_norm": 0.5940794741285327, "learning_rate": 5.653563543889433e-06, "loss": 0.0729, "step": 10227 }, { "epoch": 0.47527881040892195, "grad_norm": 3.1429103216823733, "learning_rate": 5.652817479506998e-06, "loss": 0.0345, "step": 10228 }, { "epoch": 0.4753252788104089, "grad_norm": 0.5177126090915356, "learning_rate": 5.652071400337558e-06, "loss": 0.0535, "step": 10229 }, { "epoch": 0.4753717472118959, "grad_norm": 0.6635031060255291, "learning_rate": 5.651325306398015e-06, "loss": 0.0656, "step": 10230 }, { "epoch": 0.4754182156133829, "grad_norm": 0.6258969309423927, "learning_rate": 5.6505791977052705e-06, "loss": 0.0342, "step": 10231 }, { "epoch": 0.4754646840148699, "grad_norm": 1.3643944025873567, "learning_rate": 5.64983307427622e-06, "loss": 0.0513, "step": 10232 }, { "epoch": 0.4755111524163569, "grad_norm": 0.4741142412129328, "learning_rate": 5.649086936127769e-06, "loss": 0.0374, "step": 10233 }, { "epoch": 0.47555762081784386, "grad_norm": 0.779488082512076, "learning_rate": 5.648340783276813e-06, "loss": 0.058, "step": 10234 }, { "epoch": 0.47560408921933084, "grad_norm": 0.811179245959777, "learning_rate": 5.647594615740257e-06, "loss": 0.0433, "step": 10235 }, { "epoch": 0.47565055762081787, "grad_norm": 0.5773645101548195, "learning_rate": 5.646848433535003e-06, "loss": 0.0509, "step": 10236 }, { "epoch": 0.47569702602230485, "grad_norm": 0.44168602731292517, "learning_rate": 5.646102236677948e-06, "loss": 0.039, "step": 10237 }, { "epoch": 0.4757434944237918, "grad_norm": 0.7161051295254348, "learning_rate": 5.645356025185999e-06, "loss": 0.0723, "step": 10238 }, { "epoch": 0.4757899628252788, "grad_norm": 1.0086382731759558, "learning_rate": 5.644609799076057e-06, "loss": 0.053, "step": 10239 }, { "epoch": 0.4758364312267658, "grad_norm": 0.9065093970986938, "learning_rate": 5.643863558365026e-06, "loss": 0.0718, "step": 10240 }, { "epoch": 0.4758828996282528, "grad_norm": 1.088114787116669, "learning_rate": 5.643117303069805e-06, "loss": 0.0589, "step": 10241 }, { "epoch": 0.4759293680297398, "grad_norm": 0.5822450442143118, "learning_rate": 5.642371033207301e-06, "loss": 0.0421, "step": 10242 }, { "epoch": 0.47597583643122676, "grad_norm": 0.7593846851123246, "learning_rate": 5.6416247487944185e-06, "loss": 0.0701, "step": 10243 }, { "epoch": 0.47602230483271374, "grad_norm": 0.7594836787839209, "learning_rate": 5.640878449848059e-06, "loss": 0.0663, "step": 10244 }, { "epoch": 0.47606877323420077, "grad_norm": 2.588164237369468, "learning_rate": 5.640132136385129e-06, "loss": 0.068, "step": 10245 }, { "epoch": 0.47611524163568775, "grad_norm": 0.6091550673784044, "learning_rate": 5.6393858084225305e-06, "loss": 0.0368, "step": 10246 }, { "epoch": 0.4761617100371747, "grad_norm": 0.8822943559126447, "learning_rate": 5.638639465977171e-06, "loss": 0.0741, "step": 10247 }, { "epoch": 0.4762081784386617, "grad_norm": 0.7487883280419826, "learning_rate": 5.637893109065957e-06, "loss": 0.0299, "step": 10248 }, { "epoch": 0.4762546468401487, "grad_norm": 0.5798305874402665, "learning_rate": 5.637146737705791e-06, "loss": 0.0692, "step": 10249 }, { "epoch": 0.4763011152416357, "grad_norm": 1.248789455216027, "learning_rate": 5.636400351913583e-06, "loss": 0.0675, "step": 10250 }, { "epoch": 0.4763475836431227, "grad_norm": 0.4459831769359383, "learning_rate": 5.6356539517062345e-06, "loss": 0.0427, "step": 10251 }, { "epoch": 0.47639405204460966, "grad_norm": 1.2147841486618547, "learning_rate": 5.634907537100657e-06, "loss": 0.0532, "step": 10252 }, { "epoch": 0.47644052044609664, "grad_norm": 0.5161638438502734, "learning_rate": 5.634161108113755e-06, "loss": 0.0339, "step": 10253 }, { "epoch": 0.47648698884758367, "grad_norm": 1.1900041701775825, "learning_rate": 5.6334146647624355e-06, "loss": 0.0512, "step": 10254 }, { "epoch": 0.47653345724907065, "grad_norm": 0.4235665843641909, "learning_rate": 5.632668207063608e-06, "loss": 0.0369, "step": 10255 }, { "epoch": 0.4765799256505576, "grad_norm": 1.6868078203034556, "learning_rate": 5.63192173503418e-06, "loss": 0.0778, "step": 10256 }, { "epoch": 0.4766263940520446, "grad_norm": 1.2933252939529727, "learning_rate": 5.63117524869106e-06, "loss": 0.0736, "step": 10257 }, { "epoch": 0.4766728624535316, "grad_norm": 0.8498111765586568, "learning_rate": 5.630428748051154e-06, "loss": 0.0551, "step": 10258 }, { "epoch": 0.4767193308550186, "grad_norm": 0.7642477472136832, "learning_rate": 5.629682233131375e-06, "loss": 0.0514, "step": 10259 }, { "epoch": 0.4767657992565056, "grad_norm": 2.3250907969498305, "learning_rate": 5.628935703948631e-06, "loss": 0.0702, "step": 10260 }, { "epoch": 0.47681226765799256, "grad_norm": 1.0227914384463805, "learning_rate": 5.62818916051983e-06, "loss": 0.0487, "step": 10261 }, { "epoch": 0.47685873605947954, "grad_norm": 0.4152200755415173, "learning_rate": 5.627442602861886e-06, "loss": 0.0379, "step": 10262 }, { "epoch": 0.47690520446096657, "grad_norm": 1.2419369891433598, "learning_rate": 5.626696030991704e-06, "loss": 0.0725, "step": 10263 }, { "epoch": 0.47695167286245355, "grad_norm": 0.795202121397449, "learning_rate": 5.625949444926198e-06, "loss": 0.0506, "step": 10264 }, { "epoch": 0.4769981412639405, "grad_norm": 0.7557711405030676, "learning_rate": 5.6252028446822805e-06, "loss": 0.0643, "step": 10265 }, { "epoch": 0.4770446096654275, "grad_norm": 0.5688765232458195, "learning_rate": 5.624456230276857e-06, "loss": 0.0503, "step": 10266 }, { "epoch": 0.4770910780669145, "grad_norm": 0.586525997243012, "learning_rate": 5.6237096017268445e-06, "loss": 0.0485, "step": 10267 }, { "epoch": 0.4771375464684015, "grad_norm": 5.0278187417709965, "learning_rate": 5.622962959049153e-06, "loss": 0.0821, "step": 10268 }, { "epoch": 0.4771840148698885, "grad_norm": 0.4623312764144807, "learning_rate": 5.622216302260696e-06, "loss": 0.0372, "step": 10269 }, { "epoch": 0.47723048327137546, "grad_norm": 0.7247487874613432, "learning_rate": 5.621469631378382e-06, "loss": 0.0665, "step": 10270 }, { "epoch": 0.47727695167286244, "grad_norm": 0.5506489849396828, "learning_rate": 5.620722946419129e-06, "loss": 0.0555, "step": 10271 }, { "epoch": 0.47732342007434947, "grad_norm": 0.6944063557177719, "learning_rate": 5.619976247399847e-06, "loss": 0.072, "step": 10272 }, { "epoch": 0.47736988847583645, "grad_norm": 0.36136457690976237, "learning_rate": 5.61922953433745e-06, "loss": 0.0311, "step": 10273 }, { "epoch": 0.4774163568773234, "grad_norm": 1.8768054172915016, "learning_rate": 5.618482807248855e-06, "loss": 0.1208, "step": 10274 }, { "epoch": 0.4774628252788104, "grad_norm": 0.5398589886448962, "learning_rate": 5.617736066150971e-06, "loss": 0.0447, "step": 10275 }, { "epoch": 0.4775092936802974, "grad_norm": 1.1868105465731564, "learning_rate": 5.616989311060716e-06, "loss": 0.0702, "step": 10276 }, { "epoch": 0.4775557620817844, "grad_norm": 1.58486348863723, "learning_rate": 5.6162425419950026e-06, "loss": 0.0704, "step": 10277 }, { "epoch": 0.4776022304832714, "grad_norm": 0.7951848359956872, "learning_rate": 5.615495758970747e-06, "loss": 0.0964, "step": 10278 }, { "epoch": 0.47764869888475836, "grad_norm": 0.91253035457312, "learning_rate": 5.6147489620048655e-06, "loss": 0.0745, "step": 10279 }, { "epoch": 0.47769516728624534, "grad_norm": 17.406852230537748, "learning_rate": 5.614002151114274e-06, "loss": 0.1012, "step": 10280 }, { "epoch": 0.47774163568773237, "grad_norm": 1.3088594563220832, "learning_rate": 5.613255326315887e-06, "loss": 0.0609, "step": 10281 }, { "epoch": 0.47778810408921935, "grad_norm": 0.7773066205900045, "learning_rate": 5.61250848762662e-06, "loss": 0.0558, "step": 10282 }, { "epoch": 0.4778345724907063, "grad_norm": 0.6377369223244393, "learning_rate": 5.611761635063392e-06, "loss": 0.0649, "step": 10283 }, { "epoch": 0.4778810408921933, "grad_norm": 0.4671562906471473, "learning_rate": 5.6110147686431194e-06, "loss": 0.0423, "step": 10284 }, { "epoch": 0.4779275092936803, "grad_norm": 0.8634204726882666, "learning_rate": 5.610267888382718e-06, "loss": 0.0499, "step": 10285 }, { "epoch": 0.4779739776951673, "grad_norm": 0.6424922192703713, "learning_rate": 5.609520994299109e-06, "loss": 0.0475, "step": 10286 }, { "epoch": 0.4780204460966543, "grad_norm": 0.5938369447249834, "learning_rate": 5.608774086409206e-06, "loss": 0.0315, "step": 10287 }, { "epoch": 0.47806691449814126, "grad_norm": 5.484074948397793, "learning_rate": 5.608027164729929e-06, "loss": 0.0999, "step": 10288 }, { "epoch": 0.47811338289962824, "grad_norm": 1.260362255252733, "learning_rate": 5.607280229278198e-06, "loss": 0.1042, "step": 10289 }, { "epoch": 0.47815985130111527, "grad_norm": 4.192549366056961, "learning_rate": 5.60653328007093e-06, "loss": 0.1098, "step": 10290 }, { "epoch": 0.47820631970260224, "grad_norm": 1.9296062306386448, "learning_rate": 5.605786317125045e-06, "loss": 0.0542, "step": 10291 }, { "epoch": 0.4782527881040892, "grad_norm": 4.850584514778567, "learning_rate": 5.605039340457462e-06, "loss": 0.1194, "step": 10292 }, { "epoch": 0.4782992565055762, "grad_norm": 1.8207052834463844, "learning_rate": 5.604292350085102e-06, "loss": 0.0539, "step": 10293 }, { "epoch": 0.4783457249070632, "grad_norm": 0.5593535674775455, "learning_rate": 5.603545346024882e-06, "loss": 0.0509, "step": 10294 }, { "epoch": 0.4783921933085502, "grad_norm": 0.624221729267968, "learning_rate": 5.602798328293727e-06, "loss": 0.0501, "step": 10295 }, { "epoch": 0.4784386617100372, "grad_norm": 3.2709093134595166, "learning_rate": 5.602051296908556e-06, "loss": 0.0991, "step": 10296 }, { "epoch": 0.47848513011152416, "grad_norm": 0.6885574128640797, "learning_rate": 5.601304251886287e-06, "loss": 0.0783, "step": 10297 }, { "epoch": 0.47853159851301114, "grad_norm": 0.5867034329845097, "learning_rate": 5.600557193243847e-06, "loss": 0.0522, "step": 10298 }, { "epoch": 0.47857806691449817, "grad_norm": 0.8421619011292751, "learning_rate": 5.599810120998152e-06, "loss": 0.0749, "step": 10299 }, { "epoch": 0.47862453531598514, "grad_norm": 3.3609524697719593, "learning_rate": 5.599063035166127e-06, "loss": 0.1194, "step": 10300 }, { "epoch": 0.4786710037174721, "grad_norm": 1.0511398646547194, "learning_rate": 5.598315935764695e-06, "loss": 0.0819, "step": 10301 }, { "epoch": 0.4787174721189591, "grad_norm": 0.7092696836218743, "learning_rate": 5.597568822810775e-06, "loss": 0.0536, "step": 10302 }, { "epoch": 0.4787639405204461, "grad_norm": 1.2551422668889847, "learning_rate": 5.596821696321294e-06, "loss": 0.0551, "step": 10303 }, { "epoch": 0.4788104089219331, "grad_norm": 0.3268035056464407, "learning_rate": 5.596074556313173e-06, "loss": 0.032, "step": 10304 }, { "epoch": 0.4788568773234201, "grad_norm": 0.7098553249552351, "learning_rate": 5.595327402803336e-06, "loss": 0.0695, "step": 10305 }, { "epoch": 0.47890334572490706, "grad_norm": 0.30684989224101517, "learning_rate": 5.5945802358087066e-06, "loss": 0.0318, "step": 10306 }, { "epoch": 0.47894981412639404, "grad_norm": 0.5414138621813815, "learning_rate": 5.5938330553462085e-06, "loss": 0.0368, "step": 10307 }, { "epoch": 0.47899628252788107, "grad_norm": 1.5546078780524164, "learning_rate": 5.593085861432766e-06, "loss": 0.0891, "step": 10308 }, { "epoch": 0.47904275092936804, "grad_norm": 0.5488285518746985, "learning_rate": 5.592338654085308e-06, "loss": 0.0411, "step": 10309 }, { "epoch": 0.479089219330855, "grad_norm": 1.6553706406987776, "learning_rate": 5.591591433320753e-06, "loss": 0.0854, "step": 10310 }, { "epoch": 0.479135687732342, "grad_norm": 0.9135605938406822, "learning_rate": 5.590844199156029e-06, "loss": 0.0855, "step": 10311 }, { "epoch": 0.479182156133829, "grad_norm": 0.7520206057809028, "learning_rate": 5.5900969516080635e-06, "loss": 0.0844, "step": 10312 }, { "epoch": 0.479228624535316, "grad_norm": 0.5203059630302604, "learning_rate": 5.5893496906937814e-06, "loss": 0.0533, "step": 10313 }, { "epoch": 0.479275092936803, "grad_norm": 3.3742599258476504, "learning_rate": 5.588602416430107e-06, "loss": 0.0814, "step": 10314 }, { "epoch": 0.47932156133828996, "grad_norm": 0.7609723721282475, "learning_rate": 5.587855128833971e-06, "loss": 0.0476, "step": 10315 }, { "epoch": 0.47936802973977694, "grad_norm": 0.7259352348594907, "learning_rate": 5.587107827922294e-06, "loss": 0.0616, "step": 10316 }, { "epoch": 0.47941449814126397, "grad_norm": 0.8994372828518337, "learning_rate": 5.586360513712011e-06, "loss": 0.0576, "step": 10317 }, { "epoch": 0.47946096654275094, "grad_norm": 1.5265186956030994, "learning_rate": 5.585613186220043e-06, "loss": 0.048, "step": 10318 }, { "epoch": 0.4795074349442379, "grad_norm": 2.045642276090965, "learning_rate": 5.584865845463319e-06, "loss": 0.0748, "step": 10319 }, { "epoch": 0.4795539033457249, "grad_norm": 0.5775129819463296, "learning_rate": 5.584118491458769e-06, "loss": 0.0605, "step": 10320 }, { "epoch": 0.4796003717472119, "grad_norm": 0.6831270881642886, "learning_rate": 5.58337112422332e-06, "loss": 0.0658, "step": 10321 }, { "epoch": 0.4796468401486989, "grad_norm": 0.5016149803198179, "learning_rate": 5.582623743773902e-06, "loss": 0.0496, "step": 10322 }, { "epoch": 0.4796933085501859, "grad_norm": 0.6766436652042525, "learning_rate": 5.581876350127441e-06, "loss": 0.0476, "step": 10323 }, { "epoch": 0.47973977695167286, "grad_norm": 0.45063431144754973, "learning_rate": 5.58112894330087e-06, "loss": 0.066, "step": 10324 }, { "epoch": 0.47978624535315983, "grad_norm": 0.6278084642042044, "learning_rate": 5.580381523311117e-06, "loss": 0.0453, "step": 10325 }, { "epoch": 0.47983271375464687, "grad_norm": 0.666666243907959, "learning_rate": 5.57963409017511e-06, "loss": 0.0662, "step": 10326 }, { "epoch": 0.47987918215613384, "grad_norm": 0.6760025393229447, "learning_rate": 5.578886643909784e-06, "loss": 0.0673, "step": 10327 }, { "epoch": 0.4799256505576208, "grad_norm": 0.7660440912243651, "learning_rate": 5.578139184532062e-06, "loss": 0.081, "step": 10328 }, { "epoch": 0.4799721189591078, "grad_norm": 1.8960166112740622, "learning_rate": 5.577391712058881e-06, "loss": 0.0496, "step": 10329 }, { "epoch": 0.4800185873605948, "grad_norm": 0.4188178655441435, "learning_rate": 5.5766442265071705e-06, "loss": 0.0462, "step": 10330 }, { "epoch": 0.4800650557620818, "grad_norm": 0.5200864257808057, "learning_rate": 5.575896727893861e-06, "loss": 0.056, "step": 10331 }, { "epoch": 0.4801115241635688, "grad_norm": 0.589190475619774, "learning_rate": 5.575149216235884e-06, "loss": 0.0573, "step": 10332 }, { "epoch": 0.48015799256505576, "grad_norm": 0.9230511971540435, "learning_rate": 5.574401691550173e-06, "loss": 0.0492, "step": 10333 }, { "epoch": 0.48020446096654273, "grad_norm": 0.8069958149253567, "learning_rate": 5.573654153853659e-06, "loss": 0.0717, "step": 10334 }, { "epoch": 0.48025092936802977, "grad_norm": 0.5176218871547041, "learning_rate": 5.572906603163274e-06, "loss": 0.0618, "step": 10335 }, { "epoch": 0.48029739776951674, "grad_norm": 2.032672604911199, "learning_rate": 5.572159039495951e-06, "loss": 0.0874, "step": 10336 }, { "epoch": 0.4803438661710037, "grad_norm": 0.7379653841756704, "learning_rate": 5.5714114628686255e-06, "loss": 0.0659, "step": 10337 }, { "epoch": 0.4803903345724907, "grad_norm": 0.40427598911648904, "learning_rate": 5.570663873298226e-06, "loss": 0.0401, "step": 10338 }, { "epoch": 0.4804368029739777, "grad_norm": 0.6553296241530628, "learning_rate": 5.569916270801693e-06, "loss": 0.0525, "step": 10339 }, { "epoch": 0.4804832713754647, "grad_norm": 0.8420672018779572, "learning_rate": 5.5691686553959535e-06, "loss": 0.0853, "step": 10340 }, { "epoch": 0.4805297397769517, "grad_norm": 0.9029054347101233, "learning_rate": 5.568421027097947e-06, "loss": 0.0485, "step": 10341 }, { "epoch": 0.48057620817843866, "grad_norm": 0.6147104956190056, "learning_rate": 5.567673385924605e-06, "loss": 0.0462, "step": 10342 }, { "epoch": 0.48062267657992563, "grad_norm": 1.0510139392584579, "learning_rate": 5.566925731892862e-06, "loss": 0.0936, "step": 10343 }, { "epoch": 0.48066914498141267, "grad_norm": 0.4754550293584349, "learning_rate": 5.566178065019656e-06, "loss": 0.0385, "step": 10344 }, { "epoch": 0.48071561338289964, "grad_norm": 1.0691299341676248, "learning_rate": 5.5654303853219205e-06, "loss": 0.0539, "step": 10345 }, { "epoch": 0.4807620817843866, "grad_norm": 0.7273631229454363, "learning_rate": 5.564682692816592e-06, "loss": 0.0482, "step": 10346 }, { "epoch": 0.4808085501858736, "grad_norm": 0.8940087631557252, "learning_rate": 5.563934987520606e-06, "loss": 0.0643, "step": 10347 }, { "epoch": 0.48085501858736057, "grad_norm": 12.777341057778543, "learning_rate": 5.563187269450899e-06, "loss": 0.0629, "step": 10348 }, { "epoch": 0.4809014869888476, "grad_norm": 0.5086728659821051, "learning_rate": 5.562439538624408e-06, "loss": 0.0551, "step": 10349 }, { "epoch": 0.4809479553903346, "grad_norm": 0.6189663827986948, "learning_rate": 5.561691795058069e-06, "loss": 0.0489, "step": 10350 }, { "epoch": 0.48099442379182156, "grad_norm": 0.5961943097359206, "learning_rate": 5.560944038768821e-06, "loss": 0.0392, "step": 10351 }, { "epoch": 0.48104089219330853, "grad_norm": 0.542904654495817, "learning_rate": 5.560196269773598e-06, "loss": 0.0525, "step": 10352 }, { "epoch": 0.48108736059479557, "grad_norm": 0.5628313166370086, "learning_rate": 5.559448488089341e-06, "loss": 0.0588, "step": 10353 }, { "epoch": 0.48113382899628254, "grad_norm": 0.4968543244781664, "learning_rate": 5.5587006937329864e-06, "loss": 0.0459, "step": 10354 }, { "epoch": 0.4811802973977695, "grad_norm": 0.9530139566936009, "learning_rate": 5.5579528867214726e-06, "loss": 0.0721, "step": 10355 }, { "epoch": 0.4812267657992565, "grad_norm": 0.5909870827477504, "learning_rate": 5.55720506707174e-06, "loss": 0.0662, "step": 10356 }, { "epoch": 0.48127323420074347, "grad_norm": 0.29484382050762253, "learning_rate": 5.556457234800726e-06, "loss": 0.0418, "step": 10357 }, { "epoch": 0.4813197026022305, "grad_norm": 1.1789414321010494, "learning_rate": 5.55570938992537e-06, "loss": 0.0499, "step": 10358 }, { "epoch": 0.4813661710037175, "grad_norm": 1.5504247778142362, "learning_rate": 5.55496153246261e-06, "loss": 0.0647, "step": 10359 }, { "epoch": 0.48141263940520446, "grad_norm": 0.772168333975975, "learning_rate": 5.55421366242939e-06, "loss": 0.0609, "step": 10360 }, { "epoch": 0.48145910780669143, "grad_norm": 0.58752841199151, "learning_rate": 5.553465779842645e-06, "loss": 0.0475, "step": 10361 }, { "epoch": 0.48150557620817847, "grad_norm": 0.5557806970986603, "learning_rate": 5.5527178847193185e-06, "loss": 0.0458, "step": 10362 }, { "epoch": 0.48155204460966544, "grad_norm": 0.7569799375071938, "learning_rate": 5.55196997707635e-06, "loss": 0.0346, "step": 10363 }, { "epoch": 0.4815985130111524, "grad_norm": 1.2915075853130893, "learning_rate": 5.55122205693068e-06, "loss": 0.0829, "step": 10364 }, { "epoch": 0.4816449814126394, "grad_norm": 0.6877590719104073, "learning_rate": 5.550474124299251e-06, "loss": 0.0495, "step": 10365 }, { "epoch": 0.48169144981412637, "grad_norm": 1.4799542312638014, "learning_rate": 5.549726179199003e-06, "loss": 0.0538, "step": 10366 }, { "epoch": 0.4817379182156134, "grad_norm": 1.2968879070481756, "learning_rate": 5.548978221646878e-06, "loss": 0.0711, "step": 10367 }, { "epoch": 0.4817843866171004, "grad_norm": 0.7477481995520638, "learning_rate": 5.548230251659821e-06, "loss": 0.0376, "step": 10368 }, { "epoch": 0.48183085501858736, "grad_norm": 0.34410671263645914, "learning_rate": 5.5474822692547695e-06, "loss": 0.0252, "step": 10369 }, { "epoch": 0.48187732342007433, "grad_norm": 0.5343721678252805, "learning_rate": 5.546734274448669e-06, "loss": 0.0435, "step": 10370 }, { "epoch": 0.48192379182156136, "grad_norm": 0.44902968575391133, "learning_rate": 5.545986267258462e-06, "loss": 0.0537, "step": 10371 }, { "epoch": 0.48197026022304834, "grad_norm": 1.0346673939902475, "learning_rate": 5.54523824770109e-06, "loss": 0.0508, "step": 10372 }, { "epoch": 0.4820167286245353, "grad_norm": 3.794561916348277, "learning_rate": 5.5444902157935e-06, "loss": 0.0687, "step": 10373 }, { "epoch": 0.4820631970260223, "grad_norm": 0.6756911335962084, "learning_rate": 5.543742171552633e-06, "loss": 0.0508, "step": 10374 }, { "epoch": 0.48210966542750927, "grad_norm": 0.49804600346245215, "learning_rate": 5.542994114995434e-06, "loss": 0.0404, "step": 10375 }, { "epoch": 0.4821561338289963, "grad_norm": 0.9534493997798312, "learning_rate": 5.542246046138845e-06, "loss": 0.0546, "step": 10376 }, { "epoch": 0.4822026022304833, "grad_norm": 0.8136757831451545, "learning_rate": 5.541497964999813e-06, "loss": 0.066, "step": 10377 }, { "epoch": 0.48224907063197026, "grad_norm": 0.9323524876564553, "learning_rate": 5.540749871595282e-06, "loss": 0.076, "step": 10378 }, { "epoch": 0.48229553903345723, "grad_norm": 0.759124863294453, "learning_rate": 5.540001765942198e-06, "loss": 0.0579, "step": 10379 }, { "epoch": 0.48234200743494426, "grad_norm": 5.935389912991681, "learning_rate": 5.539253648057507e-06, "loss": 0.0912, "step": 10380 }, { "epoch": 0.48238847583643124, "grad_norm": 0.8828246119074917, "learning_rate": 5.53850551795815e-06, "loss": 0.0525, "step": 10381 }, { "epoch": 0.4824349442379182, "grad_norm": 1.3829745138823766, "learning_rate": 5.537757375661077e-06, "loss": 0.0478, "step": 10382 }, { "epoch": 0.4824814126394052, "grad_norm": 0.7149553759927881, "learning_rate": 5.537009221183236e-06, "loss": 0.0617, "step": 10383 }, { "epoch": 0.48252788104089217, "grad_norm": 0.621026232992877, "learning_rate": 5.5362610545415685e-06, "loss": 0.0522, "step": 10384 }, { "epoch": 0.4825743494423792, "grad_norm": 0.5633743976863812, "learning_rate": 5.535512875753025e-06, "loss": 0.0566, "step": 10385 }, { "epoch": 0.4826208178438662, "grad_norm": 0.8471593043417159, "learning_rate": 5.534764684834551e-06, "loss": 0.0652, "step": 10386 }, { "epoch": 0.48266728624535316, "grad_norm": 0.5419003509199362, "learning_rate": 5.534016481803095e-06, "loss": 0.038, "step": 10387 }, { "epoch": 0.48271375464684013, "grad_norm": 0.9735717016260587, "learning_rate": 5.533268266675601e-06, "loss": 0.0785, "step": 10388 }, { "epoch": 0.48276022304832716, "grad_norm": 0.6908007446193466, "learning_rate": 5.532520039469021e-06, "loss": 0.0598, "step": 10389 }, { "epoch": 0.48280669144981414, "grad_norm": 1.7973823134807205, "learning_rate": 5.531771800200303e-06, "loss": 0.0881, "step": 10390 }, { "epoch": 0.4828531598513011, "grad_norm": 0.41541344369697725, "learning_rate": 5.5310235488863915e-06, "loss": 0.0691, "step": 10391 }, { "epoch": 0.4828996282527881, "grad_norm": 0.4233293172435052, "learning_rate": 5.53027528554424e-06, "loss": 0.0544, "step": 10392 }, { "epoch": 0.48294609665427507, "grad_norm": 0.5123819722380473, "learning_rate": 5.529527010190794e-06, "loss": 0.0425, "step": 10393 }, { "epoch": 0.4829925650557621, "grad_norm": 1.5407548359252485, "learning_rate": 5.528778722843002e-06, "loss": 0.067, "step": 10394 }, { "epoch": 0.4830390334572491, "grad_norm": 0.4438811929117641, "learning_rate": 5.528030423517819e-06, "loss": 0.0522, "step": 10395 }, { "epoch": 0.48308550185873605, "grad_norm": 0.8376019096019942, "learning_rate": 5.5272821122321894e-06, "loss": 0.0698, "step": 10396 }, { "epoch": 0.48313197026022303, "grad_norm": 0.42350995947756576, "learning_rate": 5.526533789003065e-06, "loss": 0.0229, "step": 10397 }, { "epoch": 0.48317843866171006, "grad_norm": 0.8927468103268904, "learning_rate": 5.5257854538473965e-06, "loss": 0.0439, "step": 10398 }, { "epoch": 0.48322490706319704, "grad_norm": 8.895379889220695, "learning_rate": 5.525037106782134e-06, "loss": 0.0815, "step": 10399 }, { "epoch": 0.483271375464684, "grad_norm": 0.5510903632575878, "learning_rate": 5.524288747824228e-06, "loss": 0.0442, "step": 10400 }, { "epoch": 0.483317843866171, "grad_norm": 0.46829118330762515, "learning_rate": 5.52354037699063e-06, "loss": 0.0362, "step": 10401 }, { "epoch": 0.48336431226765797, "grad_norm": 0.5057181008880379, "learning_rate": 5.522791994298293e-06, "loss": 0.0506, "step": 10402 }, { "epoch": 0.483410780669145, "grad_norm": 0.6576803199671466, "learning_rate": 5.522043599764165e-06, "loss": 0.049, "step": 10403 }, { "epoch": 0.483457249070632, "grad_norm": 0.713974081996278, "learning_rate": 5.521295193405202e-06, "loss": 0.0618, "step": 10404 }, { "epoch": 0.48350371747211895, "grad_norm": 0.6077133606895916, "learning_rate": 5.520546775238353e-06, "loss": 0.0354, "step": 10405 }, { "epoch": 0.48355018587360593, "grad_norm": 0.5791182203920183, "learning_rate": 5.5197983452805715e-06, "loss": 0.0592, "step": 10406 }, { "epoch": 0.48359665427509296, "grad_norm": 0.5373476548491806, "learning_rate": 5.5190499035488124e-06, "loss": 0.0476, "step": 10407 }, { "epoch": 0.48364312267657994, "grad_norm": 0.6616432947553161, "learning_rate": 5.518301450060023e-06, "loss": 0.0644, "step": 10408 }, { "epoch": 0.4836895910780669, "grad_norm": 1.4280199313388466, "learning_rate": 5.517552984831163e-06, "loss": 0.0849, "step": 10409 }, { "epoch": 0.4837360594795539, "grad_norm": 0.8194322833143011, "learning_rate": 5.5168045078791836e-06, "loss": 0.0256, "step": 10410 }, { "epoch": 0.48378252788104087, "grad_norm": 0.7495529088283894, "learning_rate": 5.516056019221035e-06, "loss": 0.069, "step": 10411 }, { "epoch": 0.4838289962825279, "grad_norm": 0.5301472060038104, "learning_rate": 5.515307518873677e-06, "loss": 0.0381, "step": 10412 }, { "epoch": 0.4838754646840149, "grad_norm": 0.5417714852862001, "learning_rate": 5.514559006854061e-06, "loss": 0.0468, "step": 10413 }, { "epoch": 0.48392193308550185, "grad_norm": 0.7008184828032712, "learning_rate": 5.513810483179143e-06, "loss": 0.0465, "step": 10414 }, { "epoch": 0.48396840148698883, "grad_norm": 0.7565998146545571, "learning_rate": 5.513061947865874e-06, "loss": 0.0753, "step": 10415 }, { "epoch": 0.48401486988847586, "grad_norm": 0.73920215269148, "learning_rate": 5.512313400931214e-06, "loss": 0.0622, "step": 10416 }, { "epoch": 0.48406133828996284, "grad_norm": 0.4426398383269267, "learning_rate": 5.511564842392115e-06, "loss": 0.0443, "step": 10417 }, { "epoch": 0.4841078066914498, "grad_norm": 0.5709195445635165, "learning_rate": 5.510816272265534e-06, "loss": 0.0423, "step": 10418 }, { "epoch": 0.4841542750929368, "grad_norm": 0.6102667218496891, "learning_rate": 5.510067690568429e-06, "loss": 0.0758, "step": 10419 }, { "epoch": 0.48420074349442377, "grad_norm": 1.5670137615464153, "learning_rate": 5.509319097317753e-06, "loss": 0.0559, "step": 10420 }, { "epoch": 0.4842472118959108, "grad_norm": 0.5555315538523167, "learning_rate": 5.508570492530461e-06, "loss": 0.0649, "step": 10421 }, { "epoch": 0.4842936802973978, "grad_norm": 0.45193407840263716, "learning_rate": 5.507821876223514e-06, "loss": 0.0439, "step": 10422 }, { "epoch": 0.48434014869888475, "grad_norm": 1.1339774955607569, "learning_rate": 5.507073248413868e-06, "loss": 0.0759, "step": 10423 }, { "epoch": 0.48438661710037173, "grad_norm": 0.6655322131106414, "learning_rate": 5.506324609118479e-06, "loss": 0.0786, "step": 10424 }, { "epoch": 0.48443308550185876, "grad_norm": 0.6358596104589269, "learning_rate": 5.505575958354303e-06, "loss": 0.061, "step": 10425 }, { "epoch": 0.48447955390334574, "grad_norm": 1.039680655265129, "learning_rate": 5.504827296138301e-06, "loss": 0.0678, "step": 10426 }, { "epoch": 0.4845260223048327, "grad_norm": 1.1471746978534614, "learning_rate": 5.504078622487429e-06, "loss": 0.0741, "step": 10427 }, { "epoch": 0.4845724907063197, "grad_norm": 0.6559623777338758, "learning_rate": 5.503329937418644e-06, "loss": 0.0446, "step": 10428 }, { "epoch": 0.48461895910780667, "grad_norm": 0.5648606133979839, "learning_rate": 5.5025812409489075e-06, "loss": 0.0476, "step": 10429 }, { "epoch": 0.4846654275092937, "grad_norm": 0.6775397351968007, "learning_rate": 5.5018325330951774e-06, "loss": 0.0572, "step": 10430 }, { "epoch": 0.4847118959107807, "grad_norm": 0.6011102382697858, "learning_rate": 5.501083813874413e-06, "loss": 0.0363, "step": 10431 }, { "epoch": 0.48475836431226765, "grad_norm": 2.0428283527905498, "learning_rate": 5.50033508330357e-06, "loss": 0.0486, "step": 10432 }, { "epoch": 0.48480483271375463, "grad_norm": 0.7170868618679359, "learning_rate": 5.499586341399613e-06, "loss": 0.0667, "step": 10433 }, { "epoch": 0.48485130111524166, "grad_norm": 0.6933850682149255, "learning_rate": 5.498837588179498e-06, "loss": 0.0643, "step": 10434 }, { "epoch": 0.48489776951672864, "grad_norm": 1.2738291018467656, "learning_rate": 5.498088823660188e-06, "loss": 0.0804, "step": 10435 }, { "epoch": 0.4849442379182156, "grad_norm": 0.6675201004825614, "learning_rate": 5.497340047858642e-06, "loss": 0.052, "step": 10436 }, { "epoch": 0.4849907063197026, "grad_norm": 0.7172065971829568, "learning_rate": 5.49659126079182e-06, "loss": 0.06, "step": 10437 }, { "epoch": 0.48503717472118957, "grad_norm": 0.6413459954681966, "learning_rate": 5.495842462476684e-06, "loss": 0.0585, "step": 10438 }, { "epoch": 0.4850836431226766, "grad_norm": 1.0570202948606877, "learning_rate": 5.495093652930193e-06, "loss": 0.058, "step": 10439 }, { "epoch": 0.4851301115241636, "grad_norm": 1.5085541063095167, "learning_rate": 5.49434483216931e-06, "loss": 0.0811, "step": 10440 }, { "epoch": 0.48517657992565055, "grad_norm": 0.7565154255363239, "learning_rate": 5.493596000210997e-06, "loss": 0.0459, "step": 10441 }, { "epoch": 0.48522304832713753, "grad_norm": 0.3061364220119672, "learning_rate": 5.4928471570722155e-06, "loss": 0.0358, "step": 10442 }, { "epoch": 0.48526951672862456, "grad_norm": 0.671766757285703, "learning_rate": 5.492098302769926e-06, "loss": 0.0523, "step": 10443 }, { "epoch": 0.48531598513011154, "grad_norm": 1.1308229281957975, "learning_rate": 5.4913494373210915e-06, "loss": 0.0557, "step": 10444 }, { "epoch": 0.4853624535315985, "grad_norm": 0.5795589299211433, "learning_rate": 5.490600560742676e-06, "loss": 0.0622, "step": 10445 }, { "epoch": 0.4854089219330855, "grad_norm": 1.1517185962628058, "learning_rate": 5.489851673051643e-06, "loss": 0.053, "step": 10446 }, { "epoch": 0.48545539033457247, "grad_norm": 0.8317816859682846, "learning_rate": 5.489102774264951e-06, "loss": 0.0679, "step": 10447 }, { "epoch": 0.4855018587360595, "grad_norm": 0.7348262341235245, "learning_rate": 5.488353864399569e-06, "loss": 0.0779, "step": 10448 }, { "epoch": 0.4855483271375465, "grad_norm": 1.044037472412565, "learning_rate": 5.487604943472457e-06, "loss": 0.0348, "step": 10449 }, { "epoch": 0.48559479553903345, "grad_norm": 0.5753291276442856, "learning_rate": 5.486856011500578e-06, "loss": 0.0443, "step": 10450 }, { "epoch": 0.48564126394052043, "grad_norm": 0.6493649066224753, "learning_rate": 5.486107068500901e-06, "loss": 0.0605, "step": 10451 }, { "epoch": 0.48568773234200746, "grad_norm": 0.4872916560820618, "learning_rate": 5.485358114490386e-06, "loss": 0.049, "step": 10452 }, { "epoch": 0.48573420074349444, "grad_norm": 0.45322581546828555, "learning_rate": 5.484609149485999e-06, "loss": 0.0281, "step": 10453 }, { "epoch": 0.4857806691449814, "grad_norm": 0.8490258259644775, "learning_rate": 5.483860173504703e-06, "loss": 0.0452, "step": 10454 }, { "epoch": 0.4858271375464684, "grad_norm": 0.5704104785825822, "learning_rate": 5.483111186563468e-06, "loss": 0.0587, "step": 10455 }, { "epoch": 0.48587360594795537, "grad_norm": 0.6748552963412893, "learning_rate": 5.482362188679253e-06, "loss": 0.0651, "step": 10456 }, { "epoch": 0.4859200743494424, "grad_norm": 0.3809917233264822, "learning_rate": 5.4816131798690285e-06, "loss": 0.048, "step": 10457 }, { "epoch": 0.4859665427509294, "grad_norm": 0.5273452810247633, "learning_rate": 5.4808641601497596e-06, "loss": 0.0537, "step": 10458 }, { "epoch": 0.48601301115241635, "grad_norm": 1.9285015830981505, "learning_rate": 5.480115129538409e-06, "loss": 0.0598, "step": 10459 }, { "epoch": 0.48605947955390333, "grad_norm": 0.43430323918405245, "learning_rate": 5.479366088051947e-06, "loss": 0.0568, "step": 10460 }, { "epoch": 0.48610594795539036, "grad_norm": 0.6696242783295107, "learning_rate": 5.478617035707337e-06, "loss": 0.0558, "step": 10461 }, { "epoch": 0.48615241635687734, "grad_norm": 0.5958206684571272, "learning_rate": 5.47786797252155e-06, "loss": 0.0479, "step": 10462 }, { "epoch": 0.4861988847583643, "grad_norm": 0.6493496583341736, "learning_rate": 5.477118898511549e-06, "loss": 0.0578, "step": 10463 }, { "epoch": 0.4862453531598513, "grad_norm": 0.6641276623035568, "learning_rate": 5.476369813694302e-06, "loss": 0.0629, "step": 10464 }, { "epoch": 0.48629182156133827, "grad_norm": 0.8303236089038221, "learning_rate": 5.475620718086778e-06, "loss": 0.0642, "step": 10465 }, { "epoch": 0.4863382899628253, "grad_norm": 0.9724286112351034, "learning_rate": 5.4748716117059456e-06, "loss": 0.066, "step": 10466 }, { "epoch": 0.4863847583643123, "grad_norm": 0.6239896193413417, "learning_rate": 5.47412249456877e-06, "loss": 0.0438, "step": 10467 }, { "epoch": 0.48643122676579925, "grad_norm": 0.7689167929099721, "learning_rate": 5.473373366692219e-06, "loss": 0.0608, "step": 10468 }, { "epoch": 0.48647769516728623, "grad_norm": 0.3957408033462398, "learning_rate": 5.472624228093266e-06, "loss": 0.0319, "step": 10469 }, { "epoch": 0.48652416356877326, "grad_norm": 0.9335972867847011, "learning_rate": 5.471875078788876e-06, "loss": 0.078, "step": 10470 }, { "epoch": 0.48657063197026024, "grad_norm": 0.4226821461730393, "learning_rate": 5.471125918796017e-06, "loss": 0.0538, "step": 10471 }, { "epoch": 0.4866171003717472, "grad_norm": 0.9107048970025634, "learning_rate": 5.470376748131664e-06, "loss": 0.0406, "step": 10472 }, { "epoch": 0.4866635687732342, "grad_norm": 0.8762418128750458, "learning_rate": 5.469627566812779e-06, "loss": 0.0755, "step": 10473 }, { "epoch": 0.48671003717472117, "grad_norm": 2.564757844929684, "learning_rate": 5.4688783748563376e-06, "loss": 0.0782, "step": 10474 }, { "epoch": 0.4867565055762082, "grad_norm": 0.7058952554365431, "learning_rate": 5.468129172279306e-06, "loss": 0.0441, "step": 10475 }, { "epoch": 0.4868029739776952, "grad_norm": 1.0426170987349161, "learning_rate": 5.467379959098657e-06, "loss": 0.07, "step": 10476 }, { "epoch": 0.48684944237918215, "grad_norm": 0.9213716874627836, "learning_rate": 5.46663073533136e-06, "loss": 0.0645, "step": 10477 }, { "epoch": 0.48689591078066913, "grad_norm": 0.5923729811434731, "learning_rate": 5.465881500994384e-06, "loss": 0.0638, "step": 10478 }, { "epoch": 0.48694237918215616, "grad_norm": 0.6712147516056721, "learning_rate": 5.465132256104704e-06, "loss": 0.054, "step": 10479 }, { "epoch": 0.48698884758364314, "grad_norm": 0.783128480923634, "learning_rate": 5.464383000679287e-06, "loss": 0.057, "step": 10480 }, { "epoch": 0.4870353159851301, "grad_norm": 1.2927261190209844, "learning_rate": 5.463633734735108e-06, "loss": 0.0497, "step": 10481 }, { "epoch": 0.4870817843866171, "grad_norm": 0.7491286034017377, "learning_rate": 5.462884458289136e-06, "loss": 0.0739, "step": 10482 }, { "epoch": 0.48712825278810407, "grad_norm": 0.866457344435774, "learning_rate": 5.462135171358344e-06, "loss": 0.0591, "step": 10483 }, { "epoch": 0.4871747211895911, "grad_norm": 0.7967953331291057, "learning_rate": 5.461385873959703e-06, "loss": 0.0586, "step": 10484 }, { "epoch": 0.4872211895910781, "grad_norm": 0.45879667773284555, "learning_rate": 5.460636566110187e-06, "loss": 0.0471, "step": 10485 }, { "epoch": 0.48726765799256505, "grad_norm": 0.6212869970821628, "learning_rate": 5.4598872478267686e-06, "loss": 0.0796, "step": 10486 }, { "epoch": 0.487314126394052, "grad_norm": 1.052175579464211, "learning_rate": 5.4591379191264205e-06, "loss": 0.0577, "step": 10487 }, { "epoch": 0.48736059479553906, "grad_norm": 1.3402014137667935, "learning_rate": 5.458388580026113e-06, "loss": 0.067, "step": 10488 }, { "epoch": 0.48740706319702604, "grad_norm": 1.6736035943614158, "learning_rate": 5.457639230542825e-06, "loss": 0.0707, "step": 10489 }, { "epoch": 0.487453531598513, "grad_norm": 0.9939950900467978, "learning_rate": 5.456889870693524e-06, "loss": 0.0645, "step": 10490 }, { "epoch": 0.4875, "grad_norm": 0.9452950996732116, "learning_rate": 5.4561405004951874e-06, "loss": 0.0575, "step": 10491 }, { "epoch": 0.48754646840148697, "grad_norm": 1.8459022320094265, "learning_rate": 5.455391119964788e-06, "loss": 0.0535, "step": 10492 }, { "epoch": 0.487592936802974, "grad_norm": 0.5720558177674938, "learning_rate": 5.454641729119301e-06, "loss": 0.0573, "step": 10493 }, { "epoch": 0.487639405204461, "grad_norm": 1.2404067179678477, "learning_rate": 5.453892327975701e-06, "loss": 0.0663, "step": 10494 }, { "epoch": 0.48768587360594795, "grad_norm": 1.804321463938783, "learning_rate": 5.453142916550961e-06, "loss": 0.0888, "step": 10495 }, { "epoch": 0.4877323420074349, "grad_norm": 0.3488497298979056, "learning_rate": 5.452393494862057e-06, "loss": 0.0353, "step": 10496 }, { "epoch": 0.48777881040892196, "grad_norm": 2.23522374560359, "learning_rate": 5.451644062925964e-06, "loss": 0.0579, "step": 10497 }, { "epoch": 0.48782527881040894, "grad_norm": 0.8222316023357139, "learning_rate": 5.450894620759658e-06, "loss": 0.0758, "step": 10498 }, { "epoch": 0.4878717472118959, "grad_norm": 0.6776354835070139, "learning_rate": 5.450145168380115e-06, "loss": 0.0578, "step": 10499 }, { "epoch": 0.4879182156133829, "grad_norm": 1.143653150952807, "learning_rate": 5.449395705804308e-06, "loss": 0.0846, "step": 10500 }, { "epoch": 0.48796468401486986, "grad_norm": 0.7142409815016082, "learning_rate": 5.448646233049219e-06, "loss": 0.055, "step": 10501 }, { "epoch": 0.4880111524163569, "grad_norm": 0.9342761986983142, "learning_rate": 5.447896750131817e-06, "loss": 0.0455, "step": 10502 }, { "epoch": 0.4880576208178439, "grad_norm": 0.8125385979970928, "learning_rate": 5.4471472570690846e-06, "loss": 0.0632, "step": 10503 }, { "epoch": 0.48810408921933085, "grad_norm": 0.6788702691773663, "learning_rate": 5.446397753877995e-06, "loss": 0.0523, "step": 10504 }, { "epoch": 0.4881505576208178, "grad_norm": 0.6026352188834042, "learning_rate": 5.445648240575526e-06, "loss": 0.0533, "step": 10505 }, { "epoch": 0.48819702602230486, "grad_norm": 0.5212519415212215, "learning_rate": 5.444898717178656e-06, "loss": 0.0397, "step": 10506 }, { "epoch": 0.48824349442379184, "grad_norm": 0.47556891903725484, "learning_rate": 5.4441491837043624e-06, "loss": 0.0707, "step": 10507 }, { "epoch": 0.4882899628252788, "grad_norm": 0.9193928458636783, "learning_rate": 5.443399640169622e-06, "loss": 0.0475, "step": 10508 }, { "epoch": 0.4883364312267658, "grad_norm": 1.1059830145970637, "learning_rate": 5.442650086591412e-06, "loss": 0.0711, "step": 10509 }, { "epoch": 0.48838289962825276, "grad_norm": 0.7298839099516465, "learning_rate": 5.441900522986712e-06, "loss": 0.0607, "step": 10510 }, { "epoch": 0.4884293680297398, "grad_norm": 0.5311046449755106, "learning_rate": 5.441150949372503e-06, "loss": 0.0397, "step": 10511 }, { "epoch": 0.4884758364312268, "grad_norm": 0.9641471364143909, "learning_rate": 5.440401365765756e-06, "loss": 0.0724, "step": 10512 }, { "epoch": 0.48852230483271375, "grad_norm": 0.580529706340093, "learning_rate": 5.439651772183459e-06, "loss": 0.0397, "step": 10513 }, { "epoch": 0.4885687732342007, "grad_norm": 0.5526377871740733, "learning_rate": 5.438902168642584e-06, "loss": 0.0714, "step": 10514 }, { "epoch": 0.48861524163568776, "grad_norm": 0.6507847602869238, "learning_rate": 5.438152555160114e-06, "loss": 0.0612, "step": 10515 }, { "epoch": 0.48866171003717473, "grad_norm": 0.8394565550938716, "learning_rate": 5.437402931753026e-06, "loss": 0.065, "step": 10516 }, { "epoch": 0.4887081784386617, "grad_norm": 0.703789422313873, "learning_rate": 5.436653298438301e-06, "loss": 0.0536, "step": 10517 }, { "epoch": 0.4887546468401487, "grad_norm": 0.8392410371790167, "learning_rate": 5.4359036552329215e-06, "loss": 0.0472, "step": 10518 }, { "epoch": 0.48880111524163566, "grad_norm": 1.037220027922172, "learning_rate": 5.435154002153864e-06, "loss": 0.0479, "step": 10519 }, { "epoch": 0.4888475836431227, "grad_norm": 2.6860563627883027, "learning_rate": 5.434404339218111e-06, "loss": 0.0637, "step": 10520 }, { "epoch": 0.4888940520446097, "grad_norm": 1.1098843427614664, "learning_rate": 5.4336546664426415e-06, "loss": 0.0642, "step": 10521 }, { "epoch": 0.48894052044609665, "grad_norm": 0.7144838420586068, "learning_rate": 5.432904983844438e-06, "loss": 0.0333, "step": 10522 }, { "epoch": 0.4889869888475836, "grad_norm": 0.40048715766090665, "learning_rate": 5.432155291440482e-06, "loss": 0.0516, "step": 10523 }, { "epoch": 0.48903345724907066, "grad_norm": 0.4583779743507837, "learning_rate": 5.431405589247751e-06, "loss": 0.051, "step": 10524 }, { "epoch": 0.48907992565055763, "grad_norm": 1.0830369501866906, "learning_rate": 5.430655877283233e-06, "loss": 0.0749, "step": 10525 }, { "epoch": 0.4891263940520446, "grad_norm": 0.2875259281570144, "learning_rate": 5.429906155563904e-06, "loss": 0.0396, "step": 10526 }, { "epoch": 0.4891728624535316, "grad_norm": 0.5944411076048929, "learning_rate": 5.429156424106748e-06, "loss": 0.0379, "step": 10527 }, { "epoch": 0.48921933085501856, "grad_norm": 0.970390499691589, "learning_rate": 5.428406682928749e-06, "loss": 0.0648, "step": 10528 }, { "epoch": 0.4892657992565056, "grad_norm": 0.5447330998253452, "learning_rate": 5.427656932046886e-06, "loss": 0.0295, "step": 10529 }, { "epoch": 0.4893122676579926, "grad_norm": 0.37705711595150876, "learning_rate": 5.426907171478143e-06, "loss": 0.0328, "step": 10530 }, { "epoch": 0.48935873605947955, "grad_norm": 0.637922676412405, "learning_rate": 5.426157401239504e-06, "loss": 0.0454, "step": 10531 }, { "epoch": 0.4894052044609665, "grad_norm": 0.6960679400017511, "learning_rate": 5.425407621347952e-06, "loss": 0.0576, "step": 10532 }, { "epoch": 0.48945167286245356, "grad_norm": 0.6201741946931558, "learning_rate": 5.424657831820468e-06, "loss": 0.0546, "step": 10533 }, { "epoch": 0.48949814126394053, "grad_norm": 0.6230903842014383, "learning_rate": 5.423908032674039e-06, "loss": 0.0538, "step": 10534 }, { "epoch": 0.4895446096654275, "grad_norm": 0.5487192483812657, "learning_rate": 5.423158223925647e-06, "loss": 0.0483, "step": 10535 }, { "epoch": 0.4895910780669145, "grad_norm": 0.857847254197147, "learning_rate": 5.422408405592273e-06, "loss": 0.0628, "step": 10536 }, { "epoch": 0.48963754646840146, "grad_norm": 0.7193075286249477, "learning_rate": 5.421658577690908e-06, "loss": 0.0456, "step": 10537 }, { "epoch": 0.4896840148698885, "grad_norm": 0.8611150735850904, "learning_rate": 5.4209087402385296e-06, "loss": 0.0461, "step": 10538 }, { "epoch": 0.48973048327137547, "grad_norm": 0.9939958638395495, "learning_rate": 5.420158893252126e-06, "loss": 0.0679, "step": 10539 }, { "epoch": 0.48977695167286245, "grad_norm": 0.5231821274578654, "learning_rate": 5.419409036748682e-06, "loss": 0.0526, "step": 10540 }, { "epoch": 0.4898234200743494, "grad_norm": 0.656765032397649, "learning_rate": 5.418659170745181e-06, "loss": 0.0448, "step": 10541 }, { "epoch": 0.48986988847583646, "grad_norm": 0.6939217014135464, "learning_rate": 5.417909295258611e-06, "loss": 0.0539, "step": 10542 }, { "epoch": 0.48991635687732343, "grad_norm": 0.8906648812025798, "learning_rate": 5.417159410305953e-06, "loss": 0.0584, "step": 10543 }, { "epoch": 0.4899628252788104, "grad_norm": 0.47340167745943074, "learning_rate": 5.416409515904196e-06, "loss": 0.0609, "step": 10544 }, { "epoch": 0.4900092936802974, "grad_norm": 2.029344542667647, "learning_rate": 5.415659612070327e-06, "loss": 0.0812, "step": 10545 }, { "epoch": 0.49005576208178436, "grad_norm": 1.2147115059954905, "learning_rate": 5.414909698821327e-06, "loss": 0.0582, "step": 10546 }, { "epoch": 0.4901022304832714, "grad_norm": 1.0848523757618194, "learning_rate": 5.4141597761741885e-06, "loss": 0.0525, "step": 10547 }, { "epoch": 0.49014869888475837, "grad_norm": 1.0563103608414548, "learning_rate": 5.413409844145894e-06, "loss": 0.067, "step": 10548 }, { "epoch": 0.49019516728624535, "grad_norm": 1.0356364189879408, "learning_rate": 5.412659902753432e-06, "loss": 0.0523, "step": 10549 }, { "epoch": 0.4902416356877323, "grad_norm": 0.6728525712910394, "learning_rate": 5.4119099520137875e-06, "loss": 0.0396, "step": 10550 }, { "epoch": 0.49028810408921936, "grad_norm": 0.639823872814524, "learning_rate": 5.411159991943951e-06, "loss": 0.0404, "step": 10551 }, { "epoch": 0.49033457249070633, "grad_norm": 0.5010407275471316, "learning_rate": 5.410410022560907e-06, "loss": 0.0496, "step": 10552 }, { "epoch": 0.4903810408921933, "grad_norm": 0.6705160452378652, "learning_rate": 5.409660043881643e-06, "loss": 0.0514, "step": 10553 }, { "epoch": 0.4904275092936803, "grad_norm": 0.9664357413761417, "learning_rate": 5.40891005592315e-06, "loss": 0.0682, "step": 10554 }, { "epoch": 0.49047397769516726, "grad_norm": 0.9656777685731213, "learning_rate": 5.408160058702412e-06, "loss": 0.0688, "step": 10555 }, { "epoch": 0.4905204460966543, "grad_norm": 0.7307702833542399, "learning_rate": 5.40741005223642e-06, "loss": 0.0546, "step": 10556 }, { "epoch": 0.49056691449814127, "grad_norm": 1.06977927881123, "learning_rate": 5.406660036542162e-06, "loss": 0.0665, "step": 10557 }, { "epoch": 0.49061338289962825, "grad_norm": 0.3900971059237017, "learning_rate": 5.405910011636625e-06, "loss": 0.025, "step": 10558 }, { "epoch": 0.4906598513011152, "grad_norm": 0.8549927088983049, "learning_rate": 5.4051599775368e-06, "loss": 0.0652, "step": 10559 }, { "epoch": 0.49070631970260226, "grad_norm": 0.8014782578480012, "learning_rate": 5.404409934259674e-06, "loss": 0.0529, "step": 10560 }, { "epoch": 0.49075278810408923, "grad_norm": 1.893829120196406, "learning_rate": 5.403659881822239e-06, "loss": 0.0768, "step": 10561 }, { "epoch": 0.4907992565055762, "grad_norm": 0.8068825761084971, "learning_rate": 5.402909820241481e-06, "loss": 0.0653, "step": 10562 }, { "epoch": 0.4908457249070632, "grad_norm": 0.5610693740813547, "learning_rate": 5.402159749534392e-06, "loss": 0.0431, "step": 10563 }, { "epoch": 0.49089219330855016, "grad_norm": 1.2188962573541817, "learning_rate": 5.401409669717963e-06, "loss": 0.0559, "step": 10564 }, { "epoch": 0.4909386617100372, "grad_norm": 0.675050380154799, "learning_rate": 5.40065958080918e-06, "loss": 0.0519, "step": 10565 }, { "epoch": 0.49098513011152417, "grad_norm": 0.5124733430079021, "learning_rate": 5.399909482825039e-06, "loss": 0.0294, "step": 10566 }, { "epoch": 0.49103159851301115, "grad_norm": 0.9173662921887937, "learning_rate": 5.399159375782524e-06, "loss": 0.0731, "step": 10567 }, { "epoch": 0.4910780669144981, "grad_norm": 0.40284655381505485, "learning_rate": 5.398409259698631e-06, "loss": 0.0545, "step": 10568 }, { "epoch": 0.49112453531598516, "grad_norm": 0.5182738403552398, "learning_rate": 5.397659134590348e-06, "loss": 0.0492, "step": 10569 }, { "epoch": 0.49117100371747213, "grad_norm": 0.47338967865115283, "learning_rate": 5.3969090004746675e-06, "loss": 0.0507, "step": 10570 }, { "epoch": 0.4912174721189591, "grad_norm": 1.2686539191492956, "learning_rate": 5.396158857368581e-06, "loss": 0.0621, "step": 10571 }, { "epoch": 0.4912639405204461, "grad_norm": 0.7260778975406548, "learning_rate": 5.395408705289078e-06, "loss": 0.048, "step": 10572 }, { "epoch": 0.49131040892193306, "grad_norm": 1.917077296234466, "learning_rate": 5.394658544253153e-06, "loss": 0.1094, "step": 10573 }, { "epoch": 0.4913568773234201, "grad_norm": 0.560116179265927, "learning_rate": 5.393908374277796e-06, "loss": 0.042, "step": 10574 }, { "epoch": 0.49140334572490707, "grad_norm": 0.6376163294618459, "learning_rate": 5.39315819538e-06, "loss": 0.0674, "step": 10575 }, { "epoch": 0.49144981412639405, "grad_norm": 0.7012469570011156, "learning_rate": 5.392408007576757e-06, "loss": 0.055, "step": 10576 }, { "epoch": 0.491496282527881, "grad_norm": 0.5968166577053438, "learning_rate": 5.391657810885058e-06, "loss": 0.0483, "step": 10577 }, { "epoch": 0.49154275092936806, "grad_norm": 0.9126274834037527, "learning_rate": 5.3909076053219e-06, "loss": 0.0799, "step": 10578 }, { "epoch": 0.49158921933085503, "grad_norm": 1.131269138075226, "learning_rate": 5.3901573909042715e-06, "loss": 0.0609, "step": 10579 }, { "epoch": 0.491635687732342, "grad_norm": 0.677323695266762, "learning_rate": 5.3894071676491685e-06, "loss": 0.0573, "step": 10580 }, { "epoch": 0.491682156133829, "grad_norm": 0.446594979465617, "learning_rate": 5.388656935573584e-06, "loss": 0.0593, "step": 10581 }, { "epoch": 0.49172862453531596, "grad_norm": 0.4945396772666422, "learning_rate": 5.387906694694509e-06, "loss": 0.0621, "step": 10582 }, { "epoch": 0.491775092936803, "grad_norm": 0.7166720908476252, "learning_rate": 5.387156445028939e-06, "loss": 0.0418, "step": 10583 }, { "epoch": 0.49182156133828997, "grad_norm": 0.8792498084434673, "learning_rate": 5.38640618659387e-06, "loss": 0.0473, "step": 10584 }, { "epoch": 0.49186802973977695, "grad_norm": 1.4375639293845655, "learning_rate": 5.385655919406293e-06, "loss": 0.0836, "step": 10585 }, { "epoch": 0.4919144981412639, "grad_norm": 1.7131182897604795, "learning_rate": 5.384905643483204e-06, "loss": 0.0806, "step": 10586 }, { "epoch": 0.49196096654275095, "grad_norm": 0.5769557276388785, "learning_rate": 5.384155358841596e-06, "loss": 0.0521, "step": 10587 }, { "epoch": 0.49200743494423793, "grad_norm": 0.5589915854078491, "learning_rate": 5.383405065498466e-06, "loss": 0.0594, "step": 10588 }, { "epoch": 0.4920539033457249, "grad_norm": 0.6570580490419556, "learning_rate": 5.382654763470807e-06, "loss": 0.0434, "step": 10589 }, { "epoch": 0.4921003717472119, "grad_norm": 0.5319396786484634, "learning_rate": 5.381904452775616e-06, "loss": 0.0589, "step": 10590 }, { "epoch": 0.49214684014869886, "grad_norm": 0.8507125035777129, "learning_rate": 5.381154133429886e-06, "loss": 0.0858, "step": 10591 }, { "epoch": 0.4921933085501859, "grad_norm": 0.487270433871745, "learning_rate": 5.380403805450613e-06, "loss": 0.0398, "step": 10592 }, { "epoch": 0.49223977695167287, "grad_norm": 0.4307759113829764, "learning_rate": 5.3796534688547965e-06, "loss": 0.0486, "step": 10593 }, { "epoch": 0.49228624535315985, "grad_norm": 0.8873734454612452, "learning_rate": 5.378903123659426e-06, "loss": 0.0733, "step": 10594 }, { "epoch": 0.4923327137546468, "grad_norm": 0.7926280775408118, "learning_rate": 5.378152769881503e-06, "loss": 0.0519, "step": 10595 }, { "epoch": 0.49237918215613385, "grad_norm": 0.6088798386437195, "learning_rate": 5.37740240753802e-06, "loss": 0.0626, "step": 10596 }, { "epoch": 0.49242565055762083, "grad_norm": 0.431783181466613, "learning_rate": 5.376652036645978e-06, "loss": 0.0521, "step": 10597 }, { "epoch": 0.4924721189591078, "grad_norm": 0.9027756360146588, "learning_rate": 5.375901657222369e-06, "loss": 0.0559, "step": 10598 }, { "epoch": 0.4925185873605948, "grad_norm": 0.6105135499165194, "learning_rate": 5.375151269284191e-06, "loss": 0.0412, "step": 10599 }, { "epoch": 0.49256505576208176, "grad_norm": 0.7350059434716588, "learning_rate": 5.3744008728484445e-06, "loss": 0.0504, "step": 10600 }, { "epoch": 0.4926115241635688, "grad_norm": 1.3422036083700026, "learning_rate": 5.373650467932122e-06, "loss": 0.0624, "step": 10601 }, { "epoch": 0.49265799256505577, "grad_norm": 0.6219632349422601, "learning_rate": 5.372900054552225e-06, "loss": 0.0459, "step": 10602 }, { "epoch": 0.49270446096654275, "grad_norm": 0.4292459560300083, "learning_rate": 5.372149632725747e-06, "loss": 0.0274, "step": 10603 }, { "epoch": 0.4927509293680297, "grad_norm": 0.5482348307847426, "learning_rate": 5.371399202469689e-06, "loss": 0.0424, "step": 10604 }, { "epoch": 0.49279739776951675, "grad_norm": 0.6716844458331146, "learning_rate": 5.37064876380105e-06, "loss": 0.0584, "step": 10605 }, { "epoch": 0.49284386617100373, "grad_norm": 1.0823197555145638, "learning_rate": 5.369898316736823e-06, "loss": 0.0496, "step": 10606 }, { "epoch": 0.4928903345724907, "grad_norm": 0.9868140726396778, "learning_rate": 5.369147861294013e-06, "loss": 0.0433, "step": 10607 }, { "epoch": 0.4929368029739777, "grad_norm": 0.8829742775745406, "learning_rate": 5.368397397489615e-06, "loss": 0.0643, "step": 10608 }, { "epoch": 0.49298327137546466, "grad_norm": 0.7648248623939495, "learning_rate": 5.367646925340626e-06, "loss": 0.0664, "step": 10609 }, { "epoch": 0.4930297397769517, "grad_norm": 1.0904714323701268, "learning_rate": 5.3668964448640495e-06, "loss": 0.0624, "step": 10610 }, { "epoch": 0.49307620817843867, "grad_norm": 3.5778495985593426, "learning_rate": 5.3661459560768805e-06, "loss": 0.054, "step": 10611 }, { "epoch": 0.49312267657992565, "grad_norm": 0.750156448425121, "learning_rate": 5.365395458996121e-06, "loss": 0.0502, "step": 10612 }, { "epoch": 0.4931691449814126, "grad_norm": 0.577975817923305, "learning_rate": 5.36464495363877e-06, "loss": 0.0408, "step": 10613 }, { "epoch": 0.49321561338289965, "grad_norm": 0.8313998509861987, "learning_rate": 5.363894440021827e-06, "loss": 0.0303, "step": 10614 }, { "epoch": 0.49326208178438663, "grad_norm": 0.5019317958075907, "learning_rate": 5.363143918162291e-06, "loss": 0.049, "step": 10615 }, { "epoch": 0.4933085501858736, "grad_norm": 0.4349695423362665, "learning_rate": 5.362393388077165e-06, "loss": 0.0505, "step": 10616 }, { "epoch": 0.4933550185873606, "grad_norm": 0.7221760796230248, "learning_rate": 5.361642849783446e-06, "loss": 0.0597, "step": 10617 }, { "epoch": 0.49340148698884756, "grad_norm": 0.5237542376305433, "learning_rate": 5.360892303298136e-06, "loss": 0.0448, "step": 10618 }, { "epoch": 0.4934479553903346, "grad_norm": 0.6150941034483554, "learning_rate": 5.360141748638237e-06, "loss": 0.0647, "step": 10619 }, { "epoch": 0.49349442379182157, "grad_norm": 0.6372217095274019, "learning_rate": 5.359391185820746e-06, "loss": 0.0563, "step": 10620 }, { "epoch": 0.49354089219330854, "grad_norm": 0.973782056474772, "learning_rate": 5.358640614862667e-06, "loss": 0.0696, "step": 10621 }, { "epoch": 0.4935873605947955, "grad_norm": 0.5411371798371337, "learning_rate": 5.357890035781001e-06, "loss": 0.0416, "step": 10622 }, { "epoch": 0.49363382899628255, "grad_norm": 0.7617096914574895, "learning_rate": 5.3571394485927484e-06, "loss": 0.0774, "step": 10623 }, { "epoch": 0.49368029739776953, "grad_norm": 0.9517760832668979, "learning_rate": 5.3563888533149135e-06, "loss": 0.0614, "step": 10624 }, { "epoch": 0.4937267657992565, "grad_norm": 2.091936315578139, "learning_rate": 5.3556382499644946e-06, "loss": 0.0738, "step": 10625 }, { "epoch": 0.4937732342007435, "grad_norm": 0.8994816457109782, "learning_rate": 5.354887638558493e-06, "loss": 0.098, "step": 10626 }, { "epoch": 0.49381970260223046, "grad_norm": 1.4992369237738152, "learning_rate": 5.354137019113916e-06, "loss": 0.066, "step": 10627 }, { "epoch": 0.4938661710037175, "grad_norm": 1.3923999903779867, "learning_rate": 5.353386391647761e-06, "loss": 0.0785, "step": 10628 }, { "epoch": 0.49391263940520447, "grad_norm": 0.7501217900037209, "learning_rate": 5.352635756177034e-06, "loss": 0.0555, "step": 10629 }, { "epoch": 0.49395910780669144, "grad_norm": 0.5050530911662748, "learning_rate": 5.3518851127187334e-06, "loss": 0.0484, "step": 10630 }, { "epoch": 0.4940055762081784, "grad_norm": 0.8761617182236706, "learning_rate": 5.351134461289867e-06, "loss": 0.0531, "step": 10631 }, { "epoch": 0.49405204460966545, "grad_norm": 0.4766498500887116, "learning_rate": 5.350383801907434e-06, "loss": 0.0549, "step": 10632 }, { "epoch": 0.49409851301115243, "grad_norm": 0.5438211130108335, "learning_rate": 5.3496331345884386e-06, "loss": 0.0456, "step": 10633 }, { "epoch": 0.4941449814126394, "grad_norm": 0.7629885150543905, "learning_rate": 5.348882459349888e-06, "loss": 0.045, "step": 10634 }, { "epoch": 0.4941914498141264, "grad_norm": 0.49973061752176096, "learning_rate": 5.348131776208779e-06, "loss": 0.0517, "step": 10635 }, { "epoch": 0.49423791821561336, "grad_norm": 1.004162404224027, "learning_rate": 5.34738108518212e-06, "loss": 0.0751, "step": 10636 }, { "epoch": 0.4942843866171004, "grad_norm": 0.46657458581195504, "learning_rate": 5.346630386286915e-06, "loss": 0.0421, "step": 10637 }, { "epoch": 0.49433085501858737, "grad_norm": 0.7702460504556727, "learning_rate": 5.345879679540164e-06, "loss": 0.0471, "step": 10638 }, { "epoch": 0.49437732342007434, "grad_norm": 1.6059071544584165, "learning_rate": 5.345128964958878e-06, "loss": 0.068, "step": 10639 }, { "epoch": 0.4944237918215613, "grad_norm": 0.8470603085278724, "learning_rate": 5.344378242560055e-06, "loss": 0.0629, "step": 10640 }, { "epoch": 0.49447026022304835, "grad_norm": 0.3415651291698192, "learning_rate": 5.343627512360704e-06, "loss": 0.0363, "step": 10641 }, { "epoch": 0.49451672862453533, "grad_norm": 1.036999345418413, "learning_rate": 5.342876774377827e-06, "loss": 0.0766, "step": 10642 }, { "epoch": 0.4945631970260223, "grad_norm": 0.6150745973498716, "learning_rate": 5.342126028628431e-06, "loss": 0.0534, "step": 10643 }, { "epoch": 0.4946096654275093, "grad_norm": 0.6144927230495764, "learning_rate": 5.34137527512952e-06, "loss": 0.0636, "step": 10644 }, { "epoch": 0.49465613382899626, "grad_norm": 0.64321700630863, "learning_rate": 5.3406245138980985e-06, "loss": 0.0586, "step": 10645 }, { "epoch": 0.4947026022304833, "grad_norm": 0.8501328819796747, "learning_rate": 5.3398737449511764e-06, "loss": 0.0559, "step": 10646 }, { "epoch": 0.49474907063197027, "grad_norm": 0.6219507304328712, "learning_rate": 5.339122968305753e-06, "loss": 0.0548, "step": 10647 }, { "epoch": 0.49479553903345724, "grad_norm": 1.3188739983074422, "learning_rate": 5.33837218397884e-06, "loss": 0.0988, "step": 10648 }, { "epoch": 0.4948420074349442, "grad_norm": 1.2602558826851409, "learning_rate": 5.3376213919874385e-06, "loss": 0.0707, "step": 10649 }, { "epoch": 0.49488847583643125, "grad_norm": 0.7574814984596677, "learning_rate": 5.336870592348557e-06, "loss": 0.0558, "step": 10650 }, { "epoch": 0.49493494423791823, "grad_norm": 0.3963082975939573, "learning_rate": 5.336119785079205e-06, "loss": 0.0296, "step": 10651 }, { "epoch": 0.4949814126394052, "grad_norm": 1.1868802377929086, "learning_rate": 5.335368970196384e-06, "loss": 0.0558, "step": 10652 }, { "epoch": 0.4950278810408922, "grad_norm": 0.43943678852634993, "learning_rate": 5.3346181477171044e-06, "loss": 0.0421, "step": 10653 }, { "epoch": 0.49507434944237916, "grad_norm": 0.5941190058197368, "learning_rate": 5.333867317658369e-06, "loss": 0.0593, "step": 10654 }, { "epoch": 0.4951208178438662, "grad_norm": 0.5749978764546265, "learning_rate": 5.333116480037189e-06, "loss": 0.075, "step": 10655 }, { "epoch": 0.49516728624535317, "grad_norm": 0.8891826914613421, "learning_rate": 5.332365634870571e-06, "loss": 0.0623, "step": 10656 }, { "epoch": 0.49521375464684014, "grad_norm": 1.755888292915942, "learning_rate": 5.33161478217552e-06, "loss": 0.0832, "step": 10657 }, { "epoch": 0.4952602230483271, "grad_norm": 0.6714813641341519, "learning_rate": 5.330863921969047e-06, "loss": 0.0737, "step": 10658 }, { "epoch": 0.49530669144981415, "grad_norm": 0.9107225959128147, "learning_rate": 5.3301130542681555e-06, "loss": 0.0561, "step": 10659 }, { "epoch": 0.49535315985130113, "grad_norm": 0.7942812330001672, "learning_rate": 5.329362179089858e-06, "loss": 0.0515, "step": 10660 }, { "epoch": 0.4953996282527881, "grad_norm": 0.7073014334694446, "learning_rate": 5.328611296451159e-06, "loss": 0.0675, "step": 10661 }, { "epoch": 0.4954460966542751, "grad_norm": 0.4913367298019677, "learning_rate": 5.327860406369068e-06, "loss": 0.0616, "step": 10662 }, { "epoch": 0.49549256505576206, "grad_norm": 0.5608883356294238, "learning_rate": 5.327109508860596e-06, "loss": 0.0603, "step": 10663 }, { "epoch": 0.4955390334572491, "grad_norm": 0.8208919066096593, "learning_rate": 5.326358603942747e-06, "loss": 0.06, "step": 10664 }, { "epoch": 0.49558550185873607, "grad_norm": 0.751209758350467, "learning_rate": 5.325607691632534e-06, "loss": 0.053, "step": 10665 }, { "epoch": 0.49563197026022304, "grad_norm": 0.6045155146588553, "learning_rate": 5.324856771946964e-06, "loss": 0.0548, "step": 10666 }, { "epoch": 0.49567843866171, "grad_norm": 0.5549206198247585, "learning_rate": 5.324105844903044e-06, "loss": 0.0459, "step": 10667 }, { "epoch": 0.49572490706319705, "grad_norm": 0.43949347402106914, "learning_rate": 5.323354910517788e-06, "loss": 0.0404, "step": 10668 }, { "epoch": 0.495771375464684, "grad_norm": 0.4593432519229554, "learning_rate": 5.322603968808201e-06, "loss": 0.0559, "step": 10669 }, { "epoch": 0.495817843866171, "grad_norm": 0.33699343317945635, "learning_rate": 5.321853019791295e-06, "loss": 0.0501, "step": 10670 }, { "epoch": 0.495864312267658, "grad_norm": 0.5545119516509354, "learning_rate": 5.3211020634840795e-06, "loss": 0.0351, "step": 10671 }, { "epoch": 0.49591078066914496, "grad_norm": 0.8079915976859617, "learning_rate": 5.320351099903565e-06, "loss": 0.0627, "step": 10672 }, { "epoch": 0.495957249070632, "grad_norm": 0.8540073018149816, "learning_rate": 5.319600129066761e-06, "loss": 0.0695, "step": 10673 }, { "epoch": 0.49600371747211897, "grad_norm": 0.4515315702449578, "learning_rate": 5.318849150990676e-06, "loss": 0.0374, "step": 10674 }, { "epoch": 0.49605018587360594, "grad_norm": 0.8840301533773376, "learning_rate": 5.318098165692325e-06, "loss": 0.068, "step": 10675 }, { "epoch": 0.4960966542750929, "grad_norm": 0.5383937442923062, "learning_rate": 5.317347173188714e-06, "loss": 0.0594, "step": 10676 }, { "epoch": 0.49614312267657995, "grad_norm": 0.5656669542509191, "learning_rate": 5.316596173496856e-06, "loss": 0.0402, "step": 10677 }, { "epoch": 0.4961895910780669, "grad_norm": 0.6270473298008749, "learning_rate": 5.315845166633761e-06, "loss": 0.0558, "step": 10678 }, { "epoch": 0.4962360594795539, "grad_norm": 0.5059145060857552, "learning_rate": 5.315094152616439e-06, "loss": 0.0457, "step": 10679 }, { "epoch": 0.4962825278810409, "grad_norm": 0.8542463922739898, "learning_rate": 5.314343131461906e-06, "loss": 0.0579, "step": 10680 }, { "epoch": 0.49632899628252786, "grad_norm": 0.6454714278798055, "learning_rate": 5.313592103187168e-06, "loss": 0.0519, "step": 10681 }, { "epoch": 0.4963754646840149, "grad_norm": 1.243925727329363, "learning_rate": 5.31284106780924e-06, "loss": 0.0563, "step": 10682 }, { "epoch": 0.49642193308550187, "grad_norm": 0.44016994122562314, "learning_rate": 5.312090025345131e-06, "loss": 0.0468, "step": 10683 }, { "epoch": 0.49646840148698884, "grad_norm": 0.4045565492238631, "learning_rate": 5.311338975811856e-06, "loss": 0.0514, "step": 10684 }, { "epoch": 0.4965148698884758, "grad_norm": 1.3919783518772864, "learning_rate": 5.310587919226425e-06, "loss": 0.0457, "step": 10685 }, { "epoch": 0.49656133828996285, "grad_norm": 0.49491395095739427, "learning_rate": 5.3098368556058496e-06, "loss": 0.0578, "step": 10686 }, { "epoch": 0.4966078066914498, "grad_norm": 1.792165128265806, "learning_rate": 5.309085784967146e-06, "loss": 0.0687, "step": 10687 }, { "epoch": 0.4966542750929368, "grad_norm": 0.9231310260257888, "learning_rate": 5.308334707327323e-06, "loss": 0.0756, "step": 10688 }, { "epoch": 0.4967007434944238, "grad_norm": 0.800382092080097, "learning_rate": 5.307583622703394e-06, "loss": 0.0584, "step": 10689 }, { "epoch": 0.49674721189591076, "grad_norm": 0.5875422218097003, "learning_rate": 5.306832531112372e-06, "loss": 0.0637, "step": 10690 }, { "epoch": 0.4967936802973978, "grad_norm": 0.618585572783622, "learning_rate": 5.30608143257127e-06, "loss": 0.0492, "step": 10691 }, { "epoch": 0.49684014869888476, "grad_norm": 0.5716723945487246, "learning_rate": 5.305330327097102e-06, "loss": 0.0397, "step": 10692 }, { "epoch": 0.49688661710037174, "grad_norm": 0.6475069883869057, "learning_rate": 5.30457921470688e-06, "loss": 0.0655, "step": 10693 }, { "epoch": 0.4969330855018587, "grad_norm": 0.8349203848998816, "learning_rate": 5.303828095417619e-06, "loss": 0.0997, "step": 10694 }, { "epoch": 0.49697955390334575, "grad_norm": 1.0792697678100056, "learning_rate": 5.3030769692463315e-06, "loss": 0.0672, "step": 10695 }, { "epoch": 0.4970260223048327, "grad_norm": 0.6338304267335167, "learning_rate": 5.302325836210031e-06, "loss": 0.0573, "step": 10696 }, { "epoch": 0.4970724907063197, "grad_norm": 0.7930038825551072, "learning_rate": 5.301574696325734e-06, "loss": 0.0549, "step": 10697 }, { "epoch": 0.4971189591078067, "grad_norm": 0.7954998377729464, "learning_rate": 5.300823549610451e-06, "loss": 0.0533, "step": 10698 }, { "epoch": 0.49716542750929366, "grad_norm": 0.6172100678583834, "learning_rate": 5.3000723960812e-06, "loss": 0.0626, "step": 10699 }, { "epoch": 0.4972118959107807, "grad_norm": 0.5468477573704561, "learning_rate": 5.29932123575499e-06, "loss": 0.0367, "step": 10700 }, { "epoch": 0.49725836431226766, "grad_norm": 0.6559035842227185, "learning_rate": 5.2985700686488415e-06, "loss": 0.0352, "step": 10701 }, { "epoch": 0.49730483271375464, "grad_norm": 1.007203176276401, "learning_rate": 5.297818894779767e-06, "loss": 0.0906, "step": 10702 }, { "epoch": 0.4973513011152416, "grad_norm": 0.7558986035421358, "learning_rate": 5.297067714164779e-06, "loss": 0.0413, "step": 10703 }, { "epoch": 0.49739776951672865, "grad_norm": 0.7770125827328271, "learning_rate": 5.296316526820897e-06, "loss": 0.0345, "step": 10704 }, { "epoch": 0.4974442379182156, "grad_norm": 1.082846447049345, "learning_rate": 5.295565332765132e-06, "loss": 0.0571, "step": 10705 }, { "epoch": 0.4974907063197026, "grad_norm": 0.6031537008383551, "learning_rate": 5.294814132014503e-06, "loss": 0.0542, "step": 10706 }, { "epoch": 0.4975371747211896, "grad_norm": 0.6709784578279689, "learning_rate": 5.294062924586022e-06, "loss": 0.0562, "step": 10707 }, { "epoch": 0.49758364312267656, "grad_norm": 1.3002507266126562, "learning_rate": 5.293311710496707e-06, "loss": 0.0831, "step": 10708 }, { "epoch": 0.4976301115241636, "grad_norm": 0.6157534672220126, "learning_rate": 5.292560489763573e-06, "loss": 0.0471, "step": 10709 }, { "epoch": 0.49767657992565056, "grad_norm": 0.5910779681615781, "learning_rate": 5.291809262403635e-06, "loss": 0.0464, "step": 10710 }, { "epoch": 0.49772304832713754, "grad_norm": 0.7812877588862942, "learning_rate": 5.291058028433912e-06, "loss": 0.0774, "step": 10711 }, { "epoch": 0.4977695167286245, "grad_norm": 0.6089058589542017, "learning_rate": 5.290306787871417e-06, "loss": 0.0565, "step": 10712 }, { "epoch": 0.49781598513011155, "grad_norm": 0.7868972101356886, "learning_rate": 5.2895555407331665e-06, "loss": 0.0708, "step": 10713 }, { "epoch": 0.4978624535315985, "grad_norm": 0.7474053850455368, "learning_rate": 5.288804287036181e-06, "loss": 0.0441, "step": 10714 }, { "epoch": 0.4979089219330855, "grad_norm": 0.488680839595133, "learning_rate": 5.288053026797472e-06, "loss": 0.0436, "step": 10715 }, { "epoch": 0.4979553903345725, "grad_norm": 1.040539572993403, "learning_rate": 5.2873017600340615e-06, "loss": 0.0775, "step": 10716 }, { "epoch": 0.49800185873605946, "grad_norm": 0.8399950039682013, "learning_rate": 5.286550486762961e-06, "loss": 0.0765, "step": 10717 }, { "epoch": 0.4980483271375465, "grad_norm": 0.666953410464684, "learning_rate": 5.285799207001192e-06, "loss": 0.0545, "step": 10718 }, { "epoch": 0.49809479553903346, "grad_norm": 0.5073293864368741, "learning_rate": 5.28504792076577e-06, "loss": 0.0545, "step": 10719 }, { "epoch": 0.49814126394052044, "grad_norm": 0.7002503586798301, "learning_rate": 5.284296628073711e-06, "loss": 0.0705, "step": 10720 }, { "epoch": 0.4981877323420074, "grad_norm": 0.8201462660780456, "learning_rate": 5.283545328942036e-06, "loss": 0.0664, "step": 10721 }, { "epoch": 0.49823420074349445, "grad_norm": 0.735724973189602, "learning_rate": 5.28279402338776e-06, "loss": 0.0703, "step": 10722 }, { "epoch": 0.4982806691449814, "grad_norm": 0.4428562115073499, "learning_rate": 5.2820427114279024e-06, "loss": 0.0635, "step": 10723 }, { "epoch": 0.4983271375464684, "grad_norm": 1.8454138134874118, "learning_rate": 5.28129139307948e-06, "loss": 0.0763, "step": 10724 }, { "epoch": 0.4983736059479554, "grad_norm": 0.6446332487924218, "learning_rate": 5.280540068359511e-06, "loss": 0.0574, "step": 10725 }, { "epoch": 0.49842007434944235, "grad_norm": 0.7002834188736222, "learning_rate": 5.2797887372850155e-06, "loss": 0.0559, "step": 10726 }, { "epoch": 0.4984665427509294, "grad_norm": 0.5988190884039377, "learning_rate": 5.279037399873009e-06, "loss": 0.049, "step": 10727 }, { "epoch": 0.49851301115241636, "grad_norm": 1.174654448906286, "learning_rate": 5.2782860561405145e-06, "loss": 0.0746, "step": 10728 }, { "epoch": 0.49855947955390334, "grad_norm": 0.6950292113419738, "learning_rate": 5.277534706104544e-06, "loss": 0.0574, "step": 10729 }, { "epoch": 0.4986059479553903, "grad_norm": 0.7421663877808921, "learning_rate": 5.2767833497821245e-06, "loss": 0.0646, "step": 10730 }, { "epoch": 0.49865241635687735, "grad_norm": 0.5632520216643447, "learning_rate": 5.276031987190269e-06, "loss": 0.0336, "step": 10731 }, { "epoch": 0.4986988847583643, "grad_norm": 0.42976345633503693, "learning_rate": 5.275280618345997e-06, "loss": 0.03, "step": 10732 }, { "epoch": 0.4987453531598513, "grad_norm": 0.9144414990407691, "learning_rate": 5.274529243266331e-06, "loss": 0.0734, "step": 10733 }, { "epoch": 0.4987918215613383, "grad_norm": 1.6523557567038278, "learning_rate": 5.273777861968288e-06, "loss": 0.0727, "step": 10734 }, { "epoch": 0.49883828996282525, "grad_norm": 0.7037019770834082, "learning_rate": 5.273026474468889e-06, "loss": 0.0502, "step": 10735 }, { "epoch": 0.4988847583643123, "grad_norm": 1.0029967508120892, "learning_rate": 5.272275080785151e-06, "loss": 0.0825, "step": 10736 }, { "epoch": 0.49893122676579926, "grad_norm": 0.7707589943309778, "learning_rate": 5.2715236809340975e-06, "loss": 0.0591, "step": 10737 }, { "epoch": 0.49897769516728624, "grad_norm": 0.821447136895567, "learning_rate": 5.270772274932747e-06, "loss": 0.0602, "step": 10738 }, { "epoch": 0.4990241635687732, "grad_norm": 0.5547116032660845, "learning_rate": 5.2700208627981185e-06, "loss": 0.0508, "step": 10739 }, { "epoch": 0.49907063197026025, "grad_norm": 0.7842745348518484, "learning_rate": 5.269269444547236e-06, "loss": 0.0702, "step": 10740 }, { "epoch": 0.4991171003717472, "grad_norm": 0.5456103560064833, "learning_rate": 5.268518020197113e-06, "loss": 0.0308, "step": 10741 }, { "epoch": 0.4991635687732342, "grad_norm": 0.7062862829086322, "learning_rate": 5.267766589764776e-06, "loss": 0.0655, "step": 10742 }, { "epoch": 0.4992100371747212, "grad_norm": 0.7895985508988462, "learning_rate": 5.267015153267246e-06, "loss": 0.0662, "step": 10743 }, { "epoch": 0.49925650557620815, "grad_norm": 1.0751028761645425, "learning_rate": 5.266263710721539e-06, "loss": 0.0701, "step": 10744 }, { "epoch": 0.4993029739776952, "grad_norm": 0.5245794281697969, "learning_rate": 5.26551226214468e-06, "loss": 0.058, "step": 10745 }, { "epoch": 0.49934944237918216, "grad_norm": 0.7666997081601357, "learning_rate": 5.264760807553688e-06, "loss": 0.0781, "step": 10746 }, { "epoch": 0.49939591078066914, "grad_norm": 0.5641565917986056, "learning_rate": 5.264009346965586e-06, "loss": 0.0513, "step": 10747 }, { "epoch": 0.4994423791821561, "grad_norm": 0.9406721165221371, "learning_rate": 5.2632578803973946e-06, "loss": 0.0787, "step": 10748 }, { "epoch": 0.49948884758364315, "grad_norm": 2.366807427371782, "learning_rate": 5.262506407866134e-06, "loss": 0.0738, "step": 10749 }, { "epoch": 0.4995353159851301, "grad_norm": 0.5799483228918381, "learning_rate": 5.261754929388828e-06, "loss": 0.0528, "step": 10750 }, { "epoch": 0.4995817843866171, "grad_norm": 0.7272881223381108, "learning_rate": 5.261003444982495e-06, "loss": 0.0355, "step": 10751 }, { "epoch": 0.4996282527881041, "grad_norm": 0.9657489109331524, "learning_rate": 5.260251954664164e-06, "loss": 0.0587, "step": 10752 }, { "epoch": 0.49967472118959105, "grad_norm": 0.9270545543519005, "learning_rate": 5.259500458450848e-06, "loss": 0.0749, "step": 10753 }, { "epoch": 0.4997211895910781, "grad_norm": 0.7335572862754781, "learning_rate": 5.258748956359575e-06, "loss": 0.0507, "step": 10754 }, { "epoch": 0.49976765799256506, "grad_norm": 0.5584204332761948, "learning_rate": 5.257997448407366e-06, "loss": 0.0507, "step": 10755 }, { "epoch": 0.49981412639405204, "grad_norm": 2.5099071802269926, "learning_rate": 5.257245934611243e-06, "loss": 0.0991, "step": 10756 }, { "epoch": 0.499860594795539, "grad_norm": 0.5536717345195085, "learning_rate": 5.256494414988229e-06, "loss": 0.0434, "step": 10757 }, { "epoch": 0.49990706319702605, "grad_norm": 0.7469961941544518, "learning_rate": 5.2557428895553465e-06, "loss": 0.0664, "step": 10758 }, { "epoch": 0.499953531598513, "grad_norm": 0.7021282012307112, "learning_rate": 5.2549913583296185e-06, "loss": 0.0568, "step": 10759 }, { "epoch": 0.5, "grad_norm": 0.9188589117574825, "learning_rate": 5.254239821328067e-06, "loss": 0.0693, "step": 10760 }, { "epoch": 0.500046468401487, "grad_norm": 0.4626058817910139, "learning_rate": 5.253488278567717e-06, "loss": 0.0491, "step": 10761 }, { "epoch": 0.500092936802974, "grad_norm": 0.5991096551609031, "learning_rate": 5.252736730065591e-06, "loss": 0.051, "step": 10762 }, { "epoch": 0.500139405204461, "grad_norm": 1.1733677249352683, "learning_rate": 5.25198517583871e-06, "loss": 0.06, "step": 10763 }, { "epoch": 0.5001858736059479, "grad_norm": 0.4337138727376302, "learning_rate": 5.251233615904102e-06, "loss": 0.0461, "step": 10764 }, { "epoch": 0.5002323420074349, "grad_norm": 0.6336554160230518, "learning_rate": 5.250482050278786e-06, "loss": 0.0564, "step": 10765 }, { "epoch": 0.500278810408922, "grad_norm": 0.6025942336360817, "learning_rate": 5.24973047897979e-06, "loss": 0.0581, "step": 10766 }, { "epoch": 0.5003252788104089, "grad_norm": 0.6877778793822626, "learning_rate": 5.248978902024135e-06, "loss": 0.0614, "step": 10767 }, { "epoch": 0.5003717472118959, "grad_norm": 0.8806051369514524, "learning_rate": 5.248227319428845e-06, "loss": 0.0586, "step": 10768 }, { "epoch": 0.500418215613383, "grad_norm": 0.5542209706967144, "learning_rate": 5.247475731210946e-06, "loss": 0.0543, "step": 10769 }, { "epoch": 0.5004646840148699, "grad_norm": 0.5204578522250188, "learning_rate": 5.246724137387461e-06, "loss": 0.0309, "step": 10770 }, { "epoch": 0.5005111524163569, "grad_norm": 0.7704692796392254, "learning_rate": 5.245972537975414e-06, "loss": 0.0462, "step": 10771 }, { "epoch": 0.5005576208178438, "grad_norm": 0.9888383225782577, "learning_rate": 5.24522093299183e-06, "loss": 0.0498, "step": 10772 }, { "epoch": 0.5006040892193309, "grad_norm": 0.6826849179988858, "learning_rate": 5.244469322453736e-06, "loss": 0.0384, "step": 10773 }, { "epoch": 0.5006505576208179, "grad_norm": 0.9954013715666389, "learning_rate": 5.243717706378153e-06, "loss": 0.064, "step": 10774 }, { "epoch": 0.5006970260223048, "grad_norm": 0.6127854975943763, "learning_rate": 5.242966084782108e-06, "loss": 0.0464, "step": 10775 }, { "epoch": 0.5007434944237918, "grad_norm": 0.6226052475812148, "learning_rate": 5.242214457682624e-06, "loss": 0.0487, "step": 10776 }, { "epoch": 0.5007899628252788, "grad_norm": 1.3040177196242706, "learning_rate": 5.2414628250967285e-06, "loss": 0.0493, "step": 10777 }, { "epoch": 0.5008364312267658, "grad_norm": 0.5466947821511383, "learning_rate": 5.240711187041446e-06, "loss": 0.0477, "step": 10778 }, { "epoch": 0.5008828996282528, "grad_norm": 0.6640325210470452, "learning_rate": 5.239959543533801e-06, "loss": 0.0577, "step": 10779 }, { "epoch": 0.5009293680297398, "grad_norm": 1.3082284015848975, "learning_rate": 5.23920789459082e-06, "loss": 0.0796, "step": 10780 }, { "epoch": 0.5009758364312268, "grad_norm": 0.6808093334433313, "learning_rate": 5.23845624022953e-06, "loss": 0.0719, "step": 10781 }, { "epoch": 0.5010223048327137, "grad_norm": 0.5337065971085942, "learning_rate": 5.237704580466953e-06, "loss": 0.061, "step": 10782 }, { "epoch": 0.5010687732342007, "grad_norm": 0.8016778358693454, "learning_rate": 5.2369529153201196e-06, "loss": 0.0513, "step": 10783 }, { "epoch": 0.5011152416356878, "grad_norm": 0.934371712918771, "learning_rate": 5.236201244806052e-06, "loss": 0.0646, "step": 10784 }, { "epoch": 0.5011617100371747, "grad_norm": 0.9992158836509115, "learning_rate": 5.235449568941776e-06, "loss": 0.0431, "step": 10785 }, { "epoch": 0.5012081784386617, "grad_norm": 1.0175053779278493, "learning_rate": 5.234697887744321e-06, "loss": 0.0583, "step": 10786 }, { "epoch": 0.5012546468401488, "grad_norm": 1.2486738546464566, "learning_rate": 5.233946201230712e-06, "loss": 0.0931, "step": 10787 }, { "epoch": 0.5013011152416357, "grad_norm": 0.8960073183662196, "learning_rate": 5.233194509417975e-06, "loss": 0.058, "step": 10788 }, { "epoch": 0.5013475836431227, "grad_norm": 0.6732226218870055, "learning_rate": 5.232442812323136e-06, "loss": 0.0589, "step": 10789 }, { "epoch": 0.5013940520446096, "grad_norm": 0.4087143662421235, "learning_rate": 5.231691109963224e-06, "loss": 0.0302, "step": 10790 }, { "epoch": 0.5014405204460967, "grad_norm": 0.582221031292301, "learning_rate": 5.230939402355265e-06, "loss": 0.0393, "step": 10791 }, { "epoch": 0.5014869888475837, "grad_norm": 0.34079703131225453, "learning_rate": 5.230187689516283e-06, "loss": 0.0485, "step": 10792 }, { "epoch": 0.5015334572490706, "grad_norm": 0.6341215679234082, "learning_rate": 5.2294359714633105e-06, "loss": 0.0629, "step": 10793 }, { "epoch": 0.5015799256505576, "grad_norm": 0.4590361115495229, "learning_rate": 5.22868424821337e-06, "loss": 0.05, "step": 10794 }, { "epoch": 0.5016263940520446, "grad_norm": 0.5984742567823497, "learning_rate": 5.2279325197834905e-06, "loss": 0.0546, "step": 10795 }, { "epoch": 0.5016728624535316, "grad_norm": 0.5945390696415843, "learning_rate": 5.227180786190701e-06, "loss": 0.042, "step": 10796 }, { "epoch": 0.5017193308550186, "grad_norm": 0.6140465715487995, "learning_rate": 5.226429047452026e-06, "loss": 0.0607, "step": 10797 }, { "epoch": 0.5017657992565056, "grad_norm": 1.0854243805197146, "learning_rate": 5.225677303584495e-06, "loss": 0.0636, "step": 10798 }, { "epoch": 0.5018122676579926, "grad_norm": 0.8118491125666403, "learning_rate": 5.224925554605137e-06, "loss": 0.0713, "step": 10799 }, { "epoch": 0.5018587360594795, "grad_norm": 0.8006610871838785, "learning_rate": 5.2241738005309776e-06, "loss": 0.0488, "step": 10800 }, { "epoch": 0.5019052044609665, "grad_norm": 0.6693149957490103, "learning_rate": 5.2234220413790435e-06, "loss": 0.0542, "step": 10801 }, { "epoch": 0.5019516728624536, "grad_norm": 0.9644813108433891, "learning_rate": 5.222670277166369e-06, "loss": 0.0543, "step": 10802 }, { "epoch": 0.5019981412639405, "grad_norm": 0.959980215239787, "learning_rate": 5.221918507909975e-06, "loss": 0.0879, "step": 10803 }, { "epoch": 0.5020446096654275, "grad_norm": 1.3307416010298858, "learning_rate": 5.221166733626895e-06, "loss": 0.0486, "step": 10804 }, { "epoch": 0.5020910780669146, "grad_norm": 1.2535824501505897, "learning_rate": 5.220414954334156e-06, "loss": 0.0551, "step": 10805 }, { "epoch": 0.5021375464684015, "grad_norm": 0.7507300774036, "learning_rate": 5.219663170048784e-06, "loss": 0.0638, "step": 10806 }, { "epoch": 0.5021840148698885, "grad_norm": 0.5251756123984394, "learning_rate": 5.218911380787812e-06, "loss": 0.0497, "step": 10807 }, { "epoch": 0.5022304832713754, "grad_norm": 0.7665507082585297, "learning_rate": 5.218159586568268e-06, "loss": 0.0564, "step": 10808 }, { "epoch": 0.5022769516728625, "grad_norm": 0.5352132242995032, "learning_rate": 5.217407787407177e-06, "loss": 0.0541, "step": 10809 }, { "epoch": 0.5023234200743495, "grad_norm": 0.5745778898018586, "learning_rate": 5.216655983321572e-06, "loss": 0.0503, "step": 10810 }, { "epoch": 0.5023698884758364, "grad_norm": 0.8887300897842552, "learning_rate": 5.215904174328481e-06, "loss": 0.0696, "step": 10811 }, { "epoch": 0.5024163568773234, "grad_norm": 0.6453165300858819, "learning_rate": 5.215152360444934e-06, "loss": 0.0658, "step": 10812 }, { "epoch": 0.5024628252788104, "grad_norm": 0.45669949185358155, "learning_rate": 5.214400541687957e-06, "loss": 0.0542, "step": 10813 }, { "epoch": 0.5025092936802974, "grad_norm": 1.1401498205818041, "learning_rate": 5.213648718074584e-06, "loss": 0.0704, "step": 10814 }, { "epoch": 0.5025557620817844, "grad_norm": 0.6633986001738224, "learning_rate": 5.212896889621844e-06, "loss": 0.0655, "step": 10815 }, { "epoch": 0.5026022304832714, "grad_norm": 0.6378804992782615, "learning_rate": 5.212145056346762e-06, "loss": 0.0567, "step": 10816 }, { "epoch": 0.5026486988847584, "grad_norm": 1.0160080152559559, "learning_rate": 5.211393218266375e-06, "loss": 0.062, "step": 10817 }, { "epoch": 0.5026951672862453, "grad_norm": 0.7636184274374868, "learning_rate": 5.210641375397706e-06, "loss": 0.071, "step": 10818 }, { "epoch": 0.5027416356877323, "grad_norm": 0.7313170661952284, "learning_rate": 5.20988952775779e-06, "loss": 0.072, "step": 10819 }, { "epoch": 0.5027881040892194, "grad_norm": 0.6751455993940422, "learning_rate": 5.209137675363654e-06, "loss": 0.0504, "step": 10820 }, { "epoch": 0.5028345724907063, "grad_norm": 3.5943739736847546, "learning_rate": 5.20838581823233e-06, "loss": 0.0626, "step": 10821 }, { "epoch": 0.5028810408921933, "grad_norm": 1.0608121851090275, "learning_rate": 5.207633956380849e-06, "loss": 0.056, "step": 10822 }, { "epoch": 0.5029275092936804, "grad_norm": 0.6667957743158865, "learning_rate": 5.206882089826238e-06, "loss": 0.0461, "step": 10823 }, { "epoch": 0.5029739776951673, "grad_norm": 0.8571707165696567, "learning_rate": 5.206130218585532e-06, "loss": 0.0821, "step": 10824 }, { "epoch": 0.5030204460966543, "grad_norm": 0.4215379539548624, "learning_rate": 5.205378342675758e-06, "loss": 0.0405, "step": 10825 }, { "epoch": 0.5030669144981412, "grad_norm": 0.6312399887364626, "learning_rate": 5.204626462113949e-06, "loss": 0.047, "step": 10826 }, { "epoch": 0.5031133828996283, "grad_norm": 0.5619013170559126, "learning_rate": 5.203874576917137e-06, "loss": 0.0535, "step": 10827 }, { "epoch": 0.5031598513011153, "grad_norm": 0.9742206009678329, "learning_rate": 5.203122687102349e-06, "loss": 0.0619, "step": 10828 }, { "epoch": 0.5032063197026022, "grad_norm": 0.8407265147924523, "learning_rate": 5.2023707926866204e-06, "loss": 0.0561, "step": 10829 }, { "epoch": 0.5032527881040892, "grad_norm": 1.6810643237689014, "learning_rate": 5.2016188936869786e-06, "loss": 0.1013, "step": 10830 }, { "epoch": 0.5032992565055762, "grad_norm": 1.0086197159528905, "learning_rate": 5.200866990120458e-06, "loss": 0.0909, "step": 10831 }, { "epoch": 0.5033457249070632, "grad_norm": 0.5374457339164797, "learning_rate": 5.200115082004088e-06, "loss": 0.0658, "step": 10832 }, { "epoch": 0.5033921933085502, "grad_norm": 1.0075045248994214, "learning_rate": 5.1993631693548994e-06, "loss": 0.072, "step": 10833 }, { "epoch": 0.5034386617100371, "grad_norm": 0.630572787592425, "learning_rate": 5.198611252189928e-06, "loss": 0.046, "step": 10834 }, { "epoch": 0.5034851301115242, "grad_norm": 0.5537727318785506, "learning_rate": 5.1978593305262e-06, "loss": 0.0539, "step": 10835 }, { "epoch": 0.5035315985130111, "grad_norm": 0.4614939100856872, "learning_rate": 5.197107404380751e-06, "loss": 0.0443, "step": 10836 }, { "epoch": 0.5035780669144981, "grad_norm": 0.8675219996797309, "learning_rate": 5.1963554737706124e-06, "loss": 0.0639, "step": 10837 }, { "epoch": 0.5036245353159852, "grad_norm": 0.5939713975361972, "learning_rate": 5.195603538712814e-06, "loss": 0.0504, "step": 10838 }, { "epoch": 0.5036710037174721, "grad_norm": 0.8029887004151671, "learning_rate": 5.194851599224392e-06, "loss": 0.0553, "step": 10839 }, { "epoch": 0.5037174721189591, "grad_norm": 0.8846908988929512, "learning_rate": 5.194099655322376e-06, "loss": 0.0812, "step": 10840 }, { "epoch": 0.5037639405204462, "grad_norm": 0.89195492541307, "learning_rate": 5.193347707023797e-06, "loss": 0.0622, "step": 10841 }, { "epoch": 0.5038104089219331, "grad_norm": 0.7048086602220437, "learning_rate": 5.192595754345688e-06, "loss": 0.0502, "step": 10842 }, { "epoch": 0.5038568773234201, "grad_norm": 0.642714970099904, "learning_rate": 5.191843797305084e-06, "loss": 0.0499, "step": 10843 }, { "epoch": 0.503903345724907, "grad_norm": 0.5767297154984533, "learning_rate": 5.191091835919017e-06, "loss": 0.0445, "step": 10844 }, { "epoch": 0.5039498141263941, "grad_norm": 1.5191434622685573, "learning_rate": 5.190339870204516e-06, "loss": 0.0423, "step": 10845 }, { "epoch": 0.5039962825278811, "grad_norm": 0.5485079438910978, "learning_rate": 5.18958790017862e-06, "loss": 0.0356, "step": 10846 }, { "epoch": 0.504042750929368, "grad_norm": 0.702855824919628, "learning_rate": 5.188835925858357e-06, "loss": 0.0539, "step": 10847 }, { "epoch": 0.504089219330855, "grad_norm": 0.446897900118479, "learning_rate": 5.1880839472607615e-06, "loss": 0.0515, "step": 10848 }, { "epoch": 0.504135687732342, "grad_norm": 1.7805140179835197, "learning_rate": 5.1873319644028664e-06, "loss": 0.0805, "step": 10849 }, { "epoch": 0.504182156133829, "grad_norm": 1.1579662475281791, "learning_rate": 5.186579977301705e-06, "loss": 0.0661, "step": 10850 }, { "epoch": 0.504228624535316, "grad_norm": 0.8104857321441084, "learning_rate": 5.185827985974311e-06, "loss": 0.0666, "step": 10851 }, { "epoch": 0.504275092936803, "grad_norm": 0.5395309315531347, "learning_rate": 5.185075990437718e-06, "loss": 0.0661, "step": 10852 }, { "epoch": 0.50432156133829, "grad_norm": 0.6878370641520669, "learning_rate": 5.184323990708959e-06, "loss": 0.041, "step": 10853 }, { "epoch": 0.5043680297397769, "grad_norm": 0.48194272580902064, "learning_rate": 5.183571986805066e-06, "loss": 0.0544, "step": 10854 }, { "epoch": 0.5044144981412639, "grad_norm": 1.0552570754875599, "learning_rate": 5.182819978743076e-06, "loss": 0.0505, "step": 10855 }, { "epoch": 0.504460966542751, "grad_norm": 2.1329099969974674, "learning_rate": 5.182067966540021e-06, "loss": 0.0584, "step": 10856 }, { "epoch": 0.5045074349442379, "grad_norm": 0.7852215554678078, "learning_rate": 5.1813159502129325e-06, "loss": 0.0662, "step": 10857 }, { "epoch": 0.5045539033457249, "grad_norm": 0.6467012343980596, "learning_rate": 5.180563929778851e-06, "loss": 0.0751, "step": 10858 }, { "epoch": 0.504600371747212, "grad_norm": 0.6061083238392426, "learning_rate": 5.179811905254804e-06, "loss": 0.0525, "step": 10859 }, { "epoch": 0.5046468401486989, "grad_norm": 0.808282290993344, "learning_rate": 5.179059876657827e-06, "loss": 0.0354, "step": 10860 }, { "epoch": 0.5046933085501859, "grad_norm": 0.4577240579451857, "learning_rate": 5.178307844004959e-06, "loss": 0.0261, "step": 10861 }, { "epoch": 0.5047397769516728, "grad_norm": 0.577601224196434, "learning_rate": 5.177555807313227e-06, "loss": 0.0476, "step": 10862 }, { "epoch": 0.5047862453531599, "grad_norm": 0.5813324946883989, "learning_rate": 5.176803766599671e-06, "loss": 0.0579, "step": 10863 }, { "epoch": 0.5048327137546469, "grad_norm": 0.973579317701439, "learning_rate": 5.176051721881323e-06, "loss": 0.045, "step": 10864 }, { "epoch": 0.5048791821561338, "grad_norm": 0.6912222295659878, "learning_rate": 5.175299673175218e-06, "loss": 0.0585, "step": 10865 }, { "epoch": 0.5049256505576208, "grad_norm": 0.7915500196348998, "learning_rate": 5.174547620498391e-06, "loss": 0.0452, "step": 10866 }, { "epoch": 0.5049721189591078, "grad_norm": 0.5036184021192748, "learning_rate": 5.173795563867877e-06, "loss": 0.0426, "step": 10867 }, { "epoch": 0.5050185873605948, "grad_norm": 0.5549362746842187, "learning_rate": 5.173043503300711e-06, "loss": 0.0444, "step": 10868 }, { "epoch": 0.5050650557620818, "grad_norm": 0.7082940606414079, "learning_rate": 5.172291438813927e-06, "loss": 0.0634, "step": 10869 }, { "epoch": 0.5051115241635687, "grad_norm": 0.5018776747379319, "learning_rate": 5.171539370424561e-06, "loss": 0.0436, "step": 10870 }, { "epoch": 0.5051579925650558, "grad_norm": 0.6201376935293639, "learning_rate": 5.170787298149646e-06, "loss": 0.0578, "step": 10871 }, { "epoch": 0.5052044609665427, "grad_norm": 0.6966823229213158, "learning_rate": 5.17003522200622e-06, "loss": 0.0762, "step": 10872 }, { "epoch": 0.5052509293680297, "grad_norm": 1.0664122950491077, "learning_rate": 5.169283142011319e-06, "loss": 0.0628, "step": 10873 }, { "epoch": 0.5052973977695168, "grad_norm": 0.8345875780980305, "learning_rate": 5.168531058181974e-06, "loss": 0.0673, "step": 10874 }, { "epoch": 0.5053438661710037, "grad_norm": 0.5392790165591972, "learning_rate": 5.167778970535225e-06, "loss": 0.043, "step": 10875 }, { "epoch": 0.5053903345724907, "grad_norm": 0.2731840681449445, "learning_rate": 5.1670268790881054e-06, "loss": 0.0183, "step": 10876 }, { "epoch": 0.5054368029739778, "grad_norm": 0.7159160720565638, "learning_rate": 5.16627478385765e-06, "loss": 0.0546, "step": 10877 }, { "epoch": 0.5054832713754647, "grad_norm": 0.7492187320479222, "learning_rate": 5.165522684860898e-06, "loss": 0.0583, "step": 10878 }, { "epoch": 0.5055297397769517, "grad_norm": 1.0770094656713278, "learning_rate": 5.164770582114881e-06, "loss": 0.044, "step": 10879 }, { "epoch": 0.5055762081784386, "grad_norm": 0.9919746241912156, "learning_rate": 5.164018475636637e-06, "loss": 0.0608, "step": 10880 }, { "epoch": 0.5056226765799257, "grad_norm": 0.8008297053133541, "learning_rate": 5.163266365443202e-06, "loss": 0.0724, "step": 10881 }, { "epoch": 0.5056691449814127, "grad_norm": 0.8709415381976995, "learning_rate": 5.162514251551613e-06, "loss": 0.0699, "step": 10882 }, { "epoch": 0.5057156133828996, "grad_norm": 0.7857219442551927, "learning_rate": 5.161762133978905e-06, "loss": 0.0722, "step": 10883 }, { "epoch": 0.5057620817843866, "grad_norm": 0.6962986894455453, "learning_rate": 5.161010012742112e-06, "loss": 0.0428, "step": 10884 }, { "epoch": 0.5058085501858736, "grad_norm": 0.7680373793455444, "learning_rate": 5.160257887858278e-06, "loss": 0.0669, "step": 10885 }, { "epoch": 0.5058550185873606, "grad_norm": 0.6132769322036518, "learning_rate": 5.159505759344429e-06, "loss": 0.0366, "step": 10886 }, { "epoch": 0.5059014869888476, "grad_norm": 0.44202089315904824, "learning_rate": 5.158753627217609e-06, "loss": 0.0423, "step": 10887 }, { "epoch": 0.5059479553903345, "grad_norm": 0.4960710065682573, "learning_rate": 5.158001491494852e-06, "loss": 0.0382, "step": 10888 }, { "epoch": 0.5059944237918216, "grad_norm": 0.7439547114596491, "learning_rate": 5.1572493521931945e-06, "loss": 0.0625, "step": 10889 }, { "epoch": 0.5060408921933085, "grad_norm": 0.6209020137291112, "learning_rate": 5.156497209329674e-06, "loss": 0.0447, "step": 10890 }, { "epoch": 0.5060873605947955, "grad_norm": 1.0754397219125633, "learning_rate": 5.155745062921327e-06, "loss": 0.0697, "step": 10891 }, { "epoch": 0.5061338289962826, "grad_norm": 0.6387294196228422, "learning_rate": 5.15499291298519e-06, "loss": 0.0581, "step": 10892 }, { "epoch": 0.5061802973977695, "grad_norm": 1.5531214774318798, "learning_rate": 5.154240759538301e-06, "loss": 0.0866, "step": 10893 }, { "epoch": 0.5062267657992565, "grad_norm": 1.007473175811014, "learning_rate": 5.153488602597696e-06, "loss": 0.0685, "step": 10894 }, { "epoch": 0.5062732342007435, "grad_norm": 0.31033235052056757, "learning_rate": 5.152736442180413e-06, "loss": 0.0452, "step": 10895 }, { "epoch": 0.5063197026022305, "grad_norm": 0.8280348675360885, "learning_rate": 5.151984278303489e-06, "loss": 0.0571, "step": 10896 }, { "epoch": 0.5063661710037175, "grad_norm": 0.7385972908747798, "learning_rate": 5.1512321109839615e-06, "loss": 0.0764, "step": 10897 }, { "epoch": 0.5064126394052044, "grad_norm": 0.8731935196382491, "learning_rate": 5.150479940238865e-06, "loss": 0.0754, "step": 10898 }, { "epoch": 0.5064591078066915, "grad_norm": 0.4924526993860414, "learning_rate": 5.149727766085242e-06, "loss": 0.0317, "step": 10899 }, { "epoch": 0.5065055762081785, "grad_norm": 0.7150519318984693, "learning_rate": 5.148975588540127e-06, "loss": 0.0564, "step": 10900 }, { "epoch": 0.5065520446096654, "grad_norm": 30.808708289754822, "learning_rate": 5.1482234076205575e-06, "loss": 0.0728, "step": 10901 }, { "epoch": 0.5065985130111524, "grad_norm": 0.5771779028303919, "learning_rate": 5.147471223343573e-06, "loss": 0.0452, "step": 10902 }, { "epoch": 0.5066449814126394, "grad_norm": 0.8093117119944522, "learning_rate": 5.146719035726209e-06, "loss": 0.077, "step": 10903 }, { "epoch": 0.5066914498141264, "grad_norm": 1.0075350473949012, "learning_rate": 5.145966844785505e-06, "loss": 0.0691, "step": 10904 }, { "epoch": 0.5067379182156134, "grad_norm": 9.302491762251403, "learning_rate": 5.1452146505384985e-06, "loss": 0.071, "step": 10905 }, { "epoch": 0.5067843866171003, "grad_norm": 0.8507662355229768, "learning_rate": 5.144462453002226e-06, "loss": 0.0664, "step": 10906 }, { "epoch": 0.5068308550185874, "grad_norm": 0.7250290020010651, "learning_rate": 5.143710252193727e-06, "loss": 0.0493, "step": 10907 }, { "epoch": 0.5068773234200743, "grad_norm": 2.0371610988285784, "learning_rate": 5.142958048130042e-06, "loss": 0.033, "step": 10908 }, { "epoch": 0.5069237918215613, "grad_norm": 0.945726283354466, "learning_rate": 5.142205840828204e-06, "loss": 0.0612, "step": 10909 }, { "epoch": 0.5069702602230484, "grad_norm": 2.279302617517916, "learning_rate": 5.141453630305254e-06, "loss": 0.0845, "step": 10910 }, { "epoch": 0.5070167286245353, "grad_norm": 0.8228719896321052, "learning_rate": 5.140701416578232e-06, "loss": 0.0463, "step": 10911 }, { "epoch": 0.5070631970260223, "grad_norm": 2.931007902473493, "learning_rate": 5.139949199664175e-06, "loss": 0.0861, "step": 10912 }, { "epoch": 0.5071096654275093, "grad_norm": 2.2915715957173655, "learning_rate": 5.13919697958012e-06, "loss": 0.0431, "step": 10913 }, { "epoch": 0.5071561338289963, "grad_norm": 1.1014964566645138, "learning_rate": 5.1384447563431084e-06, "loss": 0.079, "step": 10914 }, { "epoch": 0.5072026022304833, "grad_norm": 3.5729939201976646, "learning_rate": 5.1376925299701766e-06, "loss": 0.0878, "step": 10915 }, { "epoch": 0.5072490706319702, "grad_norm": 0.4003790231467181, "learning_rate": 5.1369403004783635e-06, "loss": 0.0477, "step": 10916 }, { "epoch": 0.5072955390334573, "grad_norm": 0.5357945265098681, "learning_rate": 5.13618806788471e-06, "loss": 0.036, "step": 10917 }, { "epoch": 0.5073420074349443, "grad_norm": 1.774734360356151, "learning_rate": 5.13543583220625e-06, "loss": 0.0701, "step": 10918 }, { "epoch": 0.5073884758364312, "grad_norm": 0.9766547878978422, "learning_rate": 5.134683593460029e-06, "loss": 0.0528, "step": 10919 }, { "epoch": 0.5074349442379182, "grad_norm": 2.2360235900228496, "learning_rate": 5.133931351663083e-06, "loss": 0.0731, "step": 10920 }, { "epoch": 0.5074814126394052, "grad_norm": 0.4447887104486724, "learning_rate": 5.13317910683245e-06, "loss": 0.0452, "step": 10921 }, { "epoch": 0.5075278810408922, "grad_norm": 0.6831660804564811, "learning_rate": 5.132426858985169e-06, "loss": 0.0561, "step": 10922 }, { "epoch": 0.5075743494423792, "grad_norm": 1.4172658054867981, "learning_rate": 5.131674608138282e-06, "loss": 0.0473, "step": 10923 }, { "epoch": 0.5076208178438661, "grad_norm": 1.0029722725965458, "learning_rate": 5.130922354308826e-06, "loss": 0.0775, "step": 10924 }, { "epoch": 0.5076672862453532, "grad_norm": 1.0426391487609021, "learning_rate": 5.1301700975138405e-06, "loss": 0.0729, "step": 10925 }, { "epoch": 0.5077137546468401, "grad_norm": 0.7425205402881551, "learning_rate": 5.129417837770366e-06, "loss": 0.049, "step": 10926 }, { "epoch": 0.5077602230483271, "grad_norm": 0.7049327678397601, "learning_rate": 5.128665575095441e-06, "loss": 0.0606, "step": 10927 }, { "epoch": 0.5078066914498142, "grad_norm": 1.0311797531119338, "learning_rate": 5.127913309506105e-06, "loss": 0.0471, "step": 10928 }, { "epoch": 0.5078531598513011, "grad_norm": 0.6283300775030155, "learning_rate": 5.127161041019398e-06, "loss": 0.0601, "step": 10929 }, { "epoch": 0.5078996282527881, "grad_norm": 0.9949843223316587, "learning_rate": 5.126408769652359e-06, "loss": 0.045, "step": 10930 }, { "epoch": 0.5079460966542751, "grad_norm": 0.5678688734891887, "learning_rate": 5.125656495422029e-06, "loss": 0.0409, "step": 10931 }, { "epoch": 0.5079925650557621, "grad_norm": 0.9386094549172521, "learning_rate": 5.124904218345448e-06, "loss": 0.0634, "step": 10932 }, { "epoch": 0.5080390334572491, "grad_norm": 1.2309248676671212, "learning_rate": 5.124151938439654e-06, "loss": 0.0524, "step": 10933 }, { "epoch": 0.508085501858736, "grad_norm": 0.9548239674278477, "learning_rate": 5.123399655721687e-06, "loss": 0.0538, "step": 10934 }, { "epoch": 0.508131970260223, "grad_norm": 0.4333362667636692, "learning_rate": 5.12264737020859e-06, "loss": 0.0442, "step": 10935 }, { "epoch": 0.5081784386617101, "grad_norm": 1.1297913389899865, "learning_rate": 5.1218950819174e-06, "loss": 0.0662, "step": 10936 }, { "epoch": 0.508224907063197, "grad_norm": 0.5366508077129489, "learning_rate": 5.1211427908651565e-06, "loss": 0.0501, "step": 10937 }, { "epoch": 0.508271375464684, "grad_norm": 0.821924265387897, "learning_rate": 5.120390497068904e-06, "loss": 0.0554, "step": 10938 }, { "epoch": 0.508317843866171, "grad_norm": 0.4029938958057256, "learning_rate": 5.119638200545678e-06, "loss": 0.0389, "step": 10939 }, { "epoch": 0.508364312267658, "grad_norm": 0.3497401167907268, "learning_rate": 5.118885901312523e-06, "loss": 0.0486, "step": 10940 }, { "epoch": 0.508410780669145, "grad_norm": 0.8121582955775108, "learning_rate": 5.1181335993864755e-06, "loss": 0.0644, "step": 10941 }, { "epoch": 0.508457249070632, "grad_norm": 2.400921708243522, "learning_rate": 5.117381294784577e-06, "loss": 0.0748, "step": 10942 }, { "epoch": 0.508503717472119, "grad_norm": 5.468739250161137, "learning_rate": 5.116628987523871e-06, "loss": 0.0809, "step": 10943 }, { "epoch": 0.5085501858736059, "grad_norm": 0.5962292035070105, "learning_rate": 5.115876677621394e-06, "loss": 0.0498, "step": 10944 }, { "epoch": 0.5085966542750929, "grad_norm": 1.1772801943778137, "learning_rate": 5.115124365094188e-06, "loss": 0.0947, "step": 10945 }, { "epoch": 0.50864312267658, "grad_norm": 0.9673285763688829, "learning_rate": 5.114372049959294e-06, "loss": 0.0835, "step": 10946 }, { "epoch": 0.5086895910780669, "grad_norm": 0.9044721375601567, "learning_rate": 5.1136197322337545e-06, "loss": 0.0508, "step": 10947 }, { "epoch": 0.5087360594795539, "grad_norm": 0.7172073273343795, "learning_rate": 5.112867411934606e-06, "loss": 0.0445, "step": 10948 }, { "epoch": 0.508782527881041, "grad_norm": 0.6905286592857736, "learning_rate": 5.112115089078894e-06, "loss": 0.0734, "step": 10949 }, { "epoch": 0.5088289962825279, "grad_norm": 0.43979591156722064, "learning_rate": 5.111362763683657e-06, "loss": 0.0425, "step": 10950 }, { "epoch": 0.5088754646840149, "grad_norm": 1.1844938822627515, "learning_rate": 5.110610435765935e-06, "loss": 0.1018, "step": 10951 }, { "epoch": 0.5089219330855018, "grad_norm": 0.5985217272912237, "learning_rate": 5.109858105342771e-06, "loss": 0.0424, "step": 10952 }, { "epoch": 0.5089684014869889, "grad_norm": 0.8138983987746893, "learning_rate": 5.109105772431205e-06, "loss": 0.0762, "step": 10953 }, { "epoch": 0.5090148698884759, "grad_norm": 1.4078876939308231, "learning_rate": 5.108353437048278e-06, "loss": 0.0769, "step": 10954 }, { "epoch": 0.5090613382899628, "grad_norm": 1.2351251434823498, "learning_rate": 5.107601099211034e-06, "loss": 0.0552, "step": 10955 }, { "epoch": 0.5091078066914498, "grad_norm": 0.4060589637083851, "learning_rate": 5.106848758936508e-06, "loss": 0.0352, "step": 10956 }, { "epoch": 0.5091542750929368, "grad_norm": 0.7342571384659139, "learning_rate": 5.106096416241748e-06, "loss": 0.0754, "step": 10957 }, { "epoch": 0.5092007434944238, "grad_norm": 0.5137541422238827, "learning_rate": 5.105344071143792e-06, "loss": 0.0304, "step": 10958 }, { "epoch": 0.5092472118959108, "grad_norm": 1.0383025031974766, "learning_rate": 5.10459172365968e-06, "loss": 0.0517, "step": 10959 }, { "epoch": 0.5092936802973977, "grad_norm": 0.6172406647786882, "learning_rate": 5.103839373806457e-06, "loss": 0.0548, "step": 10960 }, { "epoch": 0.5093401486988848, "grad_norm": 0.9280743243082492, "learning_rate": 5.103087021601163e-06, "loss": 0.0788, "step": 10961 }, { "epoch": 0.5093866171003717, "grad_norm": 2.075087484409431, "learning_rate": 5.10233466706084e-06, "loss": 0.073, "step": 10962 }, { "epoch": 0.5094330855018587, "grad_norm": 0.5989305854603371, "learning_rate": 5.101582310202527e-06, "loss": 0.0469, "step": 10963 }, { "epoch": 0.5094795539033458, "grad_norm": 1.0977734562450494, "learning_rate": 5.10082995104327e-06, "loss": 0.0481, "step": 10964 }, { "epoch": 0.5095260223048327, "grad_norm": 1.2819425255329626, "learning_rate": 5.100077589600107e-06, "loss": 0.0665, "step": 10965 }, { "epoch": 0.5095724907063197, "grad_norm": 0.6588504416993025, "learning_rate": 5.099325225890082e-06, "loss": 0.0551, "step": 10966 }, { "epoch": 0.5096189591078067, "grad_norm": 0.8556303469671408, "learning_rate": 5.0985728599302375e-06, "loss": 0.0745, "step": 10967 }, { "epoch": 0.5096654275092937, "grad_norm": 0.694506533807331, "learning_rate": 5.09782049173761e-06, "loss": 0.0573, "step": 10968 }, { "epoch": 0.5097118959107807, "grad_norm": 0.5231784377187845, "learning_rate": 5.097068121329249e-06, "loss": 0.065, "step": 10969 }, { "epoch": 0.5097583643122676, "grad_norm": 0.4606805665906997, "learning_rate": 5.096315748722193e-06, "loss": 0.0364, "step": 10970 }, { "epoch": 0.5098048327137547, "grad_norm": 0.6155169045854321, "learning_rate": 5.095563373933481e-06, "loss": 0.0425, "step": 10971 }, { "epoch": 0.5098513011152417, "grad_norm": 1.509017486893119, "learning_rate": 5.09481099698016e-06, "loss": 0.0367, "step": 10972 }, { "epoch": 0.5098977695167286, "grad_norm": 0.5817385883908365, "learning_rate": 5.0940586178792695e-06, "loss": 0.0513, "step": 10973 }, { "epoch": 0.5099442379182156, "grad_norm": 0.6695971833787359, "learning_rate": 5.093306236647853e-06, "loss": 0.0538, "step": 10974 }, { "epoch": 0.5099907063197026, "grad_norm": 1.1804260076611945, "learning_rate": 5.09255385330295e-06, "loss": 0.067, "step": 10975 }, { "epoch": 0.5100371747211896, "grad_norm": 0.5619131119378816, "learning_rate": 5.091801467861608e-06, "loss": 0.0456, "step": 10976 }, { "epoch": 0.5100836431226766, "grad_norm": 0.8885764664800488, "learning_rate": 5.091049080340865e-06, "loss": 0.0567, "step": 10977 }, { "epoch": 0.5101301115241635, "grad_norm": 0.8592366765936358, "learning_rate": 5.090296690757763e-06, "loss": 0.0686, "step": 10978 }, { "epoch": 0.5101765799256506, "grad_norm": 0.6346884108130383, "learning_rate": 5.089544299129349e-06, "loss": 0.0718, "step": 10979 }, { "epoch": 0.5102230483271375, "grad_norm": 1.0004933007437158, "learning_rate": 5.088791905472659e-06, "loss": 0.0845, "step": 10980 }, { "epoch": 0.5102695167286245, "grad_norm": 0.6838335865858234, "learning_rate": 5.088039509804741e-06, "loss": 0.0663, "step": 10981 }, { "epoch": 0.5103159851301116, "grad_norm": 0.7044791587029587, "learning_rate": 5.087287112142635e-06, "loss": 0.0809, "step": 10982 }, { "epoch": 0.5103624535315985, "grad_norm": 1.7484461965809894, "learning_rate": 5.086534712503382e-06, "loss": 0.0411, "step": 10983 }, { "epoch": 0.5104089219330855, "grad_norm": 0.6777526103112805, "learning_rate": 5.085782310904029e-06, "loss": 0.0614, "step": 10984 }, { "epoch": 0.5104553903345725, "grad_norm": 0.7011810004049212, "learning_rate": 5.085029907361616e-06, "loss": 0.0579, "step": 10985 }, { "epoch": 0.5105018587360595, "grad_norm": 0.43835186833412765, "learning_rate": 5.084277501893186e-06, "loss": 0.0575, "step": 10986 }, { "epoch": 0.5105483271375465, "grad_norm": 0.4991278354293654, "learning_rate": 5.083525094515781e-06, "loss": 0.047, "step": 10987 }, { "epoch": 0.5105947955390334, "grad_norm": 0.5858452830690245, "learning_rate": 5.082772685246446e-06, "loss": 0.0753, "step": 10988 }, { "epoch": 0.5106412639405205, "grad_norm": 0.5570406806303122, "learning_rate": 5.082020274102222e-06, "loss": 0.0516, "step": 10989 }, { "epoch": 0.5106877323420075, "grad_norm": 0.7212645895555744, "learning_rate": 5.081267861100152e-06, "loss": 0.0527, "step": 10990 }, { "epoch": 0.5107342007434944, "grad_norm": 0.6936437214516149, "learning_rate": 5.080515446257281e-06, "loss": 0.0453, "step": 10991 }, { "epoch": 0.5107806691449814, "grad_norm": 2.168527611152356, "learning_rate": 5.0797630295906484e-06, "loss": 0.0554, "step": 10992 }, { "epoch": 0.5108271375464684, "grad_norm": 0.7057623754920952, "learning_rate": 5.079010611117301e-06, "loss": 0.0557, "step": 10993 }, { "epoch": 0.5108736059479554, "grad_norm": 5.800984214831239, "learning_rate": 5.07825819085428e-06, "loss": 0.0945, "step": 10994 }, { "epoch": 0.5109200743494424, "grad_norm": 1.1011652145455948, "learning_rate": 5.077505768818626e-06, "loss": 0.055, "step": 10995 }, { "epoch": 0.5109665427509293, "grad_norm": 1.38675728333987, "learning_rate": 5.076753345027388e-06, "loss": 0.0677, "step": 10996 }, { "epoch": 0.5110130111524164, "grad_norm": 0.34684738624578304, "learning_rate": 5.0760009194976046e-06, "loss": 0.0335, "step": 10997 }, { "epoch": 0.5110594795539033, "grad_norm": 1.5426285953174625, "learning_rate": 5.07524849224632e-06, "loss": 0.0708, "step": 10998 }, { "epoch": 0.5111059479553903, "grad_norm": 0.800031823388021, "learning_rate": 5.074496063290577e-06, "loss": 0.0575, "step": 10999 }, { "epoch": 0.5111524163568774, "grad_norm": 0.457119850152787, "learning_rate": 5.073743632647422e-06, "loss": 0.0355, "step": 11000 }, { "epoch": 0.5111988847583643, "grad_norm": 0.5779472023840402, "learning_rate": 5.072991200333895e-06, "loss": 0.057, "step": 11001 }, { "epoch": 0.5112453531598513, "grad_norm": 0.6075951706120957, "learning_rate": 5.07223876636704e-06, "loss": 0.0335, "step": 11002 }, { "epoch": 0.5112918215613383, "grad_norm": 0.708500384900478, "learning_rate": 5.071486330763903e-06, "loss": 0.0436, "step": 11003 }, { "epoch": 0.5113382899628253, "grad_norm": 0.4801837799325148, "learning_rate": 5.070733893541523e-06, "loss": 0.0395, "step": 11004 }, { "epoch": 0.5113847583643123, "grad_norm": 0.8286242258227834, "learning_rate": 5.069981454716947e-06, "loss": 0.0481, "step": 11005 }, { "epoch": 0.5114312267657992, "grad_norm": 0.9753254364703238, "learning_rate": 5.069229014307217e-06, "loss": 0.0728, "step": 11006 }, { "epoch": 0.5114776951672863, "grad_norm": 1.0601031789304518, "learning_rate": 5.068476572329376e-06, "loss": 0.0831, "step": 11007 }, { "epoch": 0.5115241635687733, "grad_norm": 2.0917061104951955, "learning_rate": 5.067724128800471e-06, "loss": 0.1056, "step": 11008 }, { "epoch": 0.5115706319702602, "grad_norm": 3.5483239913442888, "learning_rate": 5.066971683737542e-06, "loss": 0.0914, "step": 11009 }, { "epoch": 0.5116171003717472, "grad_norm": 0.5726884576518972, "learning_rate": 5.066219237157634e-06, "loss": 0.043, "step": 11010 }, { "epoch": 0.5116635687732342, "grad_norm": 0.675384883541271, "learning_rate": 5.0654667890777895e-06, "loss": 0.0477, "step": 11011 }, { "epoch": 0.5117100371747212, "grad_norm": 1.0919406957618205, "learning_rate": 5.064714339515054e-06, "loss": 0.0698, "step": 11012 }, { "epoch": 0.5117565055762082, "grad_norm": 0.758006951897704, "learning_rate": 5.0639618884864705e-06, "loss": 0.0584, "step": 11013 }, { "epoch": 0.5118029739776951, "grad_norm": 0.6755823777069714, "learning_rate": 5.0632094360090844e-06, "loss": 0.0631, "step": 11014 }, { "epoch": 0.5118494423791822, "grad_norm": 0.5808078003348262, "learning_rate": 5.062456982099937e-06, "loss": 0.0558, "step": 11015 }, { "epoch": 0.5118959107806691, "grad_norm": 1.209182719527485, "learning_rate": 5.061704526776073e-06, "loss": 0.0651, "step": 11016 }, { "epoch": 0.5119423791821561, "grad_norm": 1.133654499637849, "learning_rate": 5.060952070054536e-06, "loss": 0.0493, "step": 11017 }, { "epoch": 0.5119888475836432, "grad_norm": 1.0164820570562543, "learning_rate": 5.0601996119523715e-06, "loss": 0.0893, "step": 11018 }, { "epoch": 0.5120353159851301, "grad_norm": 0.580190334865779, "learning_rate": 5.059447152486621e-06, "loss": 0.0319, "step": 11019 }, { "epoch": 0.5120817843866171, "grad_norm": 0.7214335543083479, "learning_rate": 5.058694691674332e-06, "loss": 0.0688, "step": 11020 }, { "epoch": 0.5121282527881041, "grad_norm": 5.874138026387518, "learning_rate": 5.057942229532544e-06, "loss": 0.1146, "step": 11021 }, { "epoch": 0.5121747211895911, "grad_norm": 0.5905582771120698, "learning_rate": 5.057189766078306e-06, "loss": 0.0651, "step": 11022 }, { "epoch": 0.5122211895910781, "grad_norm": 0.5245855668234448, "learning_rate": 5.056437301328658e-06, "loss": 0.0465, "step": 11023 }, { "epoch": 0.512267657992565, "grad_norm": 0.5584836785892471, "learning_rate": 5.055684835300645e-06, "loss": 0.0479, "step": 11024 }, { "epoch": 0.512314126394052, "grad_norm": 0.5181013811293949, "learning_rate": 5.054932368011314e-06, "loss": 0.045, "step": 11025 }, { "epoch": 0.5123605947955391, "grad_norm": 0.5435180807849861, "learning_rate": 5.054179899477706e-06, "loss": 0.0504, "step": 11026 }, { "epoch": 0.512407063197026, "grad_norm": 1.3384344964143353, "learning_rate": 5.053427429716867e-06, "loss": 0.0704, "step": 11027 }, { "epoch": 0.512453531598513, "grad_norm": 1.534948385318298, "learning_rate": 5.052674958745839e-06, "loss": 0.0602, "step": 11028 }, { "epoch": 0.5125, "grad_norm": 0.9148354703375136, "learning_rate": 5.051922486581669e-06, "loss": 0.0596, "step": 11029 }, { "epoch": 0.512546468401487, "grad_norm": 0.8204846055042393, "learning_rate": 5.0511700132414e-06, "loss": 0.0643, "step": 11030 }, { "epoch": 0.512592936802974, "grad_norm": 0.5513119469796672, "learning_rate": 5.050417538742075e-06, "loss": 0.0559, "step": 11031 }, { "epoch": 0.5126394052044609, "grad_norm": 0.5020712534308988, "learning_rate": 5.049665063100741e-06, "loss": 0.0565, "step": 11032 }, { "epoch": 0.512685873605948, "grad_norm": 0.6690614985152155, "learning_rate": 5.0489125863344406e-06, "loss": 0.0397, "step": 11033 }, { "epoch": 0.5127323420074349, "grad_norm": 0.6294675057766491, "learning_rate": 5.048160108460219e-06, "loss": 0.0569, "step": 11034 }, { "epoch": 0.5127788104089219, "grad_norm": 6.3942312872598, "learning_rate": 5.04740762949512e-06, "loss": 0.079, "step": 11035 }, { "epoch": 0.512825278810409, "grad_norm": 0.6985053262930406, "learning_rate": 5.0466551494561875e-06, "loss": 0.0518, "step": 11036 }, { "epoch": 0.5128717472118959, "grad_norm": 0.840432021985843, "learning_rate": 5.045902668360469e-06, "loss": 0.0614, "step": 11037 }, { "epoch": 0.5129182156133829, "grad_norm": 0.5714797263314639, "learning_rate": 5.045150186225004e-06, "loss": 0.0351, "step": 11038 }, { "epoch": 0.5129646840148699, "grad_norm": 0.7501027506720291, "learning_rate": 5.044397703066842e-06, "loss": 0.0764, "step": 11039 }, { "epoch": 0.5130111524163569, "grad_norm": 0.4935090230282626, "learning_rate": 5.043645218903024e-06, "loss": 0.0545, "step": 11040 }, { "epoch": 0.5130576208178439, "grad_norm": 0.5086612353701377, "learning_rate": 5.042892733750596e-06, "loss": 0.0286, "step": 11041 }, { "epoch": 0.5131040892193308, "grad_norm": 0.6658915960002467, "learning_rate": 5.042140247626602e-06, "loss": 0.0674, "step": 11042 }, { "epoch": 0.5131505576208178, "grad_norm": 0.4215004151792068, "learning_rate": 5.041387760548087e-06, "loss": 0.064, "step": 11043 }, { "epoch": 0.5131970260223049, "grad_norm": 0.7545090341122122, "learning_rate": 5.040635272532098e-06, "loss": 0.0557, "step": 11044 }, { "epoch": 0.5132434944237918, "grad_norm": 0.5261348294042624, "learning_rate": 5.039882783595676e-06, "loss": 0.0466, "step": 11045 }, { "epoch": 0.5132899628252788, "grad_norm": 2.6784050477217924, "learning_rate": 5.039130293755866e-06, "loss": 0.0638, "step": 11046 }, { "epoch": 0.5133364312267658, "grad_norm": 0.5266914397198322, "learning_rate": 5.038377803029715e-06, "loss": 0.0615, "step": 11047 }, { "epoch": 0.5133828996282528, "grad_norm": 2.22206440511373, "learning_rate": 5.037625311434264e-06, "loss": 0.0819, "step": 11048 }, { "epoch": 0.5134293680297398, "grad_norm": 0.6314269805326655, "learning_rate": 5.0368728189865624e-06, "loss": 0.0768, "step": 11049 }, { "epoch": 0.5134758364312267, "grad_norm": 0.4926689634600396, "learning_rate": 5.036120325703653e-06, "loss": 0.0512, "step": 11050 }, { "epoch": 0.5135223048327138, "grad_norm": 0.5526741638573871, "learning_rate": 5.035367831602579e-06, "loss": 0.0495, "step": 11051 }, { "epoch": 0.5135687732342007, "grad_norm": 1.4810430690641971, "learning_rate": 5.034615336700386e-06, "loss": 0.0674, "step": 11052 }, { "epoch": 0.5136152416356877, "grad_norm": 0.887761955740037, "learning_rate": 5.03386284101412e-06, "loss": 0.0827, "step": 11053 }, { "epoch": 0.5136617100371748, "grad_norm": 2.727930676255033, "learning_rate": 5.033110344560825e-06, "loss": 0.1313, "step": 11054 }, { "epoch": 0.5137081784386617, "grad_norm": 1.2473807861057555, "learning_rate": 5.032357847357545e-06, "loss": 0.0594, "step": 11055 }, { "epoch": 0.5137546468401487, "grad_norm": 1.1260732076393778, "learning_rate": 5.031605349421328e-06, "loss": 0.0426, "step": 11056 }, { "epoch": 0.5138011152416357, "grad_norm": 1.925025400257761, "learning_rate": 5.030852850769214e-06, "loss": 0.0749, "step": 11057 }, { "epoch": 0.5138475836431227, "grad_norm": 1.15902930027012, "learning_rate": 5.030100351418251e-06, "loss": 0.0863, "step": 11058 }, { "epoch": 0.5138940520446097, "grad_norm": 0.919415477418899, "learning_rate": 5.029347851385484e-06, "loss": 0.0823, "step": 11059 }, { "epoch": 0.5139405204460966, "grad_norm": 0.39735500440424854, "learning_rate": 5.028595350687956e-06, "loss": 0.0399, "step": 11060 }, { "epoch": 0.5139869888475836, "grad_norm": 0.870608207461779, "learning_rate": 5.027842849342715e-06, "loss": 0.0683, "step": 11061 }, { "epoch": 0.5140334572490707, "grad_norm": 1.039975025223133, "learning_rate": 5.027090347366803e-06, "loss": 0.0644, "step": 11062 }, { "epoch": 0.5140799256505576, "grad_norm": 0.7128301421348375, "learning_rate": 5.026337844777267e-06, "loss": 0.0615, "step": 11063 }, { "epoch": 0.5141263940520446, "grad_norm": 1.779918789740972, "learning_rate": 5.02558534159115e-06, "loss": 0.0777, "step": 11064 }, { "epoch": 0.5141728624535316, "grad_norm": 1.144655208952981, "learning_rate": 5.0248328378255e-06, "loss": 0.0716, "step": 11065 }, { "epoch": 0.5142193308550186, "grad_norm": 2.3001661166336627, "learning_rate": 5.024080333497359e-06, "loss": 0.0375, "step": 11066 }, { "epoch": 0.5142657992565056, "grad_norm": 4.574048925706173, "learning_rate": 5.023327828623771e-06, "loss": 0.1077, "step": 11067 }, { "epoch": 0.5143122676579925, "grad_norm": 0.5292574479526788, "learning_rate": 5.022575323221787e-06, "loss": 0.067, "step": 11068 }, { "epoch": 0.5143587360594796, "grad_norm": 0.7353579059188782, "learning_rate": 5.0218228173084445e-06, "loss": 0.0632, "step": 11069 }, { "epoch": 0.5144052044609665, "grad_norm": 0.9377860787179221, "learning_rate": 5.021070310900795e-06, "loss": 0.0469, "step": 11070 }, { "epoch": 0.5144516728624535, "grad_norm": 0.521398020503304, "learning_rate": 5.020317804015879e-06, "loss": 0.0605, "step": 11071 }, { "epoch": 0.5144981412639406, "grad_norm": 0.5989353642388099, "learning_rate": 5.019565296670744e-06, "loss": 0.0556, "step": 11072 }, { "epoch": 0.5145446096654275, "grad_norm": 1.092632008158766, "learning_rate": 5.018812788882436e-06, "loss": 0.0842, "step": 11073 }, { "epoch": 0.5145910780669145, "grad_norm": 1.0924673670742453, "learning_rate": 5.018060280667996e-06, "loss": 0.0712, "step": 11074 }, { "epoch": 0.5146375464684015, "grad_norm": 1.0415338378795231, "learning_rate": 5.017307772044473e-06, "loss": 0.0595, "step": 11075 }, { "epoch": 0.5146840148698885, "grad_norm": 0.9609220504411843, "learning_rate": 5.0165552630289105e-06, "loss": 0.0876, "step": 11076 }, { "epoch": 0.5147304832713755, "grad_norm": 0.751029409949501, "learning_rate": 5.0158027536383524e-06, "loss": 0.0691, "step": 11077 }, { "epoch": 0.5147769516728624, "grad_norm": 0.7866858703177567, "learning_rate": 5.015050243889848e-06, "loss": 0.0542, "step": 11078 }, { "epoch": 0.5148234200743494, "grad_norm": 0.48792733573072916, "learning_rate": 5.014297733800438e-06, "loss": 0.0502, "step": 11079 }, { "epoch": 0.5148698884758365, "grad_norm": 1.848752553173646, "learning_rate": 5.013545223387169e-06, "loss": 0.0803, "step": 11080 }, { "epoch": 0.5149163568773234, "grad_norm": 0.6468662685698434, "learning_rate": 5.012792712667087e-06, "loss": 0.0702, "step": 11081 }, { "epoch": 0.5149628252788104, "grad_norm": 1.8963899068003744, "learning_rate": 5.0120402016572375e-06, "loss": 0.0538, "step": 11082 }, { "epoch": 0.5150092936802974, "grad_norm": 0.9518464358319065, "learning_rate": 5.0112876903746645e-06, "loss": 0.0794, "step": 11083 }, { "epoch": 0.5150557620817844, "grad_norm": 0.8062831231832901, "learning_rate": 5.010535178836412e-06, "loss": 0.06, "step": 11084 }, { "epoch": 0.5151022304832714, "grad_norm": 0.7440177342319653, "learning_rate": 5.009782667059528e-06, "loss": 0.0629, "step": 11085 }, { "epoch": 0.5151486988847583, "grad_norm": 0.41130819913644157, "learning_rate": 5.009030155061055e-06, "loss": 0.0425, "step": 11086 }, { "epoch": 0.5151951672862454, "grad_norm": 1.3429554715118843, "learning_rate": 5.008277642858042e-06, "loss": 0.0607, "step": 11087 }, { "epoch": 0.5152416356877323, "grad_norm": 1.3630608423913888, "learning_rate": 5.00752513046753e-06, "loss": 0.0606, "step": 11088 }, { "epoch": 0.5152881040892193, "grad_norm": 0.3734519347975581, "learning_rate": 5.006772617906564e-06, "loss": 0.0343, "step": 11089 }, { "epoch": 0.5153345724907064, "grad_norm": 0.6735247929192525, "learning_rate": 5.006020105192193e-06, "loss": 0.056, "step": 11090 }, { "epoch": 0.5153810408921933, "grad_norm": 0.4955392772101335, "learning_rate": 5.0052675923414605e-06, "loss": 0.0514, "step": 11091 }, { "epoch": 0.5154275092936803, "grad_norm": 0.9795723512795376, "learning_rate": 5.0045150793714115e-06, "loss": 0.0665, "step": 11092 }, { "epoch": 0.5154739776951673, "grad_norm": 0.7970627592562309, "learning_rate": 5.003762566299091e-06, "loss": 0.0699, "step": 11093 }, { "epoch": 0.5155204460966543, "grad_norm": 0.595577853882561, "learning_rate": 5.003010053141544e-06, "loss": 0.0536, "step": 11094 }, { "epoch": 0.5155669144981413, "grad_norm": 0.7987969517312781, "learning_rate": 5.002257539915817e-06, "loss": 0.0633, "step": 11095 }, { "epoch": 0.5156133828996282, "grad_norm": 0.5099321630528991, "learning_rate": 5.001505026638953e-06, "loss": 0.0524, "step": 11096 }, { "epoch": 0.5156598513011152, "grad_norm": 0.4416457949175861, "learning_rate": 5.000752513328e-06, "loss": 0.038, "step": 11097 }, { "epoch": 0.5157063197026023, "grad_norm": 0.8082620507234234, "learning_rate": 5e-06, "loss": 0.0474, "step": 11098 }, { "epoch": 0.5157527881040892, "grad_norm": 1.135346297344322, "learning_rate": 4.999247486672001e-06, "loss": 0.0658, "step": 11099 }, { "epoch": 0.5157992565055762, "grad_norm": 0.6315850902596873, "learning_rate": 4.998494973361048e-06, "loss": 0.0569, "step": 11100 }, { "epoch": 0.5158457249070632, "grad_norm": 0.39341089020980924, "learning_rate": 4.997742460084186e-06, "loss": 0.0641, "step": 11101 }, { "epoch": 0.5158921933085502, "grad_norm": 2.3337378765962407, "learning_rate": 4.996989946858459e-06, "loss": 0.0679, "step": 11102 }, { "epoch": 0.5159386617100372, "grad_norm": 2.570773479550147, "learning_rate": 4.9962374337009116e-06, "loss": 0.0737, "step": 11103 }, { "epoch": 0.5159851301115241, "grad_norm": 0.8248764787299001, "learning_rate": 4.995484920628589e-06, "loss": 0.0535, "step": 11104 }, { "epoch": 0.5160315985130112, "grad_norm": 0.8146281877204545, "learning_rate": 4.99473240765854e-06, "loss": 0.0677, "step": 11105 }, { "epoch": 0.5160780669144981, "grad_norm": 0.7636402689395705, "learning_rate": 4.993979894807807e-06, "loss": 0.0522, "step": 11106 }, { "epoch": 0.5161245353159851, "grad_norm": 0.7838500084578619, "learning_rate": 4.993227382093438e-06, "loss": 0.059, "step": 11107 }, { "epoch": 0.5161710037174722, "grad_norm": 0.44650946641555245, "learning_rate": 4.992474869532473e-06, "loss": 0.0593, "step": 11108 }, { "epoch": 0.5162174721189591, "grad_norm": 0.8090076116444747, "learning_rate": 4.991722357141961e-06, "loss": 0.0494, "step": 11109 }, { "epoch": 0.5162639405204461, "grad_norm": 0.5544782480503694, "learning_rate": 4.990969844938946e-06, "loss": 0.0555, "step": 11110 }, { "epoch": 0.5163104089219331, "grad_norm": 1.4230738835674692, "learning_rate": 4.990217332940473e-06, "loss": 0.0748, "step": 11111 }, { "epoch": 0.5163568773234201, "grad_norm": 0.4653559360857361, "learning_rate": 4.989464821163589e-06, "loss": 0.0404, "step": 11112 }, { "epoch": 0.5164033457249071, "grad_norm": 0.3447974916351904, "learning_rate": 4.988712309625339e-06, "loss": 0.0519, "step": 11113 }, { "epoch": 0.516449814126394, "grad_norm": 0.894204584700169, "learning_rate": 4.987959798342765e-06, "loss": 0.0472, "step": 11114 }, { "epoch": 0.516496282527881, "grad_norm": 0.6623138900658102, "learning_rate": 4.987207287332914e-06, "loss": 0.0451, "step": 11115 }, { "epoch": 0.5165427509293681, "grad_norm": 1.081158140937085, "learning_rate": 4.986454776612832e-06, "loss": 0.0696, "step": 11116 }, { "epoch": 0.516589219330855, "grad_norm": 1.4353433552596284, "learning_rate": 4.9857022661995634e-06, "loss": 0.072, "step": 11117 }, { "epoch": 0.516635687732342, "grad_norm": 0.48320606935306487, "learning_rate": 4.984949756110153e-06, "loss": 0.046, "step": 11118 }, { "epoch": 0.516682156133829, "grad_norm": 0.7188097123384805, "learning_rate": 4.984197246361649e-06, "loss": 0.0652, "step": 11119 }, { "epoch": 0.516728624535316, "grad_norm": 1.0719709244194513, "learning_rate": 4.983444736971092e-06, "loss": 0.0849, "step": 11120 }, { "epoch": 0.516775092936803, "grad_norm": 1.2714911724658753, "learning_rate": 4.982692227955528e-06, "loss": 0.045, "step": 11121 }, { "epoch": 0.5168215613382899, "grad_norm": 0.7435758035215972, "learning_rate": 4.981939719332006e-06, "loss": 0.0842, "step": 11122 }, { "epoch": 0.516868029739777, "grad_norm": 1.6880673190506779, "learning_rate": 4.981187211117566e-06, "loss": 0.0788, "step": 11123 }, { "epoch": 0.5169144981412639, "grad_norm": 0.7003668883478373, "learning_rate": 4.980434703329257e-06, "loss": 0.049, "step": 11124 }, { "epoch": 0.5169609665427509, "grad_norm": 0.5033486381350464, "learning_rate": 4.9796821959841225e-06, "loss": 0.0535, "step": 11125 }, { "epoch": 0.517007434944238, "grad_norm": 1.3503695960084914, "learning_rate": 4.9789296890992065e-06, "loss": 0.0731, "step": 11126 }, { "epoch": 0.5170539033457249, "grad_norm": 1.327068093732148, "learning_rate": 4.978177182691556e-06, "loss": 0.0541, "step": 11127 }, { "epoch": 0.5171003717472119, "grad_norm": 0.6731862974375832, "learning_rate": 4.977424676778215e-06, "loss": 0.0535, "step": 11128 }, { "epoch": 0.5171468401486989, "grad_norm": 0.7148704898138141, "learning_rate": 4.976672171376229e-06, "loss": 0.058, "step": 11129 }, { "epoch": 0.5171933085501859, "grad_norm": 0.7867020118455534, "learning_rate": 4.975919666502642e-06, "loss": 0.0694, "step": 11130 }, { "epoch": 0.5172397769516729, "grad_norm": 1.1222723779346613, "learning_rate": 4.975167162174503e-06, "loss": 0.0665, "step": 11131 }, { "epoch": 0.5172862453531598, "grad_norm": 0.5853654878333575, "learning_rate": 4.974414658408851e-06, "loss": 0.0289, "step": 11132 }, { "epoch": 0.5173327137546468, "grad_norm": 0.4123759439438437, "learning_rate": 4.973662155222734e-06, "loss": 0.055, "step": 11133 }, { "epoch": 0.5173791821561339, "grad_norm": 0.5430523980095052, "learning_rate": 4.972909652633198e-06, "loss": 0.043, "step": 11134 }, { "epoch": 0.5174256505576208, "grad_norm": 0.561930374951601, "learning_rate": 4.972157150657285e-06, "loss": 0.0504, "step": 11135 }, { "epoch": 0.5174721189591078, "grad_norm": 0.505114526057084, "learning_rate": 4.971404649312044e-06, "loss": 0.0608, "step": 11136 }, { "epoch": 0.5175185873605948, "grad_norm": 0.5083253817839111, "learning_rate": 4.970652148614518e-06, "loss": 0.0354, "step": 11137 }, { "epoch": 0.5175650557620818, "grad_norm": 0.7187712736665564, "learning_rate": 4.96989964858175e-06, "loss": 0.0514, "step": 11138 }, { "epoch": 0.5176115241635688, "grad_norm": 0.6815804296385528, "learning_rate": 4.969147149230788e-06, "loss": 0.0576, "step": 11139 }, { "epoch": 0.5176579925650557, "grad_norm": 0.9714141809831123, "learning_rate": 4.968394650578674e-06, "loss": 0.0396, "step": 11140 }, { "epoch": 0.5177044609665428, "grad_norm": 0.9553044955861798, "learning_rate": 4.967642152642456e-06, "loss": 0.098, "step": 11141 }, { "epoch": 0.5177509293680297, "grad_norm": 0.704533096344174, "learning_rate": 4.966889655439176e-06, "loss": 0.0453, "step": 11142 }, { "epoch": 0.5177973977695167, "grad_norm": 0.8239846166317615, "learning_rate": 4.966137158985881e-06, "loss": 0.0607, "step": 11143 }, { "epoch": 0.5178438661710038, "grad_norm": 0.5027115152664334, "learning_rate": 4.965384663299616e-06, "loss": 0.0484, "step": 11144 }, { "epoch": 0.5178903345724907, "grad_norm": 0.985990782300343, "learning_rate": 4.9646321683974224e-06, "loss": 0.0555, "step": 11145 }, { "epoch": 0.5179368029739777, "grad_norm": 0.6707055411115597, "learning_rate": 4.9638796742963495e-06, "loss": 0.066, "step": 11146 }, { "epoch": 0.5179832713754647, "grad_norm": 0.5987879019601215, "learning_rate": 4.9631271810134375e-06, "loss": 0.049, "step": 11147 }, { "epoch": 0.5180297397769517, "grad_norm": 0.8865896440170381, "learning_rate": 4.962374688565736e-06, "loss": 0.0783, "step": 11148 }, { "epoch": 0.5180762081784387, "grad_norm": 0.5307095927209118, "learning_rate": 4.961622196970288e-06, "loss": 0.0442, "step": 11149 }, { "epoch": 0.5181226765799256, "grad_norm": 0.4450183754576366, "learning_rate": 4.9608697062441355e-06, "loss": 0.0568, "step": 11150 }, { "epoch": 0.5181691449814126, "grad_norm": 1.7261532110758946, "learning_rate": 4.960117216404326e-06, "loss": 0.0696, "step": 11151 }, { "epoch": 0.5182156133828997, "grad_norm": 0.6405300203064427, "learning_rate": 4.959364727467903e-06, "loss": 0.0711, "step": 11152 }, { "epoch": 0.5182620817843866, "grad_norm": 1.913366356806541, "learning_rate": 4.958612239451913e-06, "loss": 0.0643, "step": 11153 }, { "epoch": 0.5183085501858736, "grad_norm": 0.9931332606776488, "learning_rate": 4.957859752373398e-06, "loss": 0.0625, "step": 11154 }, { "epoch": 0.5183550185873605, "grad_norm": 0.4737338642907105, "learning_rate": 4.957107266249406e-06, "loss": 0.0516, "step": 11155 }, { "epoch": 0.5184014869888476, "grad_norm": 0.42571542486917946, "learning_rate": 4.956354781096978e-06, "loss": 0.0396, "step": 11156 }, { "epoch": 0.5184479553903346, "grad_norm": 0.45038519826505785, "learning_rate": 4.95560229693316e-06, "loss": 0.0529, "step": 11157 }, { "epoch": 0.5184944237918215, "grad_norm": 0.7633490026503003, "learning_rate": 4.954849813774997e-06, "loss": 0.0411, "step": 11158 }, { "epoch": 0.5185408921933086, "grad_norm": 0.4267787785254171, "learning_rate": 4.954097331639532e-06, "loss": 0.046, "step": 11159 }, { "epoch": 0.5185873605947955, "grad_norm": 1.4069724972100242, "learning_rate": 4.953344850543813e-06, "loss": 0.0808, "step": 11160 }, { "epoch": 0.5186338289962825, "grad_norm": 0.40497506016046486, "learning_rate": 4.952592370504882e-06, "loss": 0.0511, "step": 11161 }, { "epoch": 0.5186802973977696, "grad_norm": 0.36025681897715617, "learning_rate": 4.9518398915397825e-06, "loss": 0.0341, "step": 11162 }, { "epoch": 0.5187267657992565, "grad_norm": 0.7911033040092762, "learning_rate": 4.95108741366556e-06, "loss": 0.0726, "step": 11163 }, { "epoch": 0.5187732342007435, "grad_norm": 1.174668175867736, "learning_rate": 4.95033493689926e-06, "loss": 0.0557, "step": 11164 }, { "epoch": 0.5188197026022305, "grad_norm": 0.5961837096261161, "learning_rate": 4.949582461257926e-06, "loss": 0.0568, "step": 11165 }, { "epoch": 0.5188661710037175, "grad_norm": 0.9429162739203432, "learning_rate": 4.948829986758603e-06, "loss": 0.0709, "step": 11166 }, { "epoch": 0.5189126394052045, "grad_norm": 0.618186271292552, "learning_rate": 4.948077513418333e-06, "loss": 0.0629, "step": 11167 }, { "epoch": 0.5189591078066914, "grad_norm": 0.7337202850440756, "learning_rate": 4.947325041254163e-06, "loss": 0.0621, "step": 11168 }, { "epoch": 0.5190055762081784, "grad_norm": 0.5270695069446887, "learning_rate": 4.946572570283135e-06, "loss": 0.0347, "step": 11169 }, { "epoch": 0.5190520446096655, "grad_norm": 0.8773828402927076, "learning_rate": 4.9458201005222956e-06, "loss": 0.0735, "step": 11170 }, { "epoch": 0.5190985130111524, "grad_norm": 0.7796517550136938, "learning_rate": 4.945067631988687e-06, "loss": 0.0572, "step": 11171 }, { "epoch": 0.5191449814126394, "grad_norm": 0.6313509157424884, "learning_rate": 4.944315164699356e-06, "loss": 0.0419, "step": 11172 }, { "epoch": 0.5191914498141263, "grad_norm": 1.010165741218562, "learning_rate": 4.943562698671345e-06, "loss": 0.0786, "step": 11173 }, { "epoch": 0.5192379182156134, "grad_norm": 0.6780780181730096, "learning_rate": 4.942810233921696e-06, "loss": 0.0517, "step": 11174 }, { "epoch": 0.5192843866171004, "grad_norm": 1.0225467018901848, "learning_rate": 4.9420577704674576e-06, "loss": 0.0634, "step": 11175 }, { "epoch": 0.5193308550185873, "grad_norm": 0.43834759414050917, "learning_rate": 4.9413053083256696e-06, "loss": 0.0513, "step": 11176 }, { "epoch": 0.5193773234200744, "grad_norm": 1.5226733881278351, "learning_rate": 4.940552847513379e-06, "loss": 0.0907, "step": 11177 }, { "epoch": 0.5194237918215613, "grad_norm": 0.5606621877874985, "learning_rate": 4.939800388047632e-06, "loss": 0.0589, "step": 11178 }, { "epoch": 0.5194702602230483, "grad_norm": 2.4459925710055272, "learning_rate": 4.939047929945466e-06, "loss": 0.079, "step": 11179 }, { "epoch": 0.5195167286245354, "grad_norm": 0.6239585202046969, "learning_rate": 4.93829547322393e-06, "loss": 0.0454, "step": 11180 }, { "epoch": 0.5195631970260223, "grad_norm": 1.911033919967957, "learning_rate": 4.937543017900065e-06, "loss": 0.063, "step": 11181 }, { "epoch": 0.5196096654275093, "grad_norm": 0.5557702187990349, "learning_rate": 4.936790563990917e-06, "loss": 0.0633, "step": 11182 }, { "epoch": 0.5196561338289963, "grad_norm": 0.5489901542661966, "learning_rate": 4.9360381115135295e-06, "loss": 0.0554, "step": 11183 }, { "epoch": 0.5197026022304833, "grad_norm": 0.9750073700194068, "learning_rate": 4.935285660484948e-06, "loss": 0.0635, "step": 11184 }, { "epoch": 0.5197490706319703, "grad_norm": 1.1256537762399126, "learning_rate": 4.934533210922212e-06, "loss": 0.0598, "step": 11185 }, { "epoch": 0.5197955390334572, "grad_norm": 1.0792967355539902, "learning_rate": 4.933780762842368e-06, "loss": 0.0639, "step": 11186 }, { "epoch": 0.5198420074349442, "grad_norm": 0.5083255872856484, "learning_rate": 4.93302831626246e-06, "loss": 0.045, "step": 11187 }, { "epoch": 0.5198884758364313, "grad_norm": 0.45821030933061246, "learning_rate": 4.93227587119953e-06, "loss": 0.0538, "step": 11188 }, { "epoch": 0.5199349442379182, "grad_norm": 0.5266362774870914, "learning_rate": 4.931523427670624e-06, "loss": 0.0509, "step": 11189 }, { "epoch": 0.5199814126394052, "grad_norm": 0.444819658017, "learning_rate": 4.930770985692785e-06, "loss": 0.0654, "step": 11190 }, { "epoch": 0.5200278810408921, "grad_norm": 0.5732622538799893, "learning_rate": 4.930018545283055e-06, "loss": 0.0432, "step": 11191 }, { "epoch": 0.5200743494423792, "grad_norm": 1.8678977117430853, "learning_rate": 4.929266106458479e-06, "loss": 0.052, "step": 11192 }, { "epoch": 0.5201208178438662, "grad_norm": 0.8531593854607656, "learning_rate": 4.928513669236098e-06, "loss": 0.0357, "step": 11193 }, { "epoch": 0.5201672862453531, "grad_norm": 0.7788388620263663, "learning_rate": 4.9277612336329606e-06, "loss": 0.07, "step": 11194 }, { "epoch": 0.5202137546468402, "grad_norm": 0.8287230517665894, "learning_rate": 4.927008799666106e-06, "loss": 0.0739, "step": 11195 }, { "epoch": 0.5202602230483271, "grad_norm": 1.4283951670892594, "learning_rate": 4.9262563673525795e-06, "loss": 0.0671, "step": 11196 }, { "epoch": 0.5203066914498141, "grad_norm": 1.0720858162046576, "learning_rate": 4.9255039367094235e-06, "loss": 0.0798, "step": 11197 }, { "epoch": 0.5203531598513012, "grad_norm": 0.8259506087871825, "learning_rate": 4.924751507753681e-06, "loss": 0.0513, "step": 11198 }, { "epoch": 0.5203996282527881, "grad_norm": 1.9524504366883426, "learning_rate": 4.923999080502397e-06, "loss": 0.0837, "step": 11199 }, { "epoch": 0.5204460966542751, "grad_norm": 0.6730080396403473, "learning_rate": 4.923246654972613e-06, "loss": 0.0571, "step": 11200 }, { "epoch": 0.5204925650557621, "grad_norm": 0.38074868007602186, "learning_rate": 4.922494231181374e-06, "loss": 0.0401, "step": 11201 }, { "epoch": 0.5205390334572491, "grad_norm": 1.277881715036666, "learning_rate": 4.921741809145723e-06, "loss": 0.0615, "step": 11202 }, { "epoch": 0.5205855018587361, "grad_norm": 0.2982200062497697, "learning_rate": 4.920989388882701e-06, "loss": 0.037, "step": 11203 }, { "epoch": 0.520631970260223, "grad_norm": 0.7776943800935713, "learning_rate": 4.920236970409352e-06, "loss": 0.069, "step": 11204 }, { "epoch": 0.52067843866171, "grad_norm": 0.6142269657135182, "learning_rate": 4.91948455374272e-06, "loss": 0.0627, "step": 11205 }, { "epoch": 0.5207249070631971, "grad_norm": 1.037045348929557, "learning_rate": 4.918732138899848e-06, "loss": 0.0643, "step": 11206 }, { "epoch": 0.520771375464684, "grad_norm": 0.6840767289402646, "learning_rate": 4.917979725897778e-06, "loss": 0.0501, "step": 11207 }, { "epoch": 0.520817843866171, "grad_norm": 0.9022480822129615, "learning_rate": 4.917227314753555e-06, "loss": 0.0653, "step": 11208 }, { "epoch": 0.520864312267658, "grad_norm": 0.9884865935580432, "learning_rate": 4.91647490548422e-06, "loss": 0.0727, "step": 11209 }, { "epoch": 0.520910780669145, "grad_norm": 0.6132555869648854, "learning_rate": 4.915722498106815e-06, "loss": 0.0529, "step": 11210 }, { "epoch": 0.520957249070632, "grad_norm": 0.4952286375380561, "learning_rate": 4.914970092638385e-06, "loss": 0.0522, "step": 11211 }, { "epoch": 0.5210037174721189, "grad_norm": 0.47701731973961686, "learning_rate": 4.914217689095972e-06, "loss": 0.0394, "step": 11212 }, { "epoch": 0.521050185873606, "grad_norm": 0.6229260052082998, "learning_rate": 4.913465287496618e-06, "loss": 0.0408, "step": 11213 }, { "epoch": 0.5210966542750929, "grad_norm": 1.265073792082901, "learning_rate": 4.912712887857368e-06, "loss": 0.068, "step": 11214 }, { "epoch": 0.5211431226765799, "grad_norm": 0.8595695025292788, "learning_rate": 4.911960490195261e-06, "loss": 0.061, "step": 11215 }, { "epoch": 0.521189591078067, "grad_norm": 0.6437729926320439, "learning_rate": 4.911208094527343e-06, "loss": 0.0834, "step": 11216 }, { "epoch": 0.5212360594795539, "grad_norm": 0.9221314685056343, "learning_rate": 4.910455700870653e-06, "loss": 0.0568, "step": 11217 }, { "epoch": 0.5212825278810409, "grad_norm": 0.7191569638059124, "learning_rate": 4.909703309242238e-06, "loss": 0.0594, "step": 11218 }, { "epoch": 0.5213289962825279, "grad_norm": 0.8165294959449917, "learning_rate": 4.908950919659138e-06, "loss": 0.0637, "step": 11219 }, { "epoch": 0.5213754646840149, "grad_norm": 0.39084833476184555, "learning_rate": 4.908198532138393e-06, "loss": 0.0165, "step": 11220 }, { "epoch": 0.5214219330855019, "grad_norm": 1.1478218985058823, "learning_rate": 4.907446146697051e-06, "loss": 0.043, "step": 11221 }, { "epoch": 0.5214684014869888, "grad_norm": 0.7360116952525854, "learning_rate": 4.906693763352148e-06, "loss": 0.0552, "step": 11222 }, { "epoch": 0.5215148698884758, "grad_norm": 0.5147941770518933, "learning_rate": 4.905941382120731e-06, "loss": 0.0596, "step": 11223 }, { "epoch": 0.5215613382899629, "grad_norm": 0.5468382702726794, "learning_rate": 4.90518900301984e-06, "loss": 0.057, "step": 11224 }, { "epoch": 0.5216078066914498, "grad_norm": 0.7198619555053154, "learning_rate": 4.9044366260665206e-06, "loss": 0.0479, "step": 11225 }, { "epoch": 0.5216542750929368, "grad_norm": 0.606511576606465, "learning_rate": 4.903684251277811e-06, "loss": 0.0638, "step": 11226 }, { "epoch": 0.5217007434944237, "grad_norm": 0.694088057458585, "learning_rate": 4.902931878670752e-06, "loss": 0.0375, "step": 11227 }, { "epoch": 0.5217472118959108, "grad_norm": 0.526005207191641, "learning_rate": 4.9021795082623905e-06, "loss": 0.0379, "step": 11228 }, { "epoch": 0.5217936802973978, "grad_norm": 0.7245208845210293, "learning_rate": 4.901427140069764e-06, "loss": 0.058, "step": 11229 }, { "epoch": 0.5218401486988847, "grad_norm": 1.0642242189761608, "learning_rate": 4.900674774109918e-06, "loss": 0.0567, "step": 11230 }, { "epoch": 0.5218866171003718, "grad_norm": 0.6522291353336602, "learning_rate": 4.899922410399895e-06, "loss": 0.0646, "step": 11231 }, { "epoch": 0.5219330855018587, "grad_norm": 0.4367795350897267, "learning_rate": 4.8991700489567316e-06, "loss": 0.0707, "step": 11232 }, { "epoch": 0.5219795539033457, "grad_norm": 0.7003311701469819, "learning_rate": 4.8984176897974736e-06, "loss": 0.0677, "step": 11233 }, { "epoch": 0.5220260223048327, "grad_norm": 1.0825350956639441, "learning_rate": 4.897665332939162e-06, "loss": 0.0331, "step": 11234 }, { "epoch": 0.5220724907063197, "grad_norm": 0.7050363744448973, "learning_rate": 4.896912978398838e-06, "loss": 0.0754, "step": 11235 }, { "epoch": 0.5221189591078067, "grad_norm": 0.6072963234573312, "learning_rate": 4.896160626193543e-06, "loss": 0.0558, "step": 11236 }, { "epoch": 0.5221654275092937, "grad_norm": 0.602270151751353, "learning_rate": 4.895408276340321e-06, "loss": 0.0712, "step": 11237 }, { "epoch": 0.5222118959107807, "grad_norm": 1.133251498420982, "learning_rate": 4.894655928856211e-06, "loss": 0.0491, "step": 11238 }, { "epoch": 0.5222583643122677, "grad_norm": 0.6998053088900735, "learning_rate": 4.8939035837582536e-06, "loss": 0.0636, "step": 11239 }, { "epoch": 0.5223048327137546, "grad_norm": 0.7013804101597765, "learning_rate": 4.893151241063493e-06, "loss": 0.0755, "step": 11240 }, { "epoch": 0.5223513011152416, "grad_norm": 0.9624621532476405, "learning_rate": 4.892398900788969e-06, "loss": 0.0608, "step": 11241 }, { "epoch": 0.5223977695167287, "grad_norm": 1.2765491614953095, "learning_rate": 4.891646562951722e-06, "loss": 0.0745, "step": 11242 }, { "epoch": 0.5224442379182156, "grad_norm": 0.6147526948170482, "learning_rate": 4.890894227568797e-06, "loss": 0.0593, "step": 11243 }, { "epoch": 0.5224907063197026, "grad_norm": 0.6670574284087082, "learning_rate": 4.89014189465723e-06, "loss": 0.0481, "step": 11244 }, { "epoch": 0.5225371747211895, "grad_norm": 0.3594365904460291, "learning_rate": 4.8893895642340665e-06, "loss": 0.0613, "step": 11245 }, { "epoch": 0.5225836431226766, "grad_norm": 0.8674574626448, "learning_rate": 4.888637236316344e-06, "loss": 0.0749, "step": 11246 }, { "epoch": 0.5226301115241636, "grad_norm": 0.967512708666972, "learning_rate": 4.887884910921107e-06, "loss": 0.063, "step": 11247 }, { "epoch": 0.5226765799256505, "grad_norm": 0.5987885407292215, "learning_rate": 4.887132588065393e-06, "loss": 0.0532, "step": 11248 }, { "epoch": 0.5227230483271376, "grad_norm": 0.6733441874181867, "learning_rate": 4.886380267766247e-06, "loss": 0.0602, "step": 11249 }, { "epoch": 0.5227695167286245, "grad_norm": 0.8716163102131944, "learning_rate": 4.885627950040707e-06, "loss": 0.0857, "step": 11250 }, { "epoch": 0.5228159851301115, "grad_norm": 1.100197235731723, "learning_rate": 4.884875634905813e-06, "loss": 0.0608, "step": 11251 }, { "epoch": 0.5228624535315985, "grad_norm": 0.5245851392062303, "learning_rate": 4.884123322378608e-06, "loss": 0.0415, "step": 11252 }, { "epoch": 0.5229089219330855, "grad_norm": 0.6843973887923236, "learning_rate": 4.88337101247613e-06, "loss": 0.0714, "step": 11253 }, { "epoch": 0.5229553903345725, "grad_norm": 0.7860436899030059, "learning_rate": 4.8826187052154225e-06, "loss": 0.0516, "step": 11254 }, { "epoch": 0.5230018587360595, "grad_norm": 0.9993220759243856, "learning_rate": 4.881866400613527e-06, "loss": 0.0532, "step": 11255 }, { "epoch": 0.5230483271375465, "grad_norm": 0.617163661076342, "learning_rate": 4.881114098687479e-06, "loss": 0.0702, "step": 11256 }, { "epoch": 0.5230947955390335, "grad_norm": 0.8020428090651333, "learning_rate": 4.880361799454323e-06, "loss": 0.0513, "step": 11257 }, { "epoch": 0.5231412639405204, "grad_norm": 0.40145285526571634, "learning_rate": 4.8796095029310965e-06, "loss": 0.0404, "step": 11258 }, { "epoch": 0.5231877323420074, "grad_norm": 1.7288448079730316, "learning_rate": 4.8788572091348435e-06, "loss": 0.0926, "step": 11259 }, { "epoch": 0.5232342007434945, "grad_norm": 0.7084278993150736, "learning_rate": 4.8781049180826e-06, "loss": 0.0577, "step": 11260 }, { "epoch": 0.5232806691449814, "grad_norm": 0.4838737323505888, "learning_rate": 4.877352629791412e-06, "loss": 0.0502, "step": 11261 }, { "epoch": 0.5233271375464684, "grad_norm": 0.574603257433509, "learning_rate": 4.876600344278314e-06, "loss": 0.0648, "step": 11262 }, { "epoch": 0.5233736059479553, "grad_norm": 0.4105719574728793, "learning_rate": 4.875848061560348e-06, "loss": 0.0352, "step": 11263 }, { "epoch": 0.5234200743494424, "grad_norm": 0.4731265891930019, "learning_rate": 4.875095781654553e-06, "loss": 0.0474, "step": 11264 }, { "epoch": 0.5234665427509294, "grad_norm": 0.7521897012114294, "learning_rate": 4.874343504577971e-06, "loss": 0.058, "step": 11265 }, { "epoch": 0.5235130111524163, "grad_norm": 0.6836157630144775, "learning_rate": 4.873591230347641e-06, "loss": 0.0412, "step": 11266 }, { "epoch": 0.5235594795539034, "grad_norm": 0.6594129912635458, "learning_rate": 4.872838958980604e-06, "loss": 0.0474, "step": 11267 }, { "epoch": 0.5236059479553903, "grad_norm": 1.0858255214841461, "learning_rate": 4.8720866904938966e-06, "loss": 0.0448, "step": 11268 }, { "epoch": 0.5236524163568773, "grad_norm": 0.6383038947005969, "learning_rate": 4.871334424904561e-06, "loss": 0.0545, "step": 11269 }, { "epoch": 0.5236988847583643, "grad_norm": 0.7936985494144687, "learning_rate": 4.870582162229635e-06, "loss": 0.0655, "step": 11270 }, { "epoch": 0.5237453531598513, "grad_norm": 0.39556815220867664, "learning_rate": 4.86982990248616e-06, "loss": 0.0483, "step": 11271 }, { "epoch": 0.5237918215613383, "grad_norm": 0.6341495708987466, "learning_rate": 4.8690776456911744e-06, "loss": 0.0482, "step": 11272 }, { "epoch": 0.5238382899628253, "grad_norm": 1.2689262206854681, "learning_rate": 4.86832539186172e-06, "loss": 0.0652, "step": 11273 }, { "epoch": 0.5238847583643123, "grad_norm": 0.5879469285257698, "learning_rate": 4.867573141014832e-06, "loss": 0.0369, "step": 11274 }, { "epoch": 0.5239312267657993, "grad_norm": 0.48967962386820885, "learning_rate": 4.866820893167552e-06, "loss": 0.046, "step": 11275 }, { "epoch": 0.5239776951672862, "grad_norm": 0.5630530769920127, "learning_rate": 4.866068648336919e-06, "loss": 0.0347, "step": 11276 }, { "epoch": 0.5240241635687732, "grad_norm": 0.8914114418190143, "learning_rate": 4.865316406539971e-06, "loss": 0.0551, "step": 11277 }, { "epoch": 0.5240706319702603, "grad_norm": 0.31118537053099593, "learning_rate": 4.8645641677937505e-06, "loss": 0.0463, "step": 11278 }, { "epoch": 0.5241171003717472, "grad_norm": 2.1407812668771964, "learning_rate": 4.8638119321152935e-06, "loss": 0.0711, "step": 11279 }, { "epoch": 0.5241635687732342, "grad_norm": 0.5163690726482468, "learning_rate": 4.863059699521638e-06, "loss": 0.0561, "step": 11280 }, { "epoch": 0.5242100371747211, "grad_norm": 0.5753093608854059, "learning_rate": 4.862307470029826e-06, "loss": 0.0359, "step": 11281 }, { "epoch": 0.5242565055762082, "grad_norm": 0.6069706650261839, "learning_rate": 4.861555243656893e-06, "loss": 0.0655, "step": 11282 }, { "epoch": 0.5243029739776952, "grad_norm": 0.7089152525751105, "learning_rate": 4.860803020419881e-06, "loss": 0.0616, "step": 11283 }, { "epoch": 0.5243494423791821, "grad_norm": 0.4364779830741892, "learning_rate": 4.860050800335828e-06, "loss": 0.0394, "step": 11284 }, { "epoch": 0.5243959107806692, "grad_norm": 0.4198287092191701, "learning_rate": 4.859298583421769e-06, "loss": 0.0392, "step": 11285 }, { "epoch": 0.5244423791821561, "grad_norm": 0.8905867853611943, "learning_rate": 4.8585463696947476e-06, "loss": 0.0506, "step": 11286 }, { "epoch": 0.5244888475836431, "grad_norm": 0.8664950135182335, "learning_rate": 4.857794159171797e-06, "loss": 0.0653, "step": 11287 }, { "epoch": 0.5245353159851301, "grad_norm": 0.44343610212588935, "learning_rate": 4.857041951869961e-06, "loss": 0.0333, "step": 11288 }, { "epoch": 0.5245817843866171, "grad_norm": 0.5641613764089991, "learning_rate": 4.8562897478062735e-06, "loss": 0.0596, "step": 11289 }, { "epoch": 0.5246282527881041, "grad_norm": 0.4224369826070559, "learning_rate": 4.855537546997776e-06, "loss": 0.0468, "step": 11290 }, { "epoch": 0.5246747211895911, "grad_norm": 0.5190011107608837, "learning_rate": 4.854785349461505e-06, "loss": 0.0606, "step": 11291 }, { "epoch": 0.524721189591078, "grad_norm": 0.8927487461643512, "learning_rate": 4.854033155214497e-06, "loss": 0.0516, "step": 11292 }, { "epoch": 0.5247676579925651, "grad_norm": 1.1661308074492815, "learning_rate": 4.853280964273793e-06, "loss": 0.0635, "step": 11293 }, { "epoch": 0.524814126394052, "grad_norm": 1.7558314013716758, "learning_rate": 4.852528776656428e-06, "loss": 0.0838, "step": 11294 }, { "epoch": 0.524860594795539, "grad_norm": 0.5306144016182964, "learning_rate": 4.851776592379443e-06, "loss": 0.0375, "step": 11295 }, { "epoch": 0.5249070631970261, "grad_norm": 0.5572522913531025, "learning_rate": 4.851024411459876e-06, "loss": 0.0528, "step": 11296 }, { "epoch": 0.524953531598513, "grad_norm": 0.8779203464844236, "learning_rate": 4.8502722339147595e-06, "loss": 0.0488, "step": 11297 }, { "epoch": 0.525, "grad_norm": 0.8478947315602529, "learning_rate": 4.849520059761136e-06, "loss": 0.0548, "step": 11298 }, { "epoch": 0.5250464684014869, "grad_norm": 0.5261372335963137, "learning_rate": 4.84876788901604e-06, "loss": 0.0601, "step": 11299 }, { "epoch": 0.525092936802974, "grad_norm": 0.4742146341398984, "learning_rate": 4.848015721696513e-06, "loss": 0.0299, "step": 11300 }, { "epoch": 0.525139405204461, "grad_norm": 0.6828603821351775, "learning_rate": 4.847263557819587e-06, "loss": 0.0716, "step": 11301 }, { "epoch": 0.5251858736059479, "grad_norm": 0.6979826748144019, "learning_rate": 4.846511397402306e-06, "loss": 0.0523, "step": 11302 }, { "epoch": 0.525232342007435, "grad_norm": 0.6089439171900367, "learning_rate": 4.845759240461701e-06, "loss": 0.0481, "step": 11303 }, { "epoch": 0.5252788104089219, "grad_norm": 0.6862159140335312, "learning_rate": 4.8450070870148105e-06, "loss": 0.0581, "step": 11304 }, { "epoch": 0.5253252788104089, "grad_norm": 0.48863008028250243, "learning_rate": 4.844254937078675e-06, "loss": 0.0442, "step": 11305 }, { "epoch": 0.525371747211896, "grad_norm": 0.34628973693397164, "learning_rate": 4.843502790670327e-06, "loss": 0.0433, "step": 11306 }, { "epoch": 0.5254182156133829, "grad_norm": 0.48025577626925753, "learning_rate": 4.842750647806806e-06, "loss": 0.0432, "step": 11307 }, { "epoch": 0.5254646840148699, "grad_norm": 0.5832230207693972, "learning_rate": 4.84199850850515e-06, "loss": 0.0391, "step": 11308 }, { "epoch": 0.5255111524163569, "grad_norm": 0.5185087813070153, "learning_rate": 4.841246372782392e-06, "loss": 0.0407, "step": 11309 }, { "epoch": 0.5255576208178439, "grad_norm": 0.6520922018890318, "learning_rate": 4.840494240655572e-06, "loss": 0.0482, "step": 11310 }, { "epoch": 0.5256040892193309, "grad_norm": 0.5970159068316353, "learning_rate": 4.839742112141725e-06, "loss": 0.048, "step": 11311 }, { "epoch": 0.5256505576208178, "grad_norm": 0.3913988618014367, "learning_rate": 4.838989987257887e-06, "loss": 0.0195, "step": 11312 }, { "epoch": 0.5256970260223048, "grad_norm": 0.5856127739033388, "learning_rate": 4.838237866021096e-06, "loss": 0.0385, "step": 11313 }, { "epoch": 0.5257434944237919, "grad_norm": 0.7308514261163844, "learning_rate": 4.837485748448388e-06, "loss": 0.0554, "step": 11314 }, { "epoch": 0.5257899628252788, "grad_norm": 0.7879472645505251, "learning_rate": 4.836733634556799e-06, "loss": 0.0842, "step": 11315 }, { "epoch": 0.5258364312267658, "grad_norm": 0.5327430770036902, "learning_rate": 4.835981524363364e-06, "loss": 0.0435, "step": 11316 }, { "epoch": 0.5258828996282527, "grad_norm": 1.3188047897049409, "learning_rate": 4.835229417885121e-06, "loss": 0.0543, "step": 11317 }, { "epoch": 0.5259293680297398, "grad_norm": 1.0772534581999005, "learning_rate": 4.834477315139103e-06, "loss": 0.0702, "step": 11318 }, { "epoch": 0.5259758364312268, "grad_norm": 0.5314667311142732, "learning_rate": 4.833725216142351e-06, "loss": 0.047, "step": 11319 }, { "epoch": 0.5260223048327137, "grad_norm": 0.7311795466232751, "learning_rate": 4.832973120911897e-06, "loss": 0.0507, "step": 11320 }, { "epoch": 0.5260687732342008, "grad_norm": 0.653750596134341, "learning_rate": 4.832221029464776e-06, "loss": 0.0454, "step": 11321 }, { "epoch": 0.5261152416356877, "grad_norm": 1.855482824822267, "learning_rate": 4.831468941818027e-06, "loss": 0.076, "step": 11322 }, { "epoch": 0.5261617100371747, "grad_norm": 0.6925790689139615, "learning_rate": 4.830716857988682e-06, "loss": 0.0726, "step": 11323 }, { "epoch": 0.5262081784386617, "grad_norm": 0.5950359628392111, "learning_rate": 4.8299647779937805e-06, "loss": 0.0384, "step": 11324 }, { "epoch": 0.5262546468401487, "grad_norm": 0.5340998962685792, "learning_rate": 4.829212701850354e-06, "loss": 0.0358, "step": 11325 }, { "epoch": 0.5263011152416357, "grad_norm": 0.5517755828464995, "learning_rate": 4.828460629575441e-06, "loss": 0.0401, "step": 11326 }, { "epoch": 0.5263475836431227, "grad_norm": 0.4126904529105893, "learning_rate": 4.827708561186075e-06, "loss": 0.0432, "step": 11327 }, { "epoch": 0.5263940520446097, "grad_norm": 1.5128314519988113, "learning_rate": 4.82695649669929e-06, "loss": 0.0703, "step": 11328 }, { "epoch": 0.5264405204460967, "grad_norm": 0.4431490969938899, "learning_rate": 4.8262044361321245e-06, "loss": 0.0425, "step": 11329 }, { "epoch": 0.5264869888475836, "grad_norm": 0.9540298077463536, "learning_rate": 4.8254523795016095e-06, "loss": 0.0627, "step": 11330 }, { "epoch": 0.5265334572490706, "grad_norm": 0.7064847456218254, "learning_rate": 4.8247003268247825e-06, "loss": 0.0525, "step": 11331 }, { "epoch": 0.5265799256505577, "grad_norm": 0.41871154998031423, "learning_rate": 4.82394827811868e-06, "loss": 0.052, "step": 11332 }, { "epoch": 0.5266263940520446, "grad_norm": 0.8235680335867337, "learning_rate": 4.823196233400331e-06, "loss": 0.0779, "step": 11333 }, { "epoch": 0.5266728624535316, "grad_norm": 1.1473781821790927, "learning_rate": 4.822444192686775e-06, "loss": 0.0727, "step": 11334 }, { "epoch": 0.5267193308550185, "grad_norm": 0.41146164001228364, "learning_rate": 4.821692155995043e-06, "loss": 0.0535, "step": 11335 }, { "epoch": 0.5267657992565056, "grad_norm": 0.7431584456513634, "learning_rate": 4.820940123342173e-06, "loss": 0.0594, "step": 11336 }, { "epoch": 0.5268122676579926, "grad_norm": 0.3151418032654511, "learning_rate": 4.820188094745199e-06, "loss": 0.0371, "step": 11337 }, { "epoch": 0.5268587360594795, "grad_norm": 0.8448066420143439, "learning_rate": 4.8194360702211516e-06, "loss": 0.0682, "step": 11338 }, { "epoch": 0.5269052044609666, "grad_norm": 0.4780940041223074, "learning_rate": 4.818684049787068e-06, "loss": 0.0455, "step": 11339 }, { "epoch": 0.5269516728624535, "grad_norm": 0.3212614422805013, "learning_rate": 4.817932033459981e-06, "loss": 0.0446, "step": 11340 }, { "epoch": 0.5269981412639405, "grad_norm": 0.49240442102018894, "learning_rate": 4.817180021256926e-06, "loss": 0.0529, "step": 11341 }, { "epoch": 0.5270446096654275, "grad_norm": 0.7831201361539396, "learning_rate": 4.8164280131949345e-06, "loss": 0.0706, "step": 11342 }, { "epoch": 0.5270910780669145, "grad_norm": 0.6295106139857606, "learning_rate": 4.815676009291044e-06, "loss": 0.0605, "step": 11343 }, { "epoch": 0.5271375464684015, "grad_norm": 0.3984991893420926, "learning_rate": 4.8149240095622855e-06, "loss": 0.0398, "step": 11344 }, { "epoch": 0.5271840148698885, "grad_norm": 0.7963108755803101, "learning_rate": 4.814172014025691e-06, "loss": 0.0754, "step": 11345 }, { "epoch": 0.5272304832713754, "grad_norm": 1.067261825317337, "learning_rate": 4.813420022698297e-06, "loss": 0.0533, "step": 11346 }, { "epoch": 0.5272769516728625, "grad_norm": 0.4227523033345263, "learning_rate": 4.812668035597135e-06, "loss": 0.0455, "step": 11347 }, { "epoch": 0.5273234200743494, "grad_norm": 0.44683465411458617, "learning_rate": 4.81191605273924e-06, "loss": 0.045, "step": 11348 }, { "epoch": 0.5273698884758364, "grad_norm": 0.5165021230124242, "learning_rate": 4.8111640741416465e-06, "loss": 0.0537, "step": 11349 }, { "epoch": 0.5274163568773235, "grad_norm": 1.450186381921176, "learning_rate": 4.810412099821382e-06, "loss": 0.0551, "step": 11350 }, { "epoch": 0.5274628252788104, "grad_norm": 0.5141218498879485, "learning_rate": 4.809660129795485e-06, "loss": 0.0682, "step": 11351 }, { "epoch": 0.5275092936802974, "grad_norm": 0.7154797040782136, "learning_rate": 4.808908164080985e-06, "loss": 0.0647, "step": 11352 }, { "epoch": 0.5275557620817843, "grad_norm": 0.56866793040144, "learning_rate": 4.808156202694917e-06, "loss": 0.0502, "step": 11353 }, { "epoch": 0.5276022304832714, "grad_norm": 0.5975210033497804, "learning_rate": 4.807404245654312e-06, "loss": 0.0401, "step": 11354 }, { "epoch": 0.5276486988847584, "grad_norm": 0.9146502184882715, "learning_rate": 4.806652292976206e-06, "loss": 0.0408, "step": 11355 }, { "epoch": 0.5276951672862453, "grad_norm": 0.654108669722686, "learning_rate": 4.805900344677628e-06, "loss": 0.0704, "step": 11356 }, { "epoch": 0.5277416356877324, "grad_norm": 0.8779155280263903, "learning_rate": 4.80514840077561e-06, "loss": 0.0543, "step": 11357 }, { "epoch": 0.5277881040892193, "grad_norm": 0.5214419897443772, "learning_rate": 4.804396461287187e-06, "loss": 0.0482, "step": 11358 }, { "epoch": 0.5278345724907063, "grad_norm": 0.6551997714792005, "learning_rate": 4.803644526229388e-06, "loss": 0.0479, "step": 11359 }, { "epoch": 0.5278810408921933, "grad_norm": 0.6240751604439468, "learning_rate": 4.802892595619249e-06, "loss": 0.0639, "step": 11360 }, { "epoch": 0.5279275092936803, "grad_norm": 0.8165307832476469, "learning_rate": 4.8021406694738025e-06, "loss": 0.0794, "step": 11361 }, { "epoch": 0.5279739776951673, "grad_norm": 0.7241375759557003, "learning_rate": 4.801388747810075e-06, "loss": 0.0709, "step": 11362 }, { "epoch": 0.5280204460966543, "grad_norm": 0.5824036485234138, "learning_rate": 4.800636830645101e-06, "loss": 0.0517, "step": 11363 }, { "epoch": 0.5280669144981412, "grad_norm": 0.8063242645989231, "learning_rate": 4.799884917995914e-06, "loss": 0.0629, "step": 11364 }, { "epoch": 0.5281133828996283, "grad_norm": 0.8552096556336, "learning_rate": 4.799133009879544e-06, "loss": 0.0628, "step": 11365 }, { "epoch": 0.5281598513011152, "grad_norm": 1.111266628673094, "learning_rate": 4.798381106313021e-06, "loss": 0.0917, "step": 11366 }, { "epoch": 0.5282063197026022, "grad_norm": 0.5288410705916013, "learning_rate": 4.797629207313382e-06, "loss": 0.0498, "step": 11367 }, { "epoch": 0.5282527881040893, "grad_norm": 0.4688816147652767, "learning_rate": 4.796877312897652e-06, "loss": 0.0377, "step": 11368 }, { "epoch": 0.5282992565055762, "grad_norm": 0.6262346391681194, "learning_rate": 4.7961254230828645e-06, "loss": 0.0561, "step": 11369 }, { "epoch": 0.5283457249070632, "grad_norm": 1.6637189123971206, "learning_rate": 4.795373537886051e-06, "loss": 0.0472, "step": 11370 }, { "epoch": 0.5283921933085501, "grad_norm": 3.852489132018117, "learning_rate": 4.7946216573242416e-06, "loss": 0.0712, "step": 11371 }, { "epoch": 0.5284386617100372, "grad_norm": 0.6978846076549938, "learning_rate": 4.793869781414469e-06, "loss": 0.0465, "step": 11372 }, { "epoch": 0.5284851301115242, "grad_norm": 1.2758251113684826, "learning_rate": 4.793117910173764e-06, "loss": 0.0666, "step": 11373 }, { "epoch": 0.5285315985130111, "grad_norm": 0.7209184823232666, "learning_rate": 4.792366043619153e-06, "loss": 0.0506, "step": 11374 }, { "epoch": 0.5285780669144982, "grad_norm": 0.5142690086010928, "learning_rate": 4.7916141817676716e-06, "loss": 0.0491, "step": 11375 }, { "epoch": 0.5286245353159851, "grad_norm": 0.5876800251669502, "learning_rate": 4.7908623246363465e-06, "loss": 0.0646, "step": 11376 }, { "epoch": 0.5286710037174721, "grad_norm": 0.9987077097742706, "learning_rate": 4.790110472242212e-06, "loss": 0.0638, "step": 11377 }, { "epoch": 0.5287174721189591, "grad_norm": 0.9256421193773999, "learning_rate": 4.789358624602295e-06, "loss": 0.0718, "step": 11378 }, { "epoch": 0.5287639405204461, "grad_norm": 0.8406568998266847, "learning_rate": 4.788606781733628e-06, "loss": 0.0674, "step": 11379 }, { "epoch": 0.5288104089219331, "grad_norm": 1.5830097092620186, "learning_rate": 4.787854943653239e-06, "loss": 0.0559, "step": 11380 }, { "epoch": 0.5288568773234201, "grad_norm": 1.0367369271228384, "learning_rate": 4.787103110378158e-06, "loss": 0.0823, "step": 11381 }, { "epoch": 0.528903345724907, "grad_norm": 0.5642590551240523, "learning_rate": 4.786351281925417e-06, "loss": 0.0452, "step": 11382 }, { "epoch": 0.5289498141263941, "grad_norm": 0.7519260947489077, "learning_rate": 4.785599458312043e-06, "loss": 0.0433, "step": 11383 }, { "epoch": 0.528996282527881, "grad_norm": 0.6404092968997515, "learning_rate": 4.784847639555068e-06, "loss": 0.0564, "step": 11384 }, { "epoch": 0.529042750929368, "grad_norm": 0.402876382480945, "learning_rate": 4.784095825671521e-06, "loss": 0.0327, "step": 11385 }, { "epoch": 0.5290892193308551, "grad_norm": 0.6089379756042382, "learning_rate": 4.783344016678429e-06, "loss": 0.0582, "step": 11386 }, { "epoch": 0.529135687732342, "grad_norm": 0.5274857940736436, "learning_rate": 4.782592212592825e-06, "loss": 0.059, "step": 11387 }, { "epoch": 0.529182156133829, "grad_norm": 0.6886189205095268, "learning_rate": 4.781840413431734e-06, "loss": 0.0466, "step": 11388 }, { "epoch": 0.5292286245353159, "grad_norm": 0.42309992609398145, "learning_rate": 4.781088619212189e-06, "loss": 0.0349, "step": 11389 }, { "epoch": 0.529275092936803, "grad_norm": 0.657138942482043, "learning_rate": 4.780336829951216e-06, "loss": 0.0414, "step": 11390 }, { "epoch": 0.52932156133829, "grad_norm": 0.9486768573167121, "learning_rate": 4.7795850456658465e-06, "loss": 0.0743, "step": 11391 }, { "epoch": 0.5293680297397769, "grad_norm": 0.5917975090553035, "learning_rate": 4.778833266373107e-06, "loss": 0.0427, "step": 11392 }, { "epoch": 0.529414498141264, "grad_norm": 0.5233585027544136, "learning_rate": 4.778081492090026e-06, "loss": 0.0457, "step": 11393 }, { "epoch": 0.5294609665427509, "grad_norm": 0.40693652154174015, "learning_rate": 4.777329722833634e-06, "loss": 0.0386, "step": 11394 }, { "epoch": 0.5295074349442379, "grad_norm": 1.1093822674594278, "learning_rate": 4.776577958620956e-06, "loss": 0.0318, "step": 11395 }, { "epoch": 0.5295539033457249, "grad_norm": 13.709242332968218, "learning_rate": 4.775826199469025e-06, "loss": 0.0921, "step": 11396 }, { "epoch": 0.5296003717472119, "grad_norm": 2.1464262498118325, "learning_rate": 4.775074445394866e-06, "loss": 0.0539, "step": 11397 }, { "epoch": 0.5296468401486989, "grad_norm": 0.582389482934428, "learning_rate": 4.774322696415506e-06, "loss": 0.0629, "step": 11398 }, { "epoch": 0.5296933085501859, "grad_norm": 1.1707745643360503, "learning_rate": 4.773570952547975e-06, "loss": 0.0769, "step": 11399 }, { "epoch": 0.5297397769516728, "grad_norm": 0.7682892695011899, "learning_rate": 4.772819213809301e-06, "loss": 0.0517, "step": 11400 }, { "epoch": 0.5297862453531599, "grad_norm": 0.5517826036797934, "learning_rate": 4.77206748021651e-06, "loss": 0.0308, "step": 11401 }, { "epoch": 0.5298327137546468, "grad_norm": 0.525976718622326, "learning_rate": 4.771315751786632e-06, "loss": 0.0394, "step": 11402 }, { "epoch": 0.5298791821561338, "grad_norm": 0.5959268646406689, "learning_rate": 4.770564028536692e-06, "loss": 0.0461, "step": 11403 }, { "epoch": 0.5299256505576209, "grad_norm": 1.0023355208001745, "learning_rate": 4.769812310483718e-06, "loss": 0.0405, "step": 11404 }, { "epoch": 0.5299721189591078, "grad_norm": 2.590144826342658, "learning_rate": 4.769060597644737e-06, "loss": 0.054, "step": 11405 }, { "epoch": 0.5300185873605948, "grad_norm": 0.5245440953012045, "learning_rate": 4.768308890036778e-06, "loss": 0.0663, "step": 11406 }, { "epoch": 0.5300650557620817, "grad_norm": 0.392352197208247, "learning_rate": 4.767557187676864e-06, "loss": 0.0241, "step": 11407 }, { "epoch": 0.5301115241635688, "grad_norm": 0.5471436617031347, "learning_rate": 4.766805490582027e-06, "loss": 0.0526, "step": 11408 }, { "epoch": 0.5301579925650558, "grad_norm": 1.646447836449828, "learning_rate": 4.766053798769291e-06, "loss": 0.0834, "step": 11409 }, { "epoch": 0.5302044609665427, "grad_norm": 2.51196135394926, "learning_rate": 4.76530211225568e-06, "loss": 0.0475, "step": 11410 }, { "epoch": 0.5302509293680298, "grad_norm": 1.4020374112188474, "learning_rate": 4.764550431058225e-06, "loss": 0.0756, "step": 11411 }, { "epoch": 0.5302973977695167, "grad_norm": 0.7389073014109565, "learning_rate": 4.76379875519395e-06, "loss": 0.0567, "step": 11412 }, { "epoch": 0.5303438661710037, "grad_norm": 0.7797855286306522, "learning_rate": 4.763047084679883e-06, "loss": 0.0694, "step": 11413 }, { "epoch": 0.5303903345724907, "grad_norm": 1.391474332145084, "learning_rate": 4.7622954195330486e-06, "loss": 0.0466, "step": 11414 }, { "epoch": 0.5304368029739777, "grad_norm": 0.8918756360349568, "learning_rate": 4.761543759770472e-06, "loss": 0.0801, "step": 11415 }, { "epoch": 0.5304832713754647, "grad_norm": 0.4188603987510723, "learning_rate": 4.760792105409181e-06, "loss": 0.0473, "step": 11416 }, { "epoch": 0.5305297397769517, "grad_norm": 2.412248453168239, "learning_rate": 4.7600404564662e-06, "loss": 0.0572, "step": 11417 }, { "epoch": 0.5305762081784386, "grad_norm": 0.633031479975344, "learning_rate": 4.759288812958556e-06, "loss": 0.0698, "step": 11418 }, { "epoch": 0.5306226765799257, "grad_norm": 1.2573038715457838, "learning_rate": 4.758537174903272e-06, "loss": 0.0839, "step": 11419 }, { "epoch": 0.5306691449814126, "grad_norm": 0.8380180773989973, "learning_rate": 4.757785542317378e-06, "loss": 0.0595, "step": 11420 }, { "epoch": 0.5307156133828996, "grad_norm": 0.9577039735224558, "learning_rate": 4.757033915217895e-06, "loss": 0.0748, "step": 11421 }, { "epoch": 0.5307620817843867, "grad_norm": 1.6021027731430029, "learning_rate": 4.756282293621849e-06, "loss": 0.0748, "step": 11422 }, { "epoch": 0.5308085501858736, "grad_norm": 3.1776374761734942, "learning_rate": 4.755530677546266e-06, "loss": 0.0678, "step": 11423 }, { "epoch": 0.5308550185873606, "grad_norm": 1.011646985330407, "learning_rate": 4.75477906700817e-06, "loss": 0.056, "step": 11424 }, { "epoch": 0.5309014869888475, "grad_norm": 1.1207046610694364, "learning_rate": 4.754027462024586e-06, "loss": 0.0818, "step": 11425 }, { "epoch": 0.5309479553903346, "grad_norm": 0.893437980723258, "learning_rate": 4.753275862612541e-06, "loss": 0.0615, "step": 11426 }, { "epoch": 0.5309944237918216, "grad_norm": 0.7462265541124495, "learning_rate": 4.752524268789055e-06, "loss": 0.0489, "step": 11427 }, { "epoch": 0.5310408921933085, "grad_norm": 0.7462265541124495, "learning_rate": 4.752524268789055e-06, "loss": 0.0626, "step": 11428 }, { "epoch": 0.5310873605947956, "grad_norm": 0.5638346687216602, "learning_rate": 4.751772680571156e-06, "loss": 0.0723, "step": 11429 }, { "epoch": 0.5311338289962825, "grad_norm": 0.6615782030039924, "learning_rate": 4.751021097975866e-06, "loss": 0.058, "step": 11430 }, { "epoch": 0.5311802973977695, "grad_norm": 0.4544928623899996, "learning_rate": 4.750269521020212e-06, "loss": 0.0396, "step": 11431 }, { "epoch": 0.5312267657992565, "grad_norm": 1.057431315101357, "learning_rate": 4.749517949721214e-06, "loss": 0.0488, "step": 11432 }, { "epoch": 0.5312732342007435, "grad_norm": 0.7359200739702565, "learning_rate": 4.7487663840959004e-06, "loss": 0.043, "step": 11433 }, { "epoch": 0.5313197026022305, "grad_norm": 0.6147779345149343, "learning_rate": 4.748014824161292e-06, "loss": 0.0441, "step": 11434 }, { "epoch": 0.5313661710037175, "grad_norm": 0.5554996475530491, "learning_rate": 4.747263269934411e-06, "loss": 0.0499, "step": 11435 }, { "epoch": 0.5314126394052044, "grad_norm": 1.1597669382795344, "learning_rate": 4.746511721432285e-06, "loss": 0.0365, "step": 11436 }, { "epoch": 0.5314591078066915, "grad_norm": 0.5699911225546294, "learning_rate": 4.745760178671933e-06, "loss": 0.0611, "step": 11437 }, { "epoch": 0.5315055762081784, "grad_norm": 0.8017808591686553, "learning_rate": 4.745008641670383e-06, "loss": 0.0619, "step": 11438 }, { "epoch": 0.5315520446096654, "grad_norm": 0.669832442802855, "learning_rate": 4.744257110444656e-06, "loss": 0.063, "step": 11439 }, { "epoch": 0.5315985130111525, "grad_norm": 0.723650291790538, "learning_rate": 4.743505585011773e-06, "loss": 0.0493, "step": 11440 }, { "epoch": 0.5316449814126394, "grad_norm": 1.2452819953424745, "learning_rate": 4.742754065388759e-06, "loss": 0.0872, "step": 11441 }, { "epoch": 0.5316914498141264, "grad_norm": 0.757101063292413, "learning_rate": 4.742002551592635e-06, "loss": 0.0703, "step": 11442 }, { "epoch": 0.5317379182156133, "grad_norm": 0.602340953407072, "learning_rate": 4.741251043640426e-06, "loss": 0.0421, "step": 11443 }, { "epoch": 0.5317843866171004, "grad_norm": 0.48696481484295, "learning_rate": 4.740499541549152e-06, "loss": 0.0656, "step": 11444 }, { "epoch": 0.5318308550185874, "grad_norm": 1.1097477021760243, "learning_rate": 4.7397480453358395e-06, "loss": 0.0484, "step": 11445 }, { "epoch": 0.5318773234200743, "grad_norm": 0.30752173555334755, "learning_rate": 4.738996555017506e-06, "loss": 0.041, "step": 11446 }, { "epoch": 0.5319237918215614, "grad_norm": 1.0157287736710265, "learning_rate": 4.738245070611174e-06, "loss": 0.0632, "step": 11447 }, { "epoch": 0.5319702602230483, "grad_norm": 0.69262285651516, "learning_rate": 4.737493592133867e-06, "loss": 0.0522, "step": 11448 }, { "epoch": 0.5320167286245353, "grad_norm": 0.6606028707593149, "learning_rate": 4.736742119602606e-06, "loss": 0.0543, "step": 11449 }, { "epoch": 0.5320631970260223, "grad_norm": 0.4301455543735362, "learning_rate": 4.735990653034415e-06, "loss": 0.0566, "step": 11450 }, { "epoch": 0.5321096654275093, "grad_norm": 0.4314796947219662, "learning_rate": 4.735239192446313e-06, "loss": 0.0377, "step": 11451 }, { "epoch": 0.5321561338289963, "grad_norm": 2.51185740235636, "learning_rate": 4.734487737855322e-06, "loss": 0.0597, "step": 11452 }, { "epoch": 0.5322026022304833, "grad_norm": 0.5291475382625898, "learning_rate": 4.733736289278462e-06, "loss": 0.0598, "step": 11453 }, { "epoch": 0.5322490706319702, "grad_norm": 1.2057492372293162, "learning_rate": 4.732984846732755e-06, "loss": 0.0494, "step": 11454 }, { "epoch": 0.5322955390334573, "grad_norm": 0.6364207573476349, "learning_rate": 4.7322334102352245e-06, "loss": 0.0524, "step": 11455 }, { "epoch": 0.5323420074349442, "grad_norm": 0.42848694605344356, "learning_rate": 4.7314819798028884e-06, "loss": 0.0459, "step": 11456 }, { "epoch": 0.5323884758364312, "grad_norm": 0.5849444455746005, "learning_rate": 4.730730555452767e-06, "loss": 0.0521, "step": 11457 }, { "epoch": 0.5324349442379183, "grad_norm": 0.7132821905528602, "learning_rate": 4.729979137201882e-06, "loss": 0.064, "step": 11458 }, { "epoch": 0.5324814126394052, "grad_norm": 1.310342157568771, "learning_rate": 4.7292277250672545e-06, "loss": 0.0524, "step": 11459 }, { "epoch": 0.5325278810408922, "grad_norm": 0.849806513404067, "learning_rate": 4.728476319065904e-06, "loss": 0.0592, "step": 11460 }, { "epoch": 0.5325743494423791, "grad_norm": 2.5622295236421615, "learning_rate": 4.727724919214849e-06, "loss": 0.058, "step": 11461 }, { "epoch": 0.5326208178438662, "grad_norm": 0.5651856050597355, "learning_rate": 4.726973525531114e-06, "loss": 0.0739, "step": 11462 }, { "epoch": 0.5326672862453532, "grad_norm": 0.7083912209489316, "learning_rate": 4.726222138031715e-06, "loss": 0.0539, "step": 11463 }, { "epoch": 0.5327137546468401, "grad_norm": 0.4858743040088477, "learning_rate": 4.725470756733671e-06, "loss": 0.0515, "step": 11464 }, { "epoch": 0.5327602230483272, "grad_norm": 0.5901311877636023, "learning_rate": 4.724719381654005e-06, "loss": 0.0646, "step": 11465 }, { "epoch": 0.5328066914498141, "grad_norm": 1.586878914578828, "learning_rate": 4.7239680128097334e-06, "loss": 0.053, "step": 11466 }, { "epoch": 0.5328531598513011, "grad_norm": 0.7574433887751387, "learning_rate": 4.723216650217878e-06, "loss": 0.0548, "step": 11467 }, { "epoch": 0.5328996282527881, "grad_norm": 0.5731960296973954, "learning_rate": 4.722465293895457e-06, "loss": 0.0299, "step": 11468 }, { "epoch": 0.5329460966542751, "grad_norm": 2.090435343024099, "learning_rate": 4.721713943859488e-06, "loss": 0.0937, "step": 11469 }, { "epoch": 0.5329925650557621, "grad_norm": 0.6229006070425988, "learning_rate": 4.720962600126992e-06, "loss": 0.0387, "step": 11470 }, { "epoch": 0.5330390334572491, "grad_norm": 0.6643019794270235, "learning_rate": 4.720211262714986e-06, "loss": 0.056, "step": 11471 }, { "epoch": 0.533085501858736, "grad_norm": 0.8724826185287984, "learning_rate": 4.71945993164049e-06, "loss": 0.0508, "step": 11472 }, { "epoch": 0.5331319702602231, "grad_norm": 0.8843359145603668, "learning_rate": 4.718708606920521e-06, "loss": 0.0624, "step": 11473 }, { "epoch": 0.53317843866171, "grad_norm": 0.9148018585718091, "learning_rate": 4.7179572885721e-06, "loss": 0.0699, "step": 11474 }, { "epoch": 0.533224907063197, "grad_norm": 0.8269522135333386, "learning_rate": 4.717205976612242e-06, "loss": 0.0836, "step": 11475 }, { "epoch": 0.5332713754646841, "grad_norm": 0.6942038976281052, "learning_rate": 4.716454671057966e-06, "loss": 0.0593, "step": 11476 }, { "epoch": 0.533317843866171, "grad_norm": 0.38891654177753116, "learning_rate": 4.71570337192629e-06, "loss": 0.0536, "step": 11477 }, { "epoch": 0.533364312267658, "grad_norm": 0.9119511043902764, "learning_rate": 4.714952079234232e-06, "loss": 0.0508, "step": 11478 }, { "epoch": 0.5334107806691449, "grad_norm": 1.2179147473845322, "learning_rate": 4.71420079299881e-06, "loss": 0.0549, "step": 11479 }, { "epoch": 0.533457249070632, "grad_norm": 0.7248246822162452, "learning_rate": 4.713449513237041e-06, "loss": 0.0568, "step": 11480 }, { "epoch": 0.533503717472119, "grad_norm": 0.47280127265484645, "learning_rate": 4.712698239965942e-06, "loss": 0.0469, "step": 11481 }, { "epoch": 0.5335501858736059, "grad_norm": 1.4531594955552414, "learning_rate": 4.7119469732025296e-06, "loss": 0.0617, "step": 11482 }, { "epoch": 0.533596654275093, "grad_norm": 0.4136459335597253, "learning_rate": 4.711195712963821e-06, "loss": 0.0467, "step": 11483 }, { "epoch": 0.5336431226765799, "grad_norm": 0.6592767438309766, "learning_rate": 4.710444459266834e-06, "loss": 0.045, "step": 11484 }, { "epoch": 0.5336895910780669, "grad_norm": 0.6659398694550799, "learning_rate": 4.709693212128584e-06, "loss": 0.0442, "step": 11485 }, { "epoch": 0.5337360594795539, "grad_norm": 0.6418743026574721, "learning_rate": 4.7089419715660906e-06, "loss": 0.0445, "step": 11486 }, { "epoch": 0.5337825278810409, "grad_norm": 0.4663496646696091, "learning_rate": 4.708190737596367e-06, "loss": 0.028, "step": 11487 }, { "epoch": 0.5338289962825279, "grad_norm": 3.0268532434349233, "learning_rate": 4.707439510236429e-06, "loss": 0.0613, "step": 11488 }, { "epoch": 0.5338754646840149, "grad_norm": 0.5309979250658873, "learning_rate": 4.706688289503295e-06, "loss": 0.0436, "step": 11489 }, { "epoch": 0.5339219330855018, "grad_norm": 1.0746164985346631, "learning_rate": 4.705937075413979e-06, "loss": 0.055, "step": 11490 }, { "epoch": 0.5339684014869889, "grad_norm": 1.0586888985396632, "learning_rate": 4.705185867985498e-06, "loss": 0.0496, "step": 11491 }, { "epoch": 0.5340148698884758, "grad_norm": 0.5585084316448129, "learning_rate": 4.70443466723487e-06, "loss": 0.041, "step": 11492 }, { "epoch": 0.5340613382899628, "grad_norm": 0.7926397627648256, "learning_rate": 4.703683473179105e-06, "loss": 0.0779, "step": 11493 }, { "epoch": 0.5341078066914499, "grad_norm": 1.4777927557417199, "learning_rate": 4.7029322858352215e-06, "loss": 0.0287, "step": 11494 }, { "epoch": 0.5341542750929368, "grad_norm": 0.8111537537443891, "learning_rate": 4.7021811052202345e-06, "loss": 0.0764, "step": 11495 }, { "epoch": 0.5342007434944238, "grad_norm": 0.5168452275415901, "learning_rate": 4.701429931351159e-06, "loss": 0.0368, "step": 11496 }, { "epoch": 0.5342472118959107, "grad_norm": 0.7105708545727899, "learning_rate": 4.7006787642450096e-06, "loss": 0.0522, "step": 11497 }, { "epoch": 0.5342936802973978, "grad_norm": 0.5449971521179947, "learning_rate": 4.699927603918803e-06, "loss": 0.063, "step": 11498 }, { "epoch": 0.5343401486988848, "grad_norm": 0.828630219000253, "learning_rate": 4.699176450389551e-06, "loss": 0.0477, "step": 11499 }, { "epoch": 0.5343866171003717, "grad_norm": 1.6118347022643562, "learning_rate": 4.698425303674268e-06, "loss": 0.0675, "step": 11500 }, { "epoch": 0.5344330855018588, "grad_norm": 0.47649921850576604, "learning_rate": 4.6976741637899695e-06, "loss": 0.054, "step": 11501 }, { "epoch": 0.5344795539033457, "grad_norm": 0.6627021708955312, "learning_rate": 4.696923030753669e-06, "loss": 0.0578, "step": 11502 }, { "epoch": 0.5345260223048327, "grad_norm": 1.0457068324491499, "learning_rate": 4.696171904582382e-06, "loss": 0.0502, "step": 11503 }, { "epoch": 0.5345724907063197, "grad_norm": 0.6771347327290539, "learning_rate": 4.695420785293122e-06, "loss": 0.0623, "step": 11504 }, { "epoch": 0.5346189591078067, "grad_norm": 0.8711178124063805, "learning_rate": 4.6946696729029e-06, "loss": 0.0577, "step": 11505 }, { "epoch": 0.5346654275092937, "grad_norm": 0.8149146364375488, "learning_rate": 4.693918567428732e-06, "loss": 0.0526, "step": 11506 }, { "epoch": 0.5347118959107807, "grad_norm": 0.6291345864689379, "learning_rate": 4.693167468887629e-06, "loss": 0.0448, "step": 11507 }, { "epoch": 0.5347583643122676, "grad_norm": 0.7085214735108242, "learning_rate": 4.692416377296608e-06, "loss": 0.0472, "step": 11508 }, { "epoch": 0.5348048327137547, "grad_norm": 1.5651729371813434, "learning_rate": 4.691665292672678e-06, "loss": 0.076, "step": 11509 }, { "epoch": 0.5348513011152416, "grad_norm": 0.3878819047276632, "learning_rate": 4.690914215032856e-06, "loss": 0.0435, "step": 11510 }, { "epoch": 0.5348977695167286, "grad_norm": 2.544594256127219, "learning_rate": 4.690163144394151e-06, "loss": 0.0988, "step": 11511 }, { "epoch": 0.5349442379182157, "grad_norm": 1.729483990821016, "learning_rate": 4.689412080773576e-06, "loss": 0.0785, "step": 11512 }, { "epoch": 0.5349907063197026, "grad_norm": 1.7984228315622879, "learning_rate": 4.688661024188145e-06, "loss": 0.0851, "step": 11513 }, { "epoch": 0.5350371747211896, "grad_norm": 0.8720972665765819, "learning_rate": 4.687909974654869e-06, "loss": 0.0561, "step": 11514 }, { "epoch": 0.5350836431226765, "grad_norm": 0.7569236570485641, "learning_rate": 4.687158932190763e-06, "loss": 0.0527, "step": 11515 }, { "epoch": 0.5351301115241636, "grad_norm": 0.5062810506402292, "learning_rate": 4.686407896812834e-06, "loss": 0.042, "step": 11516 }, { "epoch": 0.5351765799256506, "grad_norm": 0.9463874704131933, "learning_rate": 4.685656868538097e-06, "loss": 0.0645, "step": 11517 }, { "epoch": 0.5352230483271375, "grad_norm": 0.5259514343267647, "learning_rate": 4.6849058473835615e-06, "loss": 0.0349, "step": 11518 }, { "epoch": 0.5352695167286246, "grad_norm": 0.5536945750601239, "learning_rate": 4.684154833366241e-06, "loss": 0.0342, "step": 11519 }, { "epoch": 0.5353159851301115, "grad_norm": 0.4213655495883972, "learning_rate": 4.683403826503146e-06, "loss": 0.0386, "step": 11520 }, { "epoch": 0.5353624535315985, "grad_norm": 0.7117401277773722, "learning_rate": 4.682652826811289e-06, "loss": 0.0569, "step": 11521 }, { "epoch": 0.5354089219330855, "grad_norm": 1.2565958982711782, "learning_rate": 4.681901834307677e-06, "loss": 0.0686, "step": 11522 }, { "epoch": 0.5354553903345725, "grad_norm": 0.6210974995873944, "learning_rate": 4.681150849009325e-06, "loss": 0.0618, "step": 11523 }, { "epoch": 0.5355018587360595, "grad_norm": 0.6053165500031255, "learning_rate": 4.6803998709332404e-06, "loss": 0.0614, "step": 11524 }, { "epoch": 0.5355483271375465, "grad_norm": 0.8995451491072168, "learning_rate": 4.679648900096436e-06, "loss": 0.0888, "step": 11525 }, { "epoch": 0.5355947955390334, "grad_norm": 0.6142913332920449, "learning_rate": 4.6788979365159205e-06, "loss": 0.0637, "step": 11526 }, { "epoch": 0.5356412639405205, "grad_norm": 1.4516223833260489, "learning_rate": 4.678146980208706e-06, "loss": 0.0586, "step": 11527 }, { "epoch": 0.5356877323420074, "grad_norm": 0.36806061223278863, "learning_rate": 4.677396031191801e-06, "loss": 0.0279, "step": 11528 }, { "epoch": 0.5357342007434944, "grad_norm": 0.7199864849253258, "learning_rate": 4.676645089482215e-06, "loss": 0.0449, "step": 11529 }, { "epoch": 0.5357806691449815, "grad_norm": 0.47933482297779456, "learning_rate": 4.675894155096957e-06, "loss": 0.0338, "step": 11530 }, { "epoch": 0.5358271375464684, "grad_norm": 0.7937301505700983, "learning_rate": 4.675143228053038e-06, "loss": 0.0515, "step": 11531 }, { "epoch": 0.5358736059479554, "grad_norm": 0.5760244374488117, "learning_rate": 4.674392308367468e-06, "loss": 0.0504, "step": 11532 }, { "epoch": 0.5359200743494423, "grad_norm": 0.48997875332151575, "learning_rate": 4.6736413960572555e-06, "loss": 0.0457, "step": 11533 }, { "epoch": 0.5359665427509294, "grad_norm": 0.637416665846553, "learning_rate": 4.6728904911394065e-06, "loss": 0.0649, "step": 11534 }, { "epoch": 0.5360130111524164, "grad_norm": 1.1214561813260122, "learning_rate": 4.672139593630933e-06, "loss": 0.0708, "step": 11535 }, { "epoch": 0.5360594795539033, "grad_norm": 0.8466997179094351, "learning_rate": 4.671388703548842e-06, "loss": 0.0669, "step": 11536 }, { "epoch": 0.5361059479553903, "grad_norm": 0.5759103415673446, "learning_rate": 4.670637820910144e-06, "loss": 0.0531, "step": 11537 }, { "epoch": 0.5361524163568773, "grad_norm": 0.5485532309760498, "learning_rate": 4.6698869457318444e-06, "loss": 0.0545, "step": 11538 }, { "epoch": 0.5361988847583643, "grad_norm": 0.7018243203564, "learning_rate": 4.669136078030956e-06, "loss": 0.0529, "step": 11539 }, { "epoch": 0.5362453531598513, "grad_norm": 0.4924616921343135, "learning_rate": 4.668385217824482e-06, "loss": 0.0403, "step": 11540 }, { "epoch": 0.5362918215613383, "grad_norm": 0.6168901209168298, "learning_rate": 4.667634365129431e-06, "loss": 0.0354, "step": 11541 }, { "epoch": 0.5363382899628253, "grad_norm": 0.3920677502671312, "learning_rate": 4.666883519962812e-06, "loss": 0.0345, "step": 11542 }, { "epoch": 0.5363847583643123, "grad_norm": 0.42628040091493585, "learning_rate": 4.666132682341631e-06, "loss": 0.0362, "step": 11543 }, { "epoch": 0.5364312267657992, "grad_norm": 0.8588053479854683, "learning_rate": 4.665381852282897e-06, "loss": 0.067, "step": 11544 }, { "epoch": 0.5364776951672863, "grad_norm": 1.627561298145664, "learning_rate": 4.664631029803618e-06, "loss": 0.0777, "step": 11545 }, { "epoch": 0.5365241635687732, "grad_norm": 0.8555566391335314, "learning_rate": 4.663880214920797e-06, "loss": 0.0629, "step": 11546 }, { "epoch": 0.5365706319702602, "grad_norm": 1.5171837578410474, "learning_rate": 4.663129407651444e-06, "loss": 0.0621, "step": 11547 }, { "epoch": 0.5366171003717473, "grad_norm": 0.9738966704435752, "learning_rate": 4.662378608012562e-06, "loss": 0.0722, "step": 11548 }, { "epoch": 0.5366635687732342, "grad_norm": 1.0255743894881666, "learning_rate": 4.661627816021162e-06, "loss": 0.0726, "step": 11549 }, { "epoch": 0.5367100371747212, "grad_norm": 0.29008976399225456, "learning_rate": 4.660877031694247e-06, "loss": 0.0194, "step": 11550 }, { "epoch": 0.5367565055762081, "grad_norm": 0.6793050123289244, "learning_rate": 4.660126255048826e-06, "loss": 0.042, "step": 11551 }, { "epoch": 0.5368029739776952, "grad_norm": 0.559749495094639, "learning_rate": 4.659375486101902e-06, "loss": 0.0591, "step": 11552 }, { "epoch": 0.5368494423791822, "grad_norm": 0.5527389048761416, "learning_rate": 4.6586247248704814e-06, "loss": 0.0599, "step": 11553 }, { "epoch": 0.5368959107806691, "grad_norm": 1.2343286240932105, "learning_rate": 4.657873971371571e-06, "loss": 0.1021, "step": 11554 }, { "epoch": 0.5369423791821561, "grad_norm": 3.116723083668863, "learning_rate": 4.657123225622173e-06, "loss": 0.077, "step": 11555 }, { "epoch": 0.5369888475836431, "grad_norm": 1.0115144285048145, "learning_rate": 4.656372487639296e-06, "loss": 0.0543, "step": 11556 }, { "epoch": 0.5370353159851301, "grad_norm": 1.461908955219439, "learning_rate": 4.655621757439946e-06, "loss": 0.0709, "step": 11557 }, { "epoch": 0.5370817843866171, "grad_norm": 0.6585459856660305, "learning_rate": 4.654871035041124e-06, "loss": 0.0501, "step": 11558 }, { "epoch": 0.537128252788104, "grad_norm": 1.0001364000621984, "learning_rate": 4.6541203204598365e-06, "loss": 0.0856, "step": 11559 }, { "epoch": 0.5371747211895911, "grad_norm": 0.6207226956239458, "learning_rate": 4.6533696137130865e-06, "loss": 0.0689, "step": 11560 }, { "epoch": 0.5372211895910781, "grad_norm": 1.5329233979836305, "learning_rate": 4.65261891481788e-06, "loss": 0.0681, "step": 11561 }, { "epoch": 0.537267657992565, "grad_norm": 1.3035647244572144, "learning_rate": 4.651868223791221e-06, "loss": 0.0474, "step": 11562 }, { "epoch": 0.5373141263940521, "grad_norm": 0.609306252602144, "learning_rate": 4.651117540650115e-06, "loss": 0.04, "step": 11563 }, { "epoch": 0.537360594795539, "grad_norm": 0.9480258537611381, "learning_rate": 4.650366865411562e-06, "loss": 0.0523, "step": 11564 }, { "epoch": 0.537407063197026, "grad_norm": 0.8675441164783323, "learning_rate": 4.649616198092567e-06, "loss": 0.0599, "step": 11565 }, { "epoch": 0.5374535315985131, "grad_norm": 1.4502949774273817, "learning_rate": 4.648865538710134e-06, "loss": 0.0636, "step": 11566 }, { "epoch": 0.5375, "grad_norm": 2.034741701367675, "learning_rate": 4.6481148872812665e-06, "loss": 0.0697, "step": 11567 }, { "epoch": 0.537546468401487, "grad_norm": 0.7245525102708512, "learning_rate": 4.647364243822967e-06, "loss": 0.0354, "step": 11568 }, { "epoch": 0.5375929368029739, "grad_norm": 1.027087848857151, "learning_rate": 4.646613608352241e-06, "loss": 0.0581, "step": 11569 }, { "epoch": 0.537639405204461, "grad_norm": 0.5872593685487839, "learning_rate": 4.645862980886086e-06, "loss": 0.0558, "step": 11570 }, { "epoch": 0.537685873605948, "grad_norm": 0.789940401280661, "learning_rate": 4.6451123614415075e-06, "loss": 0.0545, "step": 11571 }, { "epoch": 0.5377323420074349, "grad_norm": 0.4322697994155455, "learning_rate": 4.644361750035508e-06, "loss": 0.0424, "step": 11572 }, { "epoch": 0.537778810408922, "grad_norm": 0.5738170374088225, "learning_rate": 4.643611146685087e-06, "loss": 0.0472, "step": 11573 }, { "epoch": 0.5378252788104089, "grad_norm": 1.7206759435546526, "learning_rate": 4.642860551407254e-06, "loss": 0.0692, "step": 11574 }, { "epoch": 0.5378717472118959, "grad_norm": 0.5618385015657197, "learning_rate": 4.642109964219001e-06, "loss": 0.0426, "step": 11575 }, { "epoch": 0.5379182156133829, "grad_norm": 1.5892560816939003, "learning_rate": 4.641359385137335e-06, "loss": 0.0674, "step": 11576 }, { "epoch": 0.5379646840148699, "grad_norm": 0.6162100097243443, "learning_rate": 4.640608814179255e-06, "loss": 0.0483, "step": 11577 }, { "epoch": 0.5380111524163569, "grad_norm": 0.5510740465798308, "learning_rate": 4.639858251361765e-06, "loss": 0.0344, "step": 11578 }, { "epoch": 0.5380576208178439, "grad_norm": 0.6991147942009075, "learning_rate": 4.639107696701864e-06, "loss": 0.069, "step": 11579 }, { "epoch": 0.5381040892193308, "grad_norm": 0.6205702996196812, "learning_rate": 4.638357150216556e-06, "loss": 0.0392, "step": 11580 }, { "epoch": 0.5381505576208179, "grad_norm": 0.8913926580030853, "learning_rate": 4.637606611922837e-06, "loss": 0.0829, "step": 11581 }, { "epoch": 0.5381970260223048, "grad_norm": 0.7021349906511983, "learning_rate": 4.6368560818377095e-06, "loss": 0.0758, "step": 11582 }, { "epoch": 0.5382434944237918, "grad_norm": 1.3557140875806404, "learning_rate": 4.636105559978174e-06, "loss": 0.043, "step": 11583 }, { "epoch": 0.5382899628252789, "grad_norm": 0.7970162861005798, "learning_rate": 4.635355046361231e-06, "loss": 0.0412, "step": 11584 }, { "epoch": 0.5383364312267658, "grad_norm": 0.431591997607083, "learning_rate": 4.634604541003879e-06, "loss": 0.0442, "step": 11585 }, { "epoch": 0.5383828996282528, "grad_norm": 0.7843644983968605, "learning_rate": 4.633854043923122e-06, "loss": 0.0527, "step": 11586 }, { "epoch": 0.5384293680297397, "grad_norm": 0.6338580330558581, "learning_rate": 4.633103555135953e-06, "loss": 0.0384, "step": 11587 }, { "epoch": 0.5384758364312268, "grad_norm": 0.9274582051520945, "learning_rate": 4.632353074659375e-06, "loss": 0.0396, "step": 11588 }, { "epoch": 0.5385223048327138, "grad_norm": 1.2513720185197879, "learning_rate": 4.631602602510388e-06, "loss": 0.0665, "step": 11589 }, { "epoch": 0.5385687732342007, "grad_norm": 0.8840399433989067, "learning_rate": 4.630852138705988e-06, "loss": 0.0764, "step": 11590 }, { "epoch": 0.5386152416356877, "grad_norm": 0.7896729695589785, "learning_rate": 4.630101683263177e-06, "loss": 0.0633, "step": 11591 }, { "epoch": 0.5386617100371747, "grad_norm": 0.8871604162341802, "learning_rate": 4.6293512361989534e-06, "loss": 0.0634, "step": 11592 }, { "epoch": 0.5387081784386617, "grad_norm": 0.7240585553363221, "learning_rate": 4.628600797530312e-06, "loss": 0.0513, "step": 11593 }, { "epoch": 0.5387546468401487, "grad_norm": 0.6858984412698065, "learning_rate": 4.627850367274255e-06, "loss": 0.0358, "step": 11594 }, { "epoch": 0.5388011152416357, "grad_norm": 3.4452676655571066, "learning_rate": 4.6270999454477774e-06, "loss": 0.0811, "step": 11595 }, { "epoch": 0.5388475836431227, "grad_norm": 0.7919732426322966, "learning_rate": 4.626349532067879e-06, "loss": 0.0821, "step": 11596 }, { "epoch": 0.5388940520446097, "grad_norm": 1.0321761311424502, "learning_rate": 4.625599127151556e-06, "loss": 0.063, "step": 11597 }, { "epoch": 0.5389405204460966, "grad_norm": 0.8454069529737229, "learning_rate": 4.624848730715811e-06, "loss": 0.0507, "step": 11598 }, { "epoch": 0.5389869888475837, "grad_norm": 1.1222752308907358, "learning_rate": 4.6240983427776335e-06, "loss": 0.0451, "step": 11599 }, { "epoch": 0.5390334572490706, "grad_norm": 0.9259170562131752, "learning_rate": 4.623347963354024e-06, "loss": 0.0476, "step": 11600 }, { "epoch": 0.5390799256505576, "grad_norm": 0.48883945633897896, "learning_rate": 4.62259759246198e-06, "loss": 0.0357, "step": 11601 }, { "epoch": 0.5391263940520447, "grad_norm": 0.5889155443461574, "learning_rate": 4.621847230118498e-06, "loss": 0.0481, "step": 11602 }, { "epoch": 0.5391728624535316, "grad_norm": 0.6708404104276371, "learning_rate": 4.621096876340575e-06, "loss": 0.0537, "step": 11603 }, { "epoch": 0.5392193308550186, "grad_norm": 1.0801948233854561, "learning_rate": 4.620346531145207e-06, "loss": 0.0529, "step": 11604 }, { "epoch": 0.5392657992565055, "grad_norm": 0.6616872852022181, "learning_rate": 4.6195961945493874e-06, "loss": 0.0505, "step": 11605 }, { "epoch": 0.5393122676579926, "grad_norm": 0.5164871467563344, "learning_rate": 4.618845866570116e-06, "loss": 0.0336, "step": 11606 }, { "epoch": 0.5393587360594796, "grad_norm": 0.8493100281169191, "learning_rate": 4.618095547224386e-06, "loss": 0.0583, "step": 11607 }, { "epoch": 0.5394052044609665, "grad_norm": 0.6982898552983257, "learning_rate": 4.617345236529194e-06, "loss": 0.0488, "step": 11608 }, { "epoch": 0.5394516728624535, "grad_norm": 0.35041998348552766, "learning_rate": 4.616594934501535e-06, "loss": 0.0465, "step": 11609 }, { "epoch": 0.5394981412639405, "grad_norm": 1.1594848234956356, "learning_rate": 4.615844641158405e-06, "loss": 0.0636, "step": 11610 }, { "epoch": 0.5395446096654275, "grad_norm": 1.2484811537657317, "learning_rate": 4.615094356516798e-06, "loss": 0.0685, "step": 11611 }, { "epoch": 0.5395910780669145, "grad_norm": 0.7541539427734169, "learning_rate": 4.614344080593708e-06, "loss": 0.0643, "step": 11612 }, { "epoch": 0.5396375464684015, "grad_norm": 0.8115762061657859, "learning_rate": 4.613593813406132e-06, "loss": 0.0713, "step": 11613 }, { "epoch": 0.5396840148698885, "grad_norm": 0.5764139963555043, "learning_rate": 4.61284355497106e-06, "loss": 0.0401, "step": 11614 }, { "epoch": 0.5397304832713755, "grad_norm": 0.6543703010512295, "learning_rate": 4.612093305305492e-06, "loss": 0.078, "step": 11615 }, { "epoch": 0.5397769516728624, "grad_norm": 0.7474303226107507, "learning_rate": 4.611343064426419e-06, "loss": 0.0768, "step": 11616 }, { "epoch": 0.5398234200743495, "grad_norm": 0.4676879024840621, "learning_rate": 4.610592832350832e-06, "loss": 0.045, "step": 11617 }, { "epoch": 0.5398698884758364, "grad_norm": 0.7354082136608171, "learning_rate": 4.609842609095729e-06, "loss": 0.0425, "step": 11618 }, { "epoch": 0.5399163568773234, "grad_norm": 2.9182765303963785, "learning_rate": 4.6090923946781005e-06, "loss": 0.0724, "step": 11619 }, { "epoch": 0.5399628252788105, "grad_norm": 1.0139574525096242, "learning_rate": 4.608342189114942e-06, "loss": 0.0632, "step": 11620 }, { "epoch": 0.5400092936802974, "grad_norm": 2.6391437762103007, "learning_rate": 4.607591992423244e-06, "loss": 0.0867, "step": 11621 }, { "epoch": 0.5400557620817844, "grad_norm": 0.5410054402403699, "learning_rate": 4.606841804620002e-06, "loss": 0.0557, "step": 11622 }, { "epoch": 0.5401022304832713, "grad_norm": 0.6542149409746413, "learning_rate": 4.606091625722206e-06, "loss": 0.0626, "step": 11623 }, { "epoch": 0.5401486988847584, "grad_norm": 1.2128948219035502, "learning_rate": 4.605341455746848e-06, "loss": 0.0499, "step": 11624 }, { "epoch": 0.5401951672862454, "grad_norm": 1.148696594585447, "learning_rate": 4.604591294710923e-06, "loss": 0.0629, "step": 11625 }, { "epoch": 0.5402416356877323, "grad_norm": 0.5600347408214233, "learning_rate": 4.60384114263142e-06, "loss": 0.0495, "step": 11626 }, { "epoch": 0.5402881040892193, "grad_norm": 0.9469261495029845, "learning_rate": 4.603090999525333e-06, "loss": 0.0664, "step": 11627 }, { "epoch": 0.5403345724907063, "grad_norm": 0.6917930725566462, "learning_rate": 4.602340865409654e-06, "loss": 0.0443, "step": 11628 }, { "epoch": 0.5403810408921933, "grad_norm": 0.7391135922831571, "learning_rate": 4.60159074030137e-06, "loss": 0.047, "step": 11629 }, { "epoch": 0.5404275092936803, "grad_norm": 0.5070113090066559, "learning_rate": 4.6008406242174774e-06, "loss": 0.0488, "step": 11630 }, { "epoch": 0.5404739776951673, "grad_norm": 2.2494221401188765, "learning_rate": 4.600090517174963e-06, "loss": 0.0686, "step": 11631 }, { "epoch": 0.5405204460966543, "grad_norm": 0.5803463066281688, "learning_rate": 4.59934041919082e-06, "loss": 0.0387, "step": 11632 }, { "epoch": 0.5405669144981413, "grad_norm": 0.419576942934485, "learning_rate": 4.59859033028204e-06, "loss": 0.0408, "step": 11633 }, { "epoch": 0.5406133828996282, "grad_norm": 0.6906571422709785, "learning_rate": 4.597840250465609e-06, "loss": 0.0486, "step": 11634 }, { "epoch": 0.5406598513011153, "grad_norm": 1.6004829225641857, "learning_rate": 4.59709017975852e-06, "loss": 0.0609, "step": 11635 }, { "epoch": 0.5407063197026022, "grad_norm": 0.7810565139566558, "learning_rate": 4.596340118177762e-06, "loss": 0.0701, "step": 11636 }, { "epoch": 0.5407527881040892, "grad_norm": 1.3722052805641451, "learning_rate": 4.595590065740327e-06, "loss": 0.0801, "step": 11637 }, { "epoch": 0.5407992565055763, "grad_norm": 0.7941376208899954, "learning_rate": 4.594840022463201e-06, "loss": 0.059, "step": 11638 }, { "epoch": 0.5408457249070632, "grad_norm": 0.7301010658843774, "learning_rate": 4.594089988363377e-06, "loss": 0.0562, "step": 11639 }, { "epoch": 0.5408921933085502, "grad_norm": 1.8071610080101213, "learning_rate": 4.59333996345784e-06, "loss": 0.0721, "step": 11640 }, { "epoch": 0.5409386617100371, "grad_norm": 0.6468746961987727, "learning_rate": 4.592589947763581e-06, "loss": 0.0426, "step": 11641 }, { "epoch": 0.5409851301115242, "grad_norm": 1.0887972080904569, "learning_rate": 4.591839941297589e-06, "loss": 0.0524, "step": 11642 }, { "epoch": 0.5410315985130112, "grad_norm": 0.7205314362551299, "learning_rate": 4.591089944076851e-06, "loss": 0.0475, "step": 11643 }, { "epoch": 0.5410780669144981, "grad_norm": 0.35335914955582987, "learning_rate": 4.590339956118357e-06, "loss": 0.0298, "step": 11644 }, { "epoch": 0.5411245353159851, "grad_norm": 0.8237884178586684, "learning_rate": 4.5895899774390956e-06, "loss": 0.052, "step": 11645 }, { "epoch": 0.5411710037174721, "grad_norm": 0.6987781912411516, "learning_rate": 4.588840008056051e-06, "loss": 0.0699, "step": 11646 }, { "epoch": 0.5412174721189591, "grad_norm": 0.5326650576414313, "learning_rate": 4.588090047986214e-06, "loss": 0.0507, "step": 11647 }, { "epoch": 0.5412639405204461, "grad_norm": 0.42996454699598086, "learning_rate": 4.587340097246569e-06, "loss": 0.0401, "step": 11648 }, { "epoch": 0.541310408921933, "grad_norm": 1.2316001684606894, "learning_rate": 4.586590155854107e-06, "loss": 0.0551, "step": 11649 }, { "epoch": 0.5413568773234201, "grad_norm": 0.7843119752555309, "learning_rate": 4.585840223825812e-06, "loss": 0.051, "step": 11650 }, { "epoch": 0.5414033457249071, "grad_norm": 1.0961742248616442, "learning_rate": 4.585090301178674e-06, "loss": 0.0586, "step": 11651 }, { "epoch": 0.541449814126394, "grad_norm": 0.7878211851877426, "learning_rate": 4.584340387929676e-06, "loss": 0.0689, "step": 11652 }, { "epoch": 0.5414962825278811, "grad_norm": 0.5817968505626848, "learning_rate": 4.583590484095805e-06, "loss": 0.0449, "step": 11653 }, { "epoch": 0.541542750929368, "grad_norm": 0.9540420876280628, "learning_rate": 4.582840589694049e-06, "loss": 0.092, "step": 11654 }, { "epoch": 0.541589219330855, "grad_norm": 0.3969572942453808, "learning_rate": 4.582090704741391e-06, "loss": 0.0575, "step": 11655 }, { "epoch": 0.541635687732342, "grad_norm": 0.9472813266231498, "learning_rate": 4.581340829254819e-06, "loss": 0.0702, "step": 11656 }, { "epoch": 0.541682156133829, "grad_norm": 1.2863104215669197, "learning_rate": 4.58059096325132e-06, "loss": 0.059, "step": 11657 }, { "epoch": 0.541728624535316, "grad_norm": 0.6675991774842479, "learning_rate": 4.579841106747875e-06, "loss": 0.0587, "step": 11658 }, { "epoch": 0.5417750929368029, "grad_norm": 0.4906839492947495, "learning_rate": 4.579091259761472e-06, "loss": 0.0479, "step": 11659 }, { "epoch": 0.54182156133829, "grad_norm": 0.5509072600223023, "learning_rate": 4.578341422309093e-06, "loss": 0.0495, "step": 11660 }, { "epoch": 0.541868029739777, "grad_norm": 0.5012478074054131, "learning_rate": 4.577591594407727e-06, "loss": 0.0357, "step": 11661 }, { "epoch": 0.5419144981412639, "grad_norm": 0.6089153467385983, "learning_rate": 4.576841776074354e-06, "loss": 0.0552, "step": 11662 }, { "epoch": 0.5419609665427509, "grad_norm": 1.1430191283927837, "learning_rate": 4.576091967325962e-06, "loss": 0.082, "step": 11663 }, { "epoch": 0.5420074349442379, "grad_norm": 0.9285819344705573, "learning_rate": 4.5753421681795326e-06, "loss": 0.074, "step": 11664 }, { "epoch": 0.5420539033457249, "grad_norm": 0.6152783253994595, "learning_rate": 4.57459237865205e-06, "loss": 0.0485, "step": 11665 }, { "epoch": 0.5421003717472119, "grad_norm": 1.136633003456707, "learning_rate": 4.573842598760497e-06, "loss": 0.0716, "step": 11666 }, { "epoch": 0.5421468401486988, "grad_norm": 1.0812241483714213, "learning_rate": 4.573092828521857e-06, "loss": 0.0611, "step": 11667 }, { "epoch": 0.5421933085501859, "grad_norm": 1.1144428986444563, "learning_rate": 4.572343067953115e-06, "loss": 0.0826, "step": 11668 }, { "epoch": 0.5422397769516729, "grad_norm": 0.6780474527076205, "learning_rate": 4.571593317071255e-06, "loss": 0.0452, "step": 11669 }, { "epoch": 0.5422862453531598, "grad_norm": 0.684107459375677, "learning_rate": 4.5708435758932535e-06, "loss": 0.0605, "step": 11670 }, { "epoch": 0.5423327137546469, "grad_norm": 0.8562641735043502, "learning_rate": 4.570093844436098e-06, "loss": 0.0487, "step": 11671 }, { "epoch": 0.5423791821561338, "grad_norm": 0.8865095805533016, "learning_rate": 4.569344122716769e-06, "loss": 0.0674, "step": 11672 }, { "epoch": 0.5424256505576208, "grad_norm": 0.6292959066732391, "learning_rate": 4.568594410752249e-06, "loss": 0.0655, "step": 11673 }, { "epoch": 0.5424721189591079, "grad_norm": 0.5027931355100573, "learning_rate": 4.567844708559519e-06, "loss": 0.0461, "step": 11674 }, { "epoch": 0.5425185873605948, "grad_norm": 0.991812461749017, "learning_rate": 4.567095016155563e-06, "loss": 0.0664, "step": 11675 }, { "epoch": 0.5425650557620818, "grad_norm": 0.9381585199353661, "learning_rate": 4.56634533355736e-06, "loss": 0.0489, "step": 11676 }, { "epoch": 0.5426115241635687, "grad_norm": 0.35256339366476114, "learning_rate": 4.565595660781891e-06, "loss": 0.0321, "step": 11677 }, { "epoch": 0.5426579925650558, "grad_norm": 1.1773924150145767, "learning_rate": 4.564845997846137e-06, "loss": 0.0712, "step": 11678 }, { "epoch": 0.5427044609665428, "grad_norm": 1.717578387233858, "learning_rate": 4.564096344767079e-06, "loss": 0.0532, "step": 11679 }, { "epoch": 0.5427509293680297, "grad_norm": 0.5683720363169983, "learning_rate": 4.563346701561699e-06, "loss": 0.0678, "step": 11680 }, { "epoch": 0.5427973977695167, "grad_norm": 0.45748336863705735, "learning_rate": 4.562597068246976e-06, "loss": 0.0562, "step": 11681 }, { "epoch": 0.5428438661710037, "grad_norm": 0.4342364105532887, "learning_rate": 4.561847444839889e-06, "loss": 0.0576, "step": 11682 }, { "epoch": 0.5428903345724907, "grad_norm": 0.6867689251939483, "learning_rate": 4.561097831357418e-06, "loss": 0.05, "step": 11683 }, { "epoch": 0.5429368029739777, "grad_norm": 0.7024912180505175, "learning_rate": 4.560348227816543e-06, "loss": 0.0761, "step": 11684 }, { "epoch": 0.5429832713754646, "grad_norm": 0.4266283217957701, "learning_rate": 4.559598634234244e-06, "loss": 0.0441, "step": 11685 }, { "epoch": 0.5430297397769517, "grad_norm": 0.5984594059667325, "learning_rate": 4.558849050627498e-06, "loss": 0.0647, "step": 11686 }, { "epoch": 0.5430762081784387, "grad_norm": 0.7334176792118958, "learning_rate": 4.558099477013288e-06, "loss": 0.0364, "step": 11687 }, { "epoch": 0.5431226765799256, "grad_norm": 0.6446178811084011, "learning_rate": 4.557349913408589e-06, "loss": 0.0323, "step": 11688 }, { "epoch": 0.5431691449814127, "grad_norm": 0.39655122201307275, "learning_rate": 4.556600359830379e-06, "loss": 0.0381, "step": 11689 }, { "epoch": 0.5432156133828996, "grad_norm": 3.5203246068448313, "learning_rate": 4.555850816295639e-06, "loss": 0.0672, "step": 11690 }, { "epoch": 0.5432620817843866, "grad_norm": 0.3937510089706544, "learning_rate": 4.555101282821343e-06, "loss": 0.0382, "step": 11691 }, { "epoch": 0.5433085501858737, "grad_norm": 1.3749599955976348, "learning_rate": 4.554351759424475e-06, "loss": 0.0553, "step": 11692 }, { "epoch": 0.5433550185873606, "grad_norm": 1.1515141622512526, "learning_rate": 4.553602246122007e-06, "loss": 0.0885, "step": 11693 }, { "epoch": 0.5434014869888476, "grad_norm": 1.1573536067102497, "learning_rate": 4.552852742930917e-06, "loss": 0.0674, "step": 11694 }, { "epoch": 0.5434479553903345, "grad_norm": 1.4275975331760917, "learning_rate": 4.552103249868184e-06, "loss": 0.0749, "step": 11695 }, { "epoch": 0.5434944237918216, "grad_norm": 0.58005542937607, "learning_rate": 4.551353766950783e-06, "loss": 0.0626, "step": 11696 }, { "epoch": 0.5435408921933086, "grad_norm": 2.0037841857904373, "learning_rate": 4.5506042941956915e-06, "loss": 0.0513, "step": 11697 }, { "epoch": 0.5435873605947955, "grad_norm": 0.3592363895343585, "learning_rate": 4.549854831619888e-06, "loss": 0.0287, "step": 11698 }, { "epoch": 0.5436338289962825, "grad_norm": 1.627727290896326, "learning_rate": 4.549105379240343e-06, "loss": 0.0669, "step": 11699 }, { "epoch": 0.5436802973977695, "grad_norm": 0.766554006460686, "learning_rate": 4.5483559370740375e-06, "loss": 0.0629, "step": 11700 }, { "epoch": 0.5437267657992565, "grad_norm": 0.34087046853773334, "learning_rate": 4.547606505137944e-06, "loss": 0.0414, "step": 11701 }, { "epoch": 0.5437732342007435, "grad_norm": 0.524887689381396, "learning_rate": 4.546857083449041e-06, "loss": 0.06, "step": 11702 }, { "epoch": 0.5438197026022304, "grad_norm": 0.46266314631388034, "learning_rate": 4.5461076720243e-06, "loss": 0.0602, "step": 11703 }, { "epoch": 0.5438661710037175, "grad_norm": 0.8310626957441598, "learning_rate": 4.545358270880701e-06, "loss": 0.0774, "step": 11704 }, { "epoch": 0.5439126394052045, "grad_norm": 0.645217033880052, "learning_rate": 4.544608880035214e-06, "loss": 0.0544, "step": 11705 }, { "epoch": 0.5439591078066914, "grad_norm": 0.748075239905349, "learning_rate": 4.543859499504814e-06, "loss": 0.0567, "step": 11706 }, { "epoch": 0.5440055762081785, "grad_norm": 0.5880724565238207, "learning_rate": 4.543110129306478e-06, "loss": 0.036, "step": 11707 }, { "epoch": 0.5440520446096654, "grad_norm": 0.5560380531699975, "learning_rate": 4.542360769457177e-06, "loss": 0.0487, "step": 11708 }, { "epoch": 0.5440985130111524, "grad_norm": 0.6792043488107431, "learning_rate": 4.541611419973888e-06, "loss": 0.0516, "step": 11709 }, { "epoch": 0.5441449814126395, "grad_norm": 0.6324096467454446, "learning_rate": 4.540862080873583e-06, "loss": 0.0295, "step": 11710 }, { "epoch": 0.5441914498141264, "grad_norm": 0.6684437031804914, "learning_rate": 4.540112752173232e-06, "loss": 0.0618, "step": 11711 }, { "epoch": 0.5442379182156134, "grad_norm": 0.8231909992819952, "learning_rate": 4.539363433889814e-06, "loss": 0.0879, "step": 11712 }, { "epoch": 0.5442843866171003, "grad_norm": 2.5277528844019947, "learning_rate": 4.5386141260402974e-06, "loss": 0.0571, "step": 11713 }, { "epoch": 0.5443308550185874, "grad_norm": 0.7861370447950711, "learning_rate": 4.537864828641658e-06, "loss": 0.0656, "step": 11714 }, { "epoch": 0.5443773234200744, "grad_norm": 0.4769622710018042, "learning_rate": 4.5371155417108644e-06, "loss": 0.0626, "step": 11715 }, { "epoch": 0.5444237918215613, "grad_norm": 0.45340238228402385, "learning_rate": 4.5363662652648945e-06, "loss": 0.0353, "step": 11716 }, { "epoch": 0.5444702602230483, "grad_norm": 0.3946658097620499, "learning_rate": 4.535616999320714e-06, "loss": 0.0438, "step": 11717 }, { "epoch": 0.5445167286245353, "grad_norm": 0.8421391800839998, "learning_rate": 4.534867743895298e-06, "loss": 0.0621, "step": 11718 }, { "epoch": 0.5445631970260223, "grad_norm": 0.6789896294837928, "learning_rate": 4.5341184990056165e-06, "loss": 0.0446, "step": 11719 }, { "epoch": 0.5446096654275093, "grad_norm": 1.0807815422780729, "learning_rate": 4.533369264668641e-06, "loss": 0.0573, "step": 11720 }, { "epoch": 0.5446561338289962, "grad_norm": 0.6853916084679001, "learning_rate": 4.532620040901344e-06, "loss": 0.0671, "step": 11721 }, { "epoch": 0.5447026022304833, "grad_norm": 0.6846978539235027, "learning_rate": 4.531870827720695e-06, "loss": 0.0743, "step": 11722 }, { "epoch": 0.5447490706319703, "grad_norm": 0.8616478085223259, "learning_rate": 4.531121625143664e-06, "loss": 0.0489, "step": 11723 }, { "epoch": 0.5447955390334572, "grad_norm": 1.5315976791344528, "learning_rate": 4.5303724331872215e-06, "loss": 0.0748, "step": 11724 }, { "epoch": 0.5448420074349443, "grad_norm": 1.2317998563580315, "learning_rate": 4.529623251868338e-06, "loss": 0.0709, "step": 11725 }, { "epoch": 0.5448884758364312, "grad_norm": 1.0355748908802977, "learning_rate": 4.528874081203982e-06, "loss": 0.0492, "step": 11726 }, { "epoch": 0.5449349442379182, "grad_norm": 0.6698729869347161, "learning_rate": 4.528124921211125e-06, "loss": 0.0746, "step": 11727 }, { "epoch": 0.5449814126394052, "grad_norm": 0.5007340751159314, "learning_rate": 4.527375771906736e-06, "loss": 0.0403, "step": 11728 }, { "epoch": 0.5450278810408922, "grad_norm": 0.8365096449378048, "learning_rate": 4.526626633307782e-06, "loss": 0.0651, "step": 11729 }, { "epoch": 0.5450743494423792, "grad_norm": 0.4906586213175105, "learning_rate": 4.525877505431232e-06, "loss": 0.0575, "step": 11730 }, { "epoch": 0.5451208178438661, "grad_norm": 0.5482094907297027, "learning_rate": 4.525128388294057e-06, "loss": 0.0537, "step": 11731 }, { "epoch": 0.5451672862453532, "grad_norm": 0.7618168675741244, "learning_rate": 4.524379281913222e-06, "loss": 0.0582, "step": 11732 }, { "epoch": 0.5452137546468402, "grad_norm": 0.8851591372975285, "learning_rate": 4.523630186305699e-06, "loss": 0.0693, "step": 11733 }, { "epoch": 0.5452602230483271, "grad_norm": 0.8188160091317872, "learning_rate": 4.522881101488453e-06, "loss": 0.0415, "step": 11734 }, { "epoch": 0.5453066914498141, "grad_norm": 0.5885017938535775, "learning_rate": 4.522132027478452e-06, "loss": 0.0625, "step": 11735 }, { "epoch": 0.5453531598513011, "grad_norm": 1.0117260930520442, "learning_rate": 4.5213829642926635e-06, "loss": 0.0697, "step": 11736 }, { "epoch": 0.5453996282527881, "grad_norm": 1.0106692750112156, "learning_rate": 4.520633911948054e-06, "loss": 0.0546, "step": 11737 }, { "epoch": 0.5454460966542751, "grad_norm": 0.840614813626198, "learning_rate": 4.5198848704615915e-06, "loss": 0.0708, "step": 11738 }, { "epoch": 0.545492565055762, "grad_norm": 0.8417782516416018, "learning_rate": 4.519135839850241e-06, "loss": 0.0612, "step": 11739 }, { "epoch": 0.5455390334572491, "grad_norm": 0.6669032412244709, "learning_rate": 4.518386820130972e-06, "loss": 0.0603, "step": 11740 }, { "epoch": 0.5455855018587361, "grad_norm": 0.4250365868714231, "learning_rate": 4.517637811320748e-06, "loss": 0.0368, "step": 11741 }, { "epoch": 0.545631970260223, "grad_norm": 0.61737726881391, "learning_rate": 4.5168888134365335e-06, "loss": 0.0518, "step": 11742 }, { "epoch": 0.5456784386617101, "grad_norm": 0.29747560813113766, "learning_rate": 4.5161398264952975e-06, "loss": 0.0512, "step": 11743 }, { "epoch": 0.545724907063197, "grad_norm": 0.6472619957712943, "learning_rate": 4.515390850514002e-06, "loss": 0.0708, "step": 11744 }, { "epoch": 0.545771375464684, "grad_norm": 0.6466633312835723, "learning_rate": 4.514641885509615e-06, "loss": 0.0577, "step": 11745 }, { "epoch": 0.545817843866171, "grad_norm": 0.7047239127320553, "learning_rate": 4.5138929314991014e-06, "loss": 0.0577, "step": 11746 }, { "epoch": 0.545864312267658, "grad_norm": 1.7855819884018886, "learning_rate": 4.5131439884994224e-06, "loss": 0.1046, "step": 11747 }, { "epoch": 0.545910780669145, "grad_norm": 0.546011497958122, "learning_rate": 4.512395056527546e-06, "loss": 0.0583, "step": 11748 }, { "epoch": 0.5459572490706319, "grad_norm": 0.7237467588161377, "learning_rate": 4.511646135600432e-06, "loss": 0.0579, "step": 11749 }, { "epoch": 0.546003717472119, "grad_norm": 2.0466775542214184, "learning_rate": 4.51089722573505e-06, "loss": 0.0328, "step": 11750 }, { "epoch": 0.546050185873606, "grad_norm": 0.4728027340699008, "learning_rate": 4.5101483269483605e-06, "loss": 0.0368, "step": 11751 }, { "epoch": 0.5460966542750929, "grad_norm": 0.7161687908504815, "learning_rate": 4.509399439257325e-06, "loss": 0.0632, "step": 11752 }, { "epoch": 0.5461431226765799, "grad_norm": 0.42964043081298925, "learning_rate": 4.50865056267891e-06, "loss": 0.0483, "step": 11753 }, { "epoch": 0.5461895910780669, "grad_norm": 0.6863284287967153, "learning_rate": 4.5079016972300755e-06, "loss": 0.064, "step": 11754 }, { "epoch": 0.5462360594795539, "grad_norm": 0.6339086580174506, "learning_rate": 4.507152842927787e-06, "loss": 0.0519, "step": 11755 }, { "epoch": 0.5462825278810409, "grad_norm": 0.3870479649380913, "learning_rate": 4.506403999789004e-06, "loss": 0.0486, "step": 11756 }, { "epoch": 0.5463289962825278, "grad_norm": 0.39808342778380207, "learning_rate": 4.5056551678306916e-06, "loss": 0.03, "step": 11757 }, { "epoch": 0.5463754646840149, "grad_norm": 0.7447832580759907, "learning_rate": 4.504906347069809e-06, "loss": 0.0502, "step": 11758 }, { "epoch": 0.5464219330855019, "grad_norm": 0.8367904345123043, "learning_rate": 4.504157537523318e-06, "loss": 0.0582, "step": 11759 }, { "epoch": 0.5464684014869888, "grad_norm": 1.22822016759648, "learning_rate": 4.503408739208182e-06, "loss": 0.0808, "step": 11760 }, { "epoch": 0.5465148698884759, "grad_norm": 0.7305690917143997, "learning_rate": 4.5026599521413585e-06, "loss": 0.0609, "step": 11761 }, { "epoch": 0.5465613382899628, "grad_norm": 0.5388784277126043, "learning_rate": 4.501911176339813e-06, "loss": 0.051, "step": 11762 }, { "epoch": 0.5466078066914498, "grad_norm": 1.914982387626506, "learning_rate": 4.501162411820503e-06, "loss": 0.0696, "step": 11763 }, { "epoch": 0.5466542750929368, "grad_norm": 1.3885018171517534, "learning_rate": 4.500413658600389e-06, "loss": 0.0788, "step": 11764 }, { "epoch": 0.5467007434944238, "grad_norm": 0.4854364659099657, "learning_rate": 4.499664916696431e-06, "loss": 0.0591, "step": 11765 }, { "epoch": 0.5467472118959108, "grad_norm": 0.6816929638244619, "learning_rate": 4.498916186125589e-06, "loss": 0.0518, "step": 11766 }, { "epoch": 0.5467936802973977, "grad_norm": 0.6252508811704243, "learning_rate": 4.498167466904823e-06, "loss": 0.0598, "step": 11767 }, { "epoch": 0.5468401486988848, "grad_norm": 0.9708345216373778, "learning_rate": 4.4974187590510925e-06, "loss": 0.0371, "step": 11768 }, { "epoch": 0.5468866171003718, "grad_norm": 0.7046265437412783, "learning_rate": 4.496670062581357e-06, "loss": 0.0685, "step": 11769 }, { "epoch": 0.5469330855018587, "grad_norm": 0.5741236234801277, "learning_rate": 4.495921377512574e-06, "loss": 0.0573, "step": 11770 }, { "epoch": 0.5469795539033457, "grad_norm": 0.8079862837008669, "learning_rate": 4.495172703861701e-06, "loss": 0.0406, "step": 11771 }, { "epoch": 0.5470260223048327, "grad_norm": 0.6926701755968978, "learning_rate": 4.494424041645698e-06, "loss": 0.059, "step": 11772 }, { "epoch": 0.5470724907063197, "grad_norm": 1.0953516477506298, "learning_rate": 4.493675390881523e-06, "loss": 0.0783, "step": 11773 }, { "epoch": 0.5471189591078067, "grad_norm": 0.2170431228977774, "learning_rate": 4.492926751586133e-06, "loss": 0.0328, "step": 11774 }, { "epoch": 0.5471654275092936, "grad_norm": 0.8340687935499274, "learning_rate": 4.492178123776487e-06, "loss": 0.0643, "step": 11775 }, { "epoch": 0.5472118959107807, "grad_norm": 1.4611782283542265, "learning_rate": 4.4914295074695395e-06, "loss": 0.0563, "step": 11776 }, { "epoch": 0.5472583643122677, "grad_norm": 0.5302340773394375, "learning_rate": 4.49068090268225e-06, "loss": 0.0528, "step": 11777 }, { "epoch": 0.5473048327137546, "grad_norm": 0.6156044266719104, "learning_rate": 4.489932309431572e-06, "loss": 0.058, "step": 11778 }, { "epoch": 0.5473513011152417, "grad_norm": 0.4829543271565749, "learning_rate": 4.489183727734466e-06, "loss": 0.0422, "step": 11779 }, { "epoch": 0.5473977695167286, "grad_norm": 0.4859429811952383, "learning_rate": 4.488435157607885e-06, "loss": 0.0384, "step": 11780 }, { "epoch": 0.5474442379182156, "grad_norm": 0.7417284849659008, "learning_rate": 4.487686599068788e-06, "loss": 0.0642, "step": 11781 }, { "epoch": 0.5474907063197026, "grad_norm": 0.8326079555576373, "learning_rate": 4.486938052134127e-06, "loss": 0.0602, "step": 11782 }, { "epoch": 0.5475371747211896, "grad_norm": 0.5356642252944473, "learning_rate": 4.486189516820859e-06, "loss": 0.0495, "step": 11783 }, { "epoch": 0.5475836431226766, "grad_norm": 0.4802236682069522, "learning_rate": 4.485440993145941e-06, "loss": 0.0514, "step": 11784 }, { "epoch": 0.5476301115241635, "grad_norm": 0.6074329763741855, "learning_rate": 4.484692481126323e-06, "loss": 0.0687, "step": 11785 }, { "epoch": 0.5476765799256506, "grad_norm": 0.8672732927443417, "learning_rate": 4.483943980778965e-06, "loss": 0.0674, "step": 11786 }, { "epoch": 0.5477230483271376, "grad_norm": 1.2756663437349103, "learning_rate": 4.48319549212082e-06, "loss": 0.0478, "step": 11787 }, { "epoch": 0.5477695167286245, "grad_norm": 1.0535461579261132, "learning_rate": 4.482447015168838e-06, "loss": 0.0418, "step": 11788 }, { "epoch": 0.5478159851301115, "grad_norm": 0.7464136604914505, "learning_rate": 4.481698549939978e-06, "loss": 0.059, "step": 11789 }, { "epoch": 0.5478624535315985, "grad_norm": 0.41588925549194383, "learning_rate": 4.48095009645119e-06, "loss": 0.0672, "step": 11790 }, { "epoch": 0.5479089219330855, "grad_norm": 1.0039113793220769, "learning_rate": 4.480201654719429e-06, "loss": 0.0622, "step": 11791 }, { "epoch": 0.5479553903345725, "grad_norm": 0.6411213979050935, "learning_rate": 4.4794532247616476e-06, "loss": 0.0567, "step": 11792 }, { "epoch": 0.5480018587360594, "grad_norm": 0.5891471864386506, "learning_rate": 4.4787048065948e-06, "loss": 0.0529, "step": 11793 }, { "epoch": 0.5480483271375465, "grad_norm": 0.4570628204586439, "learning_rate": 4.477956400235836e-06, "loss": 0.0636, "step": 11794 }, { "epoch": 0.5480947955390335, "grad_norm": 0.5387853250365442, "learning_rate": 4.477208005701708e-06, "loss": 0.0622, "step": 11795 }, { "epoch": 0.5481412639405204, "grad_norm": 0.5160575483626241, "learning_rate": 4.476459623009371e-06, "loss": 0.0574, "step": 11796 }, { "epoch": 0.5481877323420075, "grad_norm": 0.6007061143389402, "learning_rate": 4.475711252175772e-06, "loss": 0.0388, "step": 11797 }, { "epoch": 0.5482342007434944, "grad_norm": 0.6266503434125713, "learning_rate": 4.4749628932178665e-06, "loss": 0.0437, "step": 11798 }, { "epoch": 0.5482806691449814, "grad_norm": 1.0369204397222123, "learning_rate": 4.474214546152605e-06, "loss": 0.074, "step": 11799 }, { "epoch": 0.5483271375464684, "grad_norm": 0.3730777576561651, "learning_rate": 4.473466210996936e-06, "loss": 0.0583, "step": 11800 }, { "epoch": 0.5483736059479554, "grad_norm": 0.503214614505501, "learning_rate": 4.472717887767811e-06, "loss": 0.0482, "step": 11801 }, { "epoch": 0.5484200743494424, "grad_norm": 0.8272605418141011, "learning_rate": 4.471969576482182e-06, "loss": 0.0623, "step": 11802 }, { "epoch": 0.5484665427509293, "grad_norm": 0.8594293416361098, "learning_rate": 4.471221277156997e-06, "loss": 0.0619, "step": 11803 }, { "epoch": 0.5485130111524164, "grad_norm": 0.7102817479869182, "learning_rate": 4.470472989809207e-06, "loss": 0.0829, "step": 11804 }, { "epoch": 0.5485594795539034, "grad_norm": 0.5034447926173528, "learning_rate": 4.469724714455762e-06, "loss": 0.0514, "step": 11805 }, { "epoch": 0.5486059479553903, "grad_norm": 0.6594199507256362, "learning_rate": 4.46897645111361e-06, "loss": 0.0515, "step": 11806 }, { "epoch": 0.5486524163568773, "grad_norm": 0.5891731659092111, "learning_rate": 4.468228199799699e-06, "loss": 0.0514, "step": 11807 }, { "epoch": 0.5486988847583643, "grad_norm": 0.5898250003828079, "learning_rate": 4.46747996053098e-06, "loss": 0.0411, "step": 11808 }, { "epoch": 0.5487453531598513, "grad_norm": 0.434251520813415, "learning_rate": 4.466731733324399e-06, "loss": 0.0378, "step": 11809 }, { "epoch": 0.5487918215613383, "grad_norm": 0.6005952664934476, "learning_rate": 4.465983518196908e-06, "loss": 0.0601, "step": 11810 }, { "epoch": 0.5488382899628252, "grad_norm": 0.3858498902966885, "learning_rate": 4.465235315165451e-06, "loss": 0.0422, "step": 11811 }, { "epoch": 0.5488847583643123, "grad_norm": 0.367594130577532, "learning_rate": 4.464487124246976e-06, "loss": 0.0429, "step": 11812 }, { "epoch": 0.5489312267657993, "grad_norm": 1.3660534312541612, "learning_rate": 4.463738945458432e-06, "loss": 0.0627, "step": 11813 }, { "epoch": 0.5489776951672862, "grad_norm": 1.1725149531239676, "learning_rate": 4.462990778816765e-06, "loss": 0.0575, "step": 11814 }, { "epoch": 0.5490241635687733, "grad_norm": 1.412077583804639, "learning_rate": 4.462242624338923e-06, "loss": 0.0688, "step": 11815 }, { "epoch": 0.5490706319702602, "grad_norm": 0.6542770377860154, "learning_rate": 4.461494482041852e-06, "loss": 0.0672, "step": 11816 }, { "epoch": 0.5491171003717472, "grad_norm": 0.9186421130146062, "learning_rate": 4.460746351942496e-06, "loss": 0.0652, "step": 11817 }, { "epoch": 0.5491635687732342, "grad_norm": 0.7466917625197173, "learning_rate": 4.459998234057804e-06, "loss": 0.0646, "step": 11818 }, { "epoch": 0.5492100371747212, "grad_norm": 0.617788587535979, "learning_rate": 4.4592501284047184e-06, "loss": 0.0564, "step": 11819 }, { "epoch": 0.5492565055762082, "grad_norm": 0.6579408357332887, "learning_rate": 4.4585020350001885e-06, "loss": 0.0767, "step": 11820 }, { "epoch": 0.5493029739776951, "grad_norm": 0.8750756653464551, "learning_rate": 4.457753953861156e-06, "loss": 0.0576, "step": 11821 }, { "epoch": 0.5493494423791822, "grad_norm": 0.41455407592682436, "learning_rate": 4.45700588500457e-06, "loss": 0.0411, "step": 11822 }, { "epoch": 0.5493959107806692, "grad_norm": 0.43692446449298544, "learning_rate": 4.45625782844737e-06, "loss": 0.0566, "step": 11823 }, { "epoch": 0.5494423791821561, "grad_norm": 0.777423330803599, "learning_rate": 4.455509784206501e-06, "loss": 0.0542, "step": 11824 }, { "epoch": 0.5494888475836431, "grad_norm": 1.4738974668114404, "learning_rate": 4.454761752298911e-06, "loss": 0.0647, "step": 11825 }, { "epoch": 0.5495353159851301, "grad_norm": 0.9151562696427399, "learning_rate": 4.454013732741539e-06, "loss": 0.0828, "step": 11826 }, { "epoch": 0.5495817843866171, "grad_norm": 0.41808627369301254, "learning_rate": 4.453265725551332e-06, "loss": 0.0392, "step": 11827 }, { "epoch": 0.5496282527881041, "grad_norm": 0.6449851061555706, "learning_rate": 4.452517730745233e-06, "loss": 0.0605, "step": 11828 }, { "epoch": 0.549674721189591, "grad_norm": 0.44021298066292214, "learning_rate": 4.451769748340181e-06, "loss": 0.0493, "step": 11829 }, { "epoch": 0.5497211895910781, "grad_norm": 1.0406534195717883, "learning_rate": 4.451021778353123e-06, "loss": 0.0613, "step": 11830 }, { "epoch": 0.5497676579925651, "grad_norm": 0.4436251436295024, "learning_rate": 4.450273820800998e-06, "loss": 0.0437, "step": 11831 }, { "epoch": 0.549814126394052, "grad_norm": 0.5836436967911971, "learning_rate": 4.449525875700751e-06, "loss": 0.0491, "step": 11832 }, { "epoch": 0.5498605947955391, "grad_norm": 0.5840233299840076, "learning_rate": 4.448777943069321e-06, "loss": 0.0455, "step": 11833 }, { "epoch": 0.549907063197026, "grad_norm": 0.4837333816605005, "learning_rate": 4.4480300229236525e-06, "loss": 0.054, "step": 11834 }, { "epoch": 0.549953531598513, "grad_norm": 0.531493366650986, "learning_rate": 4.447282115280684e-06, "loss": 0.0428, "step": 11835 }, { "epoch": 0.55, "grad_norm": 1.13200453660165, "learning_rate": 4.4465342201573564e-06, "loss": 0.0542, "step": 11836 }, { "epoch": 0.550046468401487, "grad_norm": 0.6654494978422145, "learning_rate": 4.445786337570612e-06, "loss": 0.0659, "step": 11837 }, { "epoch": 0.550092936802974, "grad_norm": 0.6239835553366387, "learning_rate": 4.4450384675373896e-06, "loss": 0.0465, "step": 11838 }, { "epoch": 0.5501394052044609, "grad_norm": 0.6770533810635637, "learning_rate": 4.444290610074632e-06, "loss": 0.0515, "step": 11839 }, { "epoch": 0.550185873605948, "grad_norm": 0.47494486315130163, "learning_rate": 4.443542765199276e-06, "loss": 0.0705, "step": 11840 }, { "epoch": 0.550232342007435, "grad_norm": 1.2750561886188116, "learning_rate": 4.442794932928261e-06, "loss": 0.0649, "step": 11841 }, { "epoch": 0.5502788104089219, "grad_norm": 1.1379181187089271, "learning_rate": 4.442047113278528e-06, "loss": 0.0692, "step": 11842 }, { "epoch": 0.5503252788104089, "grad_norm": 0.9821393744378097, "learning_rate": 4.441299306267014e-06, "loss": 0.0515, "step": 11843 }, { "epoch": 0.5503717472118959, "grad_norm": 0.6167735051130094, "learning_rate": 4.440551511910661e-06, "loss": 0.0495, "step": 11844 }, { "epoch": 0.5504182156133829, "grad_norm": 0.9125357982898397, "learning_rate": 4.439803730226403e-06, "loss": 0.0647, "step": 11845 }, { "epoch": 0.5504646840148699, "grad_norm": 1.0013965827132678, "learning_rate": 4.439055961231182e-06, "loss": 0.0869, "step": 11846 }, { "epoch": 0.5505111524163568, "grad_norm": 0.8875692322290262, "learning_rate": 4.438308204941933e-06, "loss": 0.0436, "step": 11847 }, { "epoch": 0.5505576208178439, "grad_norm": 0.9353663829967195, "learning_rate": 4.437560461375593e-06, "loss": 0.0712, "step": 11848 }, { "epoch": 0.5506040892193309, "grad_norm": 0.5804849389143685, "learning_rate": 4.4368127305491025e-06, "loss": 0.0641, "step": 11849 }, { "epoch": 0.5506505576208178, "grad_norm": 0.9191605623866479, "learning_rate": 4.436065012479395e-06, "loss": 0.0781, "step": 11850 }, { "epoch": 0.5506970260223049, "grad_norm": 0.7300112558970672, "learning_rate": 4.435317307183409e-06, "loss": 0.0727, "step": 11851 }, { "epoch": 0.5507434944237918, "grad_norm": 0.6867367645084068, "learning_rate": 4.434569614678081e-06, "loss": 0.0595, "step": 11852 }, { "epoch": 0.5507899628252788, "grad_norm": 1.9015424490687083, "learning_rate": 4.433821934980346e-06, "loss": 0.117, "step": 11853 }, { "epoch": 0.5508364312267658, "grad_norm": 0.39762517511894474, "learning_rate": 4.433074268107139e-06, "loss": 0.0398, "step": 11854 }, { "epoch": 0.5508828996282528, "grad_norm": 0.303425036855592, "learning_rate": 4.432326614075397e-06, "loss": 0.0435, "step": 11855 }, { "epoch": 0.5509293680297398, "grad_norm": 0.5479948930588159, "learning_rate": 4.431578972902055e-06, "loss": 0.0351, "step": 11856 }, { "epoch": 0.5509758364312267, "grad_norm": 0.40162149998024055, "learning_rate": 4.4308313446040465e-06, "loss": 0.0192, "step": 11857 }, { "epoch": 0.5510223048327137, "grad_norm": 0.4772758325540486, "learning_rate": 4.43008372919831e-06, "loss": 0.0556, "step": 11858 }, { "epoch": 0.5510687732342008, "grad_norm": 1.0483762191619719, "learning_rate": 4.4293361267017745e-06, "loss": 0.043, "step": 11859 }, { "epoch": 0.5511152416356877, "grad_norm": 0.4483076070082145, "learning_rate": 4.428588537131376e-06, "loss": 0.0474, "step": 11860 }, { "epoch": 0.5511617100371747, "grad_norm": 0.5505675873116369, "learning_rate": 4.4278409605040495e-06, "loss": 0.0557, "step": 11861 }, { "epoch": 0.5512081784386617, "grad_norm": 0.6403300647845737, "learning_rate": 4.427093396836727e-06, "loss": 0.0803, "step": 11862 }, { "epoch": 0.5512546468401487, "grad_norm": 0.4816440660788504, "learning_rate": 4.426345846146342e-06, "loss": 0.0616, "step": 11863 }, { "epoch": 0.5513011152416357, "grad_norm": 0.5098368417740635, "learning_rate": 4.4255983084498285e-06, "loss": 0.0368, "step": 11864 }, { "epoch": 0.5513475836431226, "grad_norm": 1.31565864759929, "learning_rate": 4.4248507837641165e-06, "loss": 0.0773, "step": 11865 }, { "epoch": 0.5513940520446097, "grad_norm": 0.6547317915769153, "learning_rate": 4.4241032721061406e-06, "loss": 0.0577, "step": 11866 }, { "epoch": 0.5514405204460967, "grad_norm": 0.4776168076029203, "learning_rate": 4.42335577349283e-06, "loss": 0.0372, "step": 11867 }, { "epoch": 0.5514869888475836, "grad_norm": 0.6117286663713034, "learning_rate": 4.4226082879411196e-06, "loss": 0.0638, "step": 11868 }, { "epoch": 0.5515334572490707, "grad_norm": 0.4077389973802508, "learning_rate": 4.42186081546794e-06, "loss": 0.0396, "step": 11869 }, { "epoch": 0.5515799256505576, "grad_norm": 0.7308937276050445, "learning_rate": 4.421113356090219e-06, "loss": 0.0608, "step": 11870 }, { "epoch": 0.5516263940520446, "grad_norm": 0.36690288465963894, "learning_rate": 4.4203659098248915e-06, "loss": 0.0466, "step": 11871 }, { "epoch": 0.5516728624535316, "grad_norm": 0.7588035019492593, "learning_rate": 4.419618476688885e-06, "loss": 0.0521, "step": 11872 }, { "epoch": 0.5517193308550186, "grad_norm": 0.525150673127955, "learning_rate": 4.418871056699132e-06, "loss": 0.0708, "step": 11873 }, { "epoch": 0.5517657992565056, "grad_norm": 0.6162611170942202, "learning_rate": 4.41812364987256e-06, "loss": 0.0278, "step": 11874 }, { "epoch": 0.5518122676579925, "grad_norm": 0.7340060017698642, "learning_rate": 4.417376256226101e-06, "loss": 0.0573, "step": 11875 }, { "epoch": 0.5518587360594795, "grad_norm": 0.556368496409592, "learning_rate": 4.416628875776683e-06, "loss": 0.0473, "step": 11876 }, { "epoch": 0.5519052044609666, "grad_norm": 0.6123356194289457, "learning_rate": 4.4158815085412335e-06, "loss": 0.0695, "step": 11877 }, { "epoch": 0.5519516728624535, "grad_norm": 0.5460579856561676, "learning_rate": 4.415134154536683e-06, "loss": 0.0466, "step": 11878 }, { "epoch": 0.5519981412639405, "grad_norm": 0.5460579856561676, "learning_rate": 4.415134154536683e-06, "loss": 0.0721, "step": 11879 }, { "epoch": 0.5520446096654275, "grad_norm": 0.6127061688748157, "learning_rate": 4.41438681377996e-06, "loss": 0.0581, "step": 11880 }, { "epoch": 0.5520910780669145, "grad_norm": 0.8634873207585662, "learning_rate": 4.413639486287992e-06, "loss": 0.068, "step": 11881 }, { "epoch": 0.5521375464684015, "grad_norm": 0.38767486966283093, "learning_rate": 4.412892172077707e-06, "loss": 0.024, "step": 11882 }, { "epoch": 0.5521840148698884, "grad_norm": 0.8877682899506993, "learning_rate": 4.4121448711660315e-06, "loss": 0.0618, "step": 11883 }, { "epoch": 0.5522304832713755, "grad_norm": 0.35260576733016386, "learning_rate": 4.411397583569894e-06, "loss": 0.0572, "step": 11884 }, { "epoch": 0.5522769516728625, "grad_norm": 0.514646393334724, "learning_rate": 4.410650309306219e-06, "loss": 0.067, "step": 11885 }, { "epoch": 0.5523234200743494, "grad_norm": 0.8252502534764671, "learning_rate": 4.409903048391937e-06, "loss": 0.0597, "step": 11886 }, { "epoch": 0.5523698884758365, "grad_norm": 0.6699888799131146, "learning_rate": 4.409155800843971e-06, "loss": 0.0603, "step": 11887 }, { "epoch": 0.5524163568773234, "grad_norm": 0.9770544458413655, "learning_rate": 4.408408566679249e-06, "loss": 0.0434, "step": 11888 }, { "epoch": 0.5524628252788104, "grad_norm": 0.3357228403481554, "learning_rate": 4.407661345914696e-06, "loss": 0.0266, "step": 11889 }, { "epoch": 0.5525092936802974, "grad_norm": 2.1413981752406532, "learning_rate": 4.406914138567235e-06, "loss": 0.1052, "step": 11890 }, { "epoch": 0.5525557620817844, "grad_norm": 0.49860936167703696, "learning_rate": 4.406166944653793e-06, "loss": 0.0523, "step": 11891 }, { "epoch": 0.5526022304832714, "grad_norm": 0.8596595236361176, "learning_rate": 4.405419764191295e-06, "loss": 0.0516, "step": 11892 }, { "epoch": 0.5526486988847583, "grad_norm": 0.3773590027570856, "learning_rate": 4.404672597196666e-06, "loss": 0.0441, "step": 11893 }, { "epoch": 0.5526951672862453, "grad_norm": 0.8177935349140785, "learning_rate": 4.40392544368683e-06, "loss": 0.0904, "step": 11894 }, { "epoch": 0.5527416356877324, "grad_norm": 0.4421212181967918, "learning_rate": 4.403178303678708e-06, "loss": 0.0388, "step": 11895 }, { "epoch": 0.5527881040892193, "grad_norm": 0.7774421073811844, "learning_rate": 4.402431177189227e-06, "loss": 0.0472, "step": 11896 }, { "epoch": 0.5528345724907063, "grad_norm": 0.6502246913421749, "learning_rate": 4.401684064235308e-06, "loss": 0.048, "step": 11897 }, { "epoch": 0.5528810408921933, "grad_norm": 0.5944005183994228, "learning_rate": 4.400936964833875e-06, "loss": 0.0386, "step": 11898 }, { "epoch": 0.5529275092936803, "grad_norm": 3.697633461205932, "learning_rate": 4.400189879001849e-06, "loss": 0.0671, "step": 11899 }, { "epoch": 0.5529739776951673, "grad_norm": 0.690848258357764, "learning_rate": 4.399442806756156e-06, "loss": 0.0591, "step": 11900 }, { "epoch": 0.5530204460966542, "grad_norm": 0.6530215020605384, "learning_rate": 4.398695748113714e-06, "loss": 0.0527, "step": 11901 }, { "epoch": 0.5530669144981413, "grad_norm": 0.5754838578060621, "learning_rate": 4.397948703091446e-06, "loss": 0.0393, "step": 11902 }, { "epoch": 0.5531133828996283, "grad_norm": 0.8995806564540535, "learning_rate": 4.397201671706274e-06, "loss": 0.0682, "step": 11903 }, { "epoch": 0.5531598513011152, "grad_norm": 0.6294459566526247, "learning_rate": 4.396454653975118e-06, "loss": 0.034, "step": 11904 }, { "epoch": 0.5532063197026023, "grad_norm": 0.7643567543618535, "learning_rate": 4.395707649914899e-06, "loss": 0.0643, "step": 11905 }, { "epoch": 0.5532527881040892, "grad_norm": 0.5447815620025017, "learning_rate": 4.394960659542541e-06, "loss": 0.049, "step": 11906 }, { "epoch": 0.5532992565055762, "grad_norm": 0.8466455304765088, "learning_rate": 4.394213682874957e-06, "loss": 0.0836, "step": 11907 }, { "epoch": 0.5533457249070632, "grad_norm": 0.6150682242748872, "learning_rate": 4.393466719929072e-06, "loss": 0.0315, "step": 11908 }, { "epoch": 0.5533921933085502, "grad_norm": 0.7336792632424798, "learning_rate": 4.392719770721803e-06, "loss": 0.0556, "step": 11909 }, { "epoch": 0.5534386617100372, "grad_norm": 0.6463535922959713, "learning_rate": 4.391972835270072e-06, "loss": 0.0684, "step": 11910 }, { "epoch": 0.5534851301115241, "grad_norm": 1.5481993849816156, "learning_rate": 4.391225913590795e-06, "loss": 0.0637, "step": 11911 }, { "epoch": 0.5535315985130111, "grad_norm": 1.2765923040317646, "learning_rate": 4.390479005700894e-06, "loss": 0.0581, "step": 11912 }, { "epoch": 0.5535780669144982, "grad_norm": 0.7048635898951685, "learning_rate": 4.3897321116172835e-06, "loss": 0.0548, "step": 11913 }, { "epoch": 0.5536245353159851, "grad_norm": 0.9601930890005034, "learning_rate": 4.388985231356882e-06, "loss": 0.0671, "step": 11914 }, { "epoch": 0.5536710037174721, "grad_norm": 0.7709175435524918, "learning_rate": 4.38823836493661e-06, "loss": 0.061, "step": 11915 }, { "epoch": 0.553717472118959, "grad_norm": 0.9052079623707742, "learning_rate": 4.387491512373381e-06, "loss": 0.0734, "step": 11916 }, { "epoch": 0.5537639405204461, "grad_norm": 0.563131828292102, "learning_rate": 4.386744673684114e-06, "loss": 0.0631, "step": 11917 }, { "epoch": 0.5538104089219331, "grad_norm": 0.6344731240184703, "learning_rate": 4.385997848885729e-06, "loss": 0.0447, "step": 11918 }, { "epoch": 0.55385687732342, "grad_norm": 0.786809401114085, "learning_rate": 4.385251037995135e-06, "loss": 0.0614, "step": 11919 }, { "epoch": 0.5539033457249071, "grad_norm": 0.7366648889768274, "learning_rate": 4.384504241029254e-06, "loss": 0.0457, "step": 11920 }, { "epoch": 0.5539498141263941, "grad_norm": 0.5453961226774019, "learning_rate": 4.383757458004998e-06, "loss": 0.0553, "step": 11921 }, { "epoch": 0.553996282527881, "grad_norm": 0.46696225608232445, "learning_rate": 4.383010688939286e-06, "loss": 0.0372, "step": 11922 }, { "epoch": 0.554042750929368, "grad_norm": 0.7309803871373277, "learning_rate": 4.382263933849029e-06, "loss": 0.0511, "step": 11923 }, { "epoch": 0.554089219330855, "grad_norm": 0.39556835189170275, "learning_rate": 4.381517192751147e-06, "loss": 0.0304, "step": 11924 }, { "epoch": 0.554135687732342, "grad_norm": 0.36234838070842895, "learning_rate": 4.380770465662551e-06, "loss": 0.0324, "step": 11925 }, { "epoch": 0.554182156133829, "grad_norm": 1.66001838468137, "learning_rate": 4.3800237526001535e-06, "loss": 0.0484, "step": 11926 }, { "epoch": 0.554228624535316, "grad_norm": 0.4423769687072816, "learning_rate": 4.3792770535808726e-06, "loss": 0.0319, "step": 11927 }, { "epoch": 0.554275092936803, "grad_norm": 0.7833935476186581, "learning_rate": 4.378530368621618e-06, "loss": 0.062, "step": 11928 }, { "epoch": 0.5543215613382899, "grad_norm": 0.8054791722650583, "learning_rate": 4.377783697739307e-06, "loss": 0.0633, "step": 11929 }, { "epoch": 0.554368029739777, "grad_norm": 0.7369634674608818, "learning_rate": 4.377037040950849e-06, "loss": 0.0743, "step": 11930 }, { "epoch": 0.554414498141264, "grad_norm": 0.4401171125149557, "learning_rate": 4.376290398273156e-06, "loss": 0.055, "step": 11931 }, { "epoch": 0.5544609665427509, "grad_norm": 4.114757466044234, "learning_rate": 4.375543769723144e-06, "loss": 0.0635, "step": 11932 }, { "epoch": 0.5545074349442379, "grad_norm": 0.4206437593509563, "learning_rate": 4.374797155317721e-06, "loss": 0.0441, "step": 11933 }, { "epoch": 0.5545539033457249, "grad_norm": 0.4491458222077479, "learning_rate": 4.3740505550738025e-06, "loss": 0.0542, "step": 11934 }, { "epoch": 0.5546003717472119, "grad_norm": 0.49999501833141635, "learning_rate": 4.373303969008298e-06, "loss": 0.0502, "step": 11935 }, { "epoch": 0.5546468401486989, "grad_norm": 1.0327386884595995, "learning_rate": 4.372557397138116e-06, "loss": 0.0687, "step": 11936 }, { "epoch": 0.5546933085501858, "grad_norm": 0.4836707909114053, "learning_rate": 4.3718108394801705e-06, "loss": 0.0547, "step": 11937 }, { "epoch": 0.5547397769516729, "grad_norm": 0.4657641080942818, "learning_rate": 4.37106429605137e-06, "loss": 0.0395, "step": 11938 }, { "epoch": 0.5547862453531599, "grad_norm": 0.5142881144721952, "learning_rate": 4.370317766868626e-06, "loss": 0.0387, "step": 11939 }, { "epoch": 0.5548327137546468, "grad_norm": 4.598316581572561, "learning_rate": 4.369571251948847e-06, "loss": 0.0518, "step": 11940 }, { "epoch": 0.5548791821561339, "grad_norm": 0.7699873453895836, "learning_rate": 4.3688247513089435e-06, "loss": 0.0706, "step": 11941 }, { "epoch": 0.5549256505576208, "grad_norm": 0.4518610513611264, "learning_rate": 4.368078264965823e-06, "loss": 0.0358, "step": 11942 }, { "epoch": 0.5549721189591078, "grad_norm": 1.2522830403186638, "learning_rate": 4.367331792936393e-06, "loss": 0.0702, "step": 11943 }, { "epoch": 0.5550185873605948, "grad_norm": 0.701008545912797, "learning_rate": 4.366585335237566e-06, "loss": 0.0619, "step": 11944 }, { "epoch": 0.5550650557620818, "grad_norm": 0.6954811396190326, "learning_rate": 4.365838891886247e-06, "loss": 0.0473, "step": 11945 }, { "epoch": 0.5551115241635688, "grad_norm": 0.7074920672131717, "learning_rate": 4.365092462899345e-06, "loss": 0.0609, "step": 11946 }, { "epoch": 0.5551579925650557, "grad_norm": 0.9967969720155724, "learning_rate": 4.364346048293768e-06, "loss": 0.0733, "step": 11947 }, { "epoch": 0.5552044609665427, "grad_norm": 0.6573712535673446, "learning_rate": 4.36359964808642e-06, "loss": 0.0655, "step": 11948 }, { "epoch": 0.5552509293680298, "grad_norm": 0.6405907217275796, "learning_rate": 4.36285326229421e-06, "loss": 0.0552, "step": 11949 }, { "epoch": 0.5552973977695167, "grad_norm": 1.3332302043978175, "learning_rate": 4.362106890934044e-06, "loss": 0.0817, "step": 11950 }, { "epoch": 0.5553438661710037, "grad_norm": 0.47624051154390346, "learning_rate": 4.36136053402283e-06, "loss": 0.0484, "step": 11951 }, { "epoch": 0.5553903345724907, "grad_norm": 1.2308607518527932, "learning_rate": 4.3606141915774695e-06, "loss": 0.0732, "step": 11952 }, { "epoch": 0.5554368029739777, "grad_norm": 0.6730460274783898, "learning_rate": 4.359867863614874e-06, "loss": 0.066, "step": 11953 }, { "epoch": 0.5554832713754647, "grad_norm": 0.5088961339484136, "learning_rate": 4.359121550151943e-06, "loss": 0.0378, "step": 11954 }, { "epoch": 0.5555297397769516, "grad_norm": 0.8298383249284408, "learning_rate": 4.358375251205583e-06, "loss": 0.0709, "step": 11955 }, { "epoch": 0.5555762081784387, "grad_norm": 2.3088266176824734, "learning_rate": 4.3576289667927e-06, "loss": 0.0723, "step": 11956 }, { "epoch": 0.5556226765799257, "grad_norm": 0.5067126240628181, "learning_rate": 4.356882696930195e-06, "loss": 0.0495, "step": 11957 }, { "epoch": 0.5556691449814126, "grad_norm": 0.36305842693771434, "learning_rate": 4.356136441634976e-06, "loss": 0.048, "step": 11958 }, { "epoch": 0.5557156133828997, "grad_norm": 1.4575575630651367, "learning_rate": 4.355390200923944e-06, "loss": 0.0488, "step": 11959 }, { "epoch": 0.5557620817843866, "grad_norm": 0.7323343165971563, "learning_rate": 4.3546439748140015e-06, "loss": 0.0669, "step": 11960 }, { "epoch": 0.5558085501858736, "grad_norm": 0.7758602875781198, "learning_rate": 4.353897763322053e-06, "loss": 0.0667, "step": 11961 }, { "epoch": 0.5558550185873606, "grad_norm": 0.6937711924196044, "learning_rate": 4.353151566464999e-06, "loss": 0.0562, "step": 11962 }, { "epoch": 0.5559014869888476, "grad_norm": 0.9415127894270159, "learning_rate": 4.352405384259744e-06, "loss": 0.0506, "step": 11963 }, { "epoch": 0.5559479553903346, "grad_norm": 1.1089401118818252, "learning_rate": 4.351659216723187e-06, "loss": 0.0818, "step": 11964 }, { "epoch": 0.5559944237918215, "grad_norm": 0.4884901183421368, "learning_rate": 4.350913063872234e-06, "loss": 0.0408, "step": 11965 }, { "epoch": 0.5560408921933085, "grad_norm": 0.6010719565724749, "learning_rate": 4.350166925723781e-06, "loss": 0.0492, "step": 11966 }, { "epoch": 0.5560873605947956, "grad_norm": 0.9526404582487651, "learning_rate": 4.349420802294731e-06, "loss": 0.0612, "step": 11967 }, { "epoch": 0.5561338289962825, "grad_norm": 1.0101831440323576, "learning_rate": 4.3486746936019854e-06, "loss": 0.0612, "step": 11968 }, { "epoch": 0.5561802973977695, "grad_norm": 1.0672689900175385, "learning_rate": 4.347928599662443e-06, "loss": 0.0808, "step": 11969 }, { "epoch": 0.5562267657992565, "grad_norm": 1.0089126905461088, "learning_rate": 4.347182520493004e-06, "loss": 0.062, "step": 11970 }, { "epoch": 0.5562732342007435, "grad_norm": 0.8661193649347194, "learning_rate": 4.34643645611057e-06, "loss": 0.0605, "step": 11971 }, { "epoch": 0.5563197026022305, "grad_norm": 1.848460325678484, "learning_rate": 4.345690406532036e-06, "loss": 0.0728, "step": 11972 }, { "epoch": 0.5563661710037174, "grad_norm": 0.3251865049386977, "learning_rate": 4.344944371774303e-06, "loss": 0.0324, "step": 11973 }, { "epoch": 0.5564126394052045, "grad_norm": 0.38009559599426423, "learning_rate": 4.3441983518542695e-06, "loss": 0.0396, "step": 11974 }, { "epoch": 0.5564591078066915, "grad_norm": 0.4740756351631789, "learning_rate": 4.343452346788834e-06, "loss": 0.047, "step": 11975 }, { "epoch": 0.5565055762081784, "grad_norm": 0.41985132223244975, "learning_rate": 4.342706356594892e-06, "loss": 0.055, "step": 11976 }, { "epoch": 0.5565520446096655, "grad_norm": 0.47608604575337393, "learning_rate": 4.341960381289346e-06, "loss": 0.0249, "step": 11977 }, { "epoch": 0.5565985130111524, "grad_norm": 0.6325487754945737, "learning_rate": 4.341214420889089e-06, "loss": 0.0475, "step": 11978 }, { "epoch": 0.5566449814126394, "grad_norm": 0.5088265882058374, "learning_rate": 4.340468475411018e-06, "loss": 0.0392, "step": 11979 }, { "epoch": 0.5566914498141264, "grad_norm": 1.0284081740122981, "learning_rate": 4.33972254487203e-06, "loss": 0.0469, "step": 11980 }, { "epoch": 0.5567379182156134, "grad_norm": 0.6567017709029375, "learning_rate": 4.338976629289021e-06, "loss": 0.0539, "step": 11981 }, { "epoch": 0.5567843866171004, "grad_norm": 0.5372115413487408, "learning_rate": 4.338230728678888e-06, "loss": 0.0529, "step": 11982 }, { "epoch": 0.5568308550185873, "grad_norm": 0.8134713914655184, "learning_rate": 4.337484843058527e-06, "loss": 0.0825, "step": 11983 }, { "epoch": 0.5568773234200743, "grad_norm": 1.0495646113350177, "learning_rate": 4.3367389724448285e-06, "loss": 0.076, "step": 11984 }, { "epoch": 0.5569237918215614, "grad_norm": 0.7999364323990641, "learning_rate": 4.335993116854692e-06, "loss": 0.0611, "step": 11985 }, { "epoch": 0.5569702602230483, "grad_norm": 0.7804414919157249, "learning_rate": 4.335247276305009e-06, "loss": 0.0494, "step": 11986 }, { "epoch": 0.5570167286245353, "grad_norm": 1.5670427104758418, "learning_rate": 4.334501450812676e-06, "loss": 0.0752, "step": 11987 }, { "epoch": 0.5570631970260222, "grad_norm": 0.6516079417081921, "learning_rate": 4.333755640394587e-06, "loss": 0.0772, "step": 11988 }, { "epoch": 0.5571096654275093, "grad_norm": 0.4949204355483481, "learning_rate": 4.333009845067632e-06, "loss": 0.0639, "step": 11989 }, { "epoch": 0.5571561338289963, "grad_norm": 0.9018645723874807, "learning_rate": 4.332264064848706e-06, "loss": 0.0782, "step": 11990 }, { "epoch": 0.5572026022304832, "grad_norm": 0.5741687177981328, "learning_rate": 4.331518299754701e-06, "loss": 0.0562, "step": 11991 }, { "epoch": 0.5572490706319703, "grad_norm": 0.8562941049904617, "learning_rate": 4.330772549802511e-06, "loss": 0.0611, "step": 11992 }, { "epoch": 0.5572955390334573, "grad_norm": 0.64568789428088, "learning_rate": 4.330026815009026e-06, "loss": 0.0503, "step": 11993 }, { "epoch": 0.5573420074349442, "grad_norm": 0.4889557485420872, "learning_rate": 4.329281095391142e-06, "loss": 0.0401, "step": 11994 }, { "epoch": 0.5573884758364313, "grad_norm": 0.5990210766268687, "learning_rate": 4.3285353909657454e-06, "loss": 0.0495, "step": 11995 }, { "epoch": 0.5574349442379182, "grad_norm": 0.7931259601138421, "learning_rate": 4.327789701749728e-06, "loss": 0.0547, "step": 11996 }, { "epoch": 0.5574814126394052, "grad_norm": 0.522370594973405, "learning_rate": 4.327044027759982e-06, "loss": 0.0522, "step": 11997 }, { "epoch": 0.5575278810408922, "grad_norm": 0.7173297484442908, "learning_rate": 4.326298369013396e-06, "loss": 0.0634, "step": 11998 }, { "epoch": 0.5575743494423792, "grad_norm": 0.49009110162477704, "learning_rate": 4.325552725526862e-06, "loss": 0.0393, "step": 11999 }, { "epoch": 0.5576208178438662, "grad_norm": 0.4859297128046446, "learning_rate": 4.324807097317269e-06, "loss": 0.0312, "step": 12000 }, { "epoch": 0.5576672862453531, "grad_norm": 0.95399029056606, "learning_rate": 4.324061484401506e-06, "loss": 0.0607, "step": 12001 }, { "epoch": 0.5577137546468401, "grad_norm": 0.5522460628986923, "learning_rate": 4.323315886796461e-06, "loss": 0.037, "step": 12002 }, { "epoch": 0.5577602230483272, "grad_norm": 0.48315089153259305, "learning_rate": 4.322570304519023e-06, "loss": 0.0557, "step": 12003 }, { "epoch": 0.5578066914498141, "grad_norm": 0.4135517589887547, "learning_rate": 4.321824737586081e-06, "loss": 0.0273, "step": 12004 }, { "epoch": 0.5578531598513011, "grad_norm": 0.7153842247320674, "learning_rate": 4.321079186014522e-06, "loss": 0.0554, "step": 12005 }, { "epoch": 0.557899628252788, "grad_norm": 0.4127815372353151, "learning_rate": 4.320333649821235e-06, "loss": 0.0272, "step": 12006 }, { "epoch": 0.5579460966542751, "grad_norm": 0.34452471336058493, "learning_rate": 4.319588129023107e-06, "loss": 0.0522, "step": 12007 }, { "epoch": 0.5579925650557621, "grad_norm": 0.6213131069959307, "learning_rate": 4.3188426236370215e-06, "loss": 0.0511, "step": 12008 }, { "epoch": 0.558039033457249, "grad_norm": 0.806681617446687, "learning_rate": 4.318097133679869e-06, "loss": 0.0689, "step": 12009 }, { "epoch": 0.5580855018587361, "grad_norm": 0.47313838304754824, "learning_rate": 4.317351659168533e-06, "loss": 0.0691, "step": 12010 }, { "epoch": 0.5581319702602231, "grad_norm": 0.3899151731016171, "learning_rate": 4.316606200119902e-06, "loss": 0.0349, "step": 12011 }, { "epoch": 0.55817843866171, "grad_norm": 1.9636570646901705, "learning_rate": 4.31586075655086e-06, "loss": 0.0827, "step": 12012 }, { "epoch": 0.558224907063197, "grad_norm": 0.8189694496836192, "learning_rate": 4.31511532847829e-06, "loss": 0.0808, "step": 12013 }, { "epoch": 0.558271375464684, "grad_norm": 0.7757386424199832, "learning_rate": 4.31436991591908e-06, "loss": 0.063, "step": 12014 }, { "epoch": 0.558317843866171, "grad_norm": 1.2618552806487393, "learning_rate": 4.313624518890111e-06, "loss": 0.0561, "step": 12015 }, { "epoch": 0.558364312267658, "grad_norm": 0.6484333034741405, "learning_rate": 4.312879137408271e-06, "loss": 0.068, "step": 12016 }, { "epoch": 0.558410780669145, "grad_norm": 0.7585634286187813, "learning_rate": 4.312133771490441e-06, "loss": 0.0498, "step": 12017 }, { "epoch": 0.558457249070632, "grad_norm": 1.4675340221325242, "learning_rate": 4.311388421153507e-06, "loss": 0.1016, "step": 12018 }, { "epoch": 0.5585037174721189, "grad_norm": 0.6872838989816494, "learning_rate": 4.310643086414349e-06, "loss": 0.0529, "step": 12019 }, { "epoch": 0.5585501858736059, "grad_norm": 0.6943753853673025, "learning_rate": 4.309897767289849e-06, "loss": 0.058, "step": 12020 }, { "epoch": 0.558596654275093, "grad_norm": 1.0136896759999938, "learning_rate": 4.309152463796892e-06, "loss": 0.0542, "step": 12021 }, { "epoch": 0.5586431226765799, "grad_norm": 0.6938215532541256, "learning_rate": 4.3084071759523585e-06, "loss": 0.0605, "step": 12022 }, { "epoch": 0.5586895910780669, "grad_norm": 0.5882300551725351, "learning_rate": 4.307661903773129e-06, "loss": 0.0268, "step": 12023 }, { "epoch": 0.5587360594795538, "grad_norm": 0.5385111710234632, "learning_rate": 4.3069166472760905e-06, "loss": 0.038, "step": 12024 }, { "epoch": 0.5587825278810409, "grad_norm": 0.5588576471474364, "learning_rate": 4.3061714064781156e-06, "loss": 0.0509, "step": 12025 }, { "epoch": 0.5588289962825279, "grad_norm": 0.5851812838918999, "learning_rate": 4.30542618139609e-06, "loss": 0.0734, "step": 12026 }, { "epoch": 0.5588754646840148, "grad_norm": 0.7408523059635825, "learning_rate": 4.304680972046891e-06, "loss": 0.0622, "step": 12027 }, { "epoch": 0.5589219330855019, "grad_norm": 0.7507872951098941, "learning_rate": 4.3039357784474015e-06, "loss": 0.047, "step": 12028 }, { "epoch": 0.5589684014869889, "grad_norm": 0.765619151355842, "learning_rate": 4.303190600614497e-06, "loss": 0.0541, "step": 12029 }, { "epoch": 0.5590148698884758, "grad_norm": 0.5018235119751142, "learning_rate": 4.302445438565062e-06, "loss": 0.0463, "step": 12030 }, { "epoch": 0.5590613382899629, "grad_norm": 0.7659499504536232, "learning_rate": 4.30170029231597e-06, "loss": 0.0654, "step": 12031 }, { "epoch": 0.5591078066914498, "grad_norm": 1.0602404983351958, "learning_rate": 4.300955161884101e-06, "loss": 0.0642, "step": 12032 }, { "epoch": 0.5591542750929368, "grad_norm": 0.4656329510824501, "learning_rate": 4.300210047286334e-06, "loss": 0.0589, "step": 12033 }, { "epoch": 0.5592007434944238, "grad_norm": 0.8273970955270127, "learning_rate": 4.299464948539545e-06, "loss": 0.0644, "step": 12034 }, { "epoch": 0.5592472118959108, "grad_norm": 1.1805503778305109, "learning_rate": 4.298719865660612e-06, "loss": 0.0714, "step": 12035 }, { "epoch": 0.5592936802973978, "grad_norm": 0.9140514176963905, "learning_rate": 4.297974798666416e-06, "loss": 0.0859, "step": 12036 }, { "epoch": 0.5593401486988847, "grad_norm": 0.5406507719916965, "learning_rate": 4.2972297475738256e-06, "loss": 0.0426, "step": 12037 }, { "epoch": 0.5593866171003717, "grad_norm": 1.2762871625347367, "learning_rate": 4.296484712399722e-06, "loss": 0.0645, "step": 12038 }, { "epoch": 0.5594330855018588, "grad_norm": 0.5382314074789781, "learning_rate": 4.29573969316098e-06, "loss": 0.0656, "step": 12039 }, { "epoch": 0.5594795539033457, "grad_norm": 0.47689518713702905, "learning_rate": 4.294994689874475e-06, "loss": 0.0506, "step": 12040 }, { "epoch": 0.5595260223048327, "grad_norm": 0.42253037250112907, "learning_rate": 4.294249702557084e-06, "loss": 0.0453, "step": 12041 }, { "epoch": 0.5595724907063196, "grad_norm": 0.42646195254554053, "learning_rate": 4.293504731225679e-06, "loss": 0.0561, "step": 12042 }, { "epoch": 0.5596189591078067, "grad_norm": 0.6525342078822567, "learning_rate": 4.292759775897136e-06, "loss": 0.0538, "step": 12043 }, { "epoch": 0.5596654275092937, "grad_norm": 0.5914682658681838, "learning_rate": 4.292014836588327e-06, "loss": 0.0629, "step": 12044 }, { "epoch": 0.5597118959107806, "grad_norm": 0.46644234185292127, "learning_rate": 4.291269913316127e-06, "loss": 0.0568, "step": 12045 }, { "epoch": 0.5597583643122677, "grad_norm": 0.6254890450522165, "learning_rate": 4.290525006097411e-06, "loss": 0.0371, "step": 12046 }, { "epoch": 0.5598048327137547, "grad_norm": 0.5441136530219866, "learning_rate": 4.28978011494905e-06, "loss": 0.0592, "step": 12047 }, { "epoch": 0.5598513011152416, "grad_norm": 1.4907158649584862, "learning_rate": 4.289035239887917e-06, "loss": 0.0488, "step": 12048 }, { "epoch": 0.5598977695167286, "grad_norm": 0.4900536599111669, "learning_rate": 4.288290380930883e-06, "loss": 0.0448, "step": 12049 }, { "epoch": 0.5599442379182156, "grad_norm": 0.6576312303848234, "learning_rate": 4.28754553809482e-06, "loss": 0.0543, "step": 12050 }, { "epoch": 0.5599907063197026, "grad_norm": 1.2652981042998228, "learning_rate": 4.286800711396603e-06, "loss": 0.0662, "step": 12051 }, { "epoch": 0.5600371747211896, "grad_norm": 0.687374602414308, "learning_rate": 4.286055900853098e-06, "loss": 0.0692, "step": 12052 }, { "epoch": 0.5600836431226766, "grad_norm": 0.5935617959398591, "learning_rate": 4.285311106481182e-06, "loss": 0.0531, "step": 12053 }, { "epoch": 0.5601301115241636, "grad_norm": 0.722251796156886, "learning_rate": 4.284566328297719e-06, "loss": 0.0613, "step": 12054 }, { "epoch": 0.5601765799256505, "grad_norm": 0.49247477033318376, "learning_rate": 4.283821566319581e-06, "loss": 0.0521, "step": 12055 }, { "epoch": 0.5602230483271375, "grad_norm": 0.41373579400277627, "learning_rate": 4.2830768205636385e-06, "loss": 0.0386, "step": 12056 }, { "epoch": 0.5602695167286246, "grad_norm": 0.6577721237052729, "learning_rate": 4.2823320910467606e-06, "loss": 0.038, "step": 12057 }, { "epoch": 0.5603159851301115, "grad_norm": 1.670377537959617, "learning_rate": 4.281587377785817e-06, "loss": 0.0794, "step": 12058 }, { "epoch": 0.5603624535315985, "grad_norm": 0.5020781310737634, "learning_rate": 4.280842680797677e-06, "loss": 0.0577, "step": 12059 }, { "epoch": 0.5604089219330854, "grad_norm": 0.5537389479719333, "learning_rate": 4.280098000099206e-06, "loss": 0.0465, "step": 12060 }, { "epoch": 0.5604553903345725, "grad_norm": 1.3476871050197756, "learning_rate": 4.279353335707271e-06, "loss": 0.0735, "step": 12061 }, { "epoch": 0.5605018587360595, "grad_norm": 0.7864339443288403, "learning_rate": 4.278608687638742e-06, "loss": 0.0503, "step": 12062 }, { "epoch": 0.5605483271375464, "grad_norm": 0.8022267395300371, "learning_rate": 4.277864055910487e-06, "loss": 0.0531, "step": 12063 }, { "epoch": 0.5605947955390335, "grad_norm": 0.8161875843286217, "learning_rate": 4.277119440539369e-06, "loss": 0.0472, "step": 12064 }, { "epoch": 0.5606412639405205, "grad_norm": 0.7083567375531916, "learning_rate": 4.27637484154226e-06, "loss": 0.0421, "step": 12065 }, { "epoch": 0.5606877323420074, "grad_norm": 0.7637871944093528, "learning_rate": 4.27563025893602e-06, "loss": 0.0437, "step": 12066 }, { "epoch": 0.5607342007434944, "grad_norm": 0.9301255633446577, "learning_rate": 4.274885692737516e-06, "loss": 0.057, "step": 12067 }, { "epoch": 0.5607806691449814, "grad_norm": 0.3787769344383072, "learning_rate": 4.274141142963617e-06, "loss": 0.0444, "step": 12068 }, { "epoch": 0.5608271375464684, "grad_norm": 0.8111335272766083, "learning_rate": 4.273396609631182e-06, "loss": 0.0487, "step": 12069 }, { "epoch": 0.5608736059479554, "grad_norm": 0.6857145398133291, "learning_rate": 4.272652092757081e-06, "loss": 0.0744, "step": 12070 }, { "epoch": 0.5609200743494424, "grad_norm": 1.036450913224907, "learning_rate": 4.271907592358176e-06, "loss": 0.0742, "step": 12071 }, { "epoch": 0.5609665427509294, "grad_norm": 0.41045888309216916, "learning_rate": 4.271163108451328e-06, "loss": 0.0277, "step": 12072 }, { "epoch": 0.5610130111524163, "grad_norm": 0.9382010892834015, "learning_rate": 4.2704186410534046e-06, "loss": 0.0741, "step": 12073 }, { "epoch": 0.5610594795539033, "grad_norm": 0.4751729722173012, "learning_rate": 4.269674190181265e-06, "loss": 0.0419, "step": 12074 }, { "epoch": 0.5611059479553904, "grad_norm": 1.4229772212380334, "learning_rate": 4.268929755851775e-06, "loss": 0.0993, "step": 12075 }, { "epoch": 0.5611524163568773, "grad_norm": 0.49197318075280994, "learning_rate": 4.268185338081795e-06, "loss": 0.0367, "step": 12076 }, { "epoch": 0.5611988847583643, "grad_norm": 0.639533632132614, "learning_rate": 4.267440936888188e-06, "loss": 0.0527, "step": 12077 }, { "epoch": 0.5612453531598512, "grad_norm": 0.7408121573594777, "learning_rate": 4.2666965522878145e-06, "loss": 0.0596, "step": 12078 }, { "epoch": 0.5612918215613383, "grad_norm": 0.9755483634272576, "learning_rate": 4.265952184297535e-06, "loss": 0.0625, "step": 12079 }, { "epoch": 0.5613382899628253, "grad_norm": 0.4641542364477817, "learning_rate": 4.265207832934212e-06, "loss": 0.0464, "step": 12080 }, { "epoch": 0.5613847583643122, "grad_norm": 0.517801498434007, "learning_rate": 4.264463498214704e-06, "loss": 0.0396, "step": 12081 }, { "epoch": 0.5614312267657993, "grad_norm": 1.0238098765905688, "learning_rate": 4.263719180155873e-06, "loss": 0.0679, "step": 12082 }, { "epoch": 0.5614776951672863, "grad_norm": 0.7863600738939729, "learning_rate": 4.262974878774578e-06, "loss": 0.0564, "step": 12083 }, { "epoch": 0.5615241635687732, "grad_norm": 0.42880130375554726, "learning_rate": 4.262230594087677e-06, "loss": 0.0462, "step": 12084 }, { "epoch": 0.5615706319702602, "grad_norm": 0.5847218086948929, "learning_rate": 4.261486326112029e-06, "loss": 0.0386, "step": 12085 }, { "epoch": 0.5616171003717472, "grad_norm": 1.0627391616671806, "learning_rate": 4.2607420748644925e-06, "loss": 0.0807, "step": 12086 }, { "epoch": 0.5616635687732342, "grad_norm": 0.9902288612570099, "learning_rate": 4.259997840361927e-06, "loss": 0.0729, "step": 12087 }, { "epoch": 0.5617100371747212, "grad_norm": 1.0386264015751407, "learning_rate": 4.259253622621188e-06, "loss": 0.028, "step": 12088 }, { "epoch": 0.5617565055762082, "grad_norm": 1.1502260783004647, "learning_rate": 4.2585094216591375e-06, "loss": 0.0746, "step": 12089 }, { "epoch": 0.5618029739776952, "grad_norm": 0.7330735501908457, "learning_rate": 4.257765237492627e-06, "loss": 0.0713, "step": 12090 }, { "epoch": 0.5618494423791821, "grad_norm": 0.731787424788209, "learning_rate": 4.257021070138513e-06, "loss": 0.0508, "step": 12091 }, { "epoch": 0.5618959107806691, "grad_norm": 0.7411145153294741, "learning_rate": 4.256276919613656e-06, "loss": 0.0709, "step": 12092 }, { "epoch": 0.5619423791821562, "grad_norm": 0.5798590327356571, "learning_rate": 4.255532785934908e-06, "loss": 0.0591, "step": 12093 }, { "epoch": 0.5619888475836431, "grad_norm": 0.6863816335837137, "learning_rate": 4.254788669119127e-06, "loss": 0.0573, "step": 12094 }, { "epoch": 0.5620353159851301, "grad_norm": 0.45958361985043006, "learning_rate": 4.254044569183169e-06, "loss": 0.0445, "step": 12095 }, { "epoch": 0.562081784386617, "grad_norm": 0.7131048883221509, "learning_rate": 4.253300486143884e-06, "loss": 0.0572, "step": 12096 }, { "epoch": 0.5621282527881041, "grad_norm": 1.0722494407859855, "learning_rate": 4.252556420018129e-06, "loss": 0.0937, "step": 12097 }, { "epoch": 0.5621747211895911, "grad_norm": 0.7194014140363686, "learning_rate": 4.251812370822757e-06, "loss": 0.044, "step": 12098 }, { "epoch": 0.562221189591078, "grad_norm": 0.6218660702889905, "learning_rate": 4.251068338574625e-06, "loss": 0.0619, "step": 12099 }, { "epoch": 0.5622676579925651, "grad_norm": 0.5431828070593197, "learning_rate": 4.250324323290581e-06, "loss": 0.0602, "step": 12100 }, { "epoch": 0.5623141263940521, "grad_norm": 0.8224301723090427, "learning_rate": 4.249580324987482e-06, "loss": 0.075, "step": 12101 }, { "epoch": 0.562360594795539, "grad_norm": 0.8940852138893314, "learning_rate": 4.2488363436821776e-06, "loss": 0.0642, "step": 12102 }, { "epoch": 0.562407063197026, "grad_norm": 1.564461650592424, "learning_rate": 4.248092379391519e-06, "loss": 0.0525, "step": 12103 }, { "epoch": 0.562453531598513, "grad_norm": 0.5793123695566246, "learning_rate": 4.247348432132361e-06, "loss": 0.075, "step": 12104 }, { "epoch": 0.5625, "grad_norm": 0.5768242111287024, "learning_rate": 4.246604501921553e-06, "loss": 0.0612, "step": 12105 }, { "epoch": 0.562546468401487, "grad_norm": 3.0819201432390075, "learning_rate": 4.245860588775947e-06, "loss": 0.0886, "step": 12106 }, { "epoch": 0.562592936802974, "grad_norm": 0.4390495905497012, "learning_rate": 4.245116692712391e-06, "loss": 0.035, "step": 12107 }, { "epoch": 0.562639405204461, "grad_norm": 0.5169522286199119, "learning_rate": 4.244372813747736e-06, "loss": 0.0362, "step": 12108 }, { "epoch": 0.5626858736059479, "grad_norm": 0.8582576667239257, "learning_rate": 4.243628951898832e-06, "loss": 0.0596, "step": 12109 }, { "epoch": 0.5627323420074349, "grad_norm": 5.121496226524879, "learning_rate": 4.242885107182529e-06, "loss": 0.0734, "step": 12110 }, { "epoch": 0.562778810408922, "grad_norm": 2.5147727854424224, "learning_rate": 4.242141279615675e-06, "loss": 0.0391, "step": 12111 }, { "epoch": 0.5628252788104089, "grad_norm": 0.6310181563008891, "learning_rate": 4.241397469215121e-06, "loss": 0.0547, "step": 12112 }, { "epoch": 0.5628717472118959, "grad_norm": 0.6709437975893685, "learning_rate": 4.240653675997709e-06, "loss": 0.0488, "step": 12113 }, { "epoch": 0.562918215613383, "grad_norm": 0.3949681693571137, "learning_rate": 4.239909899980293e-06, "loss": 0.0371, "step": 12114 }, { "epoch": 0.5629646840148699, "grad_norm": 1.2177146141227029, "learning_rate": 4.239166141179715e-06, "loss": 0.087, "step": 12115 }, { "epoch": 0.5630111524163569, "grad_norm": 0.49797037435383174, "learning_rate": 4.238422399612827e-06, "loss": 0.0619, "step": 12116 }, { "epoch": 0.5630576208178438, "grad_norm": 1.000551814667241, "learning_rate": 4.237678675296472e-06, "loss": 0.0682, "step": 12117 }, { "epoch": 0.5631040892193309, "grad_norm": 1.2743612958109, "learning_rate": 4.2369349682474995e-06, "loss": 0.0898, "step": 12118 }, { "epoch": 0.5631505576208179, "grad_norm": 1.0171976784357872, "learning_rate": 4.236191278482751e-06, "loss": 0.0529, "step": 12119 }, { "epoch": 0.5631970260223048, "grad_norm": 0.8538223169562763, "learning_rate": 4.235447606019074e-06, "loss": 0.0694, "step": 12120 }, { "epoch": 0.5632434944237918, "grad_norm": 0.6163121315729463, "learning_rate": 4.234703950873314e-06, "loss": 0.048, "step": 12121 }, { "epoch": 0.5632899628252788, "grad_norm": 1.2335057943770118, "learning_rate": 4.2339603130623145e-06, "loss": 0.0786, "step": 12122 }, { "epoch": 0.5633364312267658, "grad_norm": 0.4818057242936378, "learning_rate": 4.23321669260292e-06, "loss": 0.0459, "step": 12123 }, { "epoch": 0.5633828996282528, "grad_norm": 0.756452074311936, "learning_rate": 4.232473089511976e-06, "loss": 0.0529, "step": 12124 }, { "epoch": 0.5634293680297398, "grad_norm": 4.345453975263477, "learning_rate": 4.2317295038063225e-06, "loss": 0.0623, "step": 12125 }, { "epoch": 0.5634758364312268, "grad_norm": 1.0639074097292642, "learning_rate": 4.230985935502806e-06, "loss": 0.0649, "step": 12126 }, { "epoch": 0.5635223048327137, "grad_norm": 0.672977457719036, "learning_rate": 4.230242384618266e-06, "loss": 0.0576, "step": 12127 }, { "epoch": 0.5635687732342007, "grad_norm": 0.49339968629886793, "learning_rate": 4.2294988511695464e-06, "loss": 0.0356, "step": 12128 }, { "epoch": 0.5636152416356878, "grad_norm": 0.5833148710610198, "learning_rate": 4.228755335173488e-06, "loss": 0.0483, "step": 12129 }, { "epoch": 0.5636617100371747, "grad_norm": 0.47323293712388587, "learning_rate": 4.228011836646935e-06, "loss": 0.0526, "step": 12130 }, { "epoch": 0.5637081784386617, "grad_norm": 0.8125907261029621, "learning_rate": 4.227268355606726e-06, "loss": 0.0542, "step": 12131 }, { "epoch": 0.5637546468401488, "grad_norm": 0.7111827960990899, "learning_rate": 4.226524892069701e-06, "loss": 0.0627, "step": 12132 }, { "epoch": 0.5638011152416357, "grad_norm": 0.5305261853026078, "learning_rate": 4.225781446052702e-06, "loss": 0.0459, "step": 12133 }, { "epoch": 0.5638475836431227, "grad_norm": 0.9136043818119395, "learning_rate": 4.225038017572567e-06, "loss": 0.0612, "step": 12134 }, { "epoch": 0.5638940520446096, "grad_norm": 0.6476821999415041, "learning_rate": 4.224294606646137e-06, "loss": 0.0495, "step": 12135 }, { "epoch": 0.5639405204460967, "grad_norm": 0.7416101844554751, "learning_rate": 4.223551213290252e-06, "loss": 0.0754, "step": 12136 }, { "epoch": 0.5639869888475837, "grad_norm": 1.0069871023308756, "learning_rate": 4.222807837521748e-06, "loss": 0.0752, "step": 12137 }, { "epoch": 0.5640334572490706, "grad_norm": 1.043205768371329, "learning_rate": 4.222064479357464e-06, "loss": 0.0612, "step": 12138 }, { "epoch": 0.5640799256505576, "grad_norm": 0.559851022885841, "learning_rate": 4.221321138814238e-06, "loss": 0.0541, "step": 12139 }, { "epoch": 0.5641263940520446, "grad_norm": 0.3360498306338585, "learning_rate": 4.220577815908909e-06, "loss": 0.0398, "step": 12140 }, { "epoch": 0.5641728624535316, "grad_norm": 0.7573996010461542, "learning_rate": 4.219834510658311e-06, "loss": 0.048, "step": 12141 }, { "epoch": 0.5642193308550186, "grad_norm": 0.37244171128803916, "learning_rate": 4.219091223079286e-06, "loss": 0.0629, "step": 12142 }, { "epoch": 0.5642657992565056, "grad_norm": 1.0882333311944212, "learning_rate": 4.218347953188665e-06, "loss": 0.07, "step": 12143 }, { "epoch": 0.5643122676579926, "grad_norm": 0.49666636663224484, "learning_rate": 4.217604701003284e-06, "loss": 0.0483, "step": 12144 }, { "epoch": 0.5643587360594795, "grad_norm": 0.525153592765816, "learning_rate": 4.216861466539982e-06, "loss": 0.0476, "step": 12145 }, { "epoch": 0.5644052044609665, "grad_norm": 0.687018471841274, "learning_rate": 4.216118249815589e-06, "loss": 0.0653, "step": 12146 }, { "epoch": 0.5644516728624536, "grad_norm": 0.3363673202871922, "learning_rate": 4.215375050846945e-06, "loss": 0.0369, "step": 12147 }, { "epoch": 0.5644981412639405, "grad_norm": 0.574560471606629, "learning_rate": 4.2146318696508836e-06, "loss": 0.042, "step": 12148 }, { "epoch": 0.5645446096654275, "grad_norm": 0.47801377635145403, "learning_rate": 4.213888706244234e-06, "loss": 0.0606, "step": 12149 }, { "epoch": 0.5645910780669146, "grad_norm": 0.7298365005076763, "learning_rate": 4.2131455606438344e-06, "loss": 0.0657, "step": 12150 }, { "epoch": 0.5646375464684015, "grad_norm": 0.45009356653001653, "learning_rate": 4.212402432866514e-06, "loss": 0.0509, "step": 12151 }, { "epoch": 0.5646840148698885, "grad_norm": 0.42268109979065766, "learning_rate": 4.211659322929108e-06, "loss": 0.0446, "step": 12152 }, { "epoch": 0.5647304832713754, "grad_norm": 0.7740878619198737, "learning_rate": 4.210916230848449e-06, "loss": 0.0549, "step": 12153 }, { "epoch": 0.5647769516728625, "grad_norm": 0.4059534273847306, "learning_rate": 4.210173156641369e-06, "loss": 0.0388, "step": 12154 }, { "epoch": 0.5648234200743495, "grad_norm": 0.6744029906485887, "learning_rate": 4.209430100324696e-06, "loss": 0.039, "step": 12155 }, { "epoch": 0.5648698884758364, "grad_norm": 0.6839307697219106, "learning_rate": 4.208687061915264e-06, "loss": 0.051, "step": 12156 }, { "epoch": 0.5649163568773234, "grad_norm": 0.4122805648031878, "learning_rate": 4.207944041429903e-06, "loss": 0.0329, "step": 12157 }, { "epoch": 0.5649628252788104, "grad_norm": 0.6943691827569973, "learning_rate": 4.207201038885442e-06, "loss": 0.0571, "step": 12158 }, { "epoch": 0.5650092936802974, "grad_norm": 0.8436283197679182, "learning_rate": 4.206458054298713e-06, "loss": 0.084, "step": 12159 }, { "epoch": 0.5650557620817844, "grad_norm": 0.9145809321141971, "learning_rate": 4.205715087686546e-06, "loss": 0.0457, "step": 12160 }, { "epoch": 0.5651022304832714, "grad_norm": 0.8684331337076469, "learning_rate": 4.204972139065765e-06, "loss": 0.0656, "step": 12161 }, { "epoch": 0.5651486988847584, "grad_norm": 1.0274877261013138, "learning_rate": 4.204229208453204e-06, "loss": 0.0771, "step": 12162 }, { "epoch": 0.5651951672862453, "grad_norm": 0.7334943497828094, "learning_rate": 4.203486295865687e-06, "loss": 0.0672, "step": 12163 }, { "epoch": 0.5652416356877323, "grad_norm": 0.42931393987228916, "learning_rate": 4.202743401320045e-06, "loss": 0.0369, "step": 12164 }, { "epoch": 0.5652881040892194, "grad_norm": 0.47249597782266495, "learning_rate": 4.2020005248331056e-06, "loss": 0.0387, "step": 12165 }, { "epoch": 0.5653345724907063, "grad_norm": 0.8903199306864915, "learning_rate": 4.201257666421691e-06, "loss": 0.0748, "step": 12166 }, { "epoch": 0.5653810408921933, "grad_norm": 0.4963623810151176, "learning_rate": 4.200514826102633e-06, "loss": 0.0482, "step": 12167 }, { "epoch": 0.5654275092936804, "grad_norm": 0.4774898842684038, "learning_rate": 4.199772003892755e-06, "loss": 0.0552, "step": 12168 }, { "epoch": 0.5654739776951673, "grad_norm": 2.529578501398671, "learning_rate": 4.199029199808883e-06, "loss": 0.0762, "step": 12169 }, { "epoch": 0.5655204460966543, "grad_norm": 0.9267172766668554, "learning_rate": 4.198286413867843e-06, "loss": 0.0748, "step": 12170 }, { "epoch": 0.5655669144981412, "grad_norm": 0.5458807199577292, "learning_rate": 4.19754364608646e-06, "loss": 0.0473, "step": 12171 }, { "epoch": 0.5656133828996283, "grad_norm": 1.340357088712665, "learning_rate": 4.196800896481558e-06, "loss": 0.0602, "step": 12172 }, { "epoch": 0.5656598513011153, "grad_norm": 0.737284508912093, "learning_rate": 4.19605816506996e-06, "loss": 0.0647, "step": 12173 }, { "epoch": 0.5657063197026022, "grad_norm": 0.4946516345097207, "learning_rate": 4.195315451868491e-06, "loss": 0.054, "step": 12174 }, { "epoch": 0.5657527881040892, "grad_norm": 0.4606141680454046, "learning_rate": 4.194572756893973e-06, "loss": 0.033, "step": 12175 }, { "epoch": 0.5657992565055762, "grad_norm": 0.6633617452184285, "learning_rate": 4.193830080163232e-06, "loss": 0.044, "step": 12176 }, { "epoch": 0.5658457249070632, "grad_norm": 1.7850640197767988, "learning_rate": 4.193087421693089e-06, "loss": 0.0619, "step": 12177 }, { "epoch": 0.5658921933085502, "grad_norm": 0.549222068570912, "learning_rate": 4.192344781500362e-06, "loss": 0.0337, "step": 12178 }, { "epoch": 0.5659386617100371, "grad_norm": 0.30271325059281756, "learning_rate": 4.191602159601878e-06, "loss": 0.0328, "step": 12179 }, { "epoch": 0.5659851301115242, "grad_norm": 0.6176951984677187, "learning_rate": 4.190859556014453e-06, "loss": 0.0695, "step": 12180 }, { "epoch": 0.5660315985130111, "grad_norm": 0.4932933984609514, "learning_rate": 4.190116970754913e-06, "loss": 0.0481, "step": 12181 }, { "epoch": 0.5660780669144981, "grad_norm": 0.8476744143038484, "learning_rate": 4.189374403840075e-06, "loss": 0.0658, "step": 12182 }, { "epoch": 0.5661245353159852, "grad_norm": 0.6955306248958583, "learning_rate": 4.188631855286762e-06, "loss": 0.0372, "step": 12183 }, { "epoch": 0.5661710037174721, "grad_norm": 8.465381079184697, "learning_rate": 4.18788932511179e-06, "loss": 0.0553, "step": 12184 }, { "epoch": 0.5662174721189591, "grad_norm": 0.3176528204534253, "learning_rate": 4.187146813331979e-06, "loss": 0.0268, "step": 12185 }, { "epoch": 0.5662639405204462, "grad_norm": 0.7110276261393369, "learning_rate": 4.186404319964148e-06, "loss": 0.0337, "step": 12186 }, { "epoch": 0.5663104089219331, "grad_norm": 0.6351539653890579, "learning_rate": 4.185661845025115e-06, "loss": 0.0512, "step": 12187 }, { "epoch": 0.5663568773234201, "grad_norm": 0.7465464701899052, "learning_rate": 4.184919388531701e-06, "loss": 0.0536, "step": 12188 }, { "epoch": 0.566403345724907, "grad_norm": 0.6649272839527576, "learning_rate": 4.18417695050072e-06, "loss": 0.0684, "step": 12189 }, { "epoch": 0.5664498141263941, "grad_norm": 0.5678302364592929, "learning_rate": 4.183434530948988e-06, "loss": 0.049, "step": 12190 }, { "epoch": 0.5664962825278811, "grad_norm": 0.7190583156473274, "learning_rate": 4.1826921298933245e-06, "loss": 0.0442, "step": 12191 }, { "epoch": 0.566542750929368, "grad_norm": 0.423203240041561, "learning_rate": 4.181949747350544e-06, "loss": 0.0456, "step": 12192 }, { "epoch": 0.566589219330855, "grad_norm": 0.677094805822256, "learning_rate": 4.181207383337462e-06, "loss": 0.067, "step": 12193 }, { "epoch": 0.566635687732342, "grad_norm": 1.1906897223490398, "learning_rate": 4.180465037870894e-06, "loss": 0.0716, "step": 12194 }, { "epoch": 0.566682156133829, "grad_norm": 0.7811598186680048, "learning_rate": 4.1797227109676585e-06, "loss": 0.0693, "step": 12195 }, { "epoch": 0.566728624535316, "grad_norm": 0.87540051845612, "learning_rate": 4.178980402644566e-06, "loss": 0.0367, "step": 12196 }, { "epoch": 0.566775092936803, "grad_norm": 0.35992443275807595, "learning_rate": 4.178238112918429e-06, "loss": 0.0416, "step": 12197 }, { "epoch": 0.56682156133829, "grad_norm": 0.5919599381242002, "learning_rate": 4.177495841806066e-06, "loss": 0.0578, "step": 12198 }, { "epoch": 0.5668680297397769, "grad_norm": 0.5632056647113172, "learning_rate": 4.176753589324286e-06, "loss": 0.0656, "step": 12199 }, { "epoch": 0.5669144981412639, "grad_norm": 0.6839748064452941, "learning_rate": 4.176011355489905e-06, "loss": 0.0704, "step": 12200 }, { "epoch": 0.566960966542751, "grad_norm": 0.5966715459614225, "learning_rate": 4.175269140319735e-06, "loss": 0.0476, "step": 12201 }, { "epoch": 0.5670074349442379, "grad_norm": 1.8849505499550012, "learning_rate": 4.174526943830585e-06, "loss": 0.0601, "step": 12202 }, { "epoch": 0.5670539033457249, "grad_norm": 6.526794983039293, "learning_rate": 4.17378476603927e-06, "loss": 0.0795, "step": 12203 }, { "epoch": 0.567100371747212, "grad_norm": 0.24855911941974831, "learning_rate": 4.173042606962598e-06, "loss": 0.0439, "step": 12204 }, { "epoch": 0.5671468401486989, "grad_norm": 0.454183959808287, "learning_rate": 4.1723004666173825e-06, "loss": 0.04, "step": 12205 }, { "epoch": 0.5671933085501859, "grad_norm": 0.6441240473951498, "learning_rate": 4.171558345020431e-06, "loss": 0.0645, "step": 12206 }, { "epoch": 0.5672397769516728, "grad_norm": 0.46500373325216643, "learning_rate": 4.170816242188558e-06, "loss": 0.0424, "step": 12207 }, { "epoch": 0.5672862453531599, "grad_norm": 1.0226245897304163, "learning_rate": 4.170074158138568e-06, "loss": 0.039, "step": 12208 }, { "epoch": 0.5673327137546469, "grad_norm": 0.5580356172612833, "learning_rate": 4.169332092887271e-06, "loss": 0.0643, "step": 12209 }, { "epoch": 0.5673791821561338, "grad_norm": 0.42542541221138147, "learning_rate": 4.168590046451478e-06, "loss": 0.0417, "step": 12210 }, { "epoch": 0.5674256505576208, "grad_norm": 0.5197088558595895, "learning_rate": 4.167848018847994e-06, "loss": 0.0657, "step": 12211 }, { "epoch": 0.5674721189591078, "grad_norm": 0.5304077344403119, "learning_rate": 4.167106010093629e-06, "loss": 0.056, "step": 12212 }, { "epoch": 0.5675185873605948, "grad_norm": 0.7125284852431283, "learning_rate": 4.16636402020519e-06, "loss": 0.0408, "step": 12213 }, { "epoch": 0.5675650557620818, "grad_norm": 0.7371578975170235, "learning_rate": 4.165622049199482e-06, "loss": 0.0461, "step": 12214 }, { "epoch": 0.5676115241635687, "grad_norm": 1.1101037036217707, "learning_rate": 4.164880097093314e-06, "loss": 0.0578, "step": 12215 }, { "epoch": 0.5676579925650558, "grad_norm": 0.6803800662728614, "learning_rate": 4.1641381639034895e-06, "loss": 0.0563, "step": 12216 }, { "epoch": 0.5677044609665427, "grad_norm": 0.55642994803599, "learning_rate": 4.1633962496468165e-06, "loss": 0.0336, "step": 12217 }, { "epoch": 0.5677509293680297, "grad_norm": 0.4593545197061994, "learning_rate": 4.1626543543400975e-06, "loss": 0.0423, "step": 12218 }, { "epoch": 0.5677973977695168, "grad_norm": 0.6036753057507981, "learning_rate": 4.161912478000141e-06, "loss": 0.0511, "step": 12219 }, { "epoch": 0.5678438661710037, "grad_norm": 0.881495148097021, "learning_rate": 4.161170620643747e-06, "loss": 0.0605, "step": 12220 }, { "epoch": 0.5678903345724907, "grad_norm": 0.5763605856521887, "learning_rate": 4.1604287822877214e-06, "loss": 0.0539, "step": 12221 }, { "epoch": 0.5679368029739778, "grad_norm": 0.5630155650407459, "learning_rate": 4.1596869629488684e-06, "loss": 0.0606, "step": 12222 }, { "epoch": 0.5679832713754647, "grad_norm": 0.8862991235902208, "learning_rate": 4.158945162643988e-06, "loss": 0.0672, "step": 12223 }, { "epoch": 0.5680297397769517, "grad_norm": 0.6748510909766434, "learning_rate": 4.158203381389889e-06, "loss": 0.0524, "step": 12224 }, { "epoch": 0.5680762081784386, "grad_norm": 1.1269051999754722, "learning_rate": 4.157461619203367e-06, "loss": 0.0406, "step": 12225 }, { "epoch": 0.5681226765799257, "grad_norm": 0.7278486184863222, "learning_rate": 4.156719876101225e-06, "loss": 0.0692, "step": 12226 }, { "epoch": 0.5681691449814127, "grad_norm": 0.516389191472215, "learning_rate": 4.155978152100266e-06, "loss": 0.0531, "step": 12227 }, { "epoch": 0.5682156133828996, "grad_norm": 0.4517848695203378, "learning_rate": 4.1552364472172904e-06, "loss": 0.046, "step": 12228 }, { "epoch": 0.5682620817843866, "grad_norm": 0.6840415898322458, "learning_rate": 4.1544947614690984e-06, "loss": 0.0574, "step": 12229 }, { "epoch": 0.5683085501858736, "grad_norm": 0.9066199633421044, "learning_rate": 4.153753094872492e-06, "loss": 0.0659, "step": 12230 }, { "epoch": 0.5683550185873606, "grad_norm": 0.44692553307638516, "learning_rate": 4.153011447444265e-06, "loss": 0.0596, "step": 12231 }, { "epoch": 0.5684014869888476, "grad_norm": 2.3704974078878704, "learning_rate": 4.152269819201222e-06, "loss": 0.0556, "step": 12232 }, { "epoch": 0.5684479553903345, "grad_norm": 1.0006522164156264, "learning_rate": 4.151528210160159e-06, "loss": 0.0538, "step": 12233 }, { "epoch": 0.5684944237918216, "grad_norm": 0.5753787801160356, "learning_rate": 4.150786620337876e-06, "loss": 0.0338, "step": 12234 }, { "epoch": 0.5685408921933085, "grad_norm": 0.5041180072526923, "learning_rate": 4.1500450497511695e-06, "loss": 0.0587, "step": 12235 }, { "epoch": 0.5685873605947955, "grad_norm": 0.5128315869938258, "learning_rate": 4.149303498416838e-06, "loss": 0.0475, "step": 12236 }, { "epoch": 0.5686338289962826, "grad_norm": 0.40328352376375826, "learning_rate": 4.148561966351677e-06, "loss": 0.0483, "step": 12237 }, { "epoch": 0.5686802973977695, "grad_norm": 0.6636752202119084, "learning_rate": 4.147820453572482e-06, "loss": 0.0564, "step": 12238 }, { "epoch": 0.5687267657992565, "grad_norm": 1.1383725836085459, "learning_rate": 4.147078960096052e-06, "loss": 0.0542, "step": 12239 }, { "epoch": 0.5687732342007435, "grad_norm": 0.47575069787266006, "learning_rate": 4.146337485939182e-06, "loss": 0.0551, "step": 12240 }, { "epoch": 0.5688197026022305, "grad_norm": 0.4215210586710705, "learning_rate": 4.1455960311186645e-06, "loss": 0.0304, "step": 12241 }, { "epoch": 0.5688661710037175, "grad_norm": 0.3932728272890137, "learning_rate": 4.144854595651299e-06, "loss": 0.0414, "step": 12242 }, { "epoch": 0.5689126394052044, "grad_norm": 0.7067826398158269, "learning_rate": 4.144113179553874e-06, "loss": 0.0558, "step": 12243 }, { "epoch": 0.5689591078066915, "grad_norm": 0.8965354083069416, "learning_rate": 4.143371782843189e-06, "loss": 0.0635, "step": 12244 }, { "epoch": 0.5690055762081785, "grad_norm": 0.33339035986173005, "learning_rate": 4.142630405536031e-06, "loss": 0.0323, "step": 12245 }, { "epoch": 0.5690520446096654, "grad_norm": 0.6055407565123039, "learning_rate": 4.1418890476492e-06, "loss": 0.0651, "step": 12246 }, { "epoch": 0.5690985130111524, "grad_norm": 0.7189700719690127, "learning_rate": 4.141147709199483e-06, "loss": 0.0458, "step": 12247 }, { "epoch": 0.5691449814126394, "grad_norm": 0.6435796840159436, "learning_rate": 4.140406390203677e-06, "loss": 0.0711, "step": 12248 }, { "epoch": 0.5691914498141264, "grad_norm": 0.6338150214305557, "learning_rate": 4.13966509067857e-06, "loss": 0.0509, "step": 12249 }, { "epoch": 0.5692379182156134, "grad_norm": 0.7339592137812538, "learning_rate": 4.1389238106409524e-06, "loss": 0.0675, "step": 12250 }, { "epoch": 0.5692843866171003, "grad_norm": 0.6695307749388684, "learning_rate": 4.138182550107618e-06, "loss": 0.0616, "step": 12251 }, { "epoch": 0.5693308550185874, "grad_norm": 0.8080206014231545, "learning_rate": 4.137441309095355e-06, "loss": 0.0603, "step": 12252 }, { "epoch": 0.5693773234200743, "grad_norm": 0.37604916914758624, "learning_rate": 4.136700087620955e-06, "loss": 0.0412, "step": 12253 }, { "epoch": 0.5694237918215613, "grad_norm": 0.6810307009087538, "learning_rate": 4.135958885701208e-06, "loss": 0.0421, "step": 12254 }, { "epoch": 0.5694702602230484, "grad_norm": 0.5544771135905749, "learning_rate": 4.1352177033529e-06, "loss": 0.0584, "step": 12255 }, { "epoch": 0.5695167286245353, "grad_norm": 0.6128600781587876, "learning_rate": 4.134476540592821e-06, "loss": 0.0515, "step": 12256 }, { "epoch": 0.5695631970260223, "grad_norm": 0.3251601458761122, "learning_rate": 4.133735397437759e-06, "loss": 0.0344, "step": 12257 }, { "epoch": 0.5696096654275093, "grad_norm": 0.6263645957973731, "learning_rate": 4.132994273904504e-06, "loss": 0.0462, "step": 12258 }, { "epoch": 0.5696561338289963, "grad_norm": 0.5366485703670418, "learning_rate": 4.132253170009839e-06, "loss": 0.0487, "step": 12259 }, { "epoch": 0.5697026022304833, "grad_norm": 0.35793818285791384, "learning_rate": 4.131512085770555e-06, "loss": 0.0202, "step": 12260 }, { "epoch": 0.5697490706319702, "grad_norm": 0.5929876062947153, "learning_rate": 4.1307710212034354e-06, "loss": 0.0555, "step": 12261 }, { "epoch": 0.5697955390334573, "grad_norm": 0.42054016687541107, "learning_rate": 4.130029976325266e-06, "loss": 0.0472, "step": 12262 }, { "epoch": 0.5698420074349443, "grad_norm": 0.9993311568664675, "learning_rate": 4.129288951152834e-06, "loss": 0.0612, "step": 12263 }, { "epoch": 0.5698884758364312, "grad_norm": 0.2632533193066893, "learning_rate": 4.128547945702923e-06, "loss": 0.0314, "step": 12264 }, { "epoch": 0.5699349442379182, "grad_norm": 1.3339897044754543, "learning_rate": 4.1278069599923195e-06, "loss": 0.052, "step": 12265 }, { "epoch": 0.5699814126394052, "grad_norm": 0.6291388099089468, "learning_rate": 4.127065994037807e-06, "loss": 0.0565, "step": 12266 }, { "epoch": 0.5700278810408922, "grad_norm": 0.5706810396690688, "learning_rate": 4.126325047856167e-06, "loss": 0.0466, "step": 12267 }, { "epoch": 0.5700743494423792, "grad_norm": 0.539077528172689, "learning_rate": 4.1255841214641845e-06, "loss": 0.0354, "step": 12268 }, { "epoch": 0.5701208178438661, "grad_norm": 0.6857091297234909, "learning_rate": 4.124843214878641e-06, "loss": 0.0701, "step": 12269 }, { "epoch": 0.5701672862453532, "grad_norm": 0.39770550385419345, "learning_rate": 4.1241023281163205e-06, "loss": 0.0386, "step": 12270 }, { "epoch": 0.5702137546468401, "grad_norm": 0.7303297996967772, "learning_rate": 4.123361461194004e-06, "loss": 0.0579, "step": 12271 }, { "epoch": 0.5702602230483271, "grad_norm": 0.601727849059671, "learning_rate": 4.122620614128474e-06, "loss": 0.0625, "step": 12272 }, { "epoch": 0.5703066914498142, "grad_norm": 0.7326711469943705, "learning_rate": 4.1218797869365106e-06, "loss": 0.0606, "step": 12273 }, { "epoch": 0.5703531598513011, "grad_norm": 0.5496137957085, "learning_rate": 4.121138979634893e-06, "loss": 0.0652, "step": 12274 }, { "epoch": 0.5703996282527881, "grad_norm": 0.6361923151948897, "learning_rate": 4.120398192240403e-06, "loss": 0.0547, "step": 12275 }, { "epoch": 0.5704460966542751, "grad_norm": 0.5752868476576498, "learning_rate": 4.119657424769819e-06, "loss": 0.0493, "step": 12276 }, { "epoch": 0.5704925650557621, "grad_norm": 0.6158823621477685, "learning_rate": 4.118916677239922e-06, "loss": 0.0554, "step": 12277 }, { "epoch": 0.5705390334572491, "grad_norm": 0.8309555342247595, "learning_rate": 4.11817594966749e-06, "loss": 0.0664, "step": 12278 }, { "epoch": 0.570585501858736, "grad_norm": 0.4618907938921045, "learning_rate": 4.1174352420692995e-06, "loss": 0.0423, "step": 12279 }, { "epoch": 0.570631970260223, "grad_norm": 0.6324468311351673, "learning_rate": 4.1166945544621304e-06, "loss": 0.0622, "step": 12280 }, { "epoch": 0.5706784386617101, "grad_norm": 0.6537714438412446, "learning_rate": 4.115953886862758e-06, "loss": 0.0392, "step": 12281 }, { "epoch": 0.570724907063197, "grad_norm": 1.0747442078253242, "learning_rate": 4.115213239287963e-06, "loss": 0.0829, "step": 12282 }, { "epoch": 0.570771375464684, "grad_norm": 0.7320818382750487, "learning_rate": 4.114472611754518e-06, "loss": 0.0821, "step": 12283 }, { "epoch": 0.570817843866171, "grad_norm": 0.5882822336195541, "learning_rate": 4.1137320042792004e-06, "loss": 0.0571, "step": 12284 }, { "epoch": 0.570864312267658, "grad_norm": 0.6640190261936516, "learning_rate": 4.112991416878785e-06, "loss": 0.0588, "step": 12285 }, { "epoch": 0.570910780669145, "grad_norm": 0.8484676096282434, "learning_rate": 4.1122508495700476e-06, "loss": 0.0638, "step": 12286 }, { "epoch": 0.570957249070632, "grad_norm": 1.22790498294708, "learning_rate": 4.111510302369764e-06, "loss": 0.0581, "step": 12287 }, { "epoch": 0.571003717472119, "grad_norm": 1.0074258569981376, "learning_rate": 4.1107697752947055e-06, "loss": 0.0404, "step": 12288 }, { "epoch": 0.5710501858736059, "grad_norm": 0.8287923977410754, "learning_rate": 4.1100292683616495e-06, "loss": 0.0475, "step": 12289 }, { "epoch": 0.5710966542750929, "grad_norm": 0.4208076660155023, "learning_rate": 4.109288781587367e-06, "loss": 0.0413, "step": 12290 }, { "epoch": 0.57114312267658, "grad_norm": 0.8510914937741824, "learning_rate": 4.108548314988629e-06, "loss": 0.0708, "step": 12291 }, { "epoch": 0.5711895910780669, "grad_norm": 0.6090483930509852, "learning_rate": 4.1078078685822115e-06, "loss": 0.0544, "step": 12292 }, { "epoch": 0.5712360594795539, "grad_norm": 0.6366869667890659, "learning_rate": 4.107067442384884e-06, "loss": 0.0712, "step": 12293 }, { "epoch": 0.571282527881041, "grad_norm": 0.5142892710297342, "learning_rate": 4.106327036413419e-06, "loss": 0.0443, "step": 12294 }, { "epoch": 0.5713289962825279, "grad_norm": 0.6346043065971658, "learning_rate": 4.105586650684589e-06, "loss": 0.0551, "step": 12295 }, { "epoch": 0.5713754646840149, "grad_norm": 0.6131467291682848, "learning_rate": 4.10484628521516e-06, "loss": 0.0584, "step": 12296 }, { "epoch": 0.5714219330855018, "grad_norm": 0.7844300672801805, "learning_rate": 4.104105940021906e-06, "loss": 0.0548, "step": 12297 }, { "epoch": 0.5714684014869889, "grad_norm": 0.481392420677165, "learning_rate": 4.103365615121595e-06, "loss": 0.0448, "step": 12298 }, { "epoch": 0.5715148698884759, "grad_norm": 0.43148709151673686, "learning_rate": 4.1026253105309965e-06, "loss": 0.0443, "step": 12299 }, { "epoch": 0.5715613382899628, "grad_norm": 0.7231027841680511, "learning_rate": 4.101885026266879e-06, "loss": 0.0695, "step": 12300 }, { "epoch": 0.5716078066914498, "grad_norm": 0.7935294279996803, "learning_rate": 4.101144762346013e-06, "loss": 0.0592, "step": 12301 }, { "epoch": 0.5716542750929368, "grad_norm": 0.43379534751936955, "learning_rate": 4.1004045187851625e-06, "loss": 0.0464, "step": 12302 }, { "epoch": 0.5717007434944238, "grad_norm": 1.0165336641877676, "learning_rate": 4.099664295601096e-06, "loss": 0.0759, "step": 12303 }, { "epoch": 0.5717472118959108, "grad_norm": 0.549951747184248, "learning_rate": 4.098924092810581e-06, "loss": 0.0655, "step": 12304 }, { "epoch": 0.5717936802973977, "grad_norm": 0.874615282127888, "learning_rate": 4.098183910430384e-06, "loss": 0.0438, "step": 12305 }, { "epoch": 0.5718401486988848, "grad_norm": 0.6889168583772742, "learning_rate": 4.09744374847727e-06, "loss": 0.0621, "step": 12306 }, { "epoch": 0.5718866171003717, "grad_norm": 0.5778808226200621, "learning_rate": 4.096703606968007e-06, "loss": 0.0443, "step": 12307 }, { "epoch": 0.5719330855018587, "grad_norm": 0.8092133632489418, "learning_rate": 4.095963485919355e-06, "loss": 0.0519, "step": 12308 }, { "epoch": 0.5719795539033458, "grad_norm": 0.39454755458939406, "learning_rate": 4.095223385348084e-06, "loss": 0.0259, "step": 12309 }, { "epoch": 0.5720260223048327, "grad_norm": 0.46108480679611336, "learning_rate": 4.094483305270953e-06, "loss": 0.0339, "step": 12310 }, { "epoch": 0.5720724907063197, "grad_norm": 0.5374262041581009, "learning_rate": 4.093743245704731e-06, "loss": 0.0502, "step": 12311 }, { "epoch": 0.5721189591078067, "grad_norm": 0.4641613881746028, "learning_rate": 4.093003206666177e-06, "loss": 0.0586, "step": 12312 }, { "epoch": 0.5721654275092937, "grad_norm": 0.5296788638798811, "learning_rate": 4.092263188172056e-06, "loss": 0.0694, "step": 12313 }, { "epoch": 0.5722118959107807, "grad_norm": 0.5273348116301495, "learning_rate": 4.091523190239129e-06, "loss": 0.0339, "step": 12314 }, { "epoch": 0.5722583643122676, "grad_norm": 0.5280107762427315, "learning_rate": 4.090783212884155e-06, "loss": 0.059, "step": 12315 }, { "epoch": 0.5723048327137547, "grad_norm": 0.9018741535115427, "learning_rate": 4.0900432561239015e-06, "loss": 0.0621, "step": 12316 }, { "epoch": 0.5723513011152417, "grad_norm": 0.5769427316452344, "learning_rate": 4.089303319975123e-06, "loss": 0.0677, "step": 12317 }, { "epoch": 0.5723977695167286, "grad_norm": 0.32934665513793704, "learning_rate": 4.0885634044545856e-06, "loss": 0.0535, "step": 12318 }, { "epoch": 0.5724442379182156, "grad_norm": 0.47894536572067076, "learning_rate": 4.087823509579047e-06, "loss": 0.0544, "step": 12319 }, { "epoch": 0.5724907063197026, "grad_norm": 0.9911988520505632, "learning_rate": 4.087083635365263e-06, "loss": 0.0798, "step": 12320 }, { "epoch": 0.5725371747211896, "grad_norm": 0.9035645085787549, "learning_rate": 4.086343781829997e-06, "loss": 0.0562, "step": 12321 }, { "epoch": 0.5725836431226766, "grad_norm": 0.6346814444974369, "learning_rate": 4.085603948990004e-06, "loss": 0.0541, "step": 12322 }, { "epoch": 0.5726301115241635, "grad_norm": 0.6757702910892455, "learning_rate": 4.084864136862047e-06, "loss": 0.0635, "step": 12323 }, { "epoch": 0.5726765799256506, "grad_norm": 0.5748871697961511, "learning_rate": 4.084124345462877e-06, "loss": 0.0533, "step": 12324 }, { "epoch": 0.5727230483271375, "grad_norm": 0.7499409083827774, "learning_rate": 4.0833845748092586e-06, "loss": 0.0534, "step": 12325 }, { "epoch": 0.5727695167286245, "grad_norm": 0.4766246876193082, "learning_rate": 4.082644824917943e-06, "loss": 0.0451, "step": 12326 }, { "epoch": 0.5728159851301116, "grad_norm": 0.3831453999441988, "learning_rate": 4.081905095805686e-06, "loss": 0.0338, "step": 12327 }, { "epoch": 0.5728624535315985, "grad_norm": 0.951302787310459, "learning_rate": 4.081165387489246e-06, "loss": 0.0633, "step": 12328 }, { "epoch": 0.5729089219330855, "grad_norm": 0.4712586765361605, "learning_rate": 4.0804256999853755e-06, "loss": 0.0496, "step": 12329 }, { "epoch": 0.5729553903345725, "grad_norm": 0.45714759286785944, "learning_rate": 4.079686033310832e-06, "loss": 0.0611, "step": 12330 }, { "epoch": 0.5730018587360595, "grad_norm": 1.0705472061203858, "learning_rate": 4.078946387482371e-06, "loss": 0.0502, "step": 12331 }, { "epoch": 0.5730483271375465, "grad_norm": 0.7322342455129333, "learning_rate": 4.07820676251674e-06, "loss": 0.0604, "step": 12332 }, { "epoch": 0.5730947955390334, "grad_norm": 0.5467198772289594, "learning_rate": 4.077467158430698e-06, "loss": 0.0355, "step": 12333 }, { "epoch": 0.5731412639405205, "grad_norm": 0.6398644390801741, "learning_rate": 4.076727575240994e-06, "loss": 0.0563, "step": 12334 }, { "epoch": 0.5731877323420075, "grad_norm": 0.5939664172964697, "learning_rate": 4.075988012964383e-06, "loss": 0.047, "step": 12335 }, { "epoch": 0.5732342007434944, "grad_norm": 0.657760850523536, "learning_rate": 4.0752484716176155e-06, "loss": 0.0636, "step": 12336 }, { "epoch": 0.5732806691449814, "grad_norm": 0.6354206612911689, "learning_rate": 4.074508951217445e-06, "loss": 0.0614, "step": 12337 }, { "epoch": 0.5733271375464684, "grad_norm": 0.5386403696545651, "learning_rate": 4.073769451780621e-06, "loss": 0.0251, "step": 12338 }, { "epoch": 0.5733736059479554, "grad_norm": 1.2798926464067988, "learning_rate": 4.073029973323892e-06, "loss": 0.0813, "step": 12339 }, { "epoch": 0.5734200743494424, "grad_norm": 0.42909531754651503, "learning_rate": 4.07229051586401e-06, "loss": 0.0188, "step": 12340 }, { "epoch": 0.5734665427509293, "grad_norm": 0.44822770005045487, "learning_rate": 4.071551079417724e-06, "loss": 0.0464, "step": 12341 }, { "epoch": 0.5735130111524164, "grad_norm": 0.8802035153616873, "learning_rate": 4.070811664001785e-06, "loss": 0.062, "step": 12342 }, { "epoch": 0.5735594795539033, "grad_norm": 0.5294638602870908, "learning_rate": 4.070072269632939e-06, "loss": 0.0771, "step": 12343 }, { "epoch": 0.5736059479553903, "grad_norm": 0.687259075622159, "learning_rate": 4.069332896327933e-06, "loss": 0.0524, "step": 12344 }, { "epoch": 0.5736524163568774, "grad_norm": 0.37686331289422387, "learning_rate": 4.068593544103518e-06, "loss": 0.047, "step": 12345 }, { "epoch": 0.5736988847583643, "grad_norm": 0.6357018505292368, "learning_rate": 4.067854212976439e-06, "loss": 0.0574, "step": 12346 }, { "epoch": 0.5737453531598513, "grad_norm": 0.6731430781518868, "learning_rate": 4.067114902963444e-06, "loss": 0.0426, "step": 12347 }, { "epoch": 0.5737918215613383, "grad_norm": 0.6070163002818181, "learning_rate": 4.066375614081279e-06, "loss": 0.0596, "step": 12348 }, { "epoch": 0.5738382899628253, "grad_norm": 0.6852036632082235, "learning_rate": 4.065636346346687e-06, "loss": 0.0474, "step": 12349 }, { "epoch": 0.5738847583643123, "grad_norm": 0.42870581443374073, "learning_rate": 4.064897099776416e-06, "loss": 0.0389, "step": 12350 }, { "epoch": 0.5739312267657992, "grad_norm": 0.31940871314177866, "learning_rate": 4.064157874387209e-06, "loss": 0.0231, "step": 12351 }, { "epoch": 0.5739776951672863, "grad_norm": 0.5876407630312576, "learning_rate": 4.063418670195812e-06, "loss": 0.0311, "step": 12352 }, { "epoch": 0.5740241635687733, "grad_norm": 0.7834599815784539, "learning_rate": 4.062679487218966e-06, "loss": 0.0738, "step": 12353 }, { "epoch": 0.5740706319702602, "grad_norm": 0.40881311960148387, "learning_rate": 4.06194032547342e-06, "loss": 0.046, "step": 12354 }, { "epoch": 0.5741171003717472, "grad_norm": 0.6721647058410941, "learning_rate": 4.061201184975912e-06, "loss": 0.0638, "step": 12355 }, { "epoch": 0.5741635687732342, "grad_norm": 1.0097081940769947, "learning_rate": 4.060462065743183e-06, "loss": 0.0488, "step": 12356 }, { "epoch": 0.5742100371747212, "grad_norm": 0.5147705715064974, "learning_rate": 4.059722967791978e-06, "loss": 0.0646, "step": 12357 }, { "epoch": 0.5742565055762082, "grad_norm": 0.6939571010838607, "learning_rate": 4.058983891139036e-06, "loss": 0.0464, "step": 12358 }, { "epoch": 0.5743029739776951, "grad_norm": 0.8993139092290686, "learning_rate": 4.058244835801101e-06, "loss": 0.0599, "step": 12359 }, { "epoch": 0.5743494423791822, "grad_norm": 0.6234245833574734, "learning_rate": 4.057505801794913e-06, "loss": 0.0482, "step": 12360 }, { "epoch": 0.5743959107806691, "grad_norm": 0.4817836324086077, "learning_rate": 4.056766789137208e-06, "loss": 0.0404, "step": 12361 }, { "epoch": 0.5744423791821561, "grad_norm": 0.6738422571722552, "learning_rate": 4.05602779784473e-06, "loss": 0.0658, "step": 12362 }, { "epoch": 0.5744888475836432, "grad_norm": 0.3344349866298567, "learning_rate": 4.055288827934213e-06, "loss": 0.0506, "step": 12363 }, { "epoch": 0.5745353159851301, "grad_norm": 1.1943135931101592, "learning_rate": 4.0545498794224e-06, "loss": 0.0913, "step": 12364 }, { "epoch": 0.5745817843866171, "grad_norm": 0.6164658970794877, "learning_rate": 4.053810952326027e-06, "loss": 0.053, "step": 12365 }, { "epoch": 0.5746282527881041, "grad_norm": 0.46714986654011725, "learning_rate": 4.053072046661833e-06, "loss": 0.036, "step": 12366 }, { "epoch": 0.5746747211895911, "grad_norm": 0.8835388032581644, "learning_rate": 4.052333162446553e-06, "loss": 0.0771, "step": 12367 }, { "epoch": 0.5747211895910781, "grad_norm": 0.708145094173545, "learning_rate": 4.051594299696924e-06, "loss": 0.063, "step": 12368 }, { "epoch": 0.574767657992565, "grad_norm": 1.0531680789470197, "learning_rate": 4.0508554584296825e-06, "loss": 0.0594, "step": 12369 }, { "epoch": 0.574814126394052, "grad_norm": 0.50308894197331, "learning_rate": 4.050116638661563e-06, "loss": 0.0599, "step": 12370 }, { "epoch": 0.5748605947955391, "grad_norm": 0.6722317157567633, "learning_rate": 4.049377840409303e-06, "loss": 0.0647, "step": 12371 }, { "epoch": 0.574907063197026, "grad_norm": 0.6003093654600803, "learning_rate": 4.048639063689635e-06, "loss": 0.0781, "step": 12372 }, { "epoch": 0.574953531598513, "grad_norm": 0.4710963521999626, "learning_rate": 4.047900308519293e-06, "loss": 0.0398, "step": 12373 }, { "epoch": 0.575, "grad_norm": 0.6028585859249888, "learning_rate": 4.047161574915012e-06, "loss": 0.0487, "step": 12374 }, { "epoch": 0.575046468401487, "grad_norm": 0.533184370185004, "learning_rate": 4.046422862893522e-06, "loss": 0.0429, "step": 12375 }, { "epoch": 0.575092936802974, "grad_norm": 1.1634697878164482, "learning_rate": 4.045684172471559e-06, "loss": 0.0918, "step": 12376 }, { "epoch": 0.5751394052044609, "grad_norm": 0.3436718710909304, "learning_rate": 4.044945503665854e-06, "loss": 0.0454, "step": 12377 }, { "epoch": 0.575185873605948, "grad_norm": 0.5111631045792753, "learning_rate": 4.04420685649314e-06, "loss": 0.0503, "step": 12378 }, { "epoch": 0.5752323420074349, "grad_norm": 0.4630786802137441, "learning_rate": 4.0434682309701454e-06, "loss": 0.0438, "step": 12379 }, { "epoch": 0.5752788104089219, "grad_norm": 0.5249447707394155, "learning_rate": 4.042729627113601e-06, "loss": 0.0227, "step": 12380 }, { "epoch": 0.575325278810409, "grad_norm": 0.7158037781094059, "learning_rate": 4.0419910449402385e-06, "loss": 0.029, "step": 12381 }, { "epoch": 0.5753717472118959, "grad_norm": 0.8615086468051041, "learning_rate": 4.041252484466787e-06, "loss": 0.0478, "step": 12382 }, { "epoch": 0.5754182156133829, "grad_norm": 0.8871015387210032, "learning_rate": 4.040513945709977e-06, "loss": 0.0397, "step": 12383 }, { "epoch": 0.5754646840148699, "grad_norm": 0.614376927445874, "learning_rate": 4.039775428686536e-06, "loss": 0.0671, "step": 12384 }, { "epoch": 0.5755111524163569, "grad_norm": 0.7965246054573537, "learning_rate": 4.039036933413191e-06, "loss": 0.0728, "step": 12385 }, { "epoch": 0.5755576208178439, "grad_norm": 0.5992814790514677, "learning_rate": 4.038298459906672e-06, "loss": 0.0469, "step": 12386 }, { "epoch": 0.5756040892193308, "grad_norm": 0.3786641733642272, "learning_rate": 4.0375600081837035e-06, "loss": 0.0289, "step": 12387 }, { "epoch": 0.5756505576208178, "grad_norm": 0.7607465907181088, "learning_rate": 4.036821578261015e-06, "loss": 0.0673, "step": 12388 }, { "epoch": 0.5756970260223049, "grad_norm": 0.6597938313393406, "learning_rate": 4.0360831701553305e-06, "loss": 0.0514, "step": 12389 }, { "epoch": 0.5757434944237918, "grad_norm": 1.434872297778974, "learning_rate": 4.035344783883378e-06, "loss": 0.0697, "step": 12390 }, { "epoch": 0.5757899628252788, "grad_norm": 0.7468987445716603, "learning_rate": 4.034606419461882e-06, "loss": 0.0675, "step": 12391 }, { "epoch": 0.5758364312267658, "grad_norm": 0.42433818017615077, "learning_rate": 4.033868076907565e-06, "loss": 0.057, "step": 12392 }, { "epoch": 0.5758828996282528, "grad_norm": 0.5613352820027757, "learning_rate": 4.033129756237153e-06, "loss": 0.0314, "step": 12393 }, { "epoch": 0.5759293680297398, "grad_norm": 0.5204357914801258, "learning_rate": 4.032391457467369e-06, "loss": 0.056, "step": 12394 }, { "epoch": 0.5759758364312267, "grad_norm": 0.5782376792797057, "learning_rate": 4.031653180614938e-06, "loss": 0.0368, "step": 12395 }, { "epoch": 0.5760223048327138, "grad_norm": 0.4953279128885418, "learning_rate": 4.030914925696583e-06, "loss": 0.0695, "step": 12396 }, { "epoch": 0.5760687732342007, "grad_norm": 0.5604021662394881, "learning_rate": 4.030176692729022e-06, "loss": 0.0518, "step": 12397 }, { "epoch": 0.5761152416356877, "grad_norm": 0.5557194274979951, "learning_rate": 4.029438481728982e-06, "loss": 0.0628, "step": 12398 }, { "epoch": 0.5761617100371748, "grad_norm": 0.8099107038473268, "learning_rate": 4.02870029271318e-06, "loss": 0.0466, "step": 12399 }, { "epoch": 0.5762081784386617, "grad_norm": 0.44949156533056367, "learning_rate": 4.027962125698339e-06, "loss": 0.0456, "step": 12400 }, { "epoch": 0.5762546468401487, "grad_norm": 0.44203694137869715, "learning_rate": 4.027223980701181e-06, "loss": 0.0389, "step": 12401 }, { "epoch": 0.5763011152416357, "grad_norm": 0.5477929385634488, "learning_rate": 4.0264858577384215e-06, "loss": 0.0591, "step": 12402 }, { "epoch": 0.5763475836431227, "grad_norm": 1.0245777971810364, "learning_rate": 4.025747756826782e-06, "loss": 0.0782, "step": 12403 }, { "epoch": 0.5763940520446097, "grad_norm": 0.603994513538146, "learning_rate": 4.025009677982982e-06, "loss": 0.039, "step": 12404 }, { "epoch": 0.5764405204460966, "grad_norm": 1.455377461717609, "learning_rate": 4.024271621223739e-06, "loss": 0.0634, "step": 12405 }, { "epoch": 0.5764869888475836, "grad_norm": 0.5863216906667992, "learning_rate": 4.023533586565769e-06, "loss": 0.0496, "step": 12406 }, { "epoch": 0.5765334572490707, "grad_norm": 0.6025189963412058, "learning_rate": 4.022795574025793e-06, "loss": 0.0568, "step": 12407 }, { "epoch": 0.5765799256505576, "grad_norm": 0.4904164453292713, "learning_rate": 4.022057583620526e-06, "loss": 0.0515, "step": 12408 }, { "epoch": 0.5766263940520446, "grad_norm": 0.5017527834717426, "learning_rate": 4.021319615366682e-06, "loss": 0.0419, "step": 12409 }, { "epoch": 0.5766728624535316, "grad_norm": 0.6854068193100126, "learning_rate": 4.02058166928098e-06, "loss": 0.0482, "step": 12410 }, { "epoch": 0.5767193308550186, "grad_norm": 0.6942563275865925, "learning_rate": 4.019843745380133e-06, "loss": 0.0584, "step": 12411 }, { "epoch": 0.5767657992565056, "grad_norm": 0.6150801452657839, "learning_rate": 4.019105843680858e-06, "loss": 0.0457, "step": 12412 }, { "epoch": 0.5768122676579925, "grad_norm": 0.7217794481351268, "learning_rate": 4.018367964199869e-06, "loss": 0.0553, "step": 12413 }, { "epoch": 0.5768587360594796, "grad_norm": 0.7863128659646162, "learning_rate": 4.017630106953877e-06, "loss": 0.057, "step": 12414 }, { "epoch": 0.5769052044609665, "grad_norm": 0.6289472416967643, "learning_rate": 4.016892271959597e-06, "loss": 0.0469, "step": 12415 }, { "epoch": 0.5769516728624535, "grad_norm": 0.5831121912281712, "learning_rate": 4.016154459233741e-06, "loss": 0.0602, "step": 12416 }, { "epoch": 0.5769981412639406, "grad_norm": 0.6326957506675411, "learning_rate": 4.015416668793024e-06, "loss": 0.0598, "step": 12417 }, { "epoch": 0.5770446096654275, "grad_norm": 0.42859464957992477, "learning_rate": 4.014678900654154e-06, "loss": 0.0505, "step": 12418 }, { "epoch": 0.5770910780669145, "grad_norm": 0.6758292964431731, "learning_rate": 4.013941154833846e-06, "loss": 0.0563, "step": 12419 }, { "epoch": 0.5771375464684015, "grad_norm": 0.40375855032702895, "learning_rate": 4.0132034313488076e-06, "loss": 0.0344, "step": 12420 }, { "epoch": 0.5771840148698885, "grad_norm": 0.9737994717864584, "learning_rate": 4.012465730215749e-06, "loss": 0.0514, "step": 12421 }, { "epoch": 0.5772304832713755, "grad_norm": 0.6266000321057249, "learning_rate": 4.0117280514513824e-06, "loss": 0.0594, "step": 12422 }, { "epoch": 0.5772769516728624, "grad_norm": 0.5816484856535973, "learning_rate": 4.010990395072414e-06, "loss": 0.0284, "step": 12423 }, { "epoch": 0.5773234200743494, "grad_norm": 1.1055511359203067, "learning_rate": 4.010252761095556e-06, "loss": 0.0548, "step": 12424 }, { "epoch": 0.5773698884758365, "grad_norm": 0.7628220504203043, "learning_rate": 4.009515149537514e-06, "loss": 0.0363, "step": 12425 }, { "epoch": 0.5774163568773234, "grad_norm": 0.6949317014065841, "learning_rate": 4.008777560414996e-06, "loss": 0.0666, "step": 12426 }, { "epoch": 0.5774628252788104, "grad_norm": 0.42756528135281785, "learning_rate": 4.008039993744709e-06, "loss": 0.0587, "step": 12427 }, { "epoch": 0.5775092936802974, "grad_norm": 0.6317416932648858, "learning_rate": 4.00730244954336e-06, "loss": 0.0304, "step": 12428 }, { "epoch": 0.5775557620817844, "grad_norm": 0.7917723386834309, "learning_rate": 4.0065649278276565e-06, "loss": 0.0596, "step": 12429 }, { "epoch": 0.5776022304832714, "grad_norm": 0.706356999536101, "learning_rate": 4.005827428614302e-06, "loss": 0.0461, "step": 12430 }, { "epoch": 0.5776486988847583, "grad_norm": 0.6765193591580106, "learning_rate": 4.005089951920004e-06, "loss": 0.0633, "step": 12431 }, { "epoch": 0.5776951672862454, "grad_norm": 0.9784164453580052, "learning_rate": 4.004352497761465e-06, "loss": 0.0736, "step": 12432 }, { "epoch": 0.5777416356877323, "grad_norm": 0.7567855331160513, "learning_rate": 4.0036150661553885e-06, "loss": 0.0686, "step": 12433 }, { "epoch": 0.5777881040892193, "grad_norm": 0.5872424698931401, "learning_rate": 4.002877657118481e-06, "loss": 0.0582, "step": 12434 }, { "epoch": 0.5778345724907064, "grad_norm": 0.4639246452368883, "learning_rate": 4.002140270667442e-06, "loss": 0.0418, "step": 12435 }, { "epoch": 0.5778810408921933, "grad_norm": 0.691364866327581, "learning_rate": 4.001402906818979e-06, "loss": 0.0648, "step": 12436 }, { "epoch": 0.5779275092936803, "grad_norm": 0.6321352436569088, "learning_rate": 4.000665565589791e-06, "loss": 0.056, "step": 12437 }, { "epoch": 0.5779739776951673, "grad_norm": 0.5110102897113639, "learning_rate": 3.999928246996577e-06, "loss": 0.0454, "step": 12438 }, { "epoch": 0.5780204460966543, "grad_norm": 0.3978368226661209, "learning_rate": 3.999190951056044e-06, "loss": 0.0563, "step": 12439 }, { "epoch": 0.5780669144981413, "grad_norm": 0.4683706625295754, "learning_rate": 3.998453677784886e-06, "loss": 0.0518, "step": 12440 }, { "epoch": 0.5781133828996282, "grad_norm": 0.6267995108030533, "learning_rate": 3.997716427199808e-06, "loss": 0.0688, "step": 12441 }, { "epoch": 0.5781598513011152, "grad_norm": 0.756073541071844, "learning_rate": 3.9969791993175065e-06, "loss": 0.0555, "step": 12442 }, { "epoch": 0.5782063197026023, "grad_norm": 0.6345584146878832, "learning_rate": 3.996241994154685e-06, "loss": 0.0548, "step": 12443 }, { "epoch": 0.5782527881040892, "grad_norm": 0.5340065729017343, "learning_rate": 3.9955048117280366e-06, "loss": 0.0522, "step": 12444 }, { "epoch": 0.5782992565055762, "grad_norm": 0.8067755267198324, "learning_rate": 3.994767652054261e-06, "loss": 0.0449, "step": 12445 }, { "epoch": 0.5783457249070632, "grad_norm": 0.528345925395514, "learning_rate": 3.994030515150057e-06, "loss": 0.0491, "step": 12446 }, { "epoch": 0.5783921933085502, "grad_norm": 0.6818978708168485, "learning_rate": 3.993293401032119e-06, "loss": 0.0575, "step": 12447 }, { "epoch": 0.5784386617100372, "grad_norm": 0.6992925238560503, "learning_rate": 3.992556309717147e-06, "loss": 0.0678, "step": 12448 }, { "epoch": 0.5784851301115241, "grad_norm": 0.46065446481286215, "learning_rate": 3.991819241221836e-06, "loss": 0.0491, "step": 12449 }, { "epoch": 0.5785315985130112, "grad_norm": 0.4708111918352805, "learning_rate": 3.991082195562877e-06, "loss": 0.0714, "step": 12450 }, { "epoch": 0.5785780669144981, "grad_norm": 0.48079738042068265, "learning_rate": 3.99034517275697e-06, "loss": 0.0617, "step": 12451 }, { "epoch": 0.5786245353159851, "grad_norm": 1.3055257538511098, "learning_rate": 3.989608172820806e-06, "loss": 0.0686, "step": 12452 }, { "epoch": 0.5786710037174722, "grad_norm": 0.5874908991120525, "learning_rate": 3.988871195771081e-06, "loss": 0.0478, "step": 12453 }, { "epoch": 0.5787174721189591, "grad_norm": 0.502762888697454, "learning_rate": 3.988134241624487e-06, "loss": 0.0588, "step": 12454 }, { "epoch": 0.5787639405204461, "grad_norm": 0.4302288870071192, "learning_rate": 3.98739731039772e-06, "loss": 0.0456, "step": 12455 }, { "epoch": 0.5788104089219331, "grad_norm": 0.7218696053140308, "learning_rate": 3.986660402107469e-06, "loss": 0.0608, "step": 12456 }, { "epoch": 0.5788568773234201, "grad_norm": 0.925718662808942, "learning_rate": 3.985923516770424e-06, "loss": 0.078, "step": 12457 }, { "epoch": 0.5789033457249071, "grad_norm": 0.4087208527222752, "learning_rate": 3.9851866544032805e-06, "loss": 0.0356, "step": 12458 }, { "epoch": 0.578949814126394, "grad_norm": 0.8393624066302087, "learning_rate": 3.984449815022726e-06, "loss": 0.0415, "step": 12459 }, { "epoch": 0.578996282527881, "grad_norm": 0.7763009520214629, "learning_rate": 3.983712998645454e-06, "loss": 0.0703, "step": 12460 }, { "epoch": 0.5790427509293681, "grad_norm": 0.4667950880682188, "learning_rate": 3.982976205288152e-06, "loss": 0.0406, "step": 12461 }, { "epoch": 0.579089219330855, "grad_norm": 0.5192326632303581, "learning_rate": 3.9822394349675076e-06, "loss": 0.0386, "step": 12462 }, { "epoch": 0.579135687732342, "grad_norm": 0.44274974132608386, "learning_rate": 3.9815026877002125e-06, "loss": 0.0467, "step": 12463 }, { "epoch": 0.579182156133829, "grad_norm": 1.3542286268888608, "learning_rate": 3.9807659635029524e-06, "loss": 0.0674, "step": 12464 }, { "epoch": 0.579228624535316, "grad_norm": 0.5722526331819084, "learning_rate": 3.9800292623924165e-06, "loss": 0.0416, "step": 12465 }, { "epoch": 0.579275092936803, "grad_norm": 0.6986002370959253, "learning_rate": 3.979292584385293e-06, "loss": 0.0673, "step": 12466 }, { "epoch": 0.5793215613382899, "grad_norm": 0.6348334173037293, "learning_rate": 3.978555929498264e-06, "loss": 0.0589, "step": 12467 }, { "epoch": 0.579368029739777, "grad_norm": 0.9161742955664658, "learning_rate": 3.977819297748019e-06, "loss": 0.0474, "step": 12468 }, { "epoch": 0.5794144981412639, "grad_norm": 0.5184342573418832, "learning_rate": 3.9770826891512434e-06, "loss": 0.0583, "step": 12469 }, { "epoch": 0.5794609665427509, "grad_norm": 0.4095772900288243, "learning_rate": 3.976346103724621e-06, "loss": 0.0526, "step": 12470 }, { "epoch": 0.579507434944238, "grad_norm": 0.2658318095021636, "learning_rate": 3.975609541484836e-06, "loss": 0.0469, "step": 12471 }, { "epoch": 0.5795539033457249, "grad_norm": 0.4165764239593412, "learning_rate": 3.974873002448575e-06, "loss": 0.0441, "step": 12472 }, { "epoch": 0.5796003717472119, "grad_norm": 0.35933655344765386, "learning_rate": 3.9741364866325195e-06, "loss": 0.0664, "step": 12473 }, { "epoch": 0.5796468401486989, "grad_norm": 0.3940485389260641, "learning_rate": 3.97339999405335e-06, "loss": 0.0548, "step": 12474 }, { "epoch": 0.5796933085501859, "grad_norm": 0.44173628298610224, "learning_rate": 3.972663524727753e-06, "loss": 0.0499, "step": 12475 }, { "epoch": 0.5797397769516729, "grad_norm": 1.3250761886958502, "learning_rate": 3.971927078672407e-06, "loss": 0.0745, "step": 12476 }, { "epoch": 0.5797862453531598, "grad_norm": 0.8462286961234119, "learning_rate": 3.971190655903995e-06, "loss": 0.0758, "step": 12477 }, { "epoch": 0.5798327137546468, "grad_norm": 0.30780119203135525, "learning_rate": 3.9704542564392e-06, "loss": 0.0501, "step": 12478 }, { "epoch": 0.5798791821561339, "grad_norm": 0.2553619455913972, "learning_rate": 3.969717880294696e-06, "loss": 0.0389, "step": 12479 }, { "epoch": 0.5799256505576208, "grad_norm": 0.603940065730528, "learning_rate": 3.9689815274871685e-06, "loss": 0.0647, "step": 12480 }, { "epoch": 0.5799721189591078, "grad_norm": 0.6174618862334985, "learning_rate": 3.968245198033293e-06, "loss": 0.0782, "step": 12481 }, { "epoch": 0.5800185873605948, "grad_norm": 0.6628551774197776, "learning_rate": 3.967508891949751e-06, "loss": 0.0455, "step": 12482 }, { "epoch": 0.5800650557620818, "grad_norm": 0.623145767413468, "learning_rate": 3.966772609253218e-06, "loss": 0.0788, "step": 12483 }, { "epoch": 0.5801115241635688, "grad_norm": 0.9764951814635092, "learning_rate": 3.966036349960375e-06, "loss": 0.0533, "step": 12484 }, { "epoch": 0.5801579925650557, "grad_norm": 0.6400943106599143, "learning_rate": 3.965300114087895e-06, "loss": 0.0819, "step": 12485 }, { "epoch": 0.5802044609665428, "grad_norm": 0.6306031819702801, "learning_rate": 3.964563901652457e-06, "loss": 0.063, "step": 12486 }, { "epoch": 0.5802509293680297, "grad_norm": 0.47070941976746666, "learning_rate": 3.963827712670737e-06, "loss": 0.0355, "step": 12487 }, { "epoch": 0.5802973977695167, "grad_norm": 0.3787744262376501, "learning_rate": 3.9630915471594075e-06, "loss": 0.0606, "step": 12488 }, { "epoch": 0.5803438661710038, "grad_norm": 0.45033082063985025, "learning_rate": 3.962355405135149e-06, "loss": 0.052, "step": 12489 }, { "epoch": 0.5803903345724907, "grad_norm": 0.4281824842836362, "learning_rate": 3.961619286614632e-06, "loss": 0.0384, "step": 12490 }, { "epoch": 0.5804368029739777, "grad_norm": 0.5863468341049894, "learning_rate": 3.960883191614529e-06, "loss": 0.065, "step": 12491 }, { "epoch": 0.5804832713754647, "grad_norm": 0.5852447298870351, "learning_rate": 3.9601471201515175e-06, "loss": 0.0551, "step": 12492 }, { "epoch": 0.5805297397769517, "grad_norm": 0.8056553503835655, "learning_rate": 3.959411072242266e-06, "loss": 0.0509, "step": 12493 }, { "epoch": 0.5805762081784387, "grad_norm": 0.6644585396733222, "learning_rate": 3.958675047903451e-06, "loss": 0.0523, "step": 12494 }, { "epoch": 0.5806226765799256, "grad_norm": 0.48508245962527413, "learning_rate": 3.9579390471517406e-06, "loss": 0.0742, "step": 12495 }, { "epoch": 0.5806691449814126, "grad_norm": 0.7444592200013118, "learning_rate": 3.9572030700038094e-06, "loss": 0.0691, "step": 12496 }, { "epoch": 0.5807156133828997, "grad_norm": 0.2884099246208822, "learning_rate": 3.9564671164763255e-06, "loss": 0.0182, "step": 12497 }, { "epoch": 0.5807620817843866, "grad_norm": 0.6764304179657497, "learning_rate": 3.955731186585958e-06, "loss": 0.0758, "step": 12498 }, { "epoch": 0.5808085501858736, "grad_norm": 0.6184390391553765, "learning_rate": 3.95499528034938e-06, "loss": 0.0494, "step": 12499 }, { "epoch": 0.5808550185873605, "grad_norm": 0.8110834953713283, "learning_rate": 3.954259397783258e-06, "loss": 0.0343, "step": 12500 }, { "epoch": 0.5809014869888476, "grad_norm": 0.3942476980888976, "learning_rate": 3.95352353890426e-06, "loss": 0.0515, "step": 12501 }, { "epoch": 0.5809479553903346, "grad_norm": 0.36692467829206094, "learning_rate": 3.9527877037290595e-06, "loss": 0.0398, "step": 12502 }, { "epoch": 0.5809944237918215, "grad_norm": 0.7650019528458467, "learning_rate": 3.952051892274315e-06, "loss": 0.0558, "step": 12503 }, { "epoch": 0.5810408921933086, "grad_norm": 0.4526311247120132, "learning_rate": 3.951316104556701e-06, "loss": 0.0337, "step": 12504 }, { "epoch": 0.5810873605947955, "grad_norm": 0.7085847197191734, "learning_rate": 3.950580340592879e-06, "loss": 0.0696, "step": 12505 }, { "epoch": 0.5811338289962825, "grad_norm": 0.49309769874966014, "learning_rate": 3.949844600399517e-06, "loss": 0.0629, "step": 12506 }, { "epoch": 0.5811802973977696, "grad_norm": 1.5918696713264697, "learning_rate": 3.949108883993281e-06, "loss": 0.0702, "step": 12507 }, { "epoch": 0.5812267657992565, "grad_norm": 0.604366762997814, "learning_rate": 3.948373191390836e-06, "loss": 0.0587, "step": 12508 }, { "epoch": 0.5812732342007435, "grad_norm": 0.6014628849918078, "learning_rate": 3.947637522608844e-06, "loss": 0.0444, "step": 12509 }, { "epoch": 0.5813197026022305, "grad_norm": 0.7311924767514153, "learning_rate": 3.946901877663969e-06, "loss": 0.0505, "step": 12510 }, { "epoch": 0.5813661710037175, "grad_norm": 0.43078060208091623, "learning_rate": 3.9461662565728755e-06, "loss": 0.0493, "step": 12511 }, { "epoch": 0.5814126394052045, "grad_norm": 0.5610709472110517, "learning_rate": 3.945430659352224e-06, "loss": 0.073, "step": 12512 }, { "epoch": 0.5814591078066914, "grad_norm": 0.8892498370323052, "learning_rate": 3.944695086018679e-06, "loss": 0.077, "step": 12513 }, { "epoch": 0.5815055762081784, "grad_norm": 1.1247388815614012, "learning_rate": 3.943959536588903e-06, "loss": 0.0555, "step": 12514 }, { "epoch": 0.5815520446096655, "grad_norm": 0.5240786787723188, "learning_rate": 3.943224011079551e-06, "loss": 0.0524, "step": 12515 }, { "epoch": 0.5815985130111524, "grad_norm": 0.7762700037945469, "learning_rate": 3.942488509507291e-06, "loss": 0.0506, "step": 12516 }, { "epoch": 0.5816449814126394, "grad_norm": 0.8495030349720597, "learning_rate": 3.941753031888776e-06, "loss": 0.0595, "step": 12517 }, { "epoch": 0.5816914498141263, "grad_norm": 0.4564379295694925, "learning_rate": 3.94101757824067e-06, "loss": 0.0362, "step": 12518 }, { "epoch": 0.5817379182156134, "grad_norm": 0.5937012345991798, "learning_rate": 3.940282148579633e-06, "loss": 0.0407, "step": 12519 }, { "epoch": 0.5817843866171004, "grad_norm": 0.5491622484909409, "learning_rate": 3.939546742922318e-06, "loss": 0.065, "step": 12520 }, { "epoch": 0.5818308550185873, "grad_norm": 0.581997293122745, "learning_rate": 3.938811361285386e-06, "loss": 0.0569, "step": 12521 }, { "epoch": 0.5818773234200744, "grad_norm": 0.34720655594702055, "learning_rate": 3.938076003685493e-06, "loss": 0.0426, "step": 12522 }, { "epoch": 0.5819237918215613, "grad_norm": 0.8402090879302203, "learning_rate": 3.9373406701392945e-06, "loss": 0.0767, "step": 12523 }, { "epoch": 0.5819702602230483, "grad_norm": 0.7633051742839905, "learning_rate": 3.936605360663451e-06, "loss": 0.0568, "step": 12524 }, { "epoch": 0.5820167286245354, "grad_norm": 0.5557749296413715, "learning_rate": 3.9358700752746156e-06, "loss": 0.0411, "step": 12525 }, { "epoch": 0.5820631970260223, "grad_norm": 0.4584935513335404, "learning_rate": 3.935134813989443e-06, "loss": 0.0364, "step": 12526 }, { "epoch": 0.5821096654275093, "grad_norm": 1.8210215091477486, "learning_rate": 3.934399576824585e-06, "loss": 0.0955, "step": 12527 }, { "epoch": 0.5821561338289963, "grad_norm": 1.322019702455583, "learning_rate": 3.9336643637966986e-06, "loss": 0.055, "step": 12528 }, { "epoch": 0.5822026022304833, "grad_norm": 1.000983024273659, "learning_rate": 3.932929174922439e-06, "loss": 0.08, "step": 12529 }, { "epoch": 0.5822490706319703, "grad_norm": 0.4367155351127271, "learning_rate": 3.9321940102184535e-06, "loss": 0.043, "step": 12530 }, { "epoch": 0.5822955390334572, "grad_norm": 0.693760070568872, "learning_rate": 3.931458869701402e-06, "loss": 0.0558, "step": 12531 }, { "epoch": 0.5823420074349442, "grad_norm": 0.6676261674498393, "learning_rate": 3.930723753387928e-06, "loss": 0.0591, "step": 12532 }, { "epoch": 0.5823884758364313, "grad_norm": 0.5456846422251089, "learning_rate": 3.929988661294687e-06, "loss": 0.0419, "step": 12533 }, { "epoch": 0.5824349442379182, "grad_norm": 0.4684265973149128, "learning_rate": 3.92925359343833e-06, "loss": 0.0486, "step": 12534 }, { "epoch": 0.5824814126394052, "grad_norm": 0.5911685417097411, "learning_rate": 3.928518549835504e-06, "loss": 0.054, "step": 12535 }, { "epoch": 0.5825278810408921, "grad_norm": 2.025638905989661, "learning_rate": 3.927783530502862e-06, "loss": 0.0846, "step": 12536 }, { "epoch": 0.5825743494423792, "grad_norm": 0.4561714992565991, "learning_rate": 3.927048535457053e-06, "loss": 0.0493, "step": 12537 }, { "epoch": 0.5826208178438662, "grad_norm": 1.1716777858597094, "learning_rate": 3.9263135647147235e-06, "loss": 0.0887, "step": 12538 }, { "epoch": 0.5826672862453531, "grad_norm": 0.5960054637869072, "learning_rate": 3.92557861829252e-06, "loss": 0.0526, "step": 12539 }, { "epoch": 0.5827137546468402, "grad_norm": 0.8077836254623583, "learning_rate": 3.924843696207091e-06, "loss": 0.066, "step": 12540 }, { "epoch": 0.5827602230483271, "grad_norm": 0.604594110210755, "learning_rate": 3.924108798475086e-06, "loss": 0.0852, "step": 12541 }, { "epoch": 0.5828066914498141, "grad_norm": 0.747468077811614, "learning_rate": 3.923373925113147e-06, "loss": 0.0704, "step": 12542 }, { "epoch": 0.5828531598513012, "grad_norm": 0.6462810863800825, "learning_rate": 3.922639076137926e-06, "loss": 0.0573, "step": 12543 }, { "epoch": 0.5828996282527881, "grad_norm": 0.35405738937432607, "learning_rate": 3.92190425156606e-06, "loss": 0.0359, "step": 12544 }, { "epoch": 0.5829460966542751, "grad_norm": 1.1115629037954162, "learning_rate": 3.9211694514141975e-06, "loss": 0.0523, "step": 12545 }, { "epoch": 0.5829925650557621, "grad_norm": 0.7133427704565, "learning_rate": 3.920434675698983e-06, "loss": 0.0538, "step": 12546 }, { "epoch": 0.5830390334572491, "grad_norm": 0.32475818177860577, "learning_rate": 3.919699924437059e-06, "loss": 0.0307, "step": 12547 }, { "epoch": 0.5830855018587361, "grad_norm": 0.7296477534910717, "learning_rate": 3.91896519764507e-06, "loss": 0.0795, "step": 12548 }, { "epoch": 0.583131970260223, "grad_norm": 1.0100395822246482, "learning_rate": 3.918230495339656e-06, "loss": 0.06, "step": 12549 }, { "epoch": 0.58317843866171, "grad_norm": 1.3375737946697863, "learning_rate": 3.91749581753746e-06, "loss": 0.0867, "step": 12550 }, { "epoch": 0.5832249070631971, "grad_norm": 0.9710268477563828, "learning_rate": 3.9167611642551234e-06, "loss": 0.0606, "step": 12551 }, { "epoch": 0.583271375464684, "grad_norm": 0.6003323624266358, "learning_rate": 3.916026535509285e-06, "loss": 0.0568, "step": 12552 }, { "epoch": 0.583317843866171, "grad_norm": 0.7287362845293887, "learning_rate": 3.915291931316588e-06, "loss": 0.0508, "step": 12553 }, { "epoch": 0.583364312267658, "grad_norm": 1.2224452217140132, "learning_rate": 3.91455735169367e-06, "loss": 0.0597, "step": 12554 }, { "epoch": 0.583410780669145, "grad_norm": 0.5449409171386275, "learning_rate": 3.913822796657171e-06, "loss": 0.0503, "step": 12555 }, { "epoch": 0.583457249070632, "grad_norm": 0.43056153995297936, "learning_rate": 3.913088266223729e-06, "loss": 0.0473, "step": 12556 }, { "epoch": 0.5835037174721189, "grad_norm": 0.389905177180507, "learning_rate": 3.91235376040998e-06, "loss": 0.0386, "step": 12557 }, { "epoch": 0.583550185873606, "grad_norm": 0.6274591782043779, "learning_rate": 3.911619279232564e-06, "loss": 0.0659, "step": 12558 }, { "epoch": 0.5835966542750929, "grad_norm": 0.4849201005339463, "learning_rate": 3.9108848227081165e-06, "loss": 0.0391, "step": 12559 }, { "epoch": 0.5836431226765799, "grad_norm": 0.6016202756084877, "learning_rate": 3.910150390853275e-06, "loss": 0.0736, "step": 12560 }, { "epoch": 0.583689591078067, "grad_norm": 0.6408550287245716, "learning_rate": 3.909415983684675e-06, "loss": 0.0708, "step": 12561 }, { "epoch": 0.5837360594795539, "grad_norm": 1.011501016815606, "learning_rate": 3.908681601218948e-06, "loss": 0.0797, "step": 12562 }, { "epoch": 0.5837825278810409, "grad_norm": 0.7927272744946535, "learning_rate": 3.907947243472733e-06, "loss": 0.072, "step": 12563 }, { "epoch": 0.5838289962825279, "grad_norm": 0.9752147620506949, "learning_rate": 3.907212910462661e-06, "loss": 0.0606, "step": 12564 }, { "epoch": 0.5838754646840149, "grad_norm": 0.7492519307046921, "learning_rate": 3.906478602205368e-06, "loss": 0.0721, "step": 12565 }, { "epoch": 0.5839219330855019, "grad_norm": 0.6044338306096702, "learning_rate": 3.905744318717484e-06, "loss": 0.0599, "step": 12566 }, { "epoch": 0.5839684014869888, "grad_norm": 0.6208629575911322, "learning_rate": 3.905010060015646e-06, "loss": 0.064, "step": 12567 }, { "epoch": 0.5840148698884758, "grad_norm": 0.31740382416209867, "learning_rate": 3.904275826116481e-06, "loss": 0.0476, "step": 12568 }, { "epoch": 0.5840613382899629, "grad_norm": 0.5191807208011289, "learning_rate": 3.90354161703662e-06, "loss": 0.0529, "step": 12569 }, { "epoch": 0.5841078066914498, "grad_norm": 0.6442956379970273, "learning_rate": 3.902807432792698e-06, "loss": 0.0704, "step": 12570 }, { "epoch": 0.5841542750929368, "grad_norm": 0.6547508719823334, "learning_rate": 3.90207327340134e-06, "loss": 0.0583, "step": 12571 }, { "epoch": 0.5842007434944237, "grad_norm": 0.5817731632281125, "learning_rate": 3.90133913887918e-06, "loss": 0.0631, "step": 12572 }, { "epoch": 0.5842472118959108, "grad_norm": 1.0496501604225756, "learning_rate": 3.900605029242845e-06, "loss": 0.049, "step": 12573 }, { "epoch": 0.5842936802973978, "grad_norm": 0.45664951762684336, "learning_rate": 3.899870944508961e-06, "loss": 0.04, "step": 12574 }, { "epoch": 0.5843401486988847, "grad_norm": 0.32315684943122663, "learning_rate": 3.89913688469416e-06, "loss": 0.0498, "step": 12575 }, { "epoch": 0.5843866171003718, "grad_norm": 0.5564989719723977, "learning_rate": 3.898402849815065e-06, "loss": 0.0583, "step": 12576 }, { "epoch": 0.5844330855018587, "grad_norm": 0.5255182345339384, "learning_rate": 3.8976688398883064e-06, "loss": 0.0519, "step": 12577 }, { "epoch": 0.5844795539033457, "grad_norm": 0.6571458288964129, "learning_rate": 3.89693485493051e-06, "loss": 0.0644, "step": 12578 }, { "epoch": 0.5845260223048327, "grad_norm": 0.44089632796899225, "learning_rate": 3.8962008949582985e-06, "loss": 0.0392, "step": 12579 }, { "epoch": 0.5845724907063197, "grad_norm": 0.4299418575672463, "learning_rate": 3.895466959988299e-06, "loss": 0.0313, "step": 12580 }, { "epoch": 0.5846189591078067, "grad_norm": 0.625460891850274, "learning_rate": 3.8947330500371335e-06, "loss": 0.0602, "step": 12581 }, { "epoch": 0.5846654275092937, "grad_norm": 0.45706015847884773, "learning_rate": 3.8939991651214295e-06, "loss": 0.0574, "step": 12582 }, { "epoch": 0.5847118959107807, "grad_norm": 0.3888681166837853, "learning_rate": 3.8932653052578075e-06, "loss": 0.0281, "step": 12583 }, { "epoch": 0.5847583643122677, "grad_norm": 0.5064148154241698, "learning_rate": 3.892531470462894e-06, "loss": 0.0401, "step": 12584 }, { "epoch": 0.5848048327137546, "grad_norm": 0.7834309782794295, "learning_rate": 3.8917976607533055e-06, "loss": 0.0686, "step": 12585 }, { "epoch": 0.5848513011152416, "grad_norm": 0.6003906764275826, "learning_rate": 3.8910638761456665e-06, "loss": 0.0456, "step": 12586 }, { "epoch": 0.5848977695167287, "grad_norm": 0.8414339064428957, "learning_rate": 3.890330116656598e-06, "loss": 0.0725, "step": 12587 }, { "epoch": 0.5849442379182156, "grad_norm": 0.67958921047658, "learning_rate": 3.88959638230272e-06, "loss": 0.0604, "step": 12588 }, { "epoch": 0.5849907063197026, "grad_norm": 0.37721015348746917, "learning_rate": 3.8888626731006535e-06, "loss": 0.0444, "step": 12589 }, { "epoch": 0.5850371747211895, "grad_norm": 0.35937770151494297, "learning_rate": 3.888128989067017e-06, "loss": 0.0262, "step": 12590 }, { "epoch": 0.5850836431226766, "grad_norm": 0.4045240444492291, "learning_rate": 3.887395330218429e-06, "loss": 0.0544, "step": 12591 }, { "epoch": 0.5851301115241636, "grad_norm": 1.0275075861126897, "learning_rate": 3.886661696571508e-06, "loss": 0.0469, "step": 12592 }, { "epoch": 0.5851765799256505, "grad_norm": 0.5478469326641472, "learning_rate": 3.88592808814287e-06, "loss": 0.0483, "step": 12593 }, { "epoch": 0.5852230483271376, "grad_norm": 0.6438051733658335, "learning_rate": 3.8851945049491335e-06, "loss": 0.057, "step": 12594 }, { "epoch": 0.5852695167286245, "grad_norm": 0.8406858718610959, "learning_rate": 3.884460947006915e-06, "loss": 0.0758, "step": 12595 }, { "epoch": 0.5853159851301115, "grad_norm": 0.4923983257859737, "learning_rate": 3.883727414332832e-06, "loss": 0.0381, "step": 12596 }, { "epoch": 0.5853624535315985, "grad_norm": 1.6117049338631355, "learning_rate": 3.882993906943496e-06, "loss": 0.0457, "step": 12597 }, { "epoch": 0.5854089219330855, "grad_norm": 0.8711475678761276, "learning_rate": 3.882260424855523e-06, "loss": 0.0602, "step": 12598 }, { "epoch": 0.5854553903345725, "grad_norm": 0.596421774526788, "learning_rate": 3.88152696808553e-06, "loss": 0.049, "step": 12599 }, { "epoch": 0.5855018587360595, "grad_norm": 0.7792849106690437, "learning_rate": 3.880793536650126e-06, "loss": 0.0646, "step": 12600 }, { "epoch": 0.5855483271375465, "grad_norm": 0.4312778707750042, "learning_rate": 3.880060130565928e-06, "loss": 0.068, "step": 12601 }, { "epoch": 0.5855947955390335, "grad_norm": 2.052122609534111, "learning_rate": 3.879326749849548e-06, "loss": 0.0961, "step": 12602 }, { "epoch": 0.5856412639405204, "grad_norm": 0.8968697617661607, "learning_rate": 3.878593394517594e-06, "loss": 0.0594, "step": 12603 }, { "epoch": 0.5856877323420074, "grad_norm": 0.5994078688472912, "learning_rate": 3.877860064586681e-06, "loss": 0.0497, "step": 12604 }, { "epoch": 0.5857342007434945, "grad_norm": 0.4443730664261309, "learning_rate": 3.877126760073418e-06, "loss": 0.0475, "step": 12605 }, { "epoch": 0.5857806691449814, "grad_norm": 0.6335895829635317, "learning_rate": 3.876393480994416e-06, "loss": 0.0471, "step": 12606 }, { "epoch": 0.5858271375464684, "grad_norm": 0.4829033586412084, "learning_rate": 3.875660227366283e-06, "loss": 0.0508, "step": 12607 }, { "epoch": 0.5858736059479553, "grad_norm": 0.6994872693419406, "learning_rate": 3.874926999205632e-06, "loss": 0.0578, "step": 12608 }, { "epoch": 0.5859200743494424, "grad_norm": 0.5690407356858636, "learning_rate": 3.874193796529067e-06, "loss": 0.0508, "step": 12609 }, { "epoch": 0.5859665427509294, "grad_norm": 0.5462757331264101, "learning_rate": 3.8734606193531965e-06, "loss": 0.0545, "step": 12610 }, { "epoch": 0.5860130111524163, "grad_norm": 0.8042303362497032, "learning_rate": 3.8727274676946305e-06, "loss": 0.0557, "step": 12611 }, { "epoch": 0.5860594795539034, "grad_norm": 0.45708132520420564, "learning_rate": 3.871994341569972e-06, "loss": 0.0461, "step": 12612 }, { "epoch": 0.5861059479553903, "grad_norm": 0.7475857212766851, "learning_rate": 3.871261240995829e-06, "loss": 0.0635, "step": 12613 }, { "epoch": 0.5861524163568773, "grad_norm": 1.0252465277507492, "learning_rate": 3.870528165988809e-06, "loss": 0.0759, "step": 12614 }, { "epoch": 0.5861988847583643, "grad_norm": 0.8553851466869336, "learning_rate": 3.869795116565513e-06, "loss": 0.08, "step": 12615 }, { "epoch": 0.5862453531598513, "grad_norm": 0.36826529829974836, "learning_rate": 3.869062092742548e-06, "loss": 0.043, "step": 12616 }, { "epoch": 0.5862918215613383, "grad_norm": 0.8550205649991279, "learning_rate": 3.8683290945365155e-06, "loss": 0.0592, "step": 12617 }, { "epoch": 0.5863382899628253, "grad_norm": 0.6372329084162667, "learning_rate": 3.867596121964022e-06, "loss": 0.0523, "step": 12618 }, { "epoch": 0.5863847583643123, "grad_norm": 12.93577590121896, "learning_rate": 3.866863175041666e-06, "loss": 0.0423, "step": 12619 }, { "epoch": 0.5864312267657993, "grad_norm": 0.4916282330974645, "learning_rate": 3.866130253786054e-06, "loss": 0.0543, "step": 12620 }, { "epoch": 0.5864776951672862, "grad_norm": 0.4261518230132595, "learning_rate": 3.865397358213784e-06, "loss": 0.0463, "step": 12621 }, { "epoch": 0.5865241635687732, "grad_norm": 0.440704753628387, "learning_rate": 3.864664488341457e-06, "loss": 0.0622, "step": 12622 }, { "epoch": 0.5865706319702603, "grad_norm": 0.44640007256526437, "learning_rate": 3.863931644185676e-06, "loss": 0.0527, "step": 12623 }, { "epoch": 0.5866171003717472, "grad_norm": 0.3400571808773873, "learning_rate": 3.863198825763037e-06, "loss": 0.0415, "step": 12624 }, { "epoch": 0.5866635687732342, "grad_norm": 0.6262341763248553, "learning_rate": 3.862466033090143e-06, "loss": 0.0643, "step": 12625 }, { "epoch": 0.5867100371747211, "grad_norm": 1.3053985920960838, "learning_rate": 3.86173326618359e-06, "loss": 0.0556, "step": 12626 }, { "epoch": 0.5867565055762082, "grad_norm": 1.4466322849038098, "learning_rate": 3.861000525059977e-06, "loss": 0.0659, "step": 12627 }, { "epoch": 0.5868029739776952, "grad_norm": 0.5713427611156153, "learning_rate": 3.860267809735899e-06, "loss": 0.0475, "step": 12628 }, { "epoch": 0.5868494423791821, "grad_norm": 0.6709095716420094, "learning_rate": 3.859535120227956e-06, "loss": 0.0708, "step": 12629 }, { "epoch": 0.5868959107806692, "grad_norm": 0.6904070338954146, "learning_rate": 3.858802456552743e-06, "loss": 0.0559, "step": 12630 }, { "epoch": 0.5869423791821561, "grad_norm": 0.6802319064557806, "learning_rate": 3.858069818726854e-06, "loss": 0.0604, "step": 12631 }, { "epoch": 0.5869888475836431, "grad_norm": 0.5371146917940448, "learning_rate": 3.857337206766888e-06, "loss": 0.0525, "step": 12632 }, { "epoch": 0.5870353159851301, "grad_norm": 0.7073255627089452, "learning_rate": 3.8566046206894356e-06, "loss": 0.0603, "step": 12633 }, { "epoch": 0.5870817843866171, "grad_norm": 0.3769786394475085, "learning_rate": 3.855872060511091e-06, "loss": 0.0264, "step": 12634 }, { "epoch": 0.5871282527881041, "grad_norm": 0.5705877159476253, "learning_rate": 3.85513952624845e-06, "loss": 0.0503, "step": 12635 }, { "epoch": 0.5871747211895911, "grad_norm": 3.8699991044703497, "learning_rate": 3.854407017918102e-06, "loss": 0.0574, "step": 12636 }, { "epoch": 0.587221189591078, "grad_norm": 0.9392944472547499, "learning_rate": 3.8536745355366425e-06, "loss": 0.0863, "step": 12637 }, { "epoch": 0.5872676579925651, "grad_norm": 0.7152874701011225, "learning_rate": 3.852942079120661e-06, "loss": 0.077, "step": 12638 }, { "epoch": 0.587314126394052, "grad_norm": 0.526526480570071, "learning_rate": 3.852209648686748e-06, "loss": 0.0493, "step": 12639 }, { "epoch": 0.587360594795539, "grad_norm": 0.6440000397306654, "learning_rate": 3.851477244251494e-06, "loss": 0.0518, "step": 12640 }, { "epoch": 0.5874070631970261, "grad_norm": 0.719275055536515, "learning_rate": 3.85074486583149e-06, "loss": 0.0704, "step": 12641 }, { "epoch": 0.587453531598513, "grad_norm": 0.4199120463933168, "learning_rate": 3.850012513443324e-06, "loss": 0.036, "step": 12642 }, { "epoch": 0.5875, "grad_norm": 0.4332755669598194, "learning_rate": 3.849280187103586e-06, "loss": 0.0322, "step": 12643 }, { "epoch": 0.5875464684014869, "grad_norm": 0.5663333508954135, "learning_rate": 3.8485478868288615e-06, "loss": 0.0557, "step": 12644 }, { "epoch": 0.587592936802974, "grad_norm": 0.5838216864405623, "learning_rate": 3.84781561263574e-06, "loss": 0.0456, "step": 12645 }, { "epoch": 0.587639405204461, "grad_norm": 0.4656604590248768, "learning_rate": 3.847083364540806e-06, "loss": 0.0367, "step": 12646 }, { "epoch": 0.5876858736059479, "grad_norm": 0.3637042221734308, "learning_rate": 3.8463511425606496e-06, "loss": 0.0445, "step": 12647 }, { "epoch": 0.587732342007435, "grad_norm": 0.6323738032933693, "learning_rate": 3.8456189467118516e-06, "loss": 0.0578, "step": 12648 }, { "epoch": 0.5877788104089219, "grad_norm": 0.6234122363207111, "learning_rate": 3.844886777011002e-06, "loss": 0.057, "step": 12649 }, { "epoch": 0.5878252788104089, "grad_norm": 0.5807058745627589, "learning_rate": 3.844154633474681e-06, "loss": 0.0653, "step": 12650 }, { "epoch": 0.587871747211896, "grad_norm": 0.4760032654418051, "learning_rate": 3.8434225161194734e-06, "loss": 0.0567, "step": 12651 }, { "epoch": 0.5879182156133829, "grad_norm": 0.32686252490587836, "learning_rate": 3.842690424961965e-06, "loss": 0.0333, "step": 12652 }, { "epoch": 0.5879646840148699, "grad_norm": 0.7927111409390838, "learning_rate": 3.841958360018735e-06, "loss": 0.0787, "step": 12653 }, { "epoch": 0.5880111524163569, "grad_norm": 0.4496828020628707, "learning_rate": 3.841226321306367e-06, "loss": 0.0502, "step": 12654 }, { "epoch": 0.5880576208178439, "grad_norm": 0.7679353430522797, "learning_rate": 3.840494308841445e-06, "loss": 0.076, "step": 12655 }, { "epoch": 0.5881040892193309, "grad_norm": 0.36902996614738454, "learning_rate": 3.8397623226405445e-06, "loss": 0.0408, "step": 12656 }, { "epoch": 0.5881505576208178, "grad_norm": 0.5354811619732157, "learning_rate": 3.839030362720249e-06, "loss": 0.0656, "step": 12657 }, { "epoch": 0.5881970260223048, "grad_norm": 0.5854999087132252, "learning_rate": 3.838298429097137e-06, "loss": 0.0601, "step": 12658 }, { "epoch": 0.5882434944237919, "grad_norm": 0.6605568551505012, "learning_rate": 3.83756652178779e-06, "loss": 0.0623, "step": 12659 }, { "epoch": 0.5882899628252788, "grad_norm": 0.5101674576749934, "learning_rate": 3.836834640808783e-06, "loss": 0.0466, "step": 12660 }, { "epoch": 0.5883364312267658, "grad_norm": 0.6441424920303578, "learning_rate": 3.836102786176697e-06, "loss": 0.0368, "step": 12661 }, { "epoch": 0.5883828996282527, "grad_norm": 0.7320145120685079, "learning_rate": 3.835370957908108e-06, "loss": 0.0427, "step": 12662 }, { "epoch": 0.5884293680297398, "grad_norm": 0.5368929093898136, "learning_rate": 3.834639156019592e-06, "loss": 0.0634, "step": 12663 }, { "epoch": 0.5884758364312268, "grad_norm": 0.6673720723283479, "learning_rate": 3.833907380527725e-06, "loss": 0.0486, "step": 12664 }, { "epoch": 0.5885223048327137, "grad_norm": 0.2961761545513149, "learning_rate": 3.8331756314490835e-06, "loss": 0.0188, "step": 12665 }, { "epoch": 0.5885687732342008, "grad_norm": 0.49289497588120346, "learning_rate": 3.8324439088002415e-06, "loss": 0.0305, "step": 12666 }, { "epoch": 0.5886152416356877, "grad_norm": 0.3882948792797234, "learning_rate": 3.8317122125977765e-06, "loss": 0.0297, "step": 12667 }, { "epoch": 0.5886617100371747, "grad_norm": 0.8943208975671602, "learning_rate": 3.830980542858256e-06, "loss": 0.0617, "step": 12668 }, { "epoch": 0.5887081784386617, "grad_norm": 0.7797102706193165, "learning_rate": 3.830248899598259e-06, "loss": 0.0477, "step": 12669 }, { "epoch": 0.5887546468401487, "grad_norm": 0.449884815118582, "learning_rate": 3.829517282834354e-06, "loss": 0.0269, "step": 12670 }, { "epoch": 0.5888011152416357, "grad_norm": 0.5482782270238955, "learning_rate": 3.828785692583114e-06, "loss": 0.0385, "step": 12671 }, { "epoch": 0.5888475836431227, "grad_norm": 0.5458189762158847, "learning_rate": 3.828054128861112e-06, "loss": 0.0423, "step": 12672 }, { "epoch": 0.5888940520446097, "grad_norm": 0.5862463403265179, "learning_rate": 3.8273225916849185e-06, "loss": 0.0597, "step": 12673 }, { "epoch": 0.5889405204460967, "grad_norm": 0.882322759835382, "learning_rate": 3.826591081071102e-06, "loss": 0.0761, "step": 12674 }, { "epoch": 0.5889869888475836, "grad_norm": 0.5007041522911395, "learning_rate": 3.82585959703623e-06, "loss": 0.0426, "step": 12675 }, { "epoch": 0.5890334572490706, "grad_norm": 0.3823576756402706, "learning_rate": 3.825128139596877e-06, "loss": 0.0393, "step": 12676 }, { "epoch": 0.5890799256505577, "grad_norm": 0.731288677429486, "learning_rate": 3.824396708769605e-06, "loss": 0.0717, "step": 12677 }, { "epoch": 0.5891263940520446, "grad_norm": 1.0669408792690143, "learning_rate": 3.823665304570987e-06, "loss": 0.0629, "step": 12678 }, { "epoch": 0.5891728624535316, "grad_norm": 0.7754692784224445, "learning_rate": 3.822933927017589e-06, "loss": 0.059, "step": 12679 }, { "epoch": 0.5892193308550185, "grad_norm": 0.502187210603454, "learning_rate": 3.822202576125974e-06, "loss": 0.0611, "step": 12680 }, { "epoch": 0.5892657992565056, "grad_norm": 0.6912726010533498, "learning_rate": 3.821471251912712e-06, "loss": 0.0618, "step": 12681 }, { "epoch": 0.5893122676579926, "grad_norm": 1.6413554151394338, "learning_rate": 3.8207399543943646e-06, "loss": 0.0434, "step": 12682 }, { "epoch": 0.5893587360594795, "grad_norm": 0.5340698610421324, "learning_rate": 3.8200086835875e-06, "loss": 0.0502, "step": 12683 }, { "epoch": 0.5894052044609666, "grad_norm": 0.3550051821396489, "learning_rate": 3.819277439508679e-06, "loss": 0.0349, "step": 12684 }, { "epoch": 0.5894516728624535, "grad_norm": 0.4711477943259222, "learning_rate": 3.818546222174469e-06, "loss": 0.0467, "step": 12685 }, { "epoch": 0.5894981412639405, "grad_norm": 13.044043206816953, "learning_rate": 3.81781503160143e-06, "loss": 0.0559, "step": 12686 }, { "epoch": 0.5895446096654275, "grad_norm": 0.5056872816724395, "learning_rate": 3.817083867806123e-06, "loss": 0.051, "step": 12687 }, { "epoch": 0.5895910780669145, "grad_norm": 0.665595234009041, "learning_rate": 3.816352730805113e-06, "loss": 0.0547, "step": 12688 }, { "epoch": 0.5896375464684015, "grad_norm": 0.558111012898816, "learning_rate": 3.815621620614958e-06, "loss": 0.0452, "step": 12689 }, { "epoch": 0.5896840148698885, "grad_norm": 0.6233961521641294, "learning_rate": 3.814890537252221e-06, "loss": 0.0506, "step": 12690 }, { "epoch": 0.5897304832713754, "grad_norm": 0.27832200349959224, "learning_rate": 3.814159480733462e-06, "loss": 0.0431, "step": 12691 }, { "epoch": 0.5897769516728625, "grad_norm": 0.723083087711528, "learning_rate": 3.813428451075236e-06, "loss": 0.0571, "step": 12692 }, { "epoch": 0.5898234200743494, "grad_norm": 1.1449648847961815, "learning_rate": 3.8126974482941064e-06, "loss": 0.069, "step": 12693 }, { "epoch": 0.5898698884758364, "grad_norm": 0.5343772624840796, "learning_rate": 3.8119664724066275e-06, "loss": 0.0366, "step": 12694 }, { "epoch": 0.5899163568773235, "grad_norm": 0.6655502793186571, "learning_rate": 3.8112355234293596e-06, "loss": 0.058, "step": 12695 }, { "epoch": 0.5899628252788104, "grad_norm": 0.6070340574599684, "learning_rate": 3.8105046013788595e-06, "loss": 0.0699, "step": 12696 }, { "epoch": 0.5900092936802974, "grad_norm": 0.7974471645525957, "learning_rate": 3.80977370627168e-06, "loss": 0.0614, "step": 12697 }, { "epoch": 0.5900557620817843, "grad_norm": 0.7820884184899258, "learning_rate": 3.809042838124379e-06, "loss": 0.0785, "step": 12698 }, { "epoch": 0.5901022304832714, "grad_norm": 2.2945142955316205, "learning_rate": 3.808311996953511e-06, "loss": 0.0446, "step": 12699 }, { "epoch": 0.5901486988847584, "grad_norm": 0.7605467873603653, "learning_rate": 3.8075811827756314e-06, "loss": 0.0471, "step": 12700 }, { "epoch": 0.5901951672862453, "grad_norm": 1.218509294723444, "learning_rate": 3.8068503956072918e-06, "loss": 0.0804, "step": 12701 }, { "epoch": 0.5902416356877324, "grad_norm": 1.383227637409215, "learning_rate": 3.8061196354650487e-06, "loss": 0.0746, "step": 12702 }, { "epoch": 0.5902881040892193, "grad_norm": 1.5129528205854177, "learning_rate": 3.8053889023654513e-06, "loss": 0.1057, "step": 12703 }, { "epoch": 0.5903345724907063, "grad_norm": 1.8954693294925562, "learning_rate": 3.8046581963250516e-06, "loss": 0.098, "step": 12704 }, { "epoch": 0.5903810408921933, "grad_norm": 0.8300917480908296, "learning_rate": 3.803927517360403e-06, "loss": 0.0682, "step": 12705 }, { "epoch": 0.5904275092936803, "grad_norm": 0.637948238717485, "learning_rate": 3.8031968654880534e-06, "loss": 0.0723, "step": 12706 }, { "epoch": 0.5904739776951673, "grad_norm": 0.451853076668103, "learning_rate": 3.802466240724556e-06, "loss": 0.0388, "step": 12707 }, { "epoch": 0.5905204460966543, "grad_norm": 2.479369981702592, "learning_rate": 3.8017356430864594e-06, "loss": 0.0534, "step": 12708 }, { "epoch": 0.5905669144981412, "grad_norm": 0.8943941235545226, "learning_rate": 3.801005072590309e-06, "loss": 0.0562, "step": 12709 }, { "epoch": 0.5906133828996283, "grad_norm": 0.3997538412990765, "learning_rate": 3.8002745292526577e-06, "loss": 0.0349, "step": 12710 }, { "epoch": 0.5906598513011152, "grad_norm": 0.5959902259995132, "learning_rate": 3.7995440130900494e-06, "loss": 0.0612, "step": 12711 }, { "epoch": 0.5907063197026022, "grad_norm": 0.4462360519688433, "learning_rate": 3.798813524119034e-06, "loss": 0.0335, "step": 12712 }, { "epoch": 0.5907527881040893, "grad_norm": 0.6142735471386468, "learning_rate": 3.7980830623561548e-06, "loss": 0.0509, "step": 12713 }, { "epoch": 0.5907992565055762, "grad_norm": 0.6226327779139729, "learning_rate": 3.7973526278179624e-06, "loss": 0.0461, "step": 12714 }, { "epoch": 0.5908457249070632, "grad_norm": 0.5543043510494652, "learning_rate": 3.7966222205209968e-06, "loss": 0.0466, "step": 12715 }, { "epoch": 0.5908921933085501, "grad_norm": 0.848150163436912, "learning_rate": 3.7958918404818033e-06, "loss": 0.0556, "step": 12716 }, { "epoch": 0.5909386617100372, "grad_norm": 0.41751483674318346, "learning_rate": 3.7951614877169285e-06, "loss": 0.0374, "step": 12717 }, { "epoch": 0.5909851301115242, "grad_norm": 0.32967700832672336, "learning_rate": 3.794431162242913e-06, "loss": 0.0339, "step": 12718 }, { "epoch": 0.5910315985130111, "grad_norm": 0.7763036660368157, "learning_rate": 3.793700864076302e-06, "loss": 0.0535, "step": 12719 }, { "epoch": 0.5910780669144982, "grad_norm": 1.130835060089822, "learning_rate": 3.7929705932336357e-06, "loss": 0.061, "step": 12720 }, { "epoch": 0.5911245353159851, "grad_norm": 0.7093379105498271, "learning_rate": 3.7922403497314553e-06, "loss": 0.0466, "step": 12721 }, { "epoch": 0.5911710037174721, "grad_norm": 0.6000790902430994, "learning_rate": 3.791510133586302e-06, "loss": 0.0465, "step": 12722 }, { "epoch": 0.5912174721189591, "grad_norm": 0.6627564911249485, "learning_rate": 3.7907799448147155e-06, "loss": 0.0588, "step": 12723 }, { "epoch": 0.5912639405204461, "grad_norm": 0.5678161796676121, "learning_rate": 3.7900497834332374e-06, "loss": 0.0643, "step": 12724 }, { "epoch": 0.5913104089219331, "grad_norm": 0.677269718800943, "learning_rate": 3.7893196494584035e-06, "loss": 0.0602, "step": 12725 }, { "epoch": 0.5913568773234201, "grad_norm": 1.3908860999032877, "learning_rate": 3.788589542906756e-06, "loss": 0.0757, "step": 12726 }, { "epoch": 0.591403345724907, "grad_norm": 0.4178119464065854, "learning_rate": 3.7878594637948292e-06, "loss": 0.0433, "step": 12727 }, { "epoch": 0.5914498141263941, "grad_norm": 0.7850318098601582, "learning_rate": 3.78712941213916e-06, "loss": 0.058, "step": 12728 }, { "epoch": 0.591496282527881, "grad_norm": 1.5804070049650192, "learning_rate": 3.7863993879562878e-06, "loss": 0.0585, "step": 12729 }, { "epoch": 0.591542750929368, "grad_norm": 0.6303004641980419, "learning_rate": 3.785669391262745e-06, "loss": 0.0486, "step": 12730 }, { "epoch": 0.5915892193308551, "grad_norm": 0.5512904229827025, "learning_rate": 3.78493942207507e-06, "loss": 0.0508, "step": 12731 }, { "epoch": 0.591635687732342, "grad_norm": 0.3349725959056434, "learning_rate": 3.7842094804097973e-06, "loss": 0.0329, "step": 12732 }, { "epoch": 0.591682156133829, "grad_norm": 0.7419229823692084, "learning_rate": 3.783479566283457e-06, "loss": 0.0378, "step": 12733 }, { "epoch": 0.5917286245353159, "grad_norm": 1.1183620448233802, "learning_rate": 3.782749679712586e-06, "loss": 0.0602, "step": 12734 }, { "epoch": 0.591775092936803, "grad_norm": 0.8386462150955264, "learning_rate": 3.782019820713715e-06, "loss": 0.0628, "step": 12735 }, { "epoch": 0.59182156133829, "grad_norm": 1.2226135683930301, "learning_rate": 3.7812899893033785e-06, "loss": 0.0791, "step": 12736 }, { "epoch": 0.5918680297397769, "grad_norm": 0.47473944962223646, "learning_rate": 3.7805601854981042e-06, "loss": 0.0422, "step": 12737 }, { "epoch": 0.591914498141264, "grad_norm": 0.44617241124658946, "learning_rate": 3.779830409314428e-06, "loss": 0.0417, "step": 12738 }, { "epoch": 0.5919609665427509, "grad_norm": 0.527040870977162, "learning_rate": 3.7791006607688758e-06, "loss": 0.0293, "step": 12739 }, { "epoch": 0.5920074349442379, "grad_norm": 0.854252194568589, "learning_rate": 3.778370939877978e-06, "loss": 0.0627, "step": 12740 }, { "epoch": 0.5920539033457249, "grad_norm": 0.6530978105347992, "learning_rate": 3.7776412466582644e-06, "loss": 0.062, "step": 12741 }, { "epoch": 0.5921003717472119, "grad_norm": 0.46159640703936416, "learning_rate": 3.7769115811262623e-06, "loss": 0.0499, "step": 12742 }, { "epoch": 0.5921468401486989, "grad_norm": 1.2520208842353073, "learning_rate": 3.776181943298502e-06, "loss": 0.0783, "step": 12743 }, { "epoch": 0.5921933085501859, "grad_norm": 0.6638361028421832, "learning_rate": 3.775452333191509e-06, "loss": 0.0723, "step": 12744 }, { "epoch": 0.5922397769516728, "grad_norm": 0.9900817936631132, "learning_rate": 3.7747227508218077e-06, "loss": 0.0532, "step": 12745 }, { "epoch": 0.5922862453531599, "grad_norm": 0.4837639555041114, "learning_rate": 3.773993196205926e-06, "loss": 0.038, "step": 12746 }, { "epoch": 0.5923327137546468, "grad_norm": 0.842335160580527, "learning_rate": 3.7732636693603886e-06, "loss": 0.0688, "step": 12747 }, { "epoch": 0.5923791821561338, "grad_norm": 0.4048565067687814, "learning_rate": 3.772534170301721e-06, "loss": 0.0394, "step": 12748 }, { "epoch": 0.5924256505576209, "grad_norm": 0.6249340860841432, "learning_rate": 3.771804699046445e-06, "loss": 0.0539, "step": 12749 }, { "epoch": 0.5924721189591078, "grad_norm": 0.472679937135877, "learning_rate": 3.771075255611088e-06, "loss": 0.0634, "step": 12750 }, { "epoch": 0.5925185873605948, "grad_norm": 0.6330208853109603, "learning_rate": 3.7703458400121685e-06, "loss": 0.0515, "step": 12751 }, { "epoch": 0.5925650557620817, "grad_norm": 0.49382451841189373, "learning_rate": 3.7696164522662087e-06, "loss": 0.0352, "step": 12752 }, { "epoch": 0.5926115241635688, "grad_norm": 0.5467642603734051, "learning_rate": 3.7688870923897324e-06, "loss": 0.0295, "step": 12753 }, { "epoch": 0.5926579925650558, "grad_norm": 0.48853948121567187, "learning_rate": 3.768157760399258e-06, "loss": 0.0655, "step": 12754 }, { "epoch": 0.5927044609665427, "grad_norm": 0.5708782467861953, "learning_rate": 3.7674284563113088e-06, "loss": 0.0436, "step": 12755 }, { "epoch": 0.5927509293680298, "grad_norm": 0.562987905423982, "learning_rate": 3.7666991801424013e-06, "loss": 0.0508, "step": 12756 }, { "epoch": 0.5927973977695167, "grad_norm": 0.6410558713350449, "learning_rate": 3.7659699319090547e-06, "loss": 0.0657, "step": 12757 }, { "epoch": 0.5928438661710037, "grad_norm": 0.7062580003697027, "learning_rate": 3.765240711627789e-06, "loss": 0.0787, "step": 12758 }, { "epoch": 0.5928903345724907, "grad_norm": 0.2666083750830265, "learning_rate": 3.7645115193151192e-06, "loss": 0.0269, "step": 12759 }, { "epoch": 0.5929368029739777, "grad_norm": 0.9253554368600456, "learning_rate": 3.763782354987565e-06, "loss": 0.0295, "step": 12760 }, { "epoch": 0.5929832713754647, "grad_norm": 0.7680929933616162, "learning_rate": 3.763053218661643e-06, "loss": 0.0594, "step": 12761 }, { "epoch": 0.5930297397769517, "grad_norm": 0.5235819339820402, "learning_rate": 3.762324110353865e-06, "loss": 0.0507, "step": 12762 }, { "epoch": 0.5930762081784386, "grad_norm": 0.8283352619533556, "learning_rate": 3.76159503008075e-06, "loss": 0.0717, "step": 12763 }, { "epoch": 0.5931226765799257, "grad_norm": 0.41930328688811874, "learning_rate": 3.7608659778588092e-06, "loss": 0.0524, "step": 12764 }, { "epoch": 0.5931691449814126, "grad_norm": 3.784053950494908, "learning_rate": 3.7601369537045594e-06, "loss": 0.0991, "step": 12765 }, { "epoch": 0.5932156133828996, "grad_norm": 0.5544424214954264, "learning_rate": 3.7594079576345123e-06, "loss": 0.0526, "step": 12766 }, { "epoch": 0.5932620817843867, "grad_norm": 1.6110349669063753, "learning_rate": 3.758678989665182e-06, "loss": 0.0611, "step": 12767 }, { "epoch": 0.5933085501858736, "grad_norm": 0.6963591230554592, "learning_rate": 3.7579500498130777e-06, "loss": 0.0427, "step": 12768 }, { "epoch": 0.5933550185873606, "grad_norm": 0.6300884138229249, "learning_rate": 3.7572211380947114e-06, "loss": 0.0664, "step": 12769 }, { "epoch": 0.5934014869888475, "grad_norm": 0.6459883720529194, "learning_rate": 3.7564922545265957e-06, "loss": 0.063, "step": 12770 }, { "epoch": 0.5934479553903346, "grad_norm": 0.41606520843005546, "learning_rate": 3.755763399125238e-06, "loss": 0.0582, "step": 12771 }, { "epoch": 0.5934944237918216, "grad_norm": 0.5108521642183059, "learning_rate": 3.7550345719071496e-06, "loss": 0.0857, "step": 12772 }, { "epoch": 0.5935408921933085, "grad_norm": 0.4252391861542432, "learning_rate": 3.7543057728888395e-06, "loss": 0.063, "step": 12773 }, { "epoch": 0.5935873605947956, "grad_norm": 0.5358093422007082, "learning_rate": 3.753577002086813e-06, "loss": 0.0419, "step": 12774 }, { "epoch": 0.5936338289962825, "grad_norm": 1.1282712396740457, "learning_rate": 3.7528482595175796e-06, "loss": 0.055, "step": 12775 }, { "epoch": 0.5936802973977695, "grad_norm": 0.5832112966281743, "learning_rate": 3.7521195451976455e-06, "loss": 0.0519, "step": 12776 }, { "epoch": 0.5937267657992565, "grad_norm": 0.6953442315797579, "learning_rate": 3.7513908591435176e-06, "loss": 0.0458, "step": 12777 }, { "epoch": 0.5937732342007435, "grad_norm": 0.48901644230810704, "learning_rate": 3.7506622013716996e-06, "loss": 0.0522, "step": 12778 }, { "epoch": 0.5938197026022305, "grad_norm": 0.6479423397424449, "learning_rate": 3.7499335718987005e-06, "loss": 0.0482, "step": 12779 }, { "epoch": 0.5938661710037175, "grad_norm": 0.777910698685802, "learning_rate": 3.74920497074102e-06, "loss": 0.0394, "step": 12780 }, { "epoch": 0.5939126394052044, "grad_norm": 0.5157030811800043, "learning_rate": 3.748476397915163e-06, "loss": 0.0591, "step": 12781 }, { "epoch": 0.5939591078066915, "grad_norm": 0.6403567503674329, "learning_rate": 3.747747853437633e-06, "loss": 0.0286, "step": 12782 }, { "epoch": 0.5940055762081784, "grad_norm": 2.3717725842083097, "learning_rate": 3.747019337324931e-06, "loss": 0.0803, "step": 12783 }, { "epoch": 0.5940520446096654, "grad_norm": 0.5271926330682153, "learning_rate": 3.746290849593561e-06, "loss": 0.0366, "step": 12784 }, { "epoch": 0.5940985130111525, "grad_norm": 0.8535752766311963, "learning_rate": 3.7455623902600246e-06, "loss": 0.0542, "step": 12785 }, { "epoch": 0.5941449814126394, "grad_norm": 0.7594223965780913, "learning_rate": 3.744833959340818e-06, "loss": 0.0421, "step": 12786 }, { "epoch": 0.5941914498141264, "grad_norm": 0.6474543862773192, "learning_rate": 3.744105556852443e-06, "loss": 0.0538, "step": 12787 }, { "epoch": 0.5942379182156133, "grad_norm": 0.599622374682457, "learning_rate": 3.7433771828113996e-06, "loss": 0.05, "step": 12788 }, { "epoch": 0.5942843866171004, "grad_norm": 0.7663329011900911, "learning_rate": 3.742648837234185e-06, "loss": 0.0498, "step": 12789 }, { "epoch": 0.5943308550185874, "grad_norm": 1.50664427844107, "learning_rate": 3.7419205201372977e-06, "loss": 0.0559, "step": 12790 }, { "epoch": 0.5943773234200743, "grad_norm": 0.5747593449035503, "learning_rate": 3.7411922315372367e-06, "loss": 0.0454, "step": 12791 }, { "epoch": 0.5944237918215614, "grad_norm": 2.425060418851548, "learning_rate": 3.740463971450495e-06, "loss": 0.0576, "step": 12792 }, { "epoch": 0.5944702602230483, "grad_norm": 1.0864564185713759, "learning_rate": 3.7397357398935695e-06, "loss": 0.0832, "step": 12793 }, { "epoch": 0.5945167286245353, "grad_norm": 0.724042540982755, "learning_rate": 3.7390075368829572e-06, "loss": 0.0663, "step": 12794 }, { "epoch": 0.5945631970260223, "grad_norm": 0.7133358887423709, "learning_rate": 3.7382793624351494e-06, "loss": 0.0573, "step": 12795 }, { "epoch": 0.5946096654275093, "grad_norm": 0.6582074390741867, "learning_rate": 3.7375512165666435e-06, "loss": 0.0804, "step": 12796 }, { "epoch": 0.5946561338289963, "grad_norm": 0.80337341510117, "learning_rate": 3.7368230992939327e-06, "loss": 0.0686, "step": 12797 }, { "epoch": 0.5947026022304833, "grad_norm": 0.7776297573421231, "learning_rate": 3.736095010633505e-06, "loss": 0.0653, "step": 12798 }, { "epoch": 0.5947490706319702, "grad_norm": 0.9069102187321998, "learning_rate": 3.7353669506018576e-06, "loss": 0.0568, "step": 12799 }, { "epoch": 0.5947955390334573, "grad_norm": 0.6191144650983934, "learning_rate": 3.7346389192154776e-06, "loss": 0.0449, "step": 12800 }, { "epoch": 0.5948420074349442, "grad_norm": 0.6571999595097661, "learning_rate": 3.73391091649086e-06, "loss": 0.0552, "step": 12801 }, { "epoch": 0.5948884758364312, "grad_norm": 0.48482667285995945, "learning_rate": 3.733182942444491e-06, "loss": 0.0268, "step": 12802 }, { "epoch": 0.5949349442379183, "grad_norm": 1.3842067937981952, "learning_rate": 3.7324549970928636e-06, "loss": 0.0485, "step": 12803 }, { "epoch": 0.5949814126394052, "grad_norm": 0.4239753984873684, "learning_rate": 3.731727080452464e-06, "loss": 0.0404, "step": 12804 }, { "epoch": 0.5950278810408922, "grad_norm": 2.3206754226543715, "learning_rate": 3.73099919253978e-06, "loss": 0.0597, "step": 12805 }, { "epoch": 0.5950743494423791, "grad_norm": 1.0245049114459035, "learning_rate": 3.7302713333713005e-06, "loss": 0.0544, "step": 12806 }, { "epoch": 0.5951208178438662, "grad_norm": 1.0451414355702036, "learning_rate": 3.7295435029635106e-06, "loss": 0.0571, "step": 12807 }, { "epoch": 0.5951672862453532, "grad_norm": 0.9279536387325449, "learning_rate": 3.728815701332899e-06, "loss": 0.0484, "step": 12808 }, { "epoch": 0.5952137546468401, "grad_norm": 0.6507191048219084, "learning_rate": 3.7280879284959503e-06, "loss": 0.0724, "step": 12809 }, { "epoch": 0.5952602230483272, "grad_norm": 0.7522160383088605, "learning_rate": 3.7273601844691466e-06, "loss": 0.0555, "step": 12810 }, { "epoch": 0.5953066914498141, "grad_norm": 0.6303384773269346, "learning_rate": 3.726632469268976e-06, "loss": 0.0581, "step": 12811 }, { "epoch": 0.5953531598513011, "grad_norm": 0.420863816656032, "learning_rate": 3.7259047829119177e-06, "loss": 0.0509, "step": 12812 }, { "epoch": 0.5953996282527881, "grad_norm": 1.0523477564197212, "learning_rate": 3.725177125414459e-06, "loss": 0.0687, "step": 12813 }, { "epoch": 0.5954460966542751, "grad_norm": 0.6967893369341553, "learning_rate": 3.724449496793081e-06, "loss": 0.0567, "step": 12814 }, { "epoch": 0.5954925650557621, "grad_norm": 0.5664022654643017, "learning_rate": 3.7237218970642624e-06, "loss": 0.0617, "step": 12815 }, { "epoch": 0.5955390334572491, "grad_norm": 1.8238820059688836, "learning_rate": 3.722994326244487e-06, "loss": 0.0423, "step": 12816 }, { "epoch": 0.595585501858736, "grad_norm": 0.8325120345575381, "learning_rate": 3.722266784350233e-06, "loss": 0.0522, "step": 12817 }, { "epoch": 0.5956319702602231, "grad_norm": 1.2652103422704477, "learning_rate": 3.7215392713979824e-06, "loss": 0.0725, "step": 12818 }, { "epoch": 0.59567843866171, "grad_norm": 1.8747262385525052, "learning_rate": 3.7208117874042114e-06, "loss": 0.0562, "step": 12819 }, { "epoch": 0.595724907063197, "grad_norm": 0.6606363016335424, "learning_rate": 3.7200843323854026e-06, "loss": 0.0667, "step": 12820 }, { "epoch": 0.5957713754646841, "grad_norm": 1.4988541562624278, "learning_rate": 3.7193569063580295e-06, "loss": 0.0877, "step": 12821 }, { "epoch": 0.595817843866171, "grad_norm": 1.103173538822336, "learning_rate": 3.7186295093385685e-06, "loss": 0.0625, "step": 12822 }, { "epoch": 0.595864312267658, "grad_norm": 0.5604330929889602, "learning_rate": 3.717902141343499e-06, "loss": 0.0567, "step": 12823 }, { "epoch": 0.5959107806691449, "grad_norm": 0.8650720262317243, "learning_rate": 3.717174802389295e-06, "loss": 0.0617, "step": 12824 }, { "epoch": 0.595957249070632, "grad_norm": 0.44004791962832607, "learning_rate": 3.716447492492432e-06, "loss": 0.0339, "step": 12825 }, { "epoch": 0.596003717472119, "grad_norm": 1.4437441864192908, "learning_rate": 3.7157202116693864e-06, "loss": 0.0896, "step": 12826 }, { "epoch": 0.5960501858736059, "grad_norm": 0.5177932149973931, "learning_rate": 3.714992959936627e-06, "loss": 0.0527, "step": 12827 }, { "epoch": 0.596096654275093, "grad_norm": 0.4647630133374352, "learning_rate": 3.7142657373106306e-06, "loss": 0.0475, "step": 12828 }, { "epoch": 0.5961431226765799, "grad_norm": 0.979207838818383, "learning_rate": 3.7135385438078677e-06, "loss": 0.0653, "step": 12829 }, { "epoch": 0.5961895910780669, "grad_norm": 1.0023290274141092, "learning_rate": 3.7128113794448106e-06, "loss": 0.073, "step": 12830 }, { "epoch": 0.5962360594795539, "grad_norm": 0.5631509663650497, "learning_rate": 3.7120842442379303e-06, "loss": 0.0537, "step": 12831 }, { "epoch": 0.5962825278810409, "grad_norm": 0.43827405710597966, "learning_rate": 3.7113571382036995e-06, "loss": 0.0367, "step": 12832 }, { "epoch": 0.5963289962825279, "grad_norm": 0.5196423715164122, "learning_rate": 3.710630061358585e-06, "loss": 0.0555, "step": 12833 }, { "epoch": 0.5963754646840149, "grad_norm": 0.6775057057035826, "learning_rate": 3.7099030137190552e-06, "loss": 0.0557, "step": 12834 }, { "epoch": 0.5964219330855018, "grad_norm": 0.45557403910923516, "learning_rate": 3.7091759953015817e-06, "loss": 0.0473, "step": 12835 }, { "epoch": 0.5964684014869889, "grad_norm": 0.3414668796025316, "learning_rate": 3.7084490061226284e-06, "loss": 0.0412, "step": 12836 }, { "epoch": 0.5965148698884758, "grad_norm": 0.44602188761060946, "learning_rate": 3.707722046198666e-06, "loss": 0.0422, "step": 12837 }, { "epoch": 0.5965613382899628, "grad_norm": 0.852031386678494, "learning_rate": 3.7069951155461605e-06, "loss": 0.0619, "step": 12838 }, { "epoch": 0.5966078066914499, "grad_norm": 0.460159101933578, "learning_rate": 3.7062682141815743e-06, "loss": 0.0517, "step": 12839 }, { "epoch": 0.5966542750929368, "grad_norm": 0.7232143597784748, "learning_rate": 3.705541342121375e-06, "loss": 0.0407, "step": 12840 }, { "epoch": 0.5967007434944238, "grad_norm": 7.062237005200322, "learning_rate": 3.704814499382027e-06, "loss": 0.0894, "step": 12841 }, { "epoch": 0.5967472118959107, "grad_norm": 0.4365111589220228, "learning_rate": 3.7040876859799933e-06, "loss": 0.0547, "step": 12842 }, { "epoch": 0.5967936802973978, "grad_norm": 0.7434401087085845, "learning_rate": 3.7033609019317374e-06, "loss": 0.0521, "step": 12843 }, { "epoch": 0.5968401486988848, "grad_norm": 0.8023742981468273, "learning_rate": 3.7026341472537234e-06, "loss": 0.0621, "step": 12844 }, { "epoch": 0.5968866171003717, "grad_norm": 0.8920680211542689, "learning_rate": 3.701907421962411e-06, "loss": 0.0514, "step": 12845 }, { "epoch": 0.5969330855018588, "grad_norm": 1.0727478707713147, "learning_rate": 3.7011807260742606e-06, "loss": 0.0539, "step": 12846 }, { "epoch": 0.5969795539033457, "grad_norm": 0.8290733953820213, "learning_rate": 3.7004540596057335e-06, "loss": 0.0598, "step": 12847 }, { "epoch": 0.5970260223048327, "grad_norm": 0.4547280373533555, "learning_rate": 3.6997274225732906e-06, "loss": 0.0643, "step": 12848 }, { "epoch": 0.5970724907063197, "grad_norm": 0.8053852719985745, "learning_rate": 3.69900081499339e-06, "loss": 0.0569, "step": 12849 }, { "epoch": 0.5971189591078067, "grad_norm": 0.7849173609410379, "learning_rate": 3.698274236882491e-06, "loss": 0.0794, "step": 12850 }, { "epoch": 0.5971654275092937, "grad_norm": 0.5481158461981175, "learning_rate": 3.6975476882570495e-06, "loss": 0.0676, "step": 12851 }, { "epoch": 0.5972118959107807, "grad_norm": 0.7612216200071326, "learning_rate": 3.6968211691335232e-06, "loss": 0.0504, "step": 12852 }, { "epoch": 0.5972583643122676, "grad_norm": 0.3944794862389407, "learning_rate": 3.6960946795283693e-06, "loss": 0.0421, "step": 12853 }, { "epoch": 0.5973048327137547, "grad_norm": 0.3715139247345235, "learning_rate": 3.695368219458044e-06, "loss": 0.0313, "step": 12854 }, { "epoch": 0.5973513011152416, "grad_norm": 0.49561133616048964, "learning_rate": 3.694641788939e-06, "loss": 0.0406, "step": 12855 }, { "epoch": 0.5973977695167286, "grad_norm": 0.9914014196583716, "learning_rate": 3.693915387987696e-06, "loss": 0.0544, "step": 12856 }, { "epoch": 0.5974442379182157, "grad_norm": 0.42562303567993076, "learning_rate": 3.693189016620582e-06, "loss": 0.0328, "step": 12857 }, { "epoch": 0.5974907063197026, "grad_norm": 1.3949739979840163, "learning_rate": 3.692462674854111e-06, "loss": 0.064, "step": 12858 }, { "epoch": 0.5975371747211896, "grad_norm": 0.6325751090599265, "learning_rate": 3.6917363627047375e-06, "loss": 0.0761, "step": 12859 }, { "epoch": 0.5975836431226765, "grad_norm": 0.36897683414778687, "learning_rate": 3.6910100801889117e-06, "loss": 0.0344, "step": 12860 }, { "epoch": 0.5976301115241636, "grad_norm": 0.5887670524611219, "learning_rate": 3.690283827323086e-06, "loss": 0.0707, "step": 12861 }, { "epoch": 0.5976765799256506, "grad_norm": 1.162771055384633, "learning_rate": 3.6895576041237114e-06, "loss": 0.0456, "step": 12862 }, { "epoch": 0.5977230483271375, "grad_norm": 1.2780149421246811, "learning_rate": 3.6888314106072332e-06, "loss": 0.0364, "step": 12863 }, { "epoch": 0.5977695167286246, "grad_norm": 1.075475611031451, "learning_rate": 3.6881052467901056e-06, "loss": 0.088, "step": 12864 }, { "epoch": 0.5978159851301115, "grad_norm": 0.7151972499669504, "learning_rate": 3.687379112688774e-06, "loss": 0.0757, "step": 12865 }, { "epoch": 0.5978624535315985, "grad_norm": 1.3650410813760185, "learning_rate": 3.686653008319687e-06, "loss": 0.0382, "step": 12866 }, { "epoch": 0.5979089219330855, "grad_norm": 1.141134361964278, "learning_rate": 3.685926933699291e-06, "loss": 0.047, "step": 12867 }, { "epoch": 0.5979553903345725, "grad_norm": 1.0018964798808472, "learning_rate": 3.6852008888440354e-06, "loss": 0.0815, "step": 12868 }, { "epoch": 0.5980018587360595, "grad_norm": 0.5116666756264477, "learning_rate": 3.6844748737703628e-06, "loss": 0.0342, "step": 12869 }, { "epoch": 0.5980483271375465, "grad_norm": 0.8404628283764108, "learning_rate": 3.683748888494717e-06, "loss": 0.0421, "step": 12870 }, { "epoch": 0.5980947955390334, "grad_norm": 1.4940125822997168, "learning_rate": 3.6830229330335453e-06, "loss": 0.0758, "step": 12871 }, { "epoch": 0.5981412639405205, "grad_norm": 1.8195234697965135, "learning_rate": 3.6822970074032895e-06, "loss": 0.0639, "step": 12872 }, { "epoch": 0.5981877323420074, "grad_norm": 0.6429661362791557, "learning_rate": 3.6815711116203957e-06, "loss": 0.0497, "step": 12873 }, { "epoch": 0.5982342007434944, "grad_norm": 1.8254648833208438, "learning_rate": 3.6808452457013023e-06, "loss": 0.0731, "step": 12874 }, { "epoch": 0.5982806691449815, "grad_norm": 0.762210226788041, "learning_rate": 3.6801194096624515e-06, "loss": 0.0551, "step": 12875 }, { "epoch": 0.5983271375464684, "grad_norm": 0.34244203965526254, "learning_rate": 3.679393603520287e-06, "loss": 0.0412, "step": 12876 }, { "epoch": 0.5983736059479554, "grad_norm": 0.9413330162385315, "learning_rate": 3.678667827291245e-06, "loss": 0.0657, "step": 12877 }, { "epoch": 0.5984200743494423, "grad_norm": 0.8232500484948175, "learning_rate": 3.6779420809917694e-06, "loss": 0.0634, "step": 12878 }, { "epoch": 0.5984665427509294, "grad_norm": 0.5549792724106243, "learning_rate": 3.6772163646382975e-06, "loss": 0.0438, "step": 12879 }, { "epoch": 0.5985130111524164, "grad_norm": 0.8690956522846859, "learning_rate": 3.676490678247265e-06, "loss": 0.0721, "step": 12880 }, { "epoch": 0.5985594795539033, "grad_norm": 0.6905943735084152, "learning_rate": 3.6757650218351136e-06, "loss": 0.0425, "step": 12881 }, { "epoch": 0.5986059479553903, "grad_norm": 0.6126896754822576, "learning_rate": 3.6750393954182766e-06, "loss": 0.0529, "step": 12882 }, { "epoch": 0.5986524163568773, "grad_norm": 0.6302752183945204, "learning_rate": 3.674313799013193e-06, "loss": 0.0618, "step": 12883 }, { "epoch": 0.5986988847583643, "grad_norm": 0.5716473480113052, "learning_rate": 3.6735882326362958e-06, "loss": 0.0426, "step": 12884 }, { "epoch": 0.5987453531598513, "grad_norm": 0.749393334381981, "learning_rate": 3.6728626963040236e-06, "loss": 0.0467, "step": 12885 }, { "epoch": 0.5987918215613383, "grad_norm": 0.4314830163864807, "learning_rate": 3.6721371900328064e-06, "loss": 0.0535, "step": 12886 }, { "epoch": 0.5988382899628253, "grad_norm": 0.6857121386950059, "learning_rate": 3.671411713839079e-06, "loss": 0.073, "step": 12887 }, { "epoch": 0.5988847583643123, "grad_norm": 0.37182672530817507, "learning_rate": 3.6706862677392752e-06, "loss": 0.0501, "step": 12888 }, { "epoch": 0.5989312267657992, "grad_norm": 0.44577139851308933, "learning_rate": 3.6699608517498257e-06, "loss": 0.0492, "step": 12889 }, { "epoch": 0.5989776951672863, "grad_norm": 0.8385006540816612, "learning_rate": 3.669235465887164e-06, "loss": 0.0612, "step": 12890 }, { "epoch": 0.5990241635687732, "grad_norm": 0.6481725406003181, "learning_rate": 3.66851011016772e-06, "loss": 0.0434, "step": 12891 }, { "epoch": 0.5990706319702602, "grad_norm": 0.5214932928432955, "learning_rate": 3.6677847846079216e-06, "loss": 0.0628, "step": 12892 }, { "epoch": 0.5991171003717473, "grad_norm": 0.6340416870213406, "learning_rate": 3.667059489224202e-06, "loss": 0.0513, "step": 12893 }, { "epoch": 0.5991635687732342, "grad_norm": 1.5196660383377938, "learning_rate": 3.666334224032985e-06, "loss": 0.0621, "step": 12894 }, { "epoch": 0.5992100371747212, "grad_norm": 0.5181441808531465, "learning_rate": 3.6656089890507042e-06, "loss": 0.0534, "step": 12895 }, { "epoch": 0.5992565055762081, "grad_norm": 0.6564270541450641, "learning_rate": 3.6648837842937823e-06, "loss": 0.0495, "step": 12896 }, { "epoch": 0.5993029739776952, "grad_norm": 0.6051616441891488, "learning_rate": 3.6641586097786495e-06, "loss": 0.065, "step": 12897 }, { "epoch": 0.5993494423791822, "grad_norm": 0.5754023949836732, "learning_rate": 3.6634334655217296e-06, "loss": 0.0505, "step": 12898 }, { "epoch": 0.5993959107806691, "grad_norm": 0.46604101581045004, "learning_rate": 3.6627083515394478e-06, "loss": 0.0306, "step": 12899 }, { "epoch": 0.5994423791821561, "grad_norm": 0.5305895112270514, "learning_rate": 3.6619832678482305e-06, "loss": 0.0481, "step": 12900 }, { "epoch": 0.5994888475836431, "grad_norm": 0.46564171552559186, "learning_rate": 3.6612582144644983e-06, "loss": 0.0596, "step": 12901 }, { "epoch": 0.5995353159851301, "grad_norm": 0.6386995062556075, "learning_rate": 3.660533191404678e-06, "loss": 0.0549, "step": 12902 }, { "epoch": 0.5995817843866171, "grad_norm": 0.37248733396529266, "learning_rate": 3.659808198685192e-06, "loss": 0.0438, "step": 12903 }, { "epoch": 0.599628252788104, "grad_norm": 0.9825549327097386, "learning_rate": 3.6590832363224587e-06, "loss": 0.0611, "step": 12904 }, { "epoch": 0.5996747211895911, "grad_norm": 0.5883272492543619, "learning_rate": 3.6583583043329025e-06, "loss": 0.0455, "step": 12905 }, { "epoch": 0.5997211895910781, "grad_norm": 0.38709826876543113, "learning_rate": 3.657633402732941e-06, "loss": 0.0476, "step": 12906 }, { "epoch": 0.599767657992565, "grad_norm": 1.006659726176422, "learning_rate": 3.6569085315389974e-06, "loss": 0.0478, "step": 12907 }, { "epoch": 0.5998141263940521, "grad_norm": 0.36410370766849154, "learning_rate": 3.6561836907674887e-06, "loss": 0.0493, "step": 12908 }, { "epoch": 0.599860594795539, "grad_norm": 0.43889406564197664, "learning_rate": 3.6554588804348344e-06, "loss": 0.0581, "step": 12909 }, { "epoch": 0.599907063197026, "grad_norm": 0.6224443019472866, "learning_rate": 3.6547341005574514e-06, "loss": 0.0658, "step": 12910 }, { "epoch": 0.5999535315985131, "grad_norm": 0.3748005777547437, "learning_rate": 3.654009351151756e-06, "loss": 0.0496, "step": 12911 }, { "epoch": 0.6, "grad_norm": 0.47197023423328155, "learning_rate": 3.6532846322341665e-06, "loss": 0.0435, "step": 12912 }, { "epoch": 0.600046468401487, "grad_norm": 1.1323853954963279, "learning_rate": 3.6525599438210956e-06, "loss": 0.0525, "step": 12913 }, { "epoch": 0.6000929368029739, "grad_norm": 2.178217459284315, "learning_rate": 3.6518352859289617e-06, "loss": 0.0448, "step": 12914 }, { "epoch": 0.600139405204461, "grad_norm": 0.9476644834238337, "learning_rate": 3.6511106585741786e-06, "loss": 0.0596, "step": 12915 }, { "epoch": 0.600185873605948, "grad_norm": 0.7929371938564808, "learning_rate": 3.650386061773157e-06, "loss": 0.0554, "step": 12916 }, { "epoch": 0.6002323420074349, "grad_norm": 0.8381223222621792, "learning_rate": 3.649661495542312e-06, "loss": 0.0614, "step": 12917 }, { "epoch": 0.600278810408922, "grad_norm": 0.6506089293070793, "learning_rate": 3.6489369598980547e-06, "loss": 0.0528, "step": 12918 }, { "epoch": 0.6003252788104089, "grad_norm": 0.9515331794176982, "learning_rate": 3.6482124548567977e-06, "loss": 0.0583, "step": 12919 }, { "epoch": 0.6003717472118959, "grad_norm": 0.4751059302132267, "learning_rate": 3.6474879804349504e-06, "loss": 0.0367, "step": 12920 }, { "epoch": 0.6004182156133829, "grad_norm": 0.7401549963212453, "learning_rate": 3.646763536648926e-06, "loss": 0.0464, "step": 12921 }, { "epoch": 0.6004646840148699, "grad_norm": 0.7229030091001742, "learning_rate": 3.6460391235151316e-06, "loss": 0.0606, "step": 12922 }, { "epoch": 0.6005111524163569, "grad_norm": 0.9082350986049986, "learning_rate": 3.6453147410499746e-06, "loss": 0.0556, "step": 12923 }, { "epoch": 0.6005576208178439, "grad_norm": 0.7998804362426815, "learning_rate": 3.6445903892698647e-06, "loss": 0.0395, "step": 12924 }, { "epoch": 0.6006040892193308, "grad_norm": 0.5206654292045113, "learning_rate": 3.6438660681912085e-06, "loss": 0.0618, "step": 12925 }, { "epoch": 0.6006505576208179, "grad_norm": 1.1085383854857385, "learning_rate": 3.6431417778304147e-06, "loss": 0.0631, "step": 12926 }, { "epoch": 0.6006970260223048, "grad_norm": 0.46829071448736603, "learning_rate": 3.642417518203888e-06, "loss": 0.0263, "step": 12927 }, { "epoch": 0.6007434944237918, "grad_norm": 0.6171056433908011, "learning_rate": 3.6416932893280316e-06, "loss": 0.0412, "step": 12928 }, { "epoch": 0.6007899628252789, "grad_norm": 0.9389300297438057, "learning_rate": 3.640969091219253e-06, "loss": 0.0629, "step": 12929 }, { "epoch": 0.6008364312267658, "grad_norm": 0.5807132856107646, "learning_rate": 3.640244923893953e-06, "loss": 0.0557, "step": 12930 }, { "epoch": 0.6008828996282528, "grad_norm": 0.5762593541288678, "learning_rate": 3.6395207873685383e-06, "loss": 0.0367, "step": 12931 }, { "epoch": 0.6009293680297397, "grad_norm": 0.9068850432642244, "learning_rate": 3.6387966816594104e-06, "loss": 0.0701, "step": 12932 }, { "epoch": 0.6009758364312268, "grad_norm": 1.127217012360805, "learning_rate": 3.6380726067829688e-06, "loss": 0.0619, "step": 12933 }, { "epoch": 0.6010223048327138, "grad_norm": 0.5542851583466315, "learning_rate": 3.6373485627556166e-06, "loss": 0.0434, "step": 12934 }, { "epoch": 0.6010687732342007, "grad_norm": 0.528855293236807, "learning_rate": 3.6366245495937524e-06, "loss": 0.0534, "step": 12935 }, { "epoch": 0.6011152416356877, "grad_norm": 0.5709329388906447, "learning_rate": 3.635900567313778e-06, "loss": 0.0613, "step": 12936 }, { "epoch": 0.6011617100371747, "grad_norm": 0.6251096689818642, "learning_rate": 3.635176615932091e-06, "loss": 0.061, "step": 12937 }, { "epoch": 0.6012081784386617, "grad_norm": 0.5182050583639002, "learning_rate": 3.634452695465093e-06, "loss": 0.0531, "step": 12938 }, { "epoch": 0.6012546468401487, "grad_norm": 0.8408191172600215, "learning_rate": 3.633728805929176e-06, "loss": 0.0593, "step": 12939 }, { "epoch": 0.6013011152416357, "grad_norm": 0.6851082030269385, "learning_rate": 3.6330049473407394e-06, "loss": 0.0472, "step": 12940 }, { "epoch": 0.6013475836431227, "grad_norm": 0.4990999562723541, "learning_rate": 3.63228111971618e-06, "loss": 0.0564, "step": 12941 }, { "epoch": 0.6013940520446097, "grad_norm": 0.7127458379687913, "learning_rate": 3.6315573230718915e-06, "loss": 0.0691, "step": 12942 }, { "epoch": 0.6014405204460966, "grad_norm": 1.2283882446704768, "learning_rate": 3.6308335574242713e-06, "loss": 0.06, "step": 12943 }, { "epoch": 0.6014869888475837, "grad_norm": 0.5534047022404555, "learning_rate": 3.6301098227897134e-06, "loss": 0.0484, "step": 12944 }, { "epoch": 0.6015334572490706, "grad_norm": 0.6986750418760921, "learning_rate": 3.629386119184607e-06, "loss": 0.055, "step": 12945 }, { "epoch": 0.6015799256505576, "grad_norm": 0.6430842524951732, "learning_rate": 3.6286624466253496e-06, "loss": 0.0639, "step": 12946 }, { "epoch": 0.6016263940520447, "grad_norm": 1.1446220110393917, "learning_rate": 3.6279388051283294e-06, "loss": 0.0841, "step": 12947 }, { "epoch": 0.6016728624535316, "grad_norm": 2.3590246578919665, "learning_rate": 3.62721519470994e-06, "loss": 0.0978, "step": 12948 }, { "epoch": 0.6017193308550186, "grad_norm": 0.4219239630543818, "learning_rate": 3.6264916153865703e-06, "loss": 0.0264, "step": 12949 }, { "epoch": 0.6017657992565055, "grad_norm": 1.4121021121925512, "learning_rate": 3.625768067174614e-06, "loss": 0.055, "step": 12950 }, { "epoch": 0.6018122676579926, "grad_norm": 0.5705958981903573, "learning_rate": 3.6250445500904556e-06, "loss": 0.0491, "step": 12951 }, { "epoch": 0.6018587360594796, "grad_norm": 0.5725642244527925, "learning_rate": 3.6243210641504847e-06, "loss": 0.052, "step": 12952 }, { "epoch": 0.6019052044609665, "grad_norm": 0.4751413965030579, "learning_rate": 3.6235976093710902e-06, "loss": 0.0513, "step": 12953 }, { "epoch": 0.6019516728624535, "grad_norm": 0.9440880721618455, "learning_rate": 3.6228741857686578e-06, "loss": 0.0705, "step": 12954 }, { "epoch": 0.6019981412639405, "grad_norm": 2.1671629745589285, "learning_rate": 3.6221507933595756e-06, "loss": 0.0793, "step": 12955 }, { "epoch": 0.6020446096654275, "grad_norm": 0.6807087755958109, "learning_rate": 3.6214274321602283e-06, "loss": 0.0649, "step": 12956 }, { "epoch": 0.6020910780669145, "grad_norm": 1.2771753978845624, "learning_rate": 3.6207041021869992e-06, "loss": 0.1009, "step": 12957 }, { "epoch": 0.6021375464684015, "grad_norm": 0.5748285803318791, "learning_rate": 3.619980803456275e-06, "loss": 0.0366, "step": 12958 }, { "epoch": 0.6021840148698885, "grad_norm": 0.5688060210677427, "learning_rate": 3.6192575359844365e-06, "loss": 0.0504, "step": 12959 }, { "epoch": 0.6022304832713755, "grad_norm": 1.3012522508141444, "learning_rate": 3.618534299787869e-06, "loss": 0.0575, "step": 12960 }, { "epoch": 0.6022769516728624, "grad_norm": 0.5413409474659986, "learning_rate": 3.6178110948829536e-06, "loss": 0.052, "step": 12961 }, { "epoch": 0.6023234200743495, "grad_norm": 0.7335944996104535, "learning_rate": 3.617087921286073e-06, "loss": 0.0772, "step": 12962 }, { "epoch": 0.6023698884758364, "grad_norm": 0.6299485285967424, "learning_rate": 3.616364779013605e-06, "loss": 0.0611, "step": 12963 }, { "epoch": 0.6024163568773234, "grad_norm": 0.703909861018487, "learning_rate": 3.6156416680819306e-06, "loss": 0.055, "step": 12964 }, { "epoch": 0.6024628252788105, "grad_norm": 0.6311184956532278, "learning_rate": 3.61491858850743e-06, "loss": 0.0779, "step": 12965 }, { "epoch": 0.6025092936802974, "grad_norm": 0.4798763354963695, "learning_rate": 3.6141955403064806e-06, "loss": 0.0392, "step": 12966 }, { "epoch": 0.6025557620817844, "grad_norm": 0.6695441068768593, "learning_rate": 3.6134725234954615e-06, "loss": 0.0589, "step": 12967 }, { "epoch": 0.6026022304832713, "grad_norm": 0.6204707668844731, "learning_rate": 3.6127495380907503e-06, "loss": 0.0423, "step": 12968 }, { "epoch": 0.6026486988847584, "grad_norm": 0.8489421693808018, "learning_rate": 3.61202658410872e-06, "loss": 0.0733, "step": 12969 }, { "epoch": 0.6026951672862454, "grad_norm": 0.41646053751365475, "learning_rate": 3.6113036615657486e-06, "loss": 0.0503, "step": 12970 }, { "epoch": 0.6027416356877323, "grad_norm": 1.856038566164714, "learning_rate": 3.6105807704782104e-06, "loss": 0.067, "step": 12971 }, { "epoch": 0.6027881040892193, "grad_norm": 0.5922584061551348, "learning_rate": 3.6098579108624815e-06, "loss": 0.053, "step": 12972 }, { "epoch": 0.6028345724907063, "grad_norm": 0.6121404157296841, "learning_rate": 3.6091350827349327e-06, "loss": 0.0647, "step": 12973 }, { "epoch": 0.6028810408921933, "grad_norm": 0.6900374316697895, "learning_rate": 3.6084122861119404e-06, "loss": 0.0602, "step": 12974 }, { "epoch": 0.6029275092936803, "grad_norm": 0.7426068152497072, "learning_rate": 3.607689521009873e-06, "loss": 0.0669, "step": 12975 }, { "epoch": 0.6029739776951673, "grad_norm": 0.4164543263925265, "learning_rate": 3.6069667874451025e-06, "loss": 0.0319, "step": 12976 }, { "epoch": 0.6030204460966543, "grad_norm": 0.40922728002512404, "learning_rate": 3.6062440854340015e-06, "loss": 0.0523, "step": 12977 }, { "epoch": 0.6030669144981413, "grad_norm": 0.8463389746512736, "learning_rate": 3.6055214149929373e-06, "loss": 0.0593, "step": 12978 }, { "epoch": 0.6031133828996282, "grad_norm": 0.5376400997795957, "learning_rate": 3.6047987761382815e-06, "loss": 0.0389, "step": 12979 }, { "epoch": 0.6031598513011153, "grad_norm": 0.9776330765513939, "learning_rate": 3.6040761688864046e-06, "loss": 0.0617, "step": 12980 }, { "epoch": 0.6032063197026022, "grad_norm": 0.5578204655047546, "learning_rate": 3.6033535932536684e-06, "loss": 0.0629, "step": 12981 }, { "epoch": 0.6032527881040892, "grad_norm": 0.9178184974605075, "learning_rate": 3.602631049256444e-06, "loss": 0.0388, "step": 12982 }, { "epoch": 0.6032992565055763, "grad_norm": 0.5375349680127426, "learning_rate": 3.6019085369110973e-06, "loss": 0.0567, "step": 12983 }, { "epoch": 0.6033457249070632, "grad_norm": 0.4539449655970916, "learning_rate": 3.601186056233993e-06, "loss": 0.0462, "step": 12984 }, { "epoch": 0.6033921933085502, "grad_norm": 0.69781275145571, "learning_rate": 3.600463607241497e-06, "loss": 0.0624, "step": 12985 }, { "epoch": 0.6034386617100371, "grad_norm": 1.0018987440914198, "learning_rate": 3.5997411899499752e-06, "loss": 0.0586, "step": 12986 }, { "epoch": 0.6034851301115242, "grad_norm": 0.8458895475685434, "learning_rate": 3.599018804375788e-06, "loss": 0.0622, "step": 12987 }, { "epoch": 0.6035315985130112, "grad_norm": 0.604244009687717, "learning_rate": 3.598296450535298e-06, "loss": 0.0656, "step": 12988 }, { "epoch": 0.6035780669144981, "grad_norm": 0.6623027348705601, "learning_rate": 3.5975741284448696e-06, "loss": 0.0489, "step": 12989 }, { "epoch": 0.6036245353159851, "grad_norm": 0.6140986673011577, "learning_rate": 3.596851838120863e-06, "loss": 0.0805, "step": 12990 }, { "epoch": 0.6036710037174721, "grad_norm": 0.8243798889470393, "learning_rate": 3.5961295795796407e-06, "loss": 0.0304, "step": 12991 }, { "epoch": 0.6037174721189591, "grad_norm": 0.46749408238537293, "learning_rate": 3.59540735283756e-06, "loss": 0.0511, "step": 12992 }, { "epoch": 0.6037639405204461, "grad_norm": 0.7881815508606986, "learning_rate": 3.59468515791098e-06, "loss": 0.0666, "step": 12993 }, { "epoch": 0.603810408921933, "grad_norm": 0.5296499622874339, "learning_rate": 3.593962994816261e-06, "loss": 0.0486, "step": 12994 }, { "epoch": 0.6038568773234201, "grad_norm": 0.42712405358834477, "learning_rate": 3.593240863569759e-06, "loss": 0.0365, "step": 12995 }, { "epoch": 0.6039033457249071, "grad_norm": 0.6750304790447317, "learning_rate": 3.5925187641878323e-06, "loss": 0.0514, "step": 12996 }, { "epoch": 0.603949814126394, "grad_norm": 0.5906617828419977, "learning_rate": 3.59179669668684e-06, "loss": 0.0641, "step": 12997 }, { "epoch": 0.6039962825278811, "grad_norm": 0.8058128594589509, "learning_rate": 3.5910746610831313e-06, "loss": 0.0612, "step": 12998 }, { "epoch": 0.604042750929368, "grad_norm": 0.8254721693779861, "learning_rate": 3.5903526573930657e-06, "loss": 0.0774, "step": 12999 }, { "epoch": 0.604089219330855, "grad_norm": 1.337945312421832, "learning_rate": 3.5896306856329955e-06, "loss": 0.068, "step": 13000 }, { "epoch": 0.604135687732342, "grad_norm": 0.7328682896666838, "learning_rate": 3.588908745819274e-06, "loss": 0.0609, "step": 13001 }, { "epoch": 0.604182156133829, "grad_norm": 1.177684007778427, "learning_rate": 3.5881868379682554e-06, "loss": 0.0426, "step": 13002 }, { "epoch": 0.604228624535316, "grad_norm": 1.3089779536161177, "learning_rate": 3.587464962096292e-06, "loss": 0.0308, "step": 13003 }, { "epoch": 0.6042750929368029, "grad_norm": 1.1185167828705262, "learning_rate": 3.5867431182197333e-06, "loss": 0.0602, "step": 13004 }, { "epoch": 0.60432156133829, "grad_norm": 0.7369318471944807, "learning_rate": 3.586021306354929e-06, "loss": 0.056, "step": 13005 }, { "epoch": 0.604368029739777, "grad_norm": 0.7496277622916235, "learning_rate": 3.58529952651823e-06, "loss": 0.0503, "step": 13006 }, { "epoch": 0.6044144981412639, "grad_norm": 0.45038679779446583, "learning_rate": 3.5845777787259873e-06, "loss": 0.0764, "step": 13007 }, { "epoch": 0.6044609665427509, "grad_norm": 0.6805189644973552, "learning_rate": 3.5838560629945465e-06, "loss": 0.0742, "step": 13008 }, { "epoch": 0.6045074349442379, "grad_norm": 0.4641603869215451, "learning_rate": 3.583134379340259e-06, "loss": 0.0364, "step": 13009 }, { "epoch": 0.6045539033457249, "grad_norm": 0.775224386776997, "learning_rate": 3.5824127277794663e-06, "loss": 0.0727, "step": 13010 }, { "epoch": 0.6046003717472119, "grad_norm": 0.5832735555887363, "learning_rate": 3.5816911083285165e-06, "loss": 0.0469, "step": 13011 }, { "epoch": 0.6046468401486988, "grad_norm": 0.33155858353421275, "learning_rate": 3.5809695210037574e-06, "loss": 0.0421, "step": 13012 }, { "epoch": 0.6046933085501859, "grad_norm": 1.3063772732280328, "learning_rate": 3.5802479658215316e-06, "loss": 0.066, "step": 13013 }, { "epoch": 0.6047397769516729, "grad_norm": 0.9492010127329514, "learning_rate": 3.5795264427981848e-06, "loss": 0.0601, "step": 13014 }, { "epoch": 0.6047862453531598, "grad_norm": 1.2654902823252119, "learning_rate": 3.578804951950059e-06, "loss": 0.0763, "step": 13015 }, { "epoch": 0.6048327137546469, "grad_norm": 0.36792137669071395, "learning_rate": 3.578083493293496e-06, "loss": 0.0332, "step": 13016 }, { "epoch": 0.6048791821561338, "grad_norm": 0.5600314412283944, "learning_rate": 3.5773620668448384e-06, "loss": 0.0551, "step": 13017 }, { "epoch": 0.6049256505576208, "grad_norm": 1.5669580222846153, "learning_rate": 3.5766406726204274e-06, "loss": 0.0733, "step": 13018 }, { "epoch": 0.6049721189591079, "grad_norm": 0.5958304880249882, "learning_rate": 3.575919310636603e-06, "loss": 0.0769, "step": 13019 }, { "epoch": 0.6050185873605948, "grad_norm": 0.731348115230871, "learning_rate": 3.575197980909705e-06, "loss": 0.0548, "step": 13020 }, { "epoch": 0.6050650557620818, "grad_norm": 0.7644332097104677, "learning_rate": 3.5744766834560748e-06, "loss": 0.051, "step": 13021 }, { "epoch": 0.6051115241635687, "grad_norm": 0.4235992022636814, "learning_rate": 3.5737554182920454e-06, "loss": 0.0419, "step": 13022 }, { "epoch": 0.6051579925650558, "grad_norm": 0.8912929258828861, "learning_rate": 3.5730341854339563e-06, "loss": 0.0402, "step": 13023 }, { "epoch": 0.6052044609665428, "grad_norm": 0.8230180566037878, "learning_rate": 3.5723129848981456e-06, "loss": 0.0696, "step": 13024 }, { "epoch": 0.6052509293680297, "grad_norm": 0.7644566080794951, "learning_rate": 3.571591816700948e-06, "loss": 0.0654, "step": 13025 }, { "epoch": 0.6052973977695167, "grad_norm": 0.5976050106854116, "learning_rate": 3.5708706808587003e-06, "loss": 0.0401, "step": 13026 }, { "epoch": 0.6053438661710037, "grad_norm": 0.7209754986518748, "learning_rate": 3.5701495773877366e-06, "loss": 0.0427, "step": 13027 }, { "epoch": 0.6053903345724907, "grad_norm": 0.44839137218327213, "learning_rate": 3.5694285063043876e-06, "loss": 0.0447, "step": 13028 }, { "epoch": 0.6054368029739777, "grad_norm": 0.7541666976598085, "learning_rate": 3.56870746762499e-06, "loss": 0.0454, "step": 13029 }, { "epoch": 0.6054832713754646, "grad_norm": 0.5971671424457438, "learning_rate": 3.5679864613658732e-06, "loss": 0.0459, "step": 13030 }, { "epoch": 0.6055297397769517, "grad_norm": 0.4468078179132364, "learning_rate": 3.567265487543371e-06, "loss": 0.0533, "step": 13031 }, { "epoch": 0.6055762081784387, "grad_norm": 0.46219557882902484, "learning_rate": 3.5665445461738134e-06, "loss": 0.0274, "step": 13032 }, { "epoch": 0.6056226765799256, "grad_norm": 0.4672707738418958, "learning_rate": 3.5658236372735322e-06, "loss": 0.0413, "step": 13033 }, { "epoch": 0.6056691449814127, "grad_norm": 0.6160830680268031, "learning_rate": 3.5651027608588535e-06, "loss": 0.0608, "step": 13034 }, { "epoch": 0.6057156133828996, "grad_norm": 0.804832036001274, "learning_rate": 3.5643819169461068e-06, "loss": 0.0592, "step": 13035 }, { "epoch": 0.6057620817843866, "grad_norm": 0.9278275721372652, "learning_rate": 3.5636611055516214e-06, "loss": 0.0693, "step": 13036 }, { "epoch": 0.6058085501858737, "grad_norm": 0.7573446773449112, "learning_rate": 3.562940326691723e-06, "loss": 0.0624, "step": 13037 }, { "epoch": 0.6058550185873606, "grad_norm": 0.5240857409101963, "learning_rate": 3.5622195803827397e-06, "loss": 0.0538, "step": 13038 }, { "epoch": 0.6059014869888476, "grad_norm": 1.0215292855400888, "learning_rate": 3.5614988666409973e-06, "loss": 0.0648, "step": 13039 }, { "epoch": 0.6059479553903345, "grad_norm": 0.6125532275209609, "learning_rate": 3.560778185482817e-06, "loss": 0.0456, "step": 13040 }, { "epoch": 0.6059944237918216, "grad_norm": 0.7875428462402573, "learning_rate": 3.5600575369245273e-06, "loss": 0.0507, "step": 13041 }, { "epoch": 0.6060408921933086, "grad_norm": 1.2258072203572286, "learning_rate": 3.559336920982448e-06, "loss": 0.0781, "step": 13042 }, { "epoch": 0.6060873605947955, "grad_norm": 0.4567851494399339, "learning_rate": 3.5586163376729054e-06, "loss": 0.0464, "step": 13043 }, { "epoch": 0.6061338289962825, "grad_norm": 0.48120503418901284, "learning_rate": 3.557895787012219e-06, "loss": 0.0432, "step": 13044 }, { "epoch": 0.6061802973977695, "grad_norm": 0.6052464333969079, "learning_rate": 3.5571752690167115e-06, "loss": 0.0418, "step": 13045 }, { "epoch": 0.6062267657992565, "grad_norm": 0.43024148417404257, "learning_rate": 3.556454783702703e-06, "loss": 0.0556, "step": 13046 }, { "epoch": 0.6062732342007435, "grad_norm": 1.0040636095401891, "learning_rate": 3.5557343310865115e-06, "loss": 0.0537, "step": 13047 }, { "epoch": 0.6063197026022304, "grad_norm": 0.6154337722821241, "learning_rate": 3.5550139111844584e-06, "loss": 0.0725, "step": 13048 }, { "epoch": 0.6063661710037175, "grad_norm": 1.0585783963491129, "learning_rate": 3.5542935240128603e-06, "loss": 0.0799, "step": 13049 }, { "epoch": 0.6064126394052045, "grad_norm": 0.5494806821505728, "learning_rate": 3.553573169588037e-06, "loss": 0.038, "step": 13050 }, { "epoch": 0.6064591078066914, "grad_norm": 0.7066648919932429, "learning_rate": 3.5528528479263034e-06, "loss": 0.0592, "step": 13051 }, { "epoch": 0.6065055762081785, "grad_norm": 0.6573711666806895, "learning_rate": 3.552132559043975e-06, "loss": 0.0515, "step": 13052 }, { "epoch": 0.6065520446096654, "grad_norm": 0.6233115987589449, "learning_rate": 3.551412302957368e-06, "loss": 0.0524, "step": 13053 }, { "epoch": 0.6065985130111524, "grad_norm": 0.8011616362254982, "learning_rate": 3.550692079682796e-06, "loss": 0.062, "step": 13054 }, { "epoch": 0.6066449814126395, "grad_norm": 0.5843026562533903, "learning_rate": 3.5499718892365753e-06, "loss": 0.0682, "step": 13055 }, { "epoch": 0.6066914498141264, "grad_norm": 0.6738587312418692, "learning_rate": 3.5492517316350185e-06, "loss": 0.0433, "step": 13056 }, { "epoch": 0.6067379182156134, "grad_norm": 0.6070198955390732, "learning_rate": 3.5485316068944347e-06, "loss": 0.0486, "step": 13057 }, { "epoch": 0.6067843866171003, "grad_norm": 0.6759707732180799, "learning_rate": 3.5478115150311387e-06, "loss": 0.056, "step": 13058 }, { "epoch": 0.6068308550185874, "grad_norm": 0.5855508806165621, "learning_rate": 3.5470914560614393e-06, "loss": 0.0578, "step": 13059 }, { "epoch": 0.6068773234200744, "grad_norm": 0.9817468680185171, "learning_rate": 3.5463714300016483e-06, "loss": 0.0488, "step": 13060 }, { "epoch": 0.6069237918215613, "grad_norm": 0.5499539070607827, "learning_rate": 3.545651436868074e-06, "loss": 0.0647, "step": 13061 }, { "epoch": 0.6069702602230483, "grad_norm": 0.41061650134170025, "learning_rate": 3.544931476677027e-06, "loss": 0.0498, "step": 13062 }, { "epoch": 0.6070167286245353, "grad_norm": 0.6183804460559084, "learning_rate": 3.5442115494448127e-06, "loss": 0.0536, "step": 13063 }, { "epoch": 0.6070631970260223, "grad_norm": 0.43595150520345394, "learning_rate": 3.5434916551877376e-06, "loss": 0.0448, "step": 13064 }, { "epoch": 0.6071096654275093, "grad_norm": 0.4763672889140732, "learning_rate": 3.54277179392211e-06, "loss": 0.0422, "step": 13065 }, { "epoch": 0.6071561338289962, "grad_norm": 0.5308029752129623, "learning_rate": 3.542051965664235e-06, "loss": 0.051, "step": 13066 }, { "epoch": 0.6072026022304833, "grad_norm": 0.5684475690417844, "learning_rate": 3.541332170430417e-06, "loss": 0.0603, "step": 13067 }, { "epoch": 0.6072490706319703, "grad_norm": 0.6107747465124781, "learning_rate": 3.540612408236963e-06, "loss": 0.0495, "step": 13068 }, { "epoch": 0.6072955390334572, "grad_norm": 0.6301375972685019, "learning_rate": 3.5398926791001714e-06, "loss": 0.0465, "step": 13069 }, { "epoch": 0.6073420074349443, "grad_norm": 0.4865129744565994, "learning_rate": 3.5391729830363486e-06, "loss": 0.0642, "step": 13070 }, { "epoch": 0.6073884758364312, "grad_norm": 0.557348855370302, "learning_rate": 3.5384533200617938e-06, "loss": 0.0489, "step": 13071 }, { "epoch": 0.6074349442379182, "grad_norm": 0.3140999808840624, "learning_rate": 3.5377336901928106e-06, "loss": 0.0187, "step": 13072 }, { "epoch": 0.6074814126394052, "grad_norm": 1.0549243767682743, "learning_rate": 3.5370140934456977e-06, "loss": 0.0621, "step": 13073 }, { "epoch": 0.6075278810408922, "grad_norm": 0.695965850609876, "learning_rate": 3.5362945298367566e-06, "loss": 0.0463, "step": 13074 }, { "epoch": 0.6075743494423792, "grad_norm": 0.5753903929331359, "learning_rate": 3.535574999382285e-06, "loss": 0.0621, "step": 13075 }, { "epoch": 0.6076208178438661, "grad_norm": 0.3901647263112285, "learning_rate": 3.5348555020985794e-06, "loss": 0.0462, "step": 13076 }, { "epoch": 0.6076672862453532, "grad_norm": 0.34039190858145413, "learning_rate": 3.5341360380019395e-06, "loss": 0.036, "step": 13077 }, { "epoch": 0.6077137546468402, "grad_norm": 0.4048751987195985, "learning_rate": 3.5334166071086605e-06, "loss": 0.0689, "step": 13078 }, { "epoch": 0.6077602230483271, "grad_norm": 0.9001912983340007, "learning_rate": 3.53269720943504e-06, "loss": 0.0671, "step": 13079 }, { "epoch": 0.6078066914498141, "grad_norm": 0.45937228837018124, "learning_rate": 3.531977844997373e-06, "loss": 0.0505, "step": 13080 }, { "epoch": 0.6078531598513011, "grad_norm": 0.6061246754476536, "learning_rate": 3.5312585138119503e-06, "loss": 0.0387, "step": 13081 }, { "epoch": 0.6078996282527881, "grad_norm": 0.5367975301833419, "learning_rate": 3.530539215895069e-06, "loss": 0.0421, "step": 13082 }, { "epoch": 0.6079460966542751, "grad_norm": 2.1202351101123758, "learning_rate": 3.5298199512630204e-06, "loss": 0.0553, "step": 13083 }, { "epoch": 0.607992565055762, "grad_norm": 0.6651855399550767, "learning_rate": 3.5291007199320987e-06, "loss": 0.059, "step": 13084 }, { "epoch": 0.6080390334572491, "grad_norm": 0.7954107368458166, "learning_rate": 3.5283815219185926e-06, "loss": 0.0488, "step": 13085 }, { "epoch": 0.6080855018587361, "grad_norm": 0.5039438610740957, "learning_rate": 3.5276623572387954e-06, "loss": 0.0406, "step": 13086 }, { "epoch": 0.608131970260223, "grad_norm": 0.7443299912256905, "learning_rate": 3.5269432259089947e-06, "loss": 0.0566, "step": 13087 }, { "epoch": 0.6081784386617101, "grad_norm": 0.5261029036357717, "learning_rate": 3.526224127945479e-06, "loss": 0.0518, "step": 13088 }, { "epoch": 0.608224907063197, "grad_norm": 1.1132800293436365, "learning_rate": 3.5255050633645395e-06, "loss": 0.0594, "step": 13089 }, { "epoch": 0.608271375464684, "grad_norm": 0.8623742316137428, "learning_rate": 3.5247860321824605e-06, "loss": 0.0744, "step": 13090 }, { "epoch": 0.608317843866171, "grad_norm": 0.6079087418671778, "learning_rate": 3.5240670344155327e-06, "loss": 0.0676, "step": 13091 }, { "epoch": 0.608364312267658, "grad_norm": 2.336090323806217, "learning_rate": 3.5233480700800394e-06, "loss": 0.0698, "step": 13092 }, { "epoch": 0.608410780669145, "grad_norm": 0.887852704001239, "learning_rate": 3.522629139192265e-06, "loss": 0.0646, "step": 13093 }, { "epoch": 0.6084572490706319, "grad_norm": 0.6927989043285793, "learning_rate": 3.5219102417684968e-06, "loss": 0.0547, "step": 13094 }, { "epoch": 0.608503717472119, "grad_norm": 0.6778348297818502, "learning_rate": 3.5211913778250162e-06, "loss": 0.0585, "step": 13095 }, { "epoch": 0.608550185873606, "grad_norm": 0.5793486783124507, "learning_rate": 3.5204725473781088e-06, "loss": 0.0663, "step": 13096 }, { "epoch": 0.6085966542750929, "grad_norm": 1.1922526846562294, "learning_rate": 3.519753750444054e-06, "loss": 0.0541, "step": 13097 }, { "epoch": 0.6086431226765799, "grad_norm": 0.6438092166713985, "learning_rate": 3.519034987039136e-06, "loss": 0.0598, "step": 13098 }, { "epoch": 0.6086895910780669, "grad_norm": 0.7288283797116165, "learning_rate": 3.5183162571796343e-06, "loss": 0.0617, "step": 13099 }, { "epoch": 0.6087360594795539, "grad_norm": 2.5540525033640735, "learning_rate": 3.517597560881827e-06, "loss": 0.0775, "step": 13100 }, { "epoch": 0.6087825278810409, "grad_norm": 0.7601812059282999, "learning_rate": 3.516878898161997e-06, "loss": 0.0704, "step": 13101 }, { "epoch": 0.6088289962825278, "grad_norm": 0.7676977500348544, "learning_rate": 3.5161602690364195e-06, "loss": 0.0692, "step": 13102 }, { "epoch": 0.6088754646840149, "grad_norm": 0.4264161777723947, "learning_rate": 3.5154416735213743e-06, "loss": 0.0403, "step": 13103 }, { "epoch": 0.6089219330855019, "grad_norm": 0.5984244947698563, "learning_rate": 3.5147231116331386e-06, "loss": 0.0657, "step": 13104 }, { "epoch": 0.6089684014869888, "grad_norm": 0.5475536470007338, "learning_rate": 3.514004583387986e-06, "loss": 0.0561, "step": 13105 }, { "epoch": 0.6090148698884759, "grad_norm": 0.9029144746041844, "learning_rate": 3.5132860888021947e-06, "loss": 0.0543, "step": 13106 }, { "epoch": 0.6090613382899628, "grad_norm": 0.773879514982484, "learning_rate": 3.5125676278920367e-06, "loss": 0.0683, "step": 13107 }, { "epoch": 0.6091078066914498, "grad_norm": 4.968683398698197, "learning_rate": 3.511849200673789e-06, "loss": 0.0715, "step": 13108 }, { "epoch": 0.6091542750929368, "grad_norm": 0.4450364429101788, "learning_rate": 3.511130807163724e-06, "loss": 0.0432, "step": 13109 }, { "epoch": 0.6092007434944238, "grad_norm": 0.8846348597197999, "learning_rate": 3.510412447378111e-06, "loss": 0.0461, "step": 13110 }, { "epoch": 0.6092472118959108, "grad_norm": 0.5778622698920456, "learning_rate": 3.5096941213332247e-06, "loss": 0.0612, "step": 13111 }, { "epoch": 0.6092936802973977, "grad_norm": 0.34651659589231276, "learning_rate": 3.5089758290453346e-06, "loss": 0.0473, "step": 13112 }, { "epoch": 0.6093401486988848, "grad_norm": 1.8002077827680516, "learning_rate": 3.508257570530712e-06, "loss": 0.0516, "step": 13113 }, { "epoch": 0.6093866171003718, "grad_norm": 0.5151216454760678, "learning_rate": 3.507539345805624e-06, "loss": 0.0458, "step": 13114 }, { "epoch": 0.6094330855018587, "grad_norm": 0.6532962344970689, "learning_rate": 3.506821154886343e-06, "loss": 0.05, "step": 13115 }, { "epoch": 0.6094795539033457, "grad_norm": 2.6081177804387528, "learning_rate": 3.5061029977891335e-06, "loss": 0.0853, "step": 13116 }, { "epoch": 0.6095260223048327, "grad_norm": 1.259023547686595, "learning_rate": 3.505384874530262e-06, "loss": 0.0458, "step": 13117 }, { "epoch": 0.6095724907063197, "grad_norm": 0.4748978811925034, "learning_rate": 3.504666785125998e-06, "loss": 0.0301, "step": 13118 }, { "epoch": 0.6096189591078067, "grad_norm": 0.6133960806502202, "learning_rate": 3.5039487295926034e-06, "loss": 0.0524, "step": 13119 }, { "epoch": 0.6096654275092936, "grad_norm": 0.7376126418642281, "learning_rate": 3.5032307079463457e-06, "loss": 0.0383, "step": 13120 }, { "epoch": 0.6097118959107807, "grad_norm": 0.9132084004466564, "learning_rate": 3.5025127202034894e-06, "loss": 0.0702, "step": 13121 }, { "epoch": 0.6097583643122677, "grad_norm": 0.49441482585508634, "learning_rate": 3.501794766380293e-06, "loss": 0.0498, "step": 13122 }, { "epoch": 0.6098048327137546, "grad_norm": 0.5939717815681413, "learning_rate": 3.501076846493024e-06, "loss": 0.0371, "step": 13123 }, { "epoch": 0.6098513011152417, "grad_norm": 0.917732414991287, "learning_rate": 3.5003589605579404e-06, "loss": 0.0863, "step": 13124 }, { "epoch": 0.6098977695167286, "grad_norm": 1.4596366853814449, "learning_rate": 3.4996411085913056e-06, "loss": 0.0604, "step": 13125 }, { "epoch": 0.6099442379182156, "grad_norm": 0.7053727477397964, "learning_rate": 3.4989232906093784e-06, "loss": 0.0702, "step": 13126 }, { "epoch": 0.6099907063197026, "grad_norm": 0.4480658659554369, "learning_rate": 3.4982055066284198e-06, "loss": 0.0388, "step": 13127 }, { "epoch": 0.6100371747211896, "grad_norm": 0.678042556685843, "learning_rate": 3.497487756664687e-06, "loss": 0.0546, "step": 13128 }, { "epoch": 0.6100836431226766, "grad_norm": 0.5217716888687288, "learning_rate": 3.4967700407344363e-06, "loss": 0.0632, "step": 13129 }, { "epoch": 0.6101301115241635, "grad_norm": 0.7713431663333998, "learning_rate": 3.496052358853927e-06, "loss": 0.0708, "step": 13130 }, { "epoch": 0.6101765799256506, "grad_norm": 0.5924697399130575, "learning_rate": 3.495334711039414e-06, "loss": 0.052, "step": 13131 }, { "epoch": 0.6102230483271376, "grad_norm": 0.47953677823600294, "learning_rate": 3.494617097307154e-06, "loss": 0.06, "step": 13132 }, { "epoch": 0.6102695167286245, "grad_norm": 0.5485006446884326, "learning_rate": 3.4938995176734026e-06, "loss": 0.0366, "step": 13133 }, { "epoch": 0.6103159851301115, "grad_norm": 0.4625753554289432, "learning_rate": 3.4931819721544104e-06, "loss": 0.0651, "step": 13134 }, { "epoch": 0.6103624535315985, "grad_norm": 0.5126077144192979, "learning_rate": 3.4924644607664334e-06, "loss": 0.071, "step": 13135 }, { "epoch": 0.6104089219330855, "grad_norm": 0.7003622527767618, "learning_rate": 3.491746983525722e-06, "loss": 0.061, "step": 13136 }, { "epoch": 0.6104553903345725, "grad_norm": 2.04647734130643, "learning_rate": 3.4910295404485295e-06, "loss": 0.0617, "step": 13137 }, { "epoch": 0.6105018587360594, "grad_norm": 0.6138219195815009, "learning_rate": 3.4903121315511045e-06, "loss": 0.0461, "step": 13138 }, { "epoch": 0.6105483271375465, "grad_norm": 0.7462646609789939, "learning_rate": 3.489594756849702e-06, "loss": 0.0735, "step": 13139 }, { "epoch": 0.6105947955390335, "grad_norm": 1.0654348671721197, "learning_rate": 3.4888774163605667e-06, "loss": 0.0663, "step": 13140 }, { "epoch": 0.6106412639405204, "grad_norm": 0.45230360259023594, "learning_rate": 3.4881601100999463e-06, "loss": 0.0393, "step": 13141 }, { "epoch": 0.6106877323420075, "grad_norm": 0.43962260621293675, "learning_rate": 3.4874428380840926e-06, "loss": 0.0453, "step": 13142 }, { "epoch": 0.6107342007434944, "grad_norm": 0.2561833488701494, "learning_rate": 3.4867256003292495e-06, "loss": 0.0356, "step": 13143 }, { "epoch": 0.6107806691449814, "grad_norm": 0.49643012074922516, "learning_rate": 3.4860083968516647e-06, "loss": 0.0309, "step": 13144 }, { "epoch": 0.6108271375464684, "grad_norm": 0.7944301799695775, "learning_rate": 3.4852912276675843e-06, "loss": 0.08, "step": 13145 }, { "epoch": 0.6108736059479554, "grad_norm": 0.5595246884799766, "learning_rate": 3.48457409279325e-06, "loss": 0.0401, "step": 13146 }, { "epoch": 0.6109200743494424, "grad_norm": 1.1195600417408116, "learning_rate": 3.483856992244908e-06, "loss": 0.0465, "step": 13147 }, { "epoch": 0.6109665427509293, "grad_norm": 0.5920516749969162, "learning_rate": 3.4831399260388e-06, "loss": 0.0577, "step": 13148 }, { "epoch": 0.6110130111524164, "grad_norm": 1.4431930040695635, "learning_rate": 3.482422894191171e-06, "loss": 0.0527, "step": 13149 }, { "epoch": 0.6110594795539034, "grad_norm": 0.508833594461998, "learning_rate": 3.4817058967182584e-06, "loss": 0.0419, "step": 13150 }, { "epoch": 0.6111059479553903, "grad_norm": 0.9168098257949457, "learning_rate": 3.480988933636308e-06, "loss": 0.0684, "step": 13151 }, { "epoch": 0.6111524163568773, "grad_norm": 1.2367216836722625, "learning_rate": 3.4802720049615557e-06, "loss": 0.0602, "step": 13152 }, { "epoch": 0.6111988847583643, "grad_norm": 0.36355657076021336, "learning_rate": 3.479555110710241e-06, "loss": 0.0371, "step": 13153 }, { "epoch": 0.6112453531598513, "grad_norm": 1.4996994851702503, "learning_rate": 3.4788382508986046e-06, "loss": 0.0695, "step": 13154 }, { "epoch": 0.6112918215613383, "grad_norm": 0.7743760395619873, "learning_rate": 3.4781214255428815e-06, "loss": 0.0581, "step": 13155 }, { "epoch": 0.6113382899628252, "grad_norm": 0.5718765714108769, "learning_rate": 3.4774046346593106e-06, "loss": 0.0386, "step": 13156 }, { "epoch": 0.6113847583643123, "grad_norm": 0.47943302282548794, "learning_rate": 3.476687878264129e-06, "loss": 0.0334, "step": 13157 }, { "epoch": 0.6114312267657993, "grad_norm": 0.8856754687052685, "learning_rate": 3.4759711563735676e-06, "loss": 0.0586, "step": 13158 }, { "epoch": 0.6114776951672862, "grad_norm": 0.917721914272091, "learning_rate": 3.475254469003865e-06, "loss": 0.0619, "step": 13159 }, { "epoch": 0.6115241635687733, "grad_norm": 0.8303271047111679, "learning_rate": 3.4745378161712517e-06, "loss": 0.0615, "step": 13160 }, { "epoch": 0.6115706319702602, "grad_norm": 0.6468830556511208, "learning_rate": 3.473821197891964e-06, "loss": 0.0546, "step": 13161 }, { "epoch": 0.6116171003717472, "grad_norm": 1.7714060264691247, "learning_rate": 3.473104614182231e-06, "loss": 0.0724, "step": 13162 }, { "epoch": 0.6116635687732342, "grad_norm": 0.5661958653750986, "learning_rate": 3.4723880650582875e-06, "loss": 0.0501, "step": 13163 }, { "epoch": 0.6117100371747212, "grad_norm": 0.4899355535643153, "learning_rate": 3.4716715505363617e-06, "loss": 0.0305, "step": 13164 }, { "epoch": 0.6117565055762082, "grad_norm": 0.7977203022149514, "learning_rate": 3.470955070632682e-06, "loss": 0.0779, "step": 13165 }, { "epoch": 0.6118029739776951, "grad_norm": 0.659123053798879, "learning_rate": 3.4702386253634814e-06, "loss": 0.0558, "step": 13166 }, { "epoch": 0.6118494423791822, "grad_norm": 0.7732772279220477, "learning_rate": 3.469522214744984e-06, "loss": 0.0683, "step": 13167 }, { "epoch": 0.6118959107806692, "grad_norm": 0.8078820580751014, "learning_rate": 3.468805838793422e-06, "loss": 0.0547, "step": 13168 }, { "epoch": 0.6119423791821561, "grad_norm": 1.0055318806349145, "learning_rate": 3.468089497525018e-06, "loss": 0.0669, "step": 13169 }, { "epoch": 0.6119888475836431, "grad_norm": 0.771993830734245, "learning_rate": 3.467373190955998e-06, "loss": 0.047, "step": 13170 }, { "epoch": 0.6120353159851301, "grad_norm": 1.0061981885523126, "learning_rate": 3.466656919102589e-06, "loss": 0.0717, "step": 13171 }, { "epoch": 0.6120817843866171, "grad_norm": 0.4546425086508749, "learning_rate": 3.465940681981014e-06, "loss": 0.0428, "step": 13172 }, { "epoch": 0.6121282527881041, "grad_norm": 0.4884914234749878, "learning_rate": 3.465224479607498e-06, "loss": 0.0426, "step": 13173 }, { "epoch": 0.612174721189591, "grad_norm": 0.5423816605687282, "learning_rate": 3.4645083119982637e-06, "loss": 0.0515, "step": 13174 }, { "epoch": 0.6122211895910781, "grad_norm": 0.7477395524756254, "learning_rate": 3.4637921791695306e-06, "loss": 0.0416, "step": 13175 }, { "epoch": 0.6122676579925651, "grad_norm": 0.6821061321530495, "learning_rate": 3.463076081137522e-06, "loss": 0.0525, "step": 13176 }, { "epoch": 0.612314126394052, "grad_norm": 0.909298565847746, "learning_rate": 3.462360017918457e-06, "loss": 0.0462, "step": 13177 }, { "epoch": 0.6123605947955391, "grad_norm": 0.7032325168182622, "learning_rate": 3.461643989528557e-06, "loss": 0.035, "step": 13178 }, { "epoch": 0.612407063197026, "grad_norm": 0.4227058975706822, "learning_rate": 3.460927995984039e-06, "loss": 0.0708, "step": 13179 }, { "epoch": 0.612453531598513, "grad_norm": 1.3626475127635964, "learning_rate": 3.4602120373011235e-06, "loss": 0.0665, "step": 13180 }, { "epoch": 0.6125, "grad_norm": 0.6083887694171779, "learning_rate": 3.459496113496025e-06, "loss": 0.056, "step": 13181 }, { "epoch": 0.612546468401487, "grad_norm": 1.1715047303130302, "learning_rate": 3.4587802245849593e-06, "loss": 0.0801, "step": 13182 }, { "epoch": 0.612592936802974, "grad_norm": 1.3486085320016137, "learning_rate": 3.4580643705841454e-06, "loss": 0.0692, "step": 13183 }, { "epoch": 0.6126394052044609, "grad_norm": 0.5731211007060675, "learning_rate": 3.4573485515097948e-06, "loss": 0.0505, "step": 13184 }, { "epoch": 0.612685873605948, "grad_norm": 0.5972532818416438, "learning_rate": 3.456632767378124e-06, "loss": 0.0471, "step": 13185 }, { "epoch": 0.612732342007435, "grad_norm": 0.7199214999019513, "learning_rate": 3.455917018205347e-06, "loss": 0.0442, "step": 13186 }, { "epoch": 0.6127788104089219, "grad_norm": 0.5427577079208211, "learning_rate": 3.4552013040076725e-06, "loss": 0.0478, "step": 13187 }, { "epoch": 0.6128252788104089, "grad_norm": 0.9397153424799964, "learning_rate": 3.454485624801315e-06, "loss": 0.0798, "step": 13188 }, { "epoch": 0.6128717472118959, "grad_norm": 0.861624125918927, "learning_rate": 3.4537699806024838e-06, "loss": 0.053, "step": 13189 }, { "epoch": 0.6129182156133829, "grad_norm": 0.6043186329761002, "learning_rate": 3.453054371427391e-06, "loss": 0.0397, "step": 13190 }, { "epoch": 0.6129646840148699, "grad_norm": 0.45685680043727506, "learning_rate": 3.452338797292244e-06, "loss": 0.0485, "step": 13191 }, { "epoch": 0.6130111524163568, "grad_norm": 0.4683813116356584, "learning_rate": 3.451623258213254e-06, "loss": 0.0632, "step": 13192 }, { "epoch": 0.6130576208178439, "grad_norm": 0.9109821001632374, "learning_rate": 3.450907754206626e-06, "loss": 0.0687, "step": 13193 }, { "epoch": 0.6131040892193309, "grad_norm": 0.38086244703694855, "learning_rate": 3.4501922852885666e-06, "loss": 0.0404, "step": 13194 }, { "epoch": 0.6131505576208178, "grad_norm": 0.747983131816522, "learning_rate": 3.449476851475284e-06, "loss": 0.0597, "step": 13195 }, { "epoch": 0.6131970260223049, "grad_norm": 0.5385022465151666, "learning_rate": 3.4487614527829814e-06, "loss": 0.0717, "step": 13196 }, { "epoch": 0.6132434944237918, "grad_norm": 0.49567293493222425, "learning_rate": 3.4480460892278668e-06, "loss": 0.0557, "step": 13197 }, { "epoch": 0.6132899628252788, "grad_norm": 0.5395193806386102, "learning_rate": 3.4473307608261415e-06, "loss": 0.066, "step": 13198 }, { "epoch": 0.6133364312267658, "grad_norm": 0.5175766718842074, "learning_rate": 3.4466154675940065e-06, "loss": 0.0478, "step": 13199 }, { "epoch": 0.6133828996282528, "grad_norm": 0.5569220896706651, "learning_rate": 3.4459002095476675e-06, "loss": 0.0498, "step": 13200 }, { "epoch": 0.6134293680297398, "grad_norm": 0.41130181043383696, "learning_rate": 3.4451849867033232e-06, "loss": 0.042, "step": 13201 }, { "epoch": 0.6134758364312267, "grad_norm": 0.5684073045400009, "learning_rate": 3.444469799077177e-06, "loss": 0.0651, "step": 13202 }, { "epoch": 0.6135223048327137, "grad_norm": 1.3661164034620008, "learning_rate": 3.443754646685425e-06, "loss": 0.0562, "step": 13203 }, { "epoch": 0.6135687732342008, "grad_norm": 0.44164581699561967, "learning_rate": 3.443039529544271e-06, "loss": 0.0487, "step": 13204 }, { "epoch": 0.6136152416356877, "grad_norm": 0.47200615030932463, "learning_rate": 3.442324447669909e-06, "loss": 0.0776, "step": 13205 }, { "epoch": 0.6136617100371747, "grad_norm": 0.3855613435203581, "learning_rate": 3.441609401078536e-06, "loss": 0.0442, "step": 13206 }, { "epoch": 0.6137081784386617, "grad_norm": 0.5707191276167828, "learning_rate": 3.440894389786352e-06, "loss": 0.048, "step": 13207 }, { "epoch": 0.6137546468401487, "grad_norm": 0.6956126335107724, "learning_rate": 3.44017941380955e-06, "loss": 0.0634, "step": 13208 }, { "epoch": 0.6138011152416357, "grad_norm": 0.6464903409768482, "learning_rate": 3.4394644731643266e-06, "loss": 0.0787, "step": 13209 }, { "epoch": 0.6138475836431226, "grad_norm": 0.4432124251955115, "learning_rate": 3.438749567866877e-06, "loss": 0.0553, "step": 13210 }, { "epoch": 0.6138940520446097, "grad_norm": 0.4975187152025644, "learning_rate": 3.4380346979333904e-06, "loss": 0.0662, "step": 13211 }, { "epoch": 0.6139405204460967, "grad_norm": 0.5881064236271969, "learning_rate": 3.437319863380063e-06, "loss": 0.0465, "step": 13212 }, { "epoch": 0.6139869888475836, "grad_norm": 0.6784511210402153, "learning_rate": 3.4366050642230843e-06, "loss": 0.0633, "step": 13213 }, { "epoch": 0.6140334572490707, "grad_norm": 1.0329273579537301, "learning_rate": 3.4358903004786477e-06, "loss": 0.0528, "step": 13214 }, { "epoch": 0.6140799256505576, "grad_norm": 0.686376677361612, "learning_rate": 3.4351755721629404e-06, "loss": 0.054, "step": 13215 }, { "epoch": 0.6141263940520446, "grad_norm": 0.6609778777441255, "learning_rate": 3.434460879292155e-06, "loss": 0.0604, "step": 13216 }, { "epoch": 0.6141728624535316, "grad_norm": 0.7080562834218275, "learning_rate": 3.4337462218824784e-06, "loss": 0.0738, "step": 13217 }, { "epoch": 0.6142193308550186, "grad_norm": 0.8029703469120864, "learning_rate": 3.433031599950097e-06, "loss": 0.0491, "step": 13218 }, { "epoch": 0.6142657992565056, "grad_norm": 0.3814253832217513, "learning_rate": 3.4323170135112e-06, "loss": 0.0305, "step": 13219 }, { "epoch": 0.6143122676579925, "grad_norm": 0.793947784909028, "learning_rate": 3.431602462581971e-06, "loss": 0.0652, "step": 13220 }, { "epoch": 0.6143587360594795, "grad_norm": 0.43754368966181056, "learning_rate": 3.4308879471785986e-06, "loss": 0.0304, "step": 13221 }, { "epoch": 0.6144052044609666, "grad_norm": 0.40097212604303023, "learning_rate": 3.430173467317267e-06, "loss": 0.0518, "step": 13222 }, { "epoch": 0.6144516728624535, "grad_norm": 0.48025871742128123, "learning_rate": 3.4294590230141557e-06, "loss": 0.048, "step": 13223 }, { "epoch": 0.6144981412639405, "grad_norm": 0.5955830539835404, "learning_rate": 3.428744614285452e-06, "loss": 0.0568, "step": 13224 }, { "epoch": 0.6145446096654275, "grad_norm": 0.46398057521898484, "learning_rate": 3.4280302411473354e-06, "loss": 0.0534, "step": 13225 }, { "epoch": 0.6145910780669145, "grad_norm": 0.5926278370131823, "learning_rate": 3.42731590361599e-06, "loss": 0.051, "step": 13226 }, { "epoch": 0.6146375464684015, "grad_norm": 0.5799430469056315, "learning_rate": 3.426601601707594e-06, "loss": 0.0643, "step": 13227 }, { "epoch": 0.6146840148698884, "grad_norm": 0.2676931324393261, "learning_rate": 3.4258873354383273e-06, "loss": 0.0461, "step": 13228 }, { "epoch": 0.6147304832713755, "grad_norm": 0.6815530532424617, "learning_rate": 3.425173104824369e-06, "loss": 0.0362, "step": 13229 }, { "epoch": 0.6147769516728625, "grad_norm": 0.44483627665073816, "learning_rate": 3.424458909881897e-06, "loss": 0.0373, "step": 13230 }, { "epoch": 0.6148234200743494, "grad_norm": 0.7441847211931957, "learning_rate": 3.4237447506270893e-06, "loss": 0.0457, "step": 13231 }, { "epoch": 0.6148698884758365, "grad_norm": 0.6904626836095482, "learning_rate": 3.4230306270761215e-06, "loss": 0.0545, "step": 13232 }, { "epoch": 0.6149163568773234, "grad_norm": 0.41107691218126935, "learning_rate": 3.4223165392451718e-06, "loss": 0.0346, "step": 13233 }, { "epoch": 0.6149628252788104, "grad_norm": 0.7437499691381377, "learning_rate": 3.4216024871504117e-06, "loss": 0.0682, "step": 13234 }, { "epoch": 0.6150092936802974, "grad_norm": 0.5535481924041603, "learning_rate": 3.4208884708080157e-06, "loss": 0.0527, "step": 13235 }, { "epoch": 0.6150557620817844, "grad_norm": 0.3765241634071276, "learning_rate": 3.4201744902341588e-06, "loss": 0.0383, "step": 13236 }, { "epoch": 0.6151022304832714, "grad_norm": 0.7594221027839476, "learning_rate": 3.4194605454450116e-06, "loss": 0.0696, "step": 13237 }, { "epoch": 0.6151486988847583, "grad_norm": 0.36659030562656486, "learning_rate": 3.4187466364567466e-06, "loss": 0.0366, "step": 13238 }, { "epoch": 0.6151951672862453, "grad_norm": 0.6305603854619616, "learning_rate": 3.418032763285537e-06, "loss": 0.0329, "step": 13239 }, { "epoch": 0.6152416356877324, "grad_norm": 0.9561583050872783, "learning_rate": 3.417318925947547e-06, "loss": 0.0567, "step": 13240 }, { "epoch": 0.6152881040892193, "grad_norm": 0.24398221988688376, "learning_rate": 3.416605124458951e-06, "loss": 0.0261, "step": 13241 }, { "epoch": 0.6153345724907063, "grad_norm": 0.4036321727460434, "learning_rate": 3.415891358835914e-06, "loss": 0.0262, "step": 13242 }, { "epoch": 0.6153810408921933, "grad_norm": 0.7240379370523794, "learning_rate": 3.415177629094606e-06, "loss": 0.0621, "step": 13243 }, { "epoch": 0.6154275092936803, "grad_norm": 0.6927875873019654, "learning_rate": 3.4144639352511917e-06, "loss": 0.0446, "step": 13244 }, { "epoch": 0.6154739776951673, "grad_norm": 0.4504049388905375, "learning_rate": 3.41375027732184e-06, "loss": 0.0467, "step": 13245 }, { "epoch": 0.6155204460966542, "grad_norm": 0.3292441771749482, "learning_rate": 3.413036655322713e-06, "loss": 0.062, "step": 13246 }, { "epoch": 0.6155669144981413, "grad_norm": 0.43379305937571333, "learning_rate": 3.412323069269975e-06, "loss": 0.0539, "step": 13247 }, { "epoch": 0.6156133828996283, "grad_norm": 0.5261883876701898, "learning_rate": 3.4116095191797916e-06, "loss": 0.0609, "step": 13248 }, { "epoch": 0.6156598513011152, "grad_norm": 0.6271061734672186, "learning_rate": 3.4108960050683226e-06, "loss": 0.0548, "step": 13249 }, { "epoch": 0.6157063197026023, "grad_norm": 0.5148352361347233, "learning_rate": 3.4101825269517337e-06, "loss": 0.0308, "step": 13250 }, { "epoch": 0.6157527881040892, "grad_norm": 0.9654589585777242, "learning_rate": 3.409469084846184e-06, "loss": 0.0519, "step": 13251 }, { "epoch": 0.6157992565055762, "grad_norm": 0.5377131306670018, "learning_rate": 3.4087556787678322e-06, "loss": 0.0409, "step": 13252 }, { "epoch": 0.6158457249070632, "grad_norm": 0.7149231646974076, "learning_rate": 3.4080423087328397e-06, "loss": 0.0734, "step": 13253 }, { "epoch": 0.6158921933085502, "grad_norm": 0.6743180854749554, "learning_rate": 3.4073289747573633e-06, "loss": 0.0504, "step": 13254 }, { "epoch": 0.6159386617100372, "grad_norm": 0.5723649509920208, "learning_rate": 3.406615676857563e-06, "loss": 0.0485, "step": 13255 }, { "epoch": 0.6159851301115241, "grad_norm": 0.6269527270849422, "learning_rate": 3.405902415049593e-06, "loss": 0.0709, "step": 13256 }, { "epoch": 0.6160315985130111, "grad_norm": 0.7686129195015364, "learning_rate": 3.4051891893496146e-06, "loss": 0.0549, "step": 13257 }, { "epoch": 0.6160780669144982, "grad_norm": 1.1994786657100298, "learning_rate": 3.4044759997737774e-06, "loss": 0.069, "step": 13258 }, { "epoch": 0.6161245353159851, "grad_norm": 0.5485267141536805, "learning_rate": 3.4037628463382377e-06, "loss": 0.0492, "step": 13259 }, { "epoch": 0.6161710037174721, "grad_norm": 1.2917317528606487, "learning_rate": 3.4030497290591497e-06, "loss": 0.0575, "step": 13260 }, { "epoch": 0.616217472118959, "grad_norm": 1.7512691933656825, "learning_rate": 3.4023366479526652e-06, "loss": 0.0615, "step": 13261 }, { "epoch": 0.6162639405204461, "grad_norm": 0.6436518300358025, "learning_rate": 3.401623603034939e-06, "loss": 0.0663, "step": 13262 }, { "epoch": 0.6163104089219331, "grad_norm": 0.42472268667304763, "learning_rate": 3.400910594322121e-06, "loss": 0.0537, "step": 13263 }, { "epoch": 0.61635687732342, "grad_norm": 0.537111749866121, "learning_rate": 3.400197621830359e-06, "loss": 0.0471, "step": 13264 }, { "epoch": 0.6164033457249071, "grad_norm": 0.693418955119519, "learning_rate": 3.399484685575806e-06, "loss": 0.0455, "step": 13265 }, { "epoch": 0.6164498141263941, "grad_norm": 1.6793185685009053, "learning_rate": 3.3987717855746072e-06, "loss": 0.0621, "step": 13266 }, { "epoch": 0.616496282527881, "grad_norm": 0.5143333551822753, "learning_rate": 3.398058921842915e-06, "loss": 0.0649, "step": 13267 }, { "epoch": 0.616542750929368, "grad_norm": 0.724394870497037, "learning_rate": 3.397346094396873e-06, "loss": 0.0579, "step": 13268 }, { "epoch": 0.616589219330855, "grad_norm": 0.5815408038443868, "learning_rate": 3.39663330325263e-06, "loss": 0.0563, "step": 13269 }, { "epoch": 0.616635687732342, "grad_norm": 1.452127234650792, "learning_rate": 3.39592054842633e-06, "loss": 0.0648, "step": 13270 }, { "epoch": 0.616682156133829, "grad_norm": 0.702436898387338, "learning_rate": 3.3952078299341164e-06, "loss": 0.0656, "step": 13271 }, { "epoch": 0.616728624535316, "grad_norm": 0.5813664718011778, "learning_rate": 3.394495147792135e-06, "loss": 0.0304, "step": 13272 }, { "epoch": 0.616775092936803, "grad_norm": 0.42955198596097444, "learning_rate": 3.3937825020165284e-06, "loss": 0.0434, "step": 13273 }, { "epoch": 0.6168215613382899, "grad_norm": 1.317194021910271, "learning_rate": 3.3930698926234385e-06, "loss": 0.0795, "step": 13274 }, { "epoch": 0.616868029739777, "grad_norm": 0.5761107378105012, "learning_rate": 3.392357319629008e-06, "loss": 0.0489, "step": 13275 }, { "epoch": 0.616914498141264, "grad_norm": 0.34003036688873695, "learning_rate": 3.391644783049375e-06, "loss": 0.0617, "step": 13276 }, { "epoch": 0.6169609665427509, "grad_norm": 0.36560824645464207, "learning_rate": 3.3909322829006815e-06, "loss": 0.0572, "step": 13277 }, { "epoch": 0.6170074349442379, "grad_norm": 0.8463764654488664, "learning_rate": 3.390219819199063e-06, "loss": 0.0558, "step": 13278 }, { "epoch": 0.6170539033457249, "grad_norm": 0.34609725669970004, "learning_rate": 3.3895073919606626e-06, "loss": 0.0509, "step": 13279 }, { "epoch": 0.6171003717472119, "grad_norm": 0.735131128230254, "learning_rate": 3.3887950012016125e-06, "loss": 0.0715, "step": 13280 }, { "epoch": 0.6171468401486989, "grad_norm": 0.6540498549374557, "learning_rate": 3.3880826469380536e-06, "loss": 0.0474, "step": 13281 }, { "epoch": 0.6171933085501858, "grad_norm": 0.642677400438009, "learning_rate": 3.3873703291861188e-06, "loss": 0.0598, "step": 13282 }, { "epoch": 0.6172397769516729, "grad_norm": 0.7260285110411919, "learning_rate": 3.3866580479619415e-06, "loss": 0.0778, "step": 13283 }, { "epoch": 0.6172862453531599, "grad_norm": 0.5667719405291567, "learning_rate": 3.3859458032816595e-06, "loss": 0.0487, "step": 13284 }, { "epoch": 0.6173327137546468, "grad_norm": 0.6078072046297825, "learning_rate": 3.385233595161402e-06, "loss": 0.0606, "step": 13285 }, { "epoch": 0.6173791821561339, "grad_norm": 0.5802354203406893, "learning_rate": 3.384521423617305e-06, "loss": 0.0507, "step": 13286 }, { "epoch": 0.6174256505576208, "grad_norm": 0.8192860617781503, "learning_rate": 3.383809288665497e-06, "loss": 0.0672, "step": 13287 }, { "epoch": 0.6174721189591078, "grad_norm": 0.448268881541581, "learning_rate": 3.383097190322109e-06, "loss": 0.0511, "step": 13288 }, { "epoch": 0.6175185873605948, "grad_norm": 0.8072778752453547, "learning_rate": 3.382385128603273e-06, "loss": 0.0801, "step": 13289 }, { "epoch": 0.6175650557620818, "grad_norm": 0.4962374193060667, "learning_rate": 3.381673103525114e-06, "loss": 0.0463, "step": 13290 }, { "epoch": 0.6176115241635688, "grad_norm": 0.5018478052353708, "learning_rate": 3.380961115103764e-06, "loss": 0.0619, "step": 13291 }, { "epoch": 0.6176579925650557, "grad_norm": 0.5917829792782917, "learning_rate": 3.38024916335535e-06, "loss": 0.0449, "step": 13292 }, { "epoch": 0.6177044609665427, "grad_norm": 0.5740313921207564, "learning_rate": 3.3795372482959953e-06, "loss": 0.0474, "step": 13293 }, { "epoch": 0.6177509293680298, "grad_norm": 0.4131743820395078, "learning_rate": 3.3788253699418275e-06, "loss": 0.0512, "step": 13294 }, { "epoch": 0.6177973977695167, "grad_norm": 1.143126308582648, "learning_rate": 3.3781135283089716e-06, "loss": 0.0985, "step": 13295 }, { "epoch": 0.6178438661710037, "grad_norm": 0.5232597946610161, "learning_rate": 3.3774017234135516e-06, "loss": 0.051, "step": 13296 }, { "epoch": 0.6178903345724907, "grad_norm": 0.46472053982476114, "learning_rate": 3.376689955271689e-06, "loss": 0.0511, "step": 13297 }, { "epoch": 0.6179368029739777, "grad_norm": 0.6540854366449773, "learning_rate": 3.37597822389951e-06, "loss": 0.0556, "step": 13298 }, { "epoch": 0.6179832713754647, "grad_norm": 0.32864241469124045, "learning_rate": 3.3752665293131325e-06, "loss": 0.0356, "step": 13299 }, { "epoch": 0.6180297397769516, "grad_norm": 0.43787388668979155, "learning_rate": 3.374554871528677e-06, "loss": 0.0606, "step": 13300 }, { "epoch": 0.6180762081784387, "grad_norm": 0.5043886749196135, "learning_rate": 3.3738432505622653e-06, "loss": 0.0465, "step": 13301 }, { "epoch": 0.6181226765799257, "grad_norm": 0.5686149619596763, "learning_rate": 3.373131666430014e-06, "loss": 0.0533, "step": 13302 }, { "epoch": 0.6181691449814126, "grad_norm": 0.5248782062820402, "learning_rate": 3.3724201191480443e-06, "loss": 0.0391, "step": 13303 }, { "epoch": 0.6182156133828997, "grad_norm": 0.7433846183798287, "learning_rate": 3.371708608732473e-06, "loss": 0.0613, "step": 13304 }, { "epoch": 0.6182620817843866, "grad_norm": 0.6171383996776219, "learning_rate": 3.370997135199413e-06, "loss": 0.0469, "step": 13305 }, { "epoch": 0.6183085501858736, "grad_norm": 0.3976410658985189, "learning_rate": 3.370285698564984e-06, "loss": 0.049, "step": 13306 }, { "epoch": 0.6183550185873606, "grad_norm": 0.516373705898339, "learning_rate": 3.369574298845298e-06, "loss": 0.0361, "step": 13307 }, { "epoch": 0.6184014869888476, "grad_norm": 0.32509327993518317, "learning_rate": 3.3688629360564707e-06, "loss": 0.0441, "step": 13308 }, { "epoch": 0.6184479553903346, "grad_norm": 0.6752206133394666, "learning_rate": 3.368151610214614e-06, "loss": 0.04, "step": 13309 }, { "epoch": 0.6184944237918215, "grad_norm": 0.7183774316208473, "learning_rate": 3.3674403213358424e-06, "loss": 0.0631, "step": 13310 }, { "epoch": 0.6185408921933085, "grad_norm": 0.4382108136928969, "learning_rate": 3.3667290694362654e-06, "loss": 0.0563, "step": 13311 }, { "epoch": 0.6185873605947956, "grad_norm": 0.4980282977075328, "learning_rate": 3.366017854531992e-06, "loss": 0.0411, "step": 13312 }, { "epoch": 0.6186338289962825, "grad_norm": 0.7696944689062319, "learning_rate": 3.365306676639136e-06, "loss": 0.0453, "step": 13313 }, { "epoch": 0.6186802973977695, "grad_norm": 0.355645582125955, "learning_rate": 3.3645955357738027e-06, "loss": 0.0598, "step": 13314 }, { "epoch": 0.6187267657992565, "grad_norm": 0.783751169734974, "learning_rate": 3.363884431952103e-06, "loss": 0.065, "step": 13315 }, { "epoch": 0.6187732342007435, "grad_norm": 0.5809859881070354, "learning_rate": 3.3631733651901437e-06, "loss": 0.0643, "step": 13316 }, { "epoch": 0.6188197026022305, "grad_norm": 0.456852973422686, "learning_rate": 3.3624623355040285e-06, "loss": 0.0617, "step": 13317 }, { "epoch": 0.6188661710037174, "grad_norm": 0.36074470290147653, "learning_rate": 3.361751342909866e-06, "loss": 0.0403, "step": 13318 }, { "epoch": 0.6189126394052045, "grad_norm": 1.9348130052646821, "learning_rate": 3.3610403874237585e-06, "loss": 0.085, "step": 13319 }, { "epoch": 0.6189591078066915, "grad_norm": 0.37085803814443813, "learning_rate": 3.360329469061812e-06, "loss": 0.0624, "step": 13320 }, { "epoch": 0.6190055762081784, "grad_norm": 0.8135622422758991, "learning_rate": 3.359618587840128e-06, "loss": 0.0413, "step": 13321 }, { "epoch": 0.6190520446096655, "grad_norm": 0.7337336456649204, "learning_rate": 3.358907743774812e-06, "loss": 0.0605, "step": 13322 }, { "epoch": 0.6190985130111524, "grad_norm": 0.6837625249051625, "learning_rate": 3.358196936881961e-06, "loss": 0.0665, "step": 13323 }, { "epoch": 0.6191449814126394, "grad_norm": 0.6557400507860667, "learning_rate": 3.357486167177677e-06, "loss": 0.0637, "step": 13324 }, { "epoch": 0.6191914498141264, "grad_norm": 0.7691308051928755, "learning_rate": 3.3567754346780607e-06, "loss": 0.0709, "step": 13325 }, { "epoch": 0.6192379182156134, "grad_norm": 0.4714379132540272, "learning_rate": 3.3560647393992096e-06, "loss": 0.0471, "step": 13326 }, { "epoch": 0.6192843866171004, "grad_norm": 0.8561158967609058, "learning_rate": 3.355354081357223e-06, "loss": 0.0682, "step": 13327 }, { "epoch": 0.6193308550185873, "grad_norm": 0.6486169327624381, "learning_rate": 3.3546434605681983e-06, "loss": 0.0644, "step": 13328 }, { "epoch": 0.6193773234200743, "grad_norm": 0.5068569849353014, "learning_rate": 3.3539328770482296e-06, "loss": 0.0514, "step": 13329 }, { "epoch": 0.6194237918215614, "grad_norm": 0.8187697766924424, "learning_rate": 3.3532223308134136e-06, "loss": 0.082, "step": 13330 }, { "epoch": 0.6194702602230483, "grad_norm": 0.487618890542276, "learning_rate": 3.3525118218798446e-06, "loss": 0.0428, "step": 13331 }, { "epoch": 0.6195167286245353, "grad_norm": 0.4034435545758995, "learning_rate": 3.351801350263618e-06, "loss": 0.0311, "step": 13332 }, { "epoch": 0.6195631970260222, "grad_norm": 0.6081242296840014, "learning_rate": 3.351090915980824e-06, "loss": 0.0576, "step": 13333 }, { "epoch": 0.6196096654275093, "grad_norm": 0.4077358343682844, "learning_rate": 3.3503805190475587e-06, "loss": 0.0386, "step": 13334 }, { "epoch": 0.6196561338289963, "grad_norm": 0.7805561290292611, "learning_rate": 3.3496701594799095e-06, "loss": 0.066, "step": 13335 }, { "epoch": 0.6197026022304832, "grad_norm": 0.8162356345659855, "learning_rate": 3.3489598372939674e-06, "loss": 0.0671, "step": 13336 }, { "epoch": 0.6197490706319703, "grad_norm": 0.7959852951574783, "learning_rate": 3.3482495525058232e-06, "loss": 0.0678, "step": 13337 }, { "epoch": 0.6197955390334573, "grad_norm": 0.6543020664935925, "learning_rate": 3.347539305131565e-06, "loss": 0.0473, "step": 13338 }, { "epoch": 0.6198420074349442, "grad_norm": 0.7552345366025754, "learning_rate": 3.3468290951872807e-06, "loss": 0.065, "step": 13339 }, { "epoch": 0.6198884758364313, "grad_norm": 0.6551350566300721, "learning_rate": 3.346118922689059e-06, "loss": 0.0542, "step": 13340 }, { "epoch": 0.6199349442379182, "grad_norm": 0.5919302388088189, "learning_rate": 3.345408787652982e-06, "loss": 0.0509, "step": 13341 }, { "epoch": 0.6199814126394052, "grad_norm": 0.7550073098040857, "learning_rate": 3.3446986900951384e-06, "loss": 0.077, "step": 13342 }, { "epoch": 0.6200278810408922, "grad_norm": 0.47121663043649675, "learning_rate": 3.343988630031611e-06, "loss": 0.0666, "step": 13343 }, { "epoch": 0.6200743494423792, "grad_norm": 0.4611562645744878, "learning_rate": 3.343278607478485e-06, "loss": 0.0676, "step": 13344 }, { "epoch": 0.6201208178438662, "grad_norm": 0.5853506649975122, "learning_rate": 3.342568622451843e-06, "loss": 0.0615, "step": 13345 }, { "epoch": 0.6201672862453531, "grad_norm": 0.6275683786695977, "learning_rate": 3.341858674967764e-06, "loss": 0.0733, "step": 13346 }, { "epoch": 0.6202137546468401, "grad_norm": 0.7293426878471689, "learning_rate": 3.3411487650423324e-06, "loss": 0.0573, "step": 13347 }, { "epoch": 0.6202602230483272, "grad_norm": 0.6612209168554305, "learning_rate": 3.340438892691626e-06, "loss": 0.0706, "step": 13348 }, { "epoch": 0.6203066914498141, "grad_norm": 0.23861993236811943, "learning_rate": 3.3397290579317263e-06, "loss": 0.0379, "step": 13349 }, { "epoch": 0.6203531598513011, "grad_norm": 1.3705369082014862, "learning_rate": 3.3390192607787097e-06, "loss": 0.0627, "step": 13350 }, { "epoch": 0.620399628252788, "grad_norm": 1.320670108284982, "learning_rate": 3.338309501248658e-06, "loss": 0.068, "step": 13351 }, { "epoch": 0.6204460966542751, "grad_norm": 0.693076784924063, "learning_rate": 3.3375997793576433e-06, "loss": 0.0536, "step": 13352 }, { "epoch": 0.6204925650557621, "grad_norm": 0.6124665446812226, "learning_rate": 3.336890095121742e-06, "loss": 0.0451, "step": 13353 }, { "epoch": 0.620539033457249, "grad_norm": 0.489465939064749, "learning_rate": 3.3361804485570315e-06, "loss": 0.0398, "step": 13354 }, { "epoch": 0.6205855018587361, "grad_norm": 0.4634743135748698, "learning_rate": 3.335470839679584e-06, "loss": 0.0574, "step": 13355 }, { "epoch": 0.6206319702602231, "grad_norm": 0.5424364071036668, "learning_rate": 3.3347612685054753e-06, "loss": 0.0345, "step": 13356 }, { "epoch": 0.62067843866171, "grad_norm": 1.1335966687711823, "learning_rate": 3.3340517350507774e-06, "loss": 0.0531, "step": 13357 }, { "epoch": 0.620724907063197, "grad_norm": 0.48539136231370333, "learning_rate": 3.3333422393315583e-06, "loss": 0.049, "step": 13358 }, { "epoch": 0.620771375464684, "grad_norm": 0.34849221750420306, "learning_rate": 3.3326327813638937e-06, "loss": 0.043, "step": 13359 }, { "epoch": 0.620817843866171, "grad_norm": 0.5460516654777068, "learning_rate": 3.33192336116385e-06, "loss": 0.0433, "step": 13360 }, { "epoch": 0.620864312267658, "grad_norm": 0.6954576094110888, "learning_rate": 3.3312139787474986e-06, "loss": 0.0642, "step": 13361 }, { "epoch": 0.620910780669145, "grad_norm": 0.7277784848800907, "learning_rate": 3.3305046341309053e-06, "loss": 0.0604, "step": 13362 }, { "epoch": 0.620957249070632, "grad_norm": 0.5099023509400356, "learning_rate": 3.3297953273301425e-06, "loss": 0.063, "step": 13363 }, { "epoch": 0.6210037174721189, "grad_norm": 0.714012287431032, "learning_rate": 3.3290860583612715e-06, "loss": 0.0704, "step": 13364 }, { "epoch": 0.6210501858736059, "grad_norm": 1.0301635173450852, "learning_rate": 3.3283768272403592e-06, "loss": 0.0624, "step": 13365 }, { "epoch": 0.621096654275093, "grad_norm": 0.57571132099759, "learning_rate": 3.3276676339834723e-06, "loss": 0.0508, "step": 13366 }, { "epoch": 0.6211431226765799, "grad_norm": 0.6051129331413437, "learning_rate": 3.326958478606672e-06, "loss": 0.0438, "step": 13367 }, { "epoch": 0.6211895910780669, "grad_norm": 0.5776801710497373, "learning_rate": 3.3262493611260248e-06, "loss": 0.0358, "step": 13368 }, { "epoch": 0.6212360594795538, "grad_norm": 0.47011418007958333, "learning_rate": 3.325540281557592e-06, "loss": 0.0496, "step": 13369 }, { "epoch": 0.6212825278810409, "grad_norm": 0.5329543837561928, "learning_rate": 3.3248312399174322e-06, "loss": 0.0734, "step": 13370 }, { "epoch": 0.6213289962825279, "grad_norm": 0.48083040947255845, "learning_rate": 3.3241222362216086e-06, "loss": 0.0505, "step": 13371 }, { "epoch": 0.6213754646840148, "grad_norm": 0.5687164559112757, "learning_rate": 3.3234132704861786e-06, "loss": 0.0493, "step": 13372 }, { "epoch": 0.6214219330855019, "grad_norm": 0.7819282963084411, "learning_rate": 3.3227043427272043e-06, "loss": 0.0654, "step": 13373 }, { "epoch": 0.6214684014869889, "grad_norm": 0.7617167714505433, "learning_rate": 3.3219954529607413e-06, "loss": 0.0812, "step": 13374 }, { "epoch": 0.6215148698884758, "grad_norm": 0.5841597241229509, "learning_rate": 3.321286601202849e-06, "loss": 0.0394, "step": 13375 }, { "epoch": 0.6215613382899629, "grad_norm": 0.4336485031202422, "learning_rate": 3.3205777874695807e-06, "loss": 0.0352, "step": 13376 }, { "epoch": 0.6216078066914498, "grad_norm": 0.4568911483296695, "learning_rate": 3.3198690117769923e-06, "loss": 0.0548, "step": 13377 }, { "epoch": 0.6216542750929368, "grad_norm": 0.4399831063963315, "learning_rate": 3.31916027414114e-06, "loss": 0.0582, "step": 13378 }, { "epoch": 0.6217007434944238, "grad_norm": 0.39285188616403033, "learning_rate": 3.3184515745780754e-06, "loss": 0.0488, "step": 13379 }, { "epoch": 0.6217472118959108, "grad_norm": 0.49033417487472636, "learning_rate": 3.317742913103854e-06, "loss": 0.0479, "step": 13380 }, { "epoch": 0.6217936802973978, "grad_norm": 0.4931310602994512, "learning_rate": 3.3170342897345266e-06, "loss": 0.0486, "step": 13381 }, { "epoch": 0.6218401486988847, "grad_norm": 0.5965441659467527, "learning_rate": 3.316325704486142e-06, "loss": 0.0527, "step": 13382 }, { "epoch": 0.6218866171003717, "grad_norm": 0.5694426036498109, "learning_rate": 3.315617157374753e-06, "loss": 0.0405, "step": 13383 }, { "epoch": 0.6219330855018588, "grad_norm": 0.7295214658225556, "learning_rate": 3.314908648416408e-06, "loss": 0.0669, "step": 13384 }, { "epoch": 0.6219795539033457, "grad_norm": 0.679321300452415, "learning_rate": 3.314200177627156e-06, "loss": 0.0648, "step": 13385 }, { "epoch": 0.6220260223048327, "grad_norm": 0.33292636810274817, "learning_rate": 3.3134917450230433e-06, "loss": 0.0501, "step": 13386 }, { "epoch": 0.6220724907063196, "grad_norm": 0.7847339150000784, "learning_rate": 3.3127833506201194e-06, "loss": 0.0604, "step": 13387 }, { "epoch": 0.6221189591078067, "grad_norm": 0.5251459858114085, "learning_rate": 3.3120749944344277e-06, "loss": 0.0637, "step": 13388 }, { "epoch": 0.6221654275092937, "grad_norm": 0.5439080754135002, "learning_rate": 3.311366676482013e-06, "loss": 0.0774, "step": 13389 }, { "epoch": 0.6222118959107806, "grad_norm": 1.2119241789859179, "learning_rate": 3.3106583967789207e-06, "loss": 0.0427, "step": 13390 }, { "epoch": 0.6222583643122677, "grad_norm": 0.5290369011081663, "learning_rate": 3.3099501553411935e-06, "loss": 0.0489, "step": 13391 }, { "epoch": 0.6223048327137547, "grad_norm": 0.5039238639019282, "learning_rate": 3.3092419521848744e-06, "loss": 0.0418, "step": 13392 }, { "epoch": 0.6223513011152416, "grad_norm": 0.4788963753061272, "learning_rate": 3.308533787326006e-06, "loss": 0.0623, "step": 13393 }, { "epoch": 0.6223977695167286, "grad_norm": 0.7969086990473807, "learning_rate": 3.307825660780626e-06, "loss": 0.0497, "step": 13394 }, { "epoch": 0.6224442379182156, "grad_norm": 0.38957133484789197, "learning_rate": 3.307117572564775e-06, "loss": 0.0323, "step": 13395 }, { "epoch": 0.6224907063197026, "grad_norm": 0.44145930843985326, "learning_rate": 3.3064095226944936e-06, "loss": 0.0382, "step": 13396 }, { "epoch": 0.6225371747211896, "grad_norm": 0.7397097160364499, "learning_rate": 3.305701511185819e-06, "loss": 0.0596, "step": 13397 }, { "epoch": 0.6225836431226766, "grad_norm": 0.3961124765726162, "learning_rate": 3.3049935380547875e-06, "loss": 0.0483, "step": 13398 }, { "epoch": 0.6226301115241636, "grad_norm": 0.6523649030925296, "learning_rate": 3.3042856033174386e-06, "loss": 0.0483, "step": 13399 }, { "epoch": 0.6226765799256505, "grad_norm": 0.4249971797279526, "learning_rate": 3.3035777069898044e-06, "loss": 0.0504, "step": 13400 }, { "epoch": 0.6227230483271375, "grad_norm": 0.4727497022855645, "learning_rate": 3.3028698490879194e-06, "loss": 0.0487, "step": 13401 }, { "epoch": 0.6227695167286246, "grad_norm": 0.5672456446206642, "learning_rate": 3.30216202962782e-06, "loss": 0.0695, "step": 13402 }, { "epoch": 0.6228159851301115, "grad_norm": 0.3661600266541201, "learning_rate": 3.3014542486255365e-06, "loss": 0.037, "step": 13403 }, { "epoch": 0.6228624535315985, "grad_norm": 0.8563758481214468, "learning_rate": 3.3007465060971034e-06, "loss": 0.0502, "step": 13404 }, { "epoch": 0.6229089219330854, "grad_norm": 0.46916130776467113, "learning_rate": 3.3000388020585496e-06, "loss": 0.0415, "step": 13405 }, { "epoch": 0.6229553903345725, "grad_norm": 0.34322977935338694, "learning_rate": 3.299331136525905e-06, "loss": 0.0488, "step": 13406 }, { "epoch": 0.6230018587360595, "grad_norm": 0.4490282701139031, "learning_rate": 3.2986235095152007e-06, "loss": 0.0448, "step": 13407 }, { "epoch": 0.6230483271375464, "grad_norm": 0.44431271065711686, "learning_rate": 3.297915921042465e-06, "loss": 0.0362, "step": 13408 }, { "epoch": 0.6230947955390335, "grad_norm": 0.4910534078427789, "learning_rate": 3.2972083711237248e-06, "loss": 0.0434, "step": 13409 }, { "epoch": 0.6231412639405205, "grad_norm": 0.4096123130341839, "learning_rate": 3.296500859775009e-06, "loss": 0.0415, "step": 13410 }, { "epoch": 0.6231877323420074, "grad_norm": 0.39203658543321424, "learning_rate": 3.295793387012339e-06, "loss": 0.0558, "step": 13411 }, { "epoch": 0.6232342007434944, "grad_norm": 1.196492802822871, "learning_rate": 3.295085952851743e-06, "loss": 0.0588, "step": 13412 }, { "epoch": 0.6232806691449814, "grad_norm": 0.2993398893431836, "learning_rate": 3.2943785573092447e-06, "loss": 0.027, "step": 13413 }, { "epoch": 0.6233271375464684, "grad_norm": 0.42632968094933604, "learning_rate": 3.2936712004008677e-06, "loss": 0.0499, "step": 13414 }, { "epoch": 0.6233736059479554, "grad_norm": 0.4938870726457178, "learning_rate": 3.2929638821426324e-06, "loss": 0.0455, "step": 13415 }, { "epoch": 0.6234200743494424, "grad_norm": 0.5383770891746434, "learning_rate": 3.2922566025505644e-06, "loss": 0.0474, "step": 13416 }, { "epoch": 0.6234665427509294, "grad_norm": 0.6128612914949862, "learning_rate": 3.291549361640681e-06, "loss": 0.0551, "step": 13417 }, { "epoch": 0.6235130111524163, "grad_norm": 0.694837399357169, "learning_rate": 3.2908421594290006e-06, "loss": 0.0476, "step": 13418 }, { "epoch": 0.6235594795539033, "grad_norm": 0.45514427263180757, "learning_rate": 3.2901349959315466e-06, "loss": 0.0551, "step": 13419 }, { "epoch": 0.6236059479553904, "grad_norm": 0.38523132657702686, "learning_rate": 3.2894278711643322e-06, "loss": 0.052, "step": 13420 }, { "epoch": 0.6236524163568773, "grad_norm": 0.5882936250400496, "learning_rate": 3.288720785143379e-06, "loss": 0.039, "step": 13421 }, { "epoch": 0.6236988847583643, "grad_norm": 0.7642967932302287, "learning_rate": 3.2880137378847012e-06, "loss": 0.0657, "step": 13422 }, { "epoch": 0.6237453531598512, "grad_norm": 0.4243027500127062, "learning_rate": 3.2873067294043126e-06, "loss": 0.058, "step": 13423 }, { "epoch": 0.6237918215613383, "grad_norm": 0.7419004902273687, "learning_rate": 3.28659975971823e-06, "loss": 0.0515, "step": 13424 }, { "epoch": 0.6238382899628253, "grad_norm": 0.6059931481914959, "learning_rate": 3.2858928288424645e-06, "loss": 0.0422, "step": 13425 }, { "epoch": 0.6238847583643122, "grad_norm": 0.43346302849026863, "learning_rate": 3.285185936793032e-06, "loss": 0.0422, "step": 13426 }, { "epoch": 0.6239312267657993, "grad_norm": 0.4859958778803155, "learning_rate": 3.284479083585942e-06, "loss": 0.0386, "step": 13427 }, { "epoch": 0.6239776951672863, "grad_norm": 0.6892128261371792, "learning_rate": 3.2837722692372085e-06, "loss": 0.0665, "step": 13428 }, { "epoch": 0.6240241635687732, "grad_norm": 0.996087656986973, "learning_rate": 3.283065493762838e-06, "loss": 0.0801, "step": 13429 }, { "epoch": 0.6240706319702602, "grad_norm": 0.701585899535309, "learning_rate": 3.2823587571788407e-06, "loss": 0.0672, "step": 13430 }, { "epoch": 0.6241171003717472, "grad_norm": 0.6345476175046467, "learning_rate": 3.281652059501226e-06, "loss": 0.0469, "step": 13431 }, { "epoch": 0.6241635687732342, "grad_norm": 0.973901708138024, "learning_rate": 3.280945400745999e-06, "loss": 0.093, "step": 13432 }, { "epoch": 0.6242100371747212, "grad_norm": 0.5865135774164808, "learning_rate": 3.2802387809291698e-06, "loss": 0.0448, "step": 13433 }, { "epoch": 0.6242565055762082, "grad_norm": 0.9050694117238407, "learning_rate": 3.279532200066743e-06, "loss": 0.0745, "step": 13434 }, { "epoch": 0.6243029739776952, "grad_norm": 0.471400190180788, "learning_rate": 3.2788256581747204e-06, "loss": 0.0426, "step": 13435 }, { "epoch": 0.6243494423791821, "grad_norm": 0.4439370363239464, "learning_rate": 3.2781191552691093e-06, "loss": 0.0478, "step": 13436 }, { "epoch": 0.6243959107806691, "grad_norm": 0.618561437028292, "learning_rate": 3.2774126913659103e-06, "loss": 0.0631, "step": 13437 }, { "epoch": 0.6244423791821562, "grad_norm": 0.45306874617524406, "learning_rate": 3.2767062664811282e-06, "loss": 0.0457, "step": 13438 }, { "epoch": 0.6244888475836431, "grad_norm": 0.34178240451994407, "learning_rate": 3.2759998806307618e-06, "loss": 0.0354, "step": 13439 }, { "epoch": 0.6245353159851301, "grad_norm": 0.7904083951860716, "learning_rate": 3.2752935338308143e-06, "loss": 0.071, "step": 13440 }, { "epoch": 0.624581784386617, "grad_norm": 0.9804805405276112, "learning_rate": 3.2745872260972834e-06, "loss": 0.0426, "step": 13441 }, { "epoch": 0.6246282527881041, "grad_norm": 0.31537480689529074, "learning_rate": 3.273880957446166e-06, "loss": 0.0353, "step": 13442 }, { "epoch": 0.6246747211895911, "grad_norm": 0.5075490773754653, "learning_rate": 3.273174727893463e-06, "loss": 0.0295, "step": 13443 }, { "epoch": 0.624721189591078, "grad_norm": 0.6654637998550379, "learning_rate": 3.2724685374551685e-06, "loss": 0.0564, "step": 13444 }, { "epoch": 0.6247676579925651, "grad_norm": 1.3137001105797579, "learning_rate": 3.27176238614728e-06, "loss": 0.0699, "step": 13445 }, { "epoch": 0.6248141263940521, "grad_norm": 0.2855028809444982, "learning_rate": 3.2710562739857956e-06, "loss": 0.0359, "step": 13446 }, { "epoch": 0.624860594795539, "grad_norm": 0.5501480445021049, "learning_rate": 3.2703502009867026e-06, "loss": 0.0561, "step": 13447 }, { "epoch": 0.624907063197026, "grad_norm": 0.7839725632323871, "learning_rate": 3.2696441671659996e-06, "loss": 0.0598, "step": 13448 }, { "epoch": 0.624953531598513, "grad_norm": 0.8240672719642101, "learning_rate": 3.268938172539676e-06, "loss": 0.0527, "step": 13449 }, { "epoch": 0.625, "grad_norm": 0.4091606250764827, "learning_rate": 3.268232217123726e-06, "loss": 0.0483, "step": 13450 }, { "epoch": 0.625046468401487, "grad_norm": 0.42352863448135536, "learning_rate": 3.2675263009341373e-06, "loss": 0.0439, "step": 13451 }, { "epoch": 0.625092936802974, "grad_norm": 0.8929489156497931, "learning_rate": 3.2668204239869046e-06, "loss": 0.0686, "step": 13452 }, { "epoch": 0.625139405204461, "grad_norm": 0.7121412839365061, "learning_rate": 3.2661145862980113e-06, "loss": 0.0625, "step": 13453 }, { "epoch": 0.6251858736059479, "grad_norm": 1.7291696897094464, "learning_rate": 3.2654087878834472e-06, "loss": 0.0779, "step": 13454 }, { "epoch": 0.6252323420074349, "grad_norm": 1.3874125057782962, "learning_rate": 3.2647030287591997e-06, "loss": 0.0652, "step": 13455 }, { "epoch": 0.625278810408922, "grad_norm": 1.3115519028838947, "learning_rate": 3.2639973089412548e-06, "loss": 0.0567, "step": 13456 }, { "epoch": 0.6253252788104089, "grad_norm": 0.556495096368239, "learning_rate": 3.263291628445598e-06, "loss": 0.0598, "step": 13457 }, { "epoch": 0.6253717472118959, "grad_norm": 0.709963274675422, "learning_rate": 3.262585987288216e-06, "loss": 0.0568, "step": 13458 }, { "epoch": 0.625418215613383, "grad_norm": 0.6738402806113506, "learning_rate": 3.2618803854850877e-06, "loss": 0.0452, "step": 13459 }, { "epoch": 0.6254646840148699, "grad_norm": 0.7246895382705887, "learning_rate": 3.261174823052199e-06, "loss": 0.0553, "step": 13460 }, { "epoch": 0.6255111524163569, "grad_norm": 0.6199409267890769, "learning_rate": 3.2604693000055286e-06, "loss": 0.0501, "step": 13461 }, { "epoch": 0.6255576208178438, "grad_norm": 0.5602204294884489, "learning_rate": 3.2597638163610606e-06, "loss": 0.0403, "step": 13462 }, { "epoch": 0.6256040892193309, "grad_norm": 0.6775399671765325, "learning_rate": 3.259058372134776e-06, "loss": 0.0443, "step": 13463 }, { "epoch": 0.6256505576208179, "grad_norm": 0.6355010943543876, "learning_rate": 3.258352967342649e-06, "loss": 0.0489, "step": 13464 }, { "epoch": 0.6256970260223048, "grad_norm": 0.43474969123254925, "learning_rate": 3.2576476020006607e-06, "loss": 0.0595, "step": 13465 }, { "epoch": 0.6257434944237918, "grad_norm": 0.44912037307494335, "learning_rate": 3.256942276124788e-06, "loss": 0.0496, "step": 13466 }, { "epoch": 0.6257899628252788, "grad_norm": 0.6082852887531799, "learning_rate": 3.2562369897310065e-06, "loss": 0.0406, "step": 13467 }, { "epoch": 0.6258364312267658, "grad_norm": 0.570312894486024, "learning_rate": 3.255531742835294e-06, "loss": 0.0779, "step": 13468 }, { "epoch": 0.6258828996282528, "grad_norm": 0.45420727812451117, "learning_rate": 3.2548265354536246e-06, "loss": 0.0384, "step": 13469 }, { "epoch": 0.6259293680297398, "grad_norm": 0.5191824495637879, "learning_rate": 3.254121367601969e-06, "loss": 0.0383, "step": 13470 }, { "epoch": 0.6259758364312268, "grad_norm": 0.7058620046737465, "learning_rate": 3.253416239296302e-06, "loss": 0.0735, "step": 13471 }, { "epoch": 0.6260223048327137, "grad_norm": 0.6730065591003674, "learning_rate": 3.2527111505525943e-06, "loss": 0.0708, "step": 13472 }, { "epoch": 0.6260687732342007, "grad_norm": 0.5056865304700876, "learning_rate": 3.25200610138682e-06, "loss": 0.0282, "step": 13473 }, { "epoch": 0.6261152416356878, "grad_norm": 0.6063689669246961, "learning_rate": 3.2513010918149456e-06, "loss": 0.054, "step": 13474 }, { "epoch": 0.6261617100371747, "grad_norm": 0.5850746739747333, "learning_rate": 3.2505961218529457e-06, "loss": 0.0602, "step": 13475 }, { "epoch": 0.6262081784386617, "grad_norm": 0.41144566704434854, "learning_rate": 3.2498911915167815e-06, "loss": 0.0506, "step": 13476 }, { "epoch": 0.6262546468401488, "grad_norm": 0.44649578049988886, "learning_rate": 3.2491863008224245e-06, "loss": 0.0382, "step": 13477 }, { "epoch": 0.6263011152416357, "grad_norm": 0.2536545291191222, "learning_rate": 3.2484814497858397e-06, "loss": 0.0248, "step": 13478 }, { "epoch": 0.6263475836431227, "grad_norm": 0.7893302298655422, "learning_rate": 3.2477766384229937e-06, "loss": 0.0555, "step": 13479 }, { "epoch": 0.6263940520446096, "grad_norm": 0.3040782945901484, "learning_rate": 3.247071866749851e-06, "loss": 0.0329, "step": 13480 }, { "epoch": 0.6264405204460967, "grad_norm": 0.5864006034103314, "learning_rate": 3.2463671347823776e-06, "loss": 0.0474, "step": 13481 }, { "epoch": 0.6264869888475837, "grad_norm": 0.63701479922287, "learning_rate": 3.245662442536533e-06, "loss": 0.0446, "step": 13482 }, { "epoch": 0.6265334572490706, "grad_norm": 0.507010874759714, "learning_rate": 3.2449577900282793e-06, "loss": 0.0543, "step": 13483 }, { "epoch": 0.6265799256505576, "grad_norm": 0.5096892639370855, "learning_rate": 3.2442531772735784e-06, "loss": 0.0586, "step": 13484 }, { "epoch": 0.6266263940520446, "grad_norm": 0.530793251998347, "learning_rate": 3.2435486042883928e-06, "loss": 0.0357, "step": 13485 }, { "epoch": 0.6266728624535316, "grad_norm": 0.9217575019364761, "learning_rate": 3.242844071088679e-06, "loss": 0.0718, "step": 13486 }, { "epoch": 0.6267193308550186, "grad_norm": 0.40391536421099505, "learning_rate": 3.2421395776903985e-06, "loss": 0.0441, "step": 13487 }, { "epoch": 0.6267657992565056, "grad_norm": 0.8931326133881076, "learning_rate": 3.2414351241095044e-06, "loss": 0.0669, "step": 13488 }, { "epoch": 0.6268122676579926, "grad_norm": 0.656808389521873, "learning_rate": 3.240730710361955e-06, "loss": 0.056, "step": 13489 }, { "epoch": 0.6268587360594795, "grad_norm": 0.4775916117259535, "learning_rate": 3.240026336463708e-06, "loss": 0.0671, "step": 13490 }, { "epoch": 0.6269052044609665, "grad_norm": 0.46141893164843517, "learning_rate": 3.239322002430716e-06, "loss": 0.0637, "step": 13491 }, { "epoch": 0.6269516728624536, "grad_norm": 1.0579768068230726, "learning_rate": 3.238617708278934e-06, "loss": 0.0496, "step": 13492 }, { "epoch": 0.6269981412639405, "grad_norm": 0.3549978336078111, "learning_rate": 3.237913454024316e-06, "loss": 0.0558, "step": 13493 }, { "epoch": 0.6270446096654275, "grad_norm": 0.6141982679219835, "learning_rate": 3.2372092396828113e-06, "loss": 0.0603, "step": 13494 }, { "epoch": 0.6270910780669146, "grad_norm": 0.7667812398788392, "learning_rate": 3.2365050652703732e-06, "loss": 0.0471, "step": 13495 }, { "epoch": 0.6271375464684015, "grad_norm": 0.6957419388647957, "learning_rate": 3.23580093080295e-06, "loss": 0.0671, "step": 13496 }, { "epoch": 0.6271840148698885, "grad_norm": 0.393014138499802, "learning_rate": 3.2350968362964947e-06, "loss": 0.0421, "step": 13497 }, { "epoch": 0.6272304832713754, "grad_norm": 0.6076623927024053, "learning_rate": 3.2343927817669518e-06, "loss": 0.0434, "step": 13498 }, { "epoch": 0.6272769516728625, "grad_norm": 0.6442193448790063, "learning_rate": 3.233688767230273e-06, "loss": 0.0526, "step": 13499 }, { "epoch": 0.6273234200743495, "grad_norm": 0.9555351280598118, "learning_rate": 3.2329847927024018e-06, "loss": 0.06, "step": 13500 }, { "epoch": 0.6273698884758364, "grad_norm": 0.7265702606703165, "learning_rate": 3.2322808581992825e-06, "loss": 0.0581, "step": 13501 }, { "epoch": 0.6274163568773234, "grad_norm": 0.7428314935616442, "learning_rate": 3.2315769637368645e-06, "loss": 0.0514, "step": 13502 }, { "epoch": 0.6274628252788104, "grad_norm": 0.5673188148540395, "learning_rate": 3.230873109331088e-06, "loss": 0.0314, "step": 13503 }, { "epoch": 0.6275092936802974, "grad_norm": 0.4080741230707674, "learning_rate": 3.230169294997899e-06, "loss": 0.0482, "step": 13504 }, { "epoch": 0.6275557620817844, "grad_norm": 0.5571721122978617, "learning_rate": 3.229465520753239e-06, "loss": 0.0252, "step": 13505 }, { "epoch": 0.6276022304832714, "grad_norm": 0.5210138293238825, "learning_rate": 3.228761786613046e-06, "loss": 0.0586, "step": 13506 }, { "epoch": 0.6276486988847584, "grad_norm": 0.38312226381232156, "learning_rate": 3.2280580925932637e-06, "loss": 0.0339, "step": 13507 }, { "epoch": 0.6276951672862453, "grad_norm": 0.8642975645235262, "learning_rate": 3.22735443870983e-06, "loss": 0.0372, "step": 13508 }, { "epoch": 0.6277416356877323, "grad_norm": 0.9808841558466372, "learning_rate": 3.226650824978685e-06, "loss": 0.0502, "step": 13509 }, { "epoch": 0.6277881040892194, "grad_norm": 0.5719919681768937, "learning_rate": 3.2259472514157643e-06, "loss": 0.0396, "step": 13510 }, { "epoch": 0.6278345724907063, "grad_norm": 0.672182105025509, "learning_rate": 3.2252437180370065e-06, "loss": 0.0639, "step": 13511 }, { "epoch": 0.6278810408921933, "grad_norm": 0.754945493528387, "learning_rate": 3.2245402248583468e-06, "loss": 0.0596, "step": 13512 }, { "epoch": 0.6279275092936804, "grad_norm": 0.6048231614670228, "learning_rate": 3.2238367718957175e-06, "loss": 0.0637, "step": 13513 }, { "epoch": 0.6279739776951673, "grad_norm": 0.45786969539442024, "learning_rate": 3.2231333591650567e-06, "loss": 0.0628, "step": 13514 }, { "epoch": 0.6280204460966543, "grad_norm": 0.5083770264747312, "learning_rate": 3.222429986682294e-06, "loss": 0.048, "step": 13515 }, { "epoch": 0.6280669144981412, "grad_norm": 0.5597753045101117, "learning_rate": 3.2217266544633634e-06, "loss": 0.0567, "step": 13516 }, { "epoch": 0.6281133828996283, "grad_norm": 0.6479288670011757, "learning_rate": 3.221023362524197e-06, "loss": 0.0545, "step": 13517 }, { "epoch": 0.6281598513011153, "grad_norm": 0.6624739860485898, "learning_rate": 3.2203201108807226e-06, "loss": 0.0528, "step": 13518 }, { "epoch": 0.6282063197026022, "grad_norm": 0.6141203218341045, "learning_rate": 3.2196168995488706e-06, "loss": 0.0445, "step": 13519 }, { "epoch": 0.6282527881040892, "grad_norm": 0.8260869079097402, "learning_rate": 3.2189137285445694e-06, "loss": 0.0707, "step": 13520 }, { "epoch": 0.6282992565055762, "grad_norm": 0.5599278515255909, "learning_rate": 3.2182105978837473e-06, "loss": 0.0574, "step": 13521 }, { "epoch": 0.6283457249070632, "grad_norm": 0.4825268261645936, "learning_rate": 3.2175075075823313e-06, "loss": 0.046, "step": 13522 }, { "epoch": 0.6283921933085502, "grad_norm": 0.5053190311235022, "learning_rate": 3.2168044576562454e-06, "loss": 0.0413, "step": 13523 }, { "epoch": 0.6284386617100371, "grad_norm": 0.4777337532120028, "learning_rate": 3.216101448121415e-06, "loss": 0.055, "step": 13524 }, { "epoch": 0.6284851301115242, "grad_norm": 0.6920965354144799, "learning_rate": 3.2153984789937644e-06, "loss": 0.0594, "step": 13525 }, { "epoch": 0.6285315985130111, "grad_norm": 0.7869888897335203, "learning_rate": 3.214695550289217e-06, "loss": 0.0555, "step": 13526 }, { "epoch": 0.6285780669144981, "grad_norm": 0.7128766349298314, "learning_rate": 3.2139926620236927e-06, "loss": 0.0662, "step": 13527 }, { "epoch": 0.6286245353159852, "grad_norm": 0.4953614532027785, "learning_rate": 3.2132898142131176e-06, "loss": 0.0397, "step": 13528 }, { "epoch": 0.6286710037174721, "grad_norm": 0.3796144507285853, "learning_rate": 3.212587006873407e-06, "loss": 0.0446, "step": 13529 }, { "epoch": 0.6287174721189591, "grad_norm": 0.6287810497828438, "learning_rate": 3.211884240020481e-06, "loss": 0.054, "step": 13530 }, { "epoch": 0.6287639405204462, "grad_norm": 0.7814647626550277, "learning_rate": 3.21118151367026e-06, "loss": 0.0574, "step": 13531 }, { "epoch": 0.6288104089219331, "grad_norm": 0.4812009690725331, "learning_rate": 3.2104788278386596e-06, "loss": 0.0435, "step": 13532 }, { "epoch": 0.6288568773234201, "grad_norm": 1.1468930947540263, "learning_rate": 3.2097761825415975e-06, "loss": 0.0543, "step": 13533 }, { "epoch": 0.628903345724907, "grad_norm": 0.6044340859494425, "learning_rate": 3.209073577794991e-06, "loss": 0.0359, "step": 13534 }, { "epoch": 0.6289498141263941, "grad_norm": 0.5381016022966612, "learning_rate": 3.2083710136147505e-06, "loss": 0.0544, "step": 13535 }, { "epoch": 0.6289962825278811, "grad_norm": 0.6527100564743699, "learning_rate": 3.2076684900167933e-06, "loss": 0.0704, "step": 13536 }, { "epoch": 0.629042750929368, "grad_norm": 0.8328435638924122, "learning_rate": 3.2069660070170296e-06, "loss": 0.055, "step": 13537 }, { "epoch": 0.629089219330855, "grad_norm": 0.38243752735746955, "learning_rate": 3.2062635646313746e-06, "loss": 0.0381, "step": 13538 }, { "epoch": 0.629135687732342, "grad_norm": 0.42623874953518437, "learning_rate": 3.205561162875736e-06, "loss": 0.0611, "step": 13539 }, { "epoch": 0.629182156133829, "grad_norm": 0.7321436745543508, "learning_rate": 3.2048588017660286e-06, "loss": 0.062, "step": 13540 }, { "epoch": 0.629228624535316, "grad_norm": 0.36367607588440826, "learning_rate": 3.2041564813181568e-06, "loss": 0.0577, "step": 13541 }, { "epoch": 0.629275092936803, "grad_norm": 0.5032638287386311, "learning_rate": 3.203454201548031e-06, "loss": 0.0456, "step": 13542 }, { "epoch": 0.62932156133829, "grad_norm": 0.6756999088186187, "learning_rate": 3.202751962471558e-06, "loss": 0.0602, "step": 13543 }, { "epoch": 0.6293680297397769, "grad_norm": 0.47814055948448675, "learning_rate": 3.2020497641046443e-06, "loss": 0.0403, "step": 13544 }, { "epoch": 0.6294144981412639, "grad_norm": 0.9823156808584047, "learning_rate": 3.2013476064631964e-06, "loss": 0.0543, "step": 13545 }, { "epoch": 0.629460966542751, "grad_norm": 0.48440633143713024, "learning_rate": 3.20064548956312e-06, "loss": 0.0363, "step": 13546 }, { "epoch": 0.6295074349442379, "grad_norm": 0.5006897933097472, "learning_rate": 3.1999434134203146e-06, "loss": 0.0455, "step": 13547 }, { "epoch": 0.6295539033457249, "grad_norm": 0.6816851442658332, "learning_rate": 3.1992413780506858e-06, "loss": 0.0722, "step": 13548 }, { "epoch": 0.629600371747212, "grad_norm": 1.1234110863042899, "learning_rate": 3.1985393834701352e-06, "loss": 0.0782, "step": 13549 }, { "epoch": 0.6296468401486989, "grad_norm": 0.5149780477355971, "learning_rate": 3.1978374296945635e-06, "loss": 0.038, "step": 13550 }, { "epoch": 0.6296933085501859, "grad_norm": 0.7274103255026394, "learning_rate": 3.197135516739871e-06, "loss": 0.0667, "step": 13551 }, { "epoch": 0.6297397769516728, "grad_norm": 0.5181067488404054, "learning_rate": 3.1964336446219573e-06, "loss": 0.0424, "step": 13552 }, { "epoch": 0.6297862453531599, "grad_norm": 0.4870757443302362, "learning_rate": 3.1957318133567193e-06, "loss": 0.0567, "step": 13553 }, { "epoch": 0.6298327137546469, "grad_norm": 0.6635167533723174, "learning_rate": 3.195030022960054e-06, "loss": 0.0628, "step": 13554 }, { "epoch": 0.6298791821561338, "grad_norm": 0.6085416327629833, "learning_rate": 3.1943282734478585e-06, "loss": 0.055, "step": 13555 }, { "epoch": 0.6299256505576208, "grad_norm": 0.5495491252279118, "learning_rate": 3.1936265648360275e-06, "loss": 0.0356, "step": 13556 }, { "epoch": 0.6299721189591078, "grad_norm": 0.5625083862873974, "learning_rate": 3.1929248971404575e-06, "loss": 0.0624, "step": 13557 }, { "epoch": 0.6300185873605948, "grad_norm": 0.5938556231107722, "learning_rate": 3.192223270377041e-06, "loss": 0.0432, "step": 13558 }, { "epoch": 0.6300650557620818, "grad_norm": 1.0718457322029256, "learning_rate": 3.191521684561668e-06, "loss": 0.0961, "step": 13559 }, { "epoch": 0.6301115241635687, "grad_norm": 1.0495918579475363, "learning_rate": 3.1908201397102332e-06, "loss": 0.059, "step": 13560 }, { "epoch": 0.6301579925650558, "grad_norm": 0.3195986335189628, "learning_rate": 3.190118635838626e-06, "loss": 0.031, "step": 13561 }, { "epoch": 0.6302044609665427, "grad_norm": 0.5651388577000048, "learning_rate": 3.1894171729627367e-06, "loss": 0.0574, "step": 13562 }, { "epoch": 0.6302509293680297, "grad_norm": 0.6313285194481291, "learning_rate": 3.1887157510984533e-06, "loss": 0.0588, "step": 13563 }, { "epoch": 0.6302973977695168, "grad_norm": 0.6673081235708254, "learning_rate": 3.188014370261667e-06, "loss": 0.0527, "step": 13564 }, { "epoch": 0.6303438661710037, "grad_norm": 0.549245265588481, "learning_rate": 3.187313030468261e-06, "loss": 0.0675, "step": 13565 }, { "epoch": 0.6303903345724907, "grad_norm": 0.8031604894495254, "learning_rate": 3.1866117317341218e-06, "loss": 0.0707, "step": 13566 }, { "epoch": 0.6304368029739778, "grad_norm": 0.6878680606142337, "learning_rate": 3.1859104740751356e-06, "loss": 0.0626, "step": 13567 }, { "epoch": 0.6304832713754647, "grad_norm": 0.5293608282112051, "learning_rate": 3.1852092575071864e-06, "loss": 0.0472, "step": 13568 }, { "epoch": 0.6305297397769517, "grad_norm": 0.5395857540729915, "learning_rate": 3.1845080820461583e-06, "loss": 0.0615, "step": 13569 }, { "epoch": 0.6305762081784386, "grad_norm": 0.6090069824882146, "learning_rate": 3.183806947707934e-06, "loss": 0.0603, "step": 13570 }, { "epoch": 0.6306226765799257, "grad_norm": 2.378538001464157, "learning_rate": 3.1831058545083916e-06, "loss": 0.0675, "step": 13571 }, { "epoch": 0.6306691449814127, "grad_norm": 1.2235013179156968, "learning_rate": 3.1824048024634148e-06, "loss": 0.0761, "step": 13572 }, { "epoch": 0.6307156133828996, "grad_norm": 0.484789490966771, "learning_rate": 3.181703791588881e-06, "loss": 0.0405, "step": 13573 }, { "epoch": 0.6307620817843866, "grad_norm": 0.7720161638856529, "learning_rate": 3.181002821900672e-06, "loss": 0.0498, "step": 13574 }, { "epoch": 0.6308085501858736, "grad_norm": 0.8452679345421943, "learning_rate": 3.1803018934146614e-06, "loss": 0.0399, "step": 13575 }, { "epoch": 0.6308550185873606, "grad_norm": 0.526033972759835, "learning_rate": 3.1796010061467307e-06, "loss": 0.0619, "step": 13576 }, { "epoch": 0.6309014869888476, "grad_norm": 0.5780865179958942, "learning_rate": 3.1789001601127523e-06, "loss": 0.0596, "step": 13577 }, { "epoch": 0.6309479553903345, "grad_norm": 0.8505708211663093, "learning_rate": 3.1781993553286007e-06, "loss": 0.0678, "step": 13578 }, { "epoch": 0.6309944237918216, "grad_norm": 0.659315693811861, "learning_rate": 3.177498591810152e-06, "loss": 0.0679, "step": 13579 }, { "epoch": 0.6310408921933085, "grad_norm": 0.5647428706365141, "learning_rate": 3.176797869573277e-06, "loss": 0.0668, "step": 13580 }, { "epoch": 0.6310873605947955, "grad_norm": 0.659491779733421, "learning_rate": 3.176097188633852e-06, "loss": 0.0624, "step": 13581 }, { "epoch": 0.6311338289962826, "grad_norm": 0.6467678921016311, "learning_rate": 3.1753965490077438e-06, "loss": 0.0538, "step": 13582 }, { "epoch": 0.6311802973977695, "grad_norm": 0.508304072502495, "learning_rate": 3.174695950710823e-06, "loss": 0.0399, "step": 13583 }, { "epoch": 0.6312267657992565, "grad_norm": 0.5681469872141431, "learning_rate": 3.173995393758962e-06, "loss": 0.0375, "step": 13584 }, { "epoch": 0.6312732342007435, "grad_norm": 0.55684371016214, "learning_rate": 3.173294878168025e-06, "loss": 0.0594, "step": 13585 }, { "epoch": 0.6313197026022305, "grad_norm": 0.41414535591566387, "learning_rate": 3.1725944039538824e-06, "loss": 0.0596, "step": 13586 }, { "epoch": 0.6313661710037175, "grad_norm": 0.7970100524061088, "learning_rate": 3.171893971132402e-06, "loss": 0.058, "step": 13587 }, { "epoch": 0.6314126394052044, "grad_norm": 0.6618190038858542, "learning_rate": 3.1711935797194447e-06, "loss": 0.06, "step": 13588 }, { "epoch": 0.6314591078066915, "grad_norm": 0.4141943008950463, "learning_rate": 3.1704932297308788e-06, "loss": 0.0356, "step": 13589 }, { "epoch": 0.6315055762081785, "grad_norm": 0.5691110803944852, "learning_rate": 3.169792921182565e-06, "loss": 0.0542, "step": 13590 }, { "epoch": 0.6315520446096654, "grad_norm": 0.49230282506004963, "learning_rate": 3.169092654090369e-06, "loss": 0.0384, "step": 13591 }, { "epoch": 0.6315985130111524, "grad_norm": 0.4886339937802004, "learning_rate": 3.1683924284701512e-06, "loss": 0.0361, "step": 13592 }, { "epoch": 0.6316449814126394, "grad_norm": 0.5138691502350162, "learning_rate": 3.1676922443377743e-06, "loss": 0.0477, "step": 13593 }, { "epoch": 0.6316914498141264, "grad_norm": 0.5314265151568438, "learning_rate": 3.1669921017090953e-06, "loss": 0.0614, "step": 13594 }, { "epoch": 0.6317379182156134, "grad_norm": 0.6131968496945032, "learning_rate": 3.166292000599973e-06, "loss": 0.0634, "step": 13595 }, { "epoch": 0.6317843866171003, "grad_norm": 0.3958906041448583, "learning_rate": 3.1655919410262687e-06, "loss": 0.0584, "step": 13596 }, { "epoch": 0.6318308550185874, "grad_norm": 0.45380244367750355, "learning_rate": 3.1648919230038362e-06, "loss": 0.0561, "step": 13597 }, { "epoch": 0.6318773234200743, "grad_norm": 0.4606567754166818, "learning_rate": 3.164191946548534e-06, "loss": 0.0299, "step": 13598 }, { "epoch": 0.6319237918215613, "grad_norm": 0.7104958986426914, "learning_rate": 3.1634920116762175e-06, "loss": 0.0582, "step": 13599 }, { "epoch": 0.6319702602230484, "grad_norm": 0.3713471511078428, "learning_rate": 3.1627921184027375e-06, "loss": 0.0487, "step": 13600 }, { "epoch": 0.6320167286245353, "grad_norm": 0.37403480274132295, "learning_rate": 3.1620922667439514e-06, "loss": 0.0434, "step": 13601 }, { "epoch": 0.6320631970260223, "grad_norm": 0.5993115749549149, "learning_rate": 3.161392456715708e-06, "loss": 0.0724, "step": 13602 }, { "epoch": 0.6321096654275093, "grad_norm": 0.446371776302039, "learning_rate": 3.1606926883338617e-06, "loss": 0.0508, "step": 13603 }, { "epoch": 0.6321561338289963, "grad_norm": 0.5175588845224547, "learning_rate": 3.159992961614261e-06, "loss": 0.0367, "step": 13604 }, { "epoch": 0.6322026022304833, "grad_norm": 0.6600372102895202, "learning_rate": 3.1592932765727583e-06, "loss": 0.0475, "step": 13605 }, { "epoch": 0.6322490706319702, "grad_norm": 0.4187309048015886, "learning_rate": 3.158593633225199e-06, "loss": 0.0541, "step": 13606 }, { "epoch": 0.6322955390334573, "grad_norm": 0.9740274113748831, "learning_rate": 3.1578940315874317e-06, "loss": 0.0643, "step": 13607 }, { "epoch": 0.6323420074349443, "grad_norm": 0.29196201826082196, "learning_rate": 3.1571944716753034e-06, "loss": 0.0192, "step": 13608 }, { "epoch": 0.6323884758364312, "grad_norm": 0.35064060102839817, "learning_rate": 3.1564949535046597e-06, "loss": 0.0376, "step": 13609 }, { "epoch": 0.6324349442379182, "grad_norm": 0.3893819279178449, "learning_rate": 3.1557954770913457e-06, "loss": 0.0279, "step": 13610 }, { "epoch": 0.6324814126394052, "grad_norm": 0.509793813900462, "learning_rate": 3.1550960424512067e-06, "loss": 0.0363, "step": 13611 }, { "epoch": 0.6325278810408922, "grad_norm": 0.6633445243728227, "learning_rate": 3.154396649600082e-06, "loss": 0.0504, "step": 13612 }, { "epoch": 0.6325743494423792, "grad_norm": 0.7343929640037467, "learning_rate": 3.1536972985538165e-06, "loss": 0.0638, "step": 13613 }, { "epoch": 0.6326208178438661, "grad_norm": 0.6745950254784552, "learning_rate": 3.1529979893282493e-06, "loss": 0.0631, "step": 13614 }, { "epoch": 0.6326672862453532, "grad_norm": 0.6287119653678063, "learning_rate": 3.152298721939223e-06, "loss": 0.0476, "step": 13615 }, { "epoch": 0.6327137546468401, "grad_norm": 1.4436828493271796, "learning_rate": 3.1515994964025743e-06, "loss": 0.0659, "step": 13616 }, { "epoch": 0.6327602230483271, "grad_norm": 1.291893601970111, "learning_rate": 3.1509003127341443e-06, "loss": 0.0563, "step": 13617 }, { "epoch": 0.6328066914498142, "grad_norm": 0.5729994689722412, "learning_rate": 3.150201170949767e-06, "loss": 0.0542, "step": 13618 }, { "epoch": 0.6328531598513011, "grad_norm": 0.5458484444267546, "learning_rate": 3.1495020710652792e-06, "loss": 0.04, "step": 13619 }, { "epoch": 0.6328996282527881, "grad_norm": 0.4885718433044538, "learning_rate": 3.148803013096518e-06, "loss": 0.0698, "step": 13620 }, { "epoch": 0.6329460966542751, "grad_norm": 2.0115642442751582, "learning_rate": 3.1481039970593164e-06, "loss": 0.0524, "step": 13621 }, { "epoch": 0.6329925650557621, "grad_norm": 0.7179710912275231, "learning_rate": 3.1474050229695088e-06, "loss": 0.0595, "step": 13622 }, { "epoch": 0.6330390334572491, "grad_norm": 0.5430968759587822, "learning_rate": 3.146706090842928e-06, "loss": 0.0484, "step": 13623 }, { "epoch": 0.633085501858736, "grad_norm": 0.7180053138604962, "learning_rate": 3.1460072006954034e-06, "loss": 0.0592, "step": 13624 }, { "epoch": 0.633131970260223, "grad_norm": 0.6147940766819375, "learning_rate": 3.1453083525427685e-06, "loss": 0.0568, "step": 13625 }, { "epoch": 0.6331784386617101, "grad_norm": 0.7545642831538881, "learning_rate": 3.1446095464008495e-06, "loss": 0.0719, "step": 13626 }, { "epoch": 0.633224907063197, "grad_norm": 0.7591437789570155, "learning_rate": 3.1439107822854785e-06, "loss": 0.0518, "step": 13627 }, { "epoch": 0.633271375464684, "grad_norm": 0.6232947390912796, "learning_rate": 3.1432120602124803e-06, "loss": 0.0678, "step": 13628 }, { "epoch": 0.633317843866171, "grad_norm": 1.8066629990069116, "learning_rate": 3.142513380197686e-06, "loss": 0.0595, "step": 13629 }, { "epoch": 0.633364312267658, "grad_norm": 0.6682802532164924, "learning_rate": 3.1418147422569166e-06, "loss": 0.0585, "step": 13630 }, { "epoch": 0.633410780669145, "grad_norm": 0.6102036578653964, "learning_rate": 3.1411161464059988e-06, "loss": 0.0587, "step": 13631 }, { "epoch": 0.633457249070632, "grad_norm": 0.43264737964229777, "learning_rate": 3.140417592660757e-06, "loss": 0.039, "step": 13632 }, { "epoch": 0.633503717472119, "grad_norm": 0.8219266852725478, "learning_rate": 3.139719081037014e-06, "loss": 0.0708, "step": 13633 }, { "epoch": 0.6335501858736059, "grad_norm": 0.6412900335065554, "learning_rate": 3.1390206115505916e-06, "loss": 0.071, "step": 13634 }, { "epoch": 0.6335966542750929, "grad_norm": 0.48383548534833276, "learning_rate": 3.1383221842173127e-06, "loss": 0.0343, "step": 13635 }, { "epoch": 0.63364312267658, "grad_norm": 0.8546469119994166, "learning_rate": 3.137623799052993e-06, "loss": 0.049, "step": 13636 }, { "epoch": 0.6336895910780669, "grad_norm": 0.8178930777321679, "learning_rate": 3.1369254560734564e-06, "loss": 0.0581, "step": 13637 }, { "epoch": 0.6337360594795539, "grad_norm": 0.5209392703410246, "learning_rate": 3.1362271552945172e-06, "loss": 0.0516, "step": 13638 }, { "epoch": 0.633782527881041, "grad_norm": 0.6084660614082822, "learning_rate": 3.135528896731996e-06, "loss": 0.0472, "step": 13639 }, { "epoch": 0.6338289962825279, "grad_norm": 0.4939261054207609, "learning_rate": 3.1348306804017083e-06, "loss": 0.05, "step": 13640 }, { "epoch": 0.6338754646840149, "grad_norm": 1.1936292119372423, "learning_rate": 3.134132506319467e-06, "loss": 0.0495, "step": 13641 }, { "epoch": 0.6339219330855018, "grad_norm": 0.4223261827748711, "learning_rate": 3.1334343745010885e-06, "loss": 0.0535, "step": 13642 }, { "epoch": 0.6339684014869889, "grad_norm": 1.0153692086903545, "learning_rate": 3.132736284962386e-06, "loss": 0.0627, "step": 13643 }, { "epoch": 0.6340148698884759, "grad_norm": 0.6929615743551394, "learning_rate": 3.132038237719173e-06, "loss": 0.0632, "step": 13644 }, { "epoch": 0.6340613382899628, "grad_norm": 1.0186608320301458, "learning_rate": 3.1313402327872576e-06, "loss": 0.0742, "step": 13645 }, { "epoch": 0.6341078066914498, "grad_norm": 0.6403799192130486, "learning_rate": 3.1306422701824555e-06, "loss": 0.0613, "step": 13646 }, { "epoch": 0.6341542750929368, "grad_norm": 0.59590904611068, "learning_rate": 3.1299443499205725e-06, "loss": 0.0606, "step": 13647 }, { "epoch": 0.6342007434944238, "grad_norm": 0.7413092489292109, "learning_rate": 3.1292464720174163e-06, "loss": 0.0811, "step": 13648 }, { "epoch": 0.6342472118959108, "grad_norm": 0.3796020091733455, "learning_rate": 3.1285486364887986e-06, "loss": 0.0584, "step": 13649 }, { "epoch": 0.6342936802973977, "grad_norm": 0.6429957101974624, "learning_rate": 3.127850843350523e-06, "loss": 0.0503, "step": 13650 }, { "epoch": 0.6343401486988848, "grad_norm": 0.5081432717641616, "learning_rate": 3.1271530926183963e-06, "loss": 0.0521, "step": 13651 }, { "epoch": 0.6343866171003717, "grad_norm": 0.6456653219125781, "learning_rate": 3.1264553843082247e-06, "loss": 0.0718, "step": 13652 }, { "epoch": 0.6344330855018587, "grad_norm": 0.7306372951233523, "learning_rate": 3.1257577184358085e-06, "loss": 0.0827, "step": 13653 }, { "epoch": 0.6344795539033458, "grad_norm": 0.2331040429622273, "learning_rate": 3.125060095016953e-06, "loss": 0.0215, "step": 13654 }, { "epoch": 0.6345260223048327, "grad_norm": 0.3871267783874856, "learning_rate": 3.1243625140674593e-06, "loss": 0.046, "step": 13655 }, { "epoch": 0.6345724907063197, "grad_norm": 2.0144148992814306, "learning_rate": 3.12366497560313e-06, "loss": 0.0991, "step": 13656 }, { "epoch": 0.6346189591078067, "grad_norm": 0.6036852732764517, "learning_rate": 3.1229674796397626e-06, "loss": 0.0503, "step": 13657 }, { "epoch": 0.6346654275092937, "grad_norm": 0.471516271591132, "learning_rate": 3.1222700261931587e-06, "loss": 0.0628, "step": 13658 }, { "epoch": 0.6347118959107807, "grad_norm": 0.6323971572100513, "learning_rate": 3.121572615279115e-06, "loss": 0.0492, "step": 13659 }, { "epoch": 0.6347583643122676, "grad_norm": 0.44485192712073446, "learning_rate": 3.120875246913426e-06, "loss": 0.0619, "step": 13660 }, { "epoch": 0.6348048327137547, "grad_norm": 0.4532675764974262, "learning_rate": 3.120177921111893e-06, "loss": 0.0524, "step": 13661 }, { "epoch": 0.6348513011152417, "grad_norm": 0.42685938553201946, "learning_rate": 3.1194806378903065e-06, "loss": 0.0455, "step": 13662 }, { "epoch": 0.6348977695167286, "grad_norm": 0.809701704997781, "learning_rate": 3.1187833972644644e-06, "loss": 0.0684, "step": 13663 }, { "epoch": 0.6349442379182156, "grad_norm": 0.27954926549499876, "learning_rate": 3.118086199250159e-06, "loss": 0.0288, "step": 13664 }, { "epoch": 0.6349907063197026, "grad_norm": 0.40560815241348225, "learning_rate": 3.1173890438631805e-06, "loss": 0.0319, "step": 13665 }, { "epoch": 0.6350371747211896, "grad_norm": 0.44395409238136624, "learning_rate": 3.116691931119322e-06, "loss": 0.0625, "step": 13666 }, { "epoch": 0.6350836431226766, "grad_norm": 0.6413437729946705, "learning_rate": 3.115994861034373e-06, "loss": 0.0656, "step": 13667 }, { "epoch": 0.6351301115241635, "grad_norm": 0.4589451392682874, "learning_rate": 3.115297833624123e-06, "loss": 0.0653, "step": 13668 }, { "epoch": 0.6351765799256506, "grad_norm": 0.46136485797236343, "learning_rate": 3.11460084890436e-06, "loss": 0.0526, "step": 13669 }, { "epoch": 0.6352230483271375, "grad_norm": 0.5653110104547877, "learning_rate": 3.113903906890875e-06, "loss": 0.0565, "step": 13670 }, { "epoch": 0.6352695167286245, "grad_norm": 0.5701966388625004, "learning_rate": 3.11320700759945e-06, "loss": 0.0643, "step": 13671 }, { "epoch": 0.6353159851301116, "grad_norm": 0.621124164264914, "learning_rate": 3.112510151045871e-06, "loss": 0.0472, "step": 13672 }, { "epoch": 0.6353624535315985, "grad_norm": 0.6443875059428326, "learning_rate": 3.1118133372459254e-06, "loss": 0.0639, "step": 13673 }, { "epoch": 0.6354089219330855, "grad_norm": 3.1594399532630786, "learning_rate": 3.111116566215393e-06, "loss": 0.0521, "step": 13674 }, { "epoch": 0.6354553903345725, "grad_norm": 0.6700373408046484, "learning_rate": 3.11041983797006e-06, "loss": 0.0671, "step": 13675 }, { "epoch": 0.6355018587360595, "grad_norm": 0.5454375115123675, "learning_rate": 3.109723152525708e-06, "loss": 0.0657, "step": 13676 }, { "epoch": 0.6355483271375465, "grad_norm": 0.9931618544027954, "learning_rate": 3.1090265098981126e-06, "loss": 0.0604, "step": 13677 }, { "epoch": 0.6355947955390334, "grad_norm": 0.3454191170322532, "learning_rate": 3.1083299101030583e-06, "loss": 0.039, "step": 13678 }, { "epoch": 0.6356412639405205, "grad_norm": 0.7766543791477115, "learning_rate": 3.107633353156322e-06, "loss": 0.0503, "step": 13679 }, { "epoch": 0.6356877323420075, "grad_norm": 0.3014605296046089, "learning_rate": 3.106936839073683e-06, "loss": 0.0378, "step": 13680 }, { "epoch": 0.6357342007434944, "grad_norm": 0.5803914582991744, "learning_rate": 3.106240367870916e-06, "loss": 0.073, "step": 13681 }, { "epoch": 0.6357806691449814, "grad_norm": 0.5343325567471475, "learning_rate": 3.1055439395638e-06, "loss": 0.0509, "step": 13682 }, { "epoch": 0.6358271375464684, "grad_norm": 0.49029471002115405, "learning_rate": 3.1048475541681054e-06, "loss": 0.0501, "step": 13683 }, { "epoch": 0.6358736059479554, "grad_norm": 0.704807163238952, "learning_rate": 3.1041512116996086e-06, "loss": 0.0475, "step": 13684 }, { "epoch": 0.6359200743494424, "grad_norm": 0.4277687756470383, "learning_rate": 3.1034549121740826e-06, "loss": 0.0676, "step": 13685 }, { "epoch": 0.6359665427509293, "grad_norm": 0.5026544007526199, "learning_rate": 3.102758655607298e-06, "loss": 0.0311, "step": 13686 }, { "epoch": 0.6360130111524164, "grad_norm": 0.4046659404021472, "learning_rate": 3.102062442015028e-06, "loss": 0.0503, "step": 13687 }, { "epoch": 0.6360594795539033, "grad_norm": 0.47549550406130037, "learning_rate": 3.1013662714130423e-06, "loss": 0.0692, "step": 13688 }, { "epoch": 0.6361059479553903, "grad_norm": 0.5418515551417418, "learning_rate": 3.100670143817106e-06, "loss": 0.0528, "step": 13689 }, { "epoch": 0.6361524163568774, "grad_norm": 0.6471658949191219, "learning_rate": 3.0999740592429917e-06, "loss": 0.0407, "step": 13690 }, { "epoch": 0.6361988847583643, "grad_norm": 0.8952010032815718, "learning_rate": 3.099278017706463e-06, "loss": 0.0584, "step": 13691 }, { "epoch": 0.6362453531598513, "grad_norm": 0.24608024437811793, "learning_rate": 3.0985820192232897e-06, "loss": 0.0249, "step": 13692 }, { "epoch": 0.6362918215613383, "grad_norm": 1.2814218453503678, "learning_rate": 3.0978860638092332e-06, "loss": 0.0798, "step": 13693 }, { "epoch": 0.6363382899628253, "grad_norm": 0.4098167683224143, "learning_rate": 3.097190151480061e-06, "loss": 0.059, "step": 13694 }, { "epoch": 0.6363847583643123, "grad_norm": 0.7424725781836998, "learning_rate": 3.096494282251534e-06, "loss": 0.0618, "step": 13695 }, { "epoch": 0.6364312267657992, "grad_norm": 0.2507341360021941, "learning_rate": 3.095798456139414e-06, "loss": 0.0345, "step": 13696 }, { "epoch": 0.6364776951672863, "grad_norm": 0.4839681681922309, "learning_rate": 3.0951026731594634e-06, "loss": 0.0341, "step": 13697 }, { "epoch": 0.6365241635687733, "grad_norm": 0.48186828631825007, "learning_rate": 3.0944069333274418e-06, "loss": 0.0457, "step": 13698 }, { "epoch": 0.6365706319702602, "grad_norm": 0.6622453454546584, "learning_rate": 3.093711236659111e-06, "loss": 0.0489, "step": 13699 }, { "epoch": 0.6366171003717472, "grad_norm": 0.34245011604860376, "learning_rate": 3.093015583170226e-06, "loss": 0.04, "step": 13700 }, { "epoch": 0.6366635687732342, "grad_norm": 0.6941705359589146, "learning_rate": 3.0923199728765436e-06, "loss": 0.0453, "step": 13701 }, { "epoch": 0.6367100371747212, "grad_norm": 0.506226918058872, "learning_rate": 3.091624405793823e-06, "loss": 0.0602, "step": 13702 }, { "epoch": 0.6367565055762082, "grad_norm": 0.5925593888624314, "learning_rate": 3.090928881937817e-06, "loss": 0.0615, "step": 13703 }, { "epoch": 0.6368029739776951, "grad_norm": 0.6870111507093905, "learning_rate": 3.090233401324282e-06, "loss": 0.0577, "step": 13704 }, { "epoch": 0.6368494423791822, "grad_norm": 0.6478000501558977, "learning_rate": 3.089537963968971e-06, "loss": 0.0701, "step": 13705 }, { "epoch": 0.6368959107806691, "grad_norm": 0.5705359521298918, "learning_rate": 3.0888425698876343e-06, "loss": 0.0576, "step": 13706 }, { "epoch": 0.6369423791821561, "grad_norm": 0.7261535741171913, "learning_rate": 3.088147219096026e-06, "loss": 0.0491, "step": 13707 }, { "epoch": 0.6369888475836432, "grad_norm": 0.43370903028816976, "learning_rate": 3.0874519116098945e-06, "loss": 0.0385, "step": 13708 }, { "epoch": 0.6370353159851301, "grad_norm": 0.694736438374257, "learning_rate": 3.0867566474449907e-06, "loss": 0.0574, "step": 13709 }, { "epoch": 0.6370817843866171, "grad_norm": 0.6821309176621506, "learning_rate": 3.0860614266170614e-06, "loss": 0.0593, "step": 13710 }, { "epoch": 0.6371282527881041, "grad_norm": 0.8488801625403289, "learning_rate": 3.0853662491418573e-06, "loss": 0.0408, "step": 13711 }, { "epoch": 0.6371747211895911, "grad_norm": 0.46188693715119283, "learning_rate": 3.0846711150351216e-06, "loss": 0.0421, "step": 13712 }, { "epoch": 0.6372211895910781, "grad_norm": 0.612408394846058, "learning_rate": 3.0839760243126004e-06, "loss": 0.0498, "step": 13713 }, { "epoch": 0.637267657992565, "grad_norm": 0.38316791601904004, "learning_rate": 3.0832809769900394e-06, "loss": 0.0495, "step": 13714 }, { "epoch": 0.637314126394052, "grad_norm": 0.9696650587714696, "learning_rate": 3.082585973083181e-06, "loss": 0.0663, "step": 13715 }, { "epoch": 0.6373605947955391, "grad_norm": 0.4837062421598778, "learning_rate": 3.0818910126077695e-06, "loss": 0.0569, "step": 13716 }, { "epoch": 0.637407063197026, "grad_norm": 0.6284199253845325, "learning_rate": 3.081196095579546e-06, "loss": 0.0718, "step": 13717 }, { "epoch": 0.637453531598513, "grad_norm": 0.5065016843825535, "learning_rate": 3.0805012220142483e-06, "loss": 0.0623, "step": 13718 }, { "epoch": 0.6375, "grad_norm": 0.5319484310979323, "learning_rate": 3.07980639192762e-06, "loss": 0.0411, "step": 13719 }, { "epoch": 0.637546468401487, "grad_norm": 0.6332718397455731, "learning_rate": 3.0791116053353958e-06, "loss": 0.0563, "step": 13720 }, { "epoch": 0.637592936802974, "grad_norm": 0.5189681724455949, "learning_rate": 3.078416862253317e-06, "loss": 0.0547, "step": 13721 }, { "epoch": 0.6376394052044609, "grad_norm": 0.7150281196129316, "learning_rate": 3.0777221626971178e-06, "loss": 0.0616, "step": 13722 }, { "epoch": 0.637685873605948, "grad_norm": 0.7198657012768833, "learning_rate": 3.077027506682536e-06, "loss": 0.0548, "step": 13723 }, { "epoch": 0.6377323420074349, "grad_norm": 0.521033255685171, "learning_rate": 3.0763328942253046e-06, "loss": 0.0451, "step": 13724 }, { "epoch": 0.6377788104089219, "grad_norm": 0.6628648104257189, "learning_rate": 3.075638325341157e-06, "loss": 0.0606, "step": 13725 }, { "epoch": 0.637825278810409, "grad_norm": 0.4804718298023197, "learning_rate": 3.0749438000458275e-06, "loss": 0.0586, "step": 13726 }, { "epoch": 0.6378717472118959, "grad_norm": 0.6696710435877002, "learning_rate": 3.074249318355046e-06, "loss": 0.0528, "step": 13727 }, { "epoch": 0.6379182156133829, "grad_norm": 0.5874906162634681, "learning_rate": 3.073554880284545e-06, "loss": 0.0598, "step": 13728 }, { "epoch": 0.6379646840148699, "grad_norm": 0.4335963905536728, "learning_rate": 3.072860485850055e-06, "loss": 0.045, "step": 13729 }, { "epoch": 0.6380111524163569, "grad_norm": 0.4964985546085811, "learning_rate": 3.072166135067301e-06, "loss": 0.0625, "step": 13730 }, { "epoch": 0.6380576208178439, "grad_norm": 1.0486268517107116, "learning_rate": 3.071471827952014e-06, "loss": 0.0771, "step": 13731 }, { "epoch": 0.6381040892193308, "grad_norm": 0.3184683595591831, "learning_rate": 3.070777564519919e-06, "loss": 0.0544, "step": 13732 }, { "epoch": 0.6381505576208178, "grad_norm": 0.6643157670221704, "learning_rate": 3.070083344786744e-06, "loss": 0.0744, "step": 13733 }, { "epoch": 0.6381970260223049, "grad_norm": 0.6806099811655482, "learning_rate": 3.069389168768211e-06, "loss": 0.0577, "step": 13734 }, { "epoch": 0.6382434944237918, "grad_norm": 0.6032466584035734, "learning_rate": 3.068695036480047e-06, "loss": 0.0633, "step": 13735 }, { "epoch": 0.6382899628252788, "grad_norm": 0.35261655051220103, "learning_rate": 3.0680009479379724e-06, "loss": 0.0304, "step": 13736 }, { "epoch": 0.6383364312267658, "grad_norm": 0.760137066630292, "learning_rate": 3.0673069031577086e-06, "loss": 0.0658, "step": 13737 }, { "epoch": 0.6383828996282528, "grad_norm": 0.690186273258053, "learning_rate": 3.0666129021549782e-06, "loss": 0.0491, "step": 13738 }, { "epoch": 0.6384293680297398, "grad_norm": 0.4317450321579969, "learning_rate": 3.0659189449455e-06, "loss": 0.0384, "step": 13739 }, { "epoch": 0.6384758364312267, "grad_norm": 0.42797600943491176, "learning_rate": 3.065225031544994e-06, "loss": 0.0389, "step": 13740 }, { "epoch": 0.6385223048327138, "grad_norm": 0.5404075149032741, "learning_rate": 3.064531161969179e-06, "loss": 0.0469, "step": 13741 }, { "epoch": 0.6385687732342007, "grad_norm": 0.43841792393405166, "learning_rate": 3.063837336233767e-06, "loss": 0.0387, "step": 13742 }, { "epoch": 0.6386152416356877, "grad_norm": 0.3826889930526199, "learning_rate": 3.0631435543544792e-06, "loss": 0.0425, "step": 13743 }, { "epoch": 0.6386617100371748, "grad_norm": 0.4052716292829631, "learning_rate": 3.062449816347027e-06, "loss": 0.0399, "step": 13744 }, { "epoch": 0.6387081784386617, "grad_norm": 0.390446621471973, "learning_rate": 3.0617561222271263e-06, "loss": 0.0347, "step": 13745 }, { "epoch": 0.6387546468401487, "grad_norm": 0.4917340851094035, "learning_rate": 3.061062472010489e-06, "loss": 0.0418, "step": 13746 }, { "epoch": 0.6388011152416357, "grad_norm": 0.5599812655735826, "learning_rate": 3.0603688657128295e-06, "loss": 0.0639, "step": 13747 }, { "epoch": 0.6388475836431227, "grad_norm": 0.5027731168933681, "learning_rate": 3.059675303349855e-06, "loss": 0.0551, "step": 13748 }, { "epoch": 0.6388940520446097, "grad_norm": 3.6287340156230847, "learning_rate": 3.058981784937276e-06, "loss": 0.0593, "step": 13749 }, { "epoch": 0.6389405204460966, "grad_norm": 0.429399933486676, "learning_rate": 3.0582883104908045e-06, "loss": 0.0439, "step": 13750 }, { "epoch": 0.6389869888475836, "grad_norm": 0.648174703314231, "learning_rate": 3.0575948800261447e-06, "loss": 0.0791, "step": 13751 }, { "epoch": 0.6390334572490707, "grad_norm": 0.6431131446324537, "learning_rate": 3.056901493559006e-06, "loss": 0.0578, "step": 13752 }, { "epoch": 0.6390799256505576, "grad_norm": 0.6728843227517382, "learning_rate": 3.0562081511050944e-06, "loss": 0.058, "step": 13753 }, { "epoch": 0.6391263940520446, "grad_norm": 0.3560929357743938, "learning_rate": 3.0555148526801127e-06, "loss": 0.0397, "step": 13754 }, { "epoch": 0.6391728624535316, "grad_norm": 0.4753087664363648, "learning_rate": 3.0548215982997664e-06, "loss": 0.0392, "step": 13755 }, { "epoch": 0.6392193308550186, "grad_norm": 0.6702226875532389, "learning_rate": 3.0541283879797578e-06, "loss": 0.0608, "step": 13756 }, { "epoch": 0.6392657992565056, "grad_norm": 0.8452280452925349, "learning_rate": 3.05343522173579e-06, "loss": 0.0651, "step": 13757 }, { "epoch": 0.6393122676579925, "grad_norm": 0.3106689583015533, "learning_rate": 3.0527420995835637e-06, "loss": 0.0412, "step": 13758 }, { "epoch": 0.6393587360594796, "grad_norm": 0.4717306835772735, "learning_rate": 3.0520490215387765e-06, "loss": 0.0399, "step": 13759 }, { "epoch": 0.6394052044609665, "grad_norm": 0.5751564105425393, "learning_rate": 3.0513559876171307e-06, "loss": 0.0572, "step": 13760 }, { "epoch": 0.6394516728624535, "grad_norm": 0.5471197724655547, "learning_rate": 3.050662997834321e-06, "loss": 0.0534, "step": 13761 }, { "epoch": 0.6394981412639406, "grad_norm": 0.39547883924085153, "learning_rate": 3.0499700522060473e-06, "loss": 0.035, "step": 13762 }, { "epoch": 0.6395446096654275, "grad_norm": 0.5388268616511331, "learning_rate": 3.0492771507480034e-06, "loss": 0.0463, "step": 13763 }, { "epoch": 0.6395910780669145, "grad_norm": 0.426837862764517, "learning_rate": 3.048584293475887e-06, "loss": 0.0368, "step": 13764 }, { "epoch": 0.6396375464684015, "grad_norm": 0.423908895117407, "learning_rate": 3.047891480405389e-06, "loss": 0.0658, "step": 13765 }, { "epoch": 0.6396840148698885, "grad_norm": 0.47458517686355434, "learning_rate": 3.0471987115522027e-06, "loss": 0.0467, "step": 13766 }, { "epoch": 0.6397304832713755, "grad_norm": 0.41697368397960327, "learning_rate": 3.0465059869320224e-06, "loss": 0.04, "step": 13767 }, { "epoch": 0.6397769516728624, "grad_norm": 0.544464530496631, "learning_rate": 3.0458133065605362e-06, "loss": 0.044, "step": 13768 }, { "epoch": 0.6398234200743494, "grad_norm": 0.3888856501969643, "learning_rate": 3.045120670453436e-06, "loss": 0.0415, "step": 13769 }, { "epoch": 0.6398698884758365, "grad_norm": 0.56247719807957, "learning_rate": 3.0444280786264114e-06, "loss": 0.0625, "step": 13770 }, { "epoch": 0.6399163568773234, "grad_norm": 0.4283029249952658, "learning_rate": 3.043735531095147e-06, "loss": 0.0437, "step": 13771 }, { "epoch": 0.6399628252788104, "grad_norm": 0.5525693058651994, "learning_rate": 3.043043027875332e-06, "loss": 0.052, "step": 13772 }, { "epoch": 0.6400092936802974, "grad_norm": 0.9636008427299336, "learning_rate": 3.042350568982652e-06, "loss": 0.067, "step": 13773 }, { "epoch": 0.6400557620817844, "grad_norm": 0.4589994452014483, "learning_rate": 3.041658154432793e-06, "loss": 0.04, "step": 13774 }, { "epoch": 0.6401022304832714, "grad_norm": 0.6223858070650671, "learning_rate": 3.0409657842414363e-06, "loss": 0.0554, "step": 13775 }, { "epoch": 0.6401486988847583, "grad_norm": 0.3960153155874531, "learning_rate": 3.0402734584242686e-06, "loss": 0.0508, "step": 13776 }, { "epoch": 0.6401951672862454, "grad_norm": 0.5061261333202173, "learning_rate": 3.039581176996969e-06, "loss": 0.0442, "step": 13777 }, { "epoch": 0.6402416356877323, "grad_norm": 0.43625849111964404, "learning_rate": 3.038888939975218e-06, "loss": 0.056, "step": 13778 }, { "epoch": 0.6402881040892193, "grad_norm": 0.6873645630577645, "learning_rate": 3.0381967473746975e-06, "loss": 0.0597, "step": 13779 }, { "epoch": 0.6403345724907064, "grad_norm": 0.630225767896628, "learning_rate": 3.037504599211084e-06, "loss": 0.0583, "step": 13780 }, { "epoch": 0.6403810408921933, "grad_norm": 0.2187435791029458, "learning_rate": 3.036812495500058e-06, "loss": 0.044, "step": 13781 }, { "epoch": 0.6404275092936803, "grad_norm": 0.6779011524975642, "learning_rate": 3.0361204362572973e-06, "loss": 0.0463, "step": 13782 }, { "epoch": 0.6404739776951673, "grad_norm": 0.35472493039207986, "learning_rate": 3.035428421498472e-06, "loss": 0.0413, "step": 13783 }, { "epoch": 0.6405204460966543, "grad_norm": 0.4815894930429889, "learning_rate": 3.0347364512392624e-06, "loss": 0.0413, "step": 13784 }, { "epoch": 0.6405669144981413, "grad_norm": 0.33061769492778087, "learning_rate": 3.03404452549534e-06, "loss": 0.0477, "step": 13785 }, { "epoch": 0.6406133828996282, "grad_norm": 0.5676734732976347, "learning_rate": 3.033352644282378e-06, "loss": 0.0599, "step": 13786 }, { "epoch": 0.6406598513011152, "grad_norm": 0.6851279740738861, "learning_rate": 3.0326608076160483e-06, "loss": 0.0274, "step": 13787 }, { "epoch": 0.6407063197026023, "grad_norm": 0.5548737965384072, "learning_rate": 3.031969015512024e-06, "loss": 0.0356, "step": 13788 }, { "epoch": 0.6407527881040892, "grad_norm": 0.6043249826556519, "learning_rate": 3.0312772679859712e-06, "loss": 0.0496, "step": 13789 }, { "epoch": 0.6407992565055762, "grad_norm": 0.6186126307494593, "learning_rate": 3.03058556505356e-06, "loss": 0.0439, "step": 13790 }, { "epoch": 0.6408457249070632, "grad_norm": 0.5300198853478671, "learning_rate": 3.0298939067304593e-06, "loss": 0.053, "step": 13791 }, { "epoch": 0.6408921933085502, "grad_norm": 0.41688055122819667, "learning_rate": 3.029202293032334e-06, "loss": 0.0398, "step": 13792 }, { "epoch": 0.6409386617100372, "grad_norm": 0.9211443013870932, "learning_rate": 3.0285107239748523e-06, "loss": 0.0459, "step": 13793 }, { "epoch": 0.6409851301115241, "grad_norm": 0.458015154430968, "learning_rate": 3.0278191995736784e-06, "loss": 0.0532, "step": 13794 }, { "epoch": 0.6410315985130112, "grad_norm": 0.8917958777229914, "learning_rate": 3.0271277198444737e-06, "loss": 0.0538, "step": 13795 }, { "epoch": 0.6410780669144981, "grad_norm": 0.49770451319709125, "learning_rate": 3.026436284802903e-06, "loss": 0.0451, "step": 13796 }, { "epoch": 0.6411245353159851, "grad_norm": 0.5385798705745191, "learning_rate": 3.025744894464627e-06, "loss": 0.0545, "step": 13797 }, { "epoch": 0.6411710037174722, "grad_norm": 0.41894734664844163, "learning_rate": 3.0250535488453077e-06, "loss": 0.0462, "step": 13798 }, { "epoch": 0.6412174721189591, "grad_norm": 0.6564434037177055, "learning_rate": 3.024362247960603e-06, "loss": 0.0524, "step": 13799 }, { "epoch": 0.6412639405204461, "grad_norm": 0.49209113697650314, "learning_rate": 3.0236709918261746e-06, "loss": 0.0356, "step": 13800 }, { "epoch": 0.6413104089219331, "grad_norm": 0.6588705759371256, "learning_rate": 3.0229797804576778e-06, "loss": 0.0594, "step": 13801 }, { "epoch": 0.6413568773234201, "grad_norm": 0.4757078587193669, "learning_rate": 3.0222886138707684e-06, "loss": 0.0497, "step": 13802 }, { "epoch": 0.6414033457249071, "grad_norm": 0.4767313319545236, "learning_rate": 3.021597492081105e-06, "loss": 0.0599, "step": 13803 }, { "epoch": 0.641449814126394, "grad_norm": 0.5907979572571843, "learning_rate": 3.0209064151043386e-06, "loss": 0.0595, "step": 13804 }, { "epoch": 0.641496282527881, "grad_norm": 0.6176745492977314, "learning_rate": 3.0202153829561266e-06, "loss": 0.0565, "step": 13805 }, { "epoch": 0.6415427509293681, "grad_norm": 0.4738071724978368, "learning_rate": 3.0195243956521206e-06, "loss": 0.0623, "step": 13806 }, { "epoch": 0.641589219330855, "grad_norm": 0.4740466762770889, "learning_rate": 3.01883345320797e-06, "loss": 0.0596, "step": 13807 }, { "epoch": 0.641635687732342, "grad_norm": 0.49447845349104114, "learning_rate": 3.018142555639327e-06, "loss": 0.05, "step": 13808 }, { "epoch": 0.641682156133829, "grad_norm": 0.7178619378363752, "learning_rate": 3.0174517029618405e-06, "loss": 0.0468, "step": 13809 }, { "epoch": 0.641728624535316, "grad_norm": 0.6055995818782169, "learning_rate": 3.0167608951911603e-06, "loss": 0.0743, "step": 13810 }, { "epoch": 0.641775092936803, "grad_norm": 0.860399383046201, "learning_rate": 3.016070132342932e-06, "loss": 0.0757, "step": 13811 }, { "epoch": 0.6418215613382899, "grad_norm": 0.842428862674761, "learning_rate": 3.015379414432806e-06, "loss": 0.0499, "step": 13812 }, { "epoch": 0.641868029739777, "grad_norm": 0.7020603535413836, "learning_rate": 3.0146887414764233e-06, "loss": 0.0416, "step": 13813 }, { "epoch": 0.6419144981412639, "grad_norm": 0.6423324267783185, "learning_rate": 3.0139981134894293e-06, "loss": 0.0627, "step": 13814 }, { "epoch": 0.6419609665427509, "grad_norm": 0.3303439543054319, "learning_rate": 3.013307530487469e-06, "loss": 0.0228, "step": 13815 }, { "epoch": 0.642007434944238, "grad_norm": 0.6391284421600462, "learning_rate": 3.0126169924861838e-06, "loss": 0.0605, "step": 13816 }, { "epoch": 0.6420539033457249, "grad_norm": 0.45413184540350293, "learning_rate": 3.0119264995012165e-06, "loss": 0.0541, "step": 13817 }, { "epoch": 0.6421003717472119, "grad_norm": 0.5879288397600922, "learning_rate": 3.0112360515482066e-06, "loss": 0.0708, "step": 13818 }, { "epoch": 0.6421468401486989, "grad_norm": 0.5126909164996457, "learning_rate": 3.0105456486427907e-06, "loss": 0.0659, "step": 13819 }, { "epoch": 0.6421933085501859, "grad_norm": 0.507047076599497, "learning_rate": 3.0098552908006117e-06, "loss": 0.0466, "step": 13820 }, { "epoch": 0.6422397769516729, "grad_norm": 0.7366418848239904, "learning_rate": 3.009164978037304e-06, "loss": 0.0634, "step": 13821 }, { "epoch": 0.6422862453531598, "grad_norm": 0.6569823621293447, "learning_rate": 3.008474710368505e-06, "loss": 0.0654, "step": 13822 }, { "epoch": 0.6423327137546468, "grad_norm": 1.333961280940885, "learning_rate": 3.0077844878098524e-06, "loss": 0.0769, "step": 13823 }, { "epoch": 0.6423791821561339, "grad_norm": 0.7452399990396511, "learning_rate": 3.0070943103769744e-06, "loss": 0.0389, "step": 13824 }, { "epoch": 0.6424256505576208, "grad_norm": 0.5931978868054978, "learning_rate": 3.0064041780855095e-06, "loss": 0.0429, "step": 13825 }, { "epoch": 0.6424721189591078, "grad_norm": 0.6398379811079392, "learning_rate": 3.0057140909510875e-06, "loss": 0.0579, "step": 13826 }, { "epoch": 0.6425185873605948, "grad_norm": 0.45611404279662854, "learning_rate": 3.0050240489893404e-06, "loss": 0.0433, "step": 13827 }, { "epoch": 0.6425650557620818, "grad_norm": 0.5117542216720403, "learning_rate": 3.004334052215898e-06, "loss": 0.0526, "step": 13828 }, { "epoch": 0.6426115241635688, "grad_norm": 0.8250063030839774, "learning_rate": 3.0036441006463914e-06, "loss": 0.0427, "step": 13829 }, { "epoch": 0.6426579925650557, "grad_norm": 0.520967165487309, "learning_rate": 3.0029541942964462e-06, "loss": 0.0297, "step": 13830 }, { "epoch": 0.6427044609665428, "grad_norm": 0.6845527869536652, "learning_rate": 3.002264333181689e-06, "loss": 0.0487, "step": 13831 }, { "epoch": 0.6427509293680297, "grad_norm": 0.22206074071695298, "learning_rate": 3.001574517317749e-06, "loss": 0.0268, "step": 13832 }, { "epoch": 0.6427973977695167, "grad_norm": 0.32172532723094105, "learning_rate": 3.0008847467202472e-06, "loss": 0.0413, "step": 13833 }, { "epoch": 0.6428438661710038, "grad_norm": 0.4768168641143614, "learning_rate": 3.0001950214048124e-06, "loss": 0.0446, "step": 13834 }, { "epoch": 0.6428903345724907, "grad_norm": 0.5910168111512243, "learning_rate": 2.9995053413870657e-06, "loss": 0.0566, "step": 13835 }, { "epoch": 0.6429368029739777, "grad_norm": 0.4133289900440708, "learning_rate": 2.9988157066826255e-06, "loss": 0.0477, "step": 13836 }, { "epoch": 0.6429832713754647, "grad_norm": 0.8912972654142998, "learning_rate": 2.9981261173071175e-06, "loss": 0.0712, "step": 13837 }, { "epoch": 0.6430297397769517, "grad_norm": 1.1011204817563947, "learning_rate": 2.997436573276159e-06, "loss": 0.0521, "step": 13838 }, { "epoch": 0.6430762081784387, "grad_norm": 1.3829721693634849, "learning_rate": 2.9967470746053705e-06, "loss": 0.059, "step": 13839 }, { "epoch": 0.6431226765799256, "grad_norm": 0.5056653366155107, "learning_rate": 2.996057621310369e-06, "loss": 0.0408, "step": 13840 }, { "epoch": 0.6431691449814126, "grad_norm": 0.5981221954476689, "learning_rate": 2.995368213406772e-06, "loss": 0.0511, "step": 13841 }, { "epoch": 0.6432156133828997, "grad_norm": 0.4557248043713705, "learning_rate": 2.994678850910195e-06, "loss": 0.0666, "step": 13842 }, { "epoch": 0.6432620817843866, "grad_norm": 0.4298252845932561, "learning_rate": 2.993989533836251e-06, "loss": 0.0482, "step": 13843 }, { "epoch": 0.6433085501858736, "grad_norm": 0.5347449874638484, "learning_rate": 2.9933002622005563e-06, "loss": 0.0451, "step": 13844 }, { "epoch": 0.6433550185873605, "grad_norm": 1.0683658438100916, "learning_rate": 2.992611036018722e-06, "loss": 0.0587, "step": 13845 }, { "epoch": 0.6434014869888476, "grad_norm": 0.4484401160228189, "learning_rate": 2.991921855306361e-06, "loss": 0.0397, "step": 13846 }, { "epoch": 0.6434479553903346, "grad_norm": 7.274002428756042, "learning_rate": 2.991232720079085e-06, "loss": 0.0695, "step": 13847 }, { "epoch": 0.6434944237918215, "grad_norm": 0.5793338556199848, "learning_rate": 2.9905436303525e-06, "loss": 0.0564, "step": 13848 }, { "epoch": 0.6435408921933086, "grad_norm": 0.7564373856495443, "learning_rate": 2.9898545861422174e-06, "loss": 0.061, "step": 13849 }, { "epoch": 0.6435873605947955, "grad_norm": 0.9568983842104959, "learning_rate": 2.989165587463843e-06, "loss": 0.086, "step": 13850 }, { "epoch": 0.6436338289962825, "grad_norm": 0.33444488126677596, "learning_rate": 2.9884766343329856e-06, "loss": 0.0309, "step": 13851 }, { "epoch": 0.6436802973977696, "grad_norm": 0.48557856102202623, "learning_rate": 2.9877877267652487e-06, "loss": 0.0516, "step": 13852 }, { "epoch": 0.6437267657992565, "grad_norm": 1.0176868700201507, "learning_rate": 2.9870988647762395e-06, "loss": 0.0418, "step": 13853 }, { "epoch": 0.6437732342007435, "grad_norm": 0.8159507239363303, "learning_rate": 2.986410048381558e-06, "loss": 0.0605, "step": 13854 }, { "epoch": 0.6438197026022305, "grad_norm": 1.1093926279368769, "learning_rate": 2.985721277596808e-06, "loss": 0.0849, "step": 13855 }, { "epoch": 0.6438661710037175, "grad_norm": 0.6785879389568311, "learning_rate": 2.985032552437592e-06, "loss": 0.0438, "step": 13856 }, { "epoch": 0.6439126394052045, "grad_norm": 1.6780003463925108, "learning_rate": 2.9843438729195083e-06, "loss": 0.0539, "step": 13857 }, { "epoch": 0.6439591078066914, "grad_norm": 0.6363614586594986, "learning_rate": 2.983655239058158e-06, "loss": 0.0421, "step": 13858 }, { "epoch": 0.6440055762081784, "grad_norm": 0.6084927313623185, "learning_rate": 2.9829666508691397e-06, "loss": 0.0401, "step": 13859 }, { "epoch": 0.6440520446096655, "grad_norm": 0.8754530319889594, "learning_rate": 2.9822781083680473e-06, "loss": 0.0755, "step": 13860 }, { "epoch": 0.6440985130111524, "grad_norm": 0.46983358000601644, "learning_rate": 2.981589611570481e-06, "loss": 0.0554, "step": 13861 }, { "epoch": 0.6441449814126394, "grad_norm": 0.7058648642445429, "learning_rate": 2.9809011604920335e-06, "loss": 0.0556, "step": 13862 }, { "epoch": 0.6441914498141263, "grad_norm": 0.4251969755707493, "learning_rate": 2.9802127551483007e-06, "loss": 0.0341, "step": 13863 }, { "epoch": 0.6442379182156134, "grad_norm": 0.2933461598395522, "learning_rate": 2.9795243955548736e-06, "loss": 0.0283, "step": 13864 }, { "epoch": 0.6442843866171004, "grad_norm": 1.2016590283302941, "learning_rate": 2.978836081727347e-06, "loss": 0.0734, "step": 13865 }, { "epoch": 0.6443308550185873, "grad_norm": 0.4298983282391589, "learning_rate": 2.97814781368131e-06, "loss": 0.0795, "step": 13866 }, { "epoch": 0.6443773234200744, "grad_norm": 0.5657592734386334, "learning_rate": 2.977459591432352e-06, "loss": 0.0396, "step": 13867 }, { "epoch": 0.6444237918215613, "grad_norm": 0.849793847674059, "learning_rate": 2.9767714149960636e-06, "loss": 0.0518, "step": 13868 }, { "epoch": 0.6444702602230483, "grad_norm": 0.4103366343172432, "learning_rate": 2.976083284388031e-06, "loss": 0.0346, "step": 13869 }, { "epoch": 0.6445167286245354, "grad_norm": 0.3578775030859303, "learning_rate": 2.9753951996238427e-06, "loss": 0.0359, "step": 13870 }, { "epoch": 0.6445631970260223, "grad_norm": 0.6196308627655438, "learning_rate": 2.974707160719086e-06, "loss": 0.0577, "step": 13871 }, { "epoch": 0.6446096654275093, "grad_norm": 0.5038590724889948, "learning_rate": 2.9740191676893417e-06, "loss": 0.0656, "step": 13872 }, { "epoch": 0.6446561338289963, "grad_norm": 0.3919492735023735, "learning_rate": 2.973331220550196e-06, "loss": 0.0447, "step": 13873 }, { "epoch": 0.6447026022304833, "grad_norm": 0.673794215679616, "learning_rate": 2.9726433193172298e-06, "loss": 0.0535, "step": 13874 }, { "epoch": 0.6447490706319703, "grad_norm": 0.6430468528827457, "learning_rate": 2.971955464006028e-06, "loss": 0.0562, "step": 13875 }, { "epoch": 0.6447955390334572, "grad_norm": 0.4749569485449305, "learning_rate": 2.9712676546321708e-06, "loss": 0.0474, "step": 13876 }, { "epoch": 0.6448420074349442, "grad_norm": 0.5046758810891216, "learning_rate": 2.970579891211233e-06, "loss": 0.048, "step": 13877 }, { "epoch": 0.6448884758364313, "grad_norm": 0.7093548172395564, "learning_rate": 2.9698921737587984e-06, "loss": 0.0692, "step": 13878 }, { "epoch": 0.6449349442379182, "grad_norm": 0.27165813022602847, "learning_rate": 2.9692045022904416e-06, "loss": 0.0381, "step": 13879 }, { "epoch": 0.6449814126394052, "grad_norm": 0.4755428399479956, "learning_rate": 2.968516876821741e-06, "loss": 0.0573, "step": 13880 }, { "epoch": 0.6450278810408921, "grad_norm": 0.35437480521091014, "learning_rate": 2.967829297368271e-06, "loss": 0.0488, "step": 13881 }, { "epoch": 0.6450743494423792, "grad_norm": 0.538499796748949, "learning_rate": 2.967141763945608e-06, "loss": 0.0428, "step": 13882 }, { "epoch": 0.6451208178438662, "grad_norm": 0.53265693225464, "learning_rate": 2.966454276569322e-06, "loss": 0.0442, "step": 13883 }, { "epoch": 0.6451672862453531, "grad_norm": 0.4684108524984185, "learning_rate": 2.9657668352549863e-06, "loss": 0.0596, "step": 13884 }, { "epoch": 0.6452137546468402, "grad_norm": 0.4065811576646523, "learning_rate": 2.9650794400181737e-06, "loss": 0.0459, "step": 13885 }, { "epoch": 0.6452602230483271, "grad_norm": 0.359408784604363, "learning_rate": 2.964392090874452e-06, "loss": 0.0455, "step": 13886 }, { "epoch": 0.6453066914498141, "grad_norm": 0.359408784604363, "learning_rate": 2.964392090874452e-06, "loss": 0.064, "step": 13887 }, { "epoch": 0.6453531598513012, "grad_norm": 0.4337823179110427, "learning_rate": 2.9637047878393933e-06, "loss": 0.0438, "step": 13888 }, { "epoch": 0.6453996282527881, "grad_norm": 0.6597714590463712, "learning_rate": 2.963017530928566e-06, "loss": 0.038, "step": 13889 }, { "epoch": 0.6454460966542751, "grad_norm": 0.8334409413771745, "learning_rate": 2.9623303201575327e-06, "loss": 0.0706, "step": 13890 }, { "epoch": 0.6454925650557621, "grad_norm": 0.47152965966511345, "learning_rate": 2.9616431555418636e-06, "loss": 0.0563, "step": 13891 }, { "epoch": 0.6455390334572491, "grad_norm": 0.4759847520976067, "learning_rate": 2.9609560370971206e-06, "loss": 0.0409, "step": 13892 }, { "epoch": 0.6455855018587361, "grad_norm": 0.5184313866889868, "learning_rate": 2.9602689648388717e-06, "loss": 0.0545, "step": 13893 }, { "epoch": 0.645631970260223, "grad_norm": 0.5028135971579408, "learning_rate": 2.9595819387826753e-06, "loss": 0.0518, "step": 13894 }, { "epoch": 0.64567843866171, "grad_norm": 0.8758319906581475, "learning_rate": 2.9588949589440986e-06, "loss": 0.0384, "step": 13895 }, { "epoch": 0.6457249070631971, "grad_norm": 0.8847365538609161, "learning_rate": 2.958208025338697e-06, "loss": 0.0323, "step": 13896 }, { "epoch": 0.645771375464684, "grad_norm": 0.5667312518009131, "learning_rate": 2.957521137982032e-06, "loss": 0.0482, "step": 13897 }, { "epoch": 0.645817843866171, "grad_norm": 0.4763482123595753, "learning_rate": 2.9568342968896647e-06, "loss": 0.0412, "step": 13898 }, { "epoch": 0.645864312267658, "grad_norm": 0.7087174937148945, "learning_rate": 2.9561475020771496e-06, "loss": 0.052, "step": 13899 }, { "epoch": 0.645910780669145, "grad_norm": 0.339781022559052, "learning_rate": 2.9554607535600456e-06, "loss": 0.0535, "step": 13900 }, { "epoch": 0.645957249070632, "grad_norm": 0.2561358030585908, "learning_rate": 2.9547740513539082e-06, "loss": 0.0371, "step": 13901 }, { "epoch": 0.6460037174721189, "grad_norm": 0.5127249010713839, "learning_rate": 2.9540873954742892e-06, "loss": 0.0357, "step": 13902 }, { "epoch": 0.646050185873606, "grad_norm": 0.3990285239244932, "learning_rate": 2.9534007859367453e-06, "loss": 0.0478, "step": 13903 }, { "epoch": 0.6460966542750929, "grad_norm": 0.49125093392301994, "learning_rate": 2.9527142227568263e-06, "loss": 0.0553, "step": 13904 }, { "epoch": 0.6461431226765799, "grad_norm": 0.3909803360142752, "learning_rate": 2.952027705950086e-06, "loss": 0.0392, "step": 13905 }, { "epoch": 0.646189591078067, "grad_norm": 0.5334250255080452, "learning_rate": 2.9513412355320724e-06, "loss": 0.0611, "step": 13906 }, { "epoch": 0.6462360594795539, "grad_norm": 0.680610507908506, "learning_rate": 2.9506548115183376e-06, "loss": 0.0513, "step": 13907 }, { "epoch": 0.6462825278810409, "grad_norm": 1.3337791290089114, "learning_rate": 2.9499684339244283e-06, "loss": 0.0714, "step": 13908 }, { "epoch": 0.6463289962825279, "grad_norm": 1.798364226568595, "learning_rate": 2.9492821027658897e-06, "loss": 0.0814, "step": 13909 }, { "epoch": 0.6463754646840149, "grad_norm": 0.5449293640292314, "learning_rate": 2.9485958180582707e-06, "loss": 0.0722, "step": 13910 }, { "epoch": 0.6464219330855019, "grad_norm": 0.999878668401447, "learning_rate": 2.947909579817115e-06, "loss": 0.0802, "step": 13911 }, { "epoch": 0.6464684014869888, "grad_norm": 0.7331863711623822, "learning_rate": 2.947223388057968e-06, "loss": 0.0538, "step": 13912 }, { "epoch": 0.6465148698884758, "grad_norm": 0.7502070535020791, "learning_rate": 2.9465372427963735e-06, "loss": 0.0525, "step": 13913 }, { "epoch": 0.6465613382899629, "grad_norm": 0.5539967905117431, "learning_rate": 2.9458511440478686e-06, "loss": 0.0571, "step": 13914 }, { "epoch": 0.6466078066914498, "grad_norm": 0.5408362930751689, "learning_rate": 2.9451650918279983e-06, "loss": 0.0519, "step": 13915 }, { "epoch": 0.6466542750929368, "grad_norm": 0.55699490185263, "learning_rate": 2.944479086152301e-06, "loss": 0.0722, "step": 13916 }, { "epoch": 0.6467007434944237, "grad_norm": 0.6279222237350862, "learning_rate": 2.9437931270363164e-06, "loss": 0.0643, "step": 13917 }, { "epoch": 0.6467472118959108, "grad_norm": 0.48804488369032584, "learning_rate": 2.9431072144955807e-06, "loss": 0.0625, "step": 13918 }, { "epoch": 0.6467936802973978, "grad_norm": 0.7410358793556636, "learning_rate": 2.942421348545633e-06, "loss": 0.058, "step": 13919 }, { "epoch": 0.6468401486988847, "grad_norm": 0.4600805231411638, "learning_rate": 2.9417355292020065e-06, "loss": 0.0441, "step": 13920 }, { "epoch": 0.6468866171003718, "grad_norm": 0.6647008787783055, "learning_rate": 2.941049756480236e-06, "loss": 0.0738, "step": 13921 }, { "epoch": 0.6469330855018587, "grad_norm": 0.5444357135900815, "learning_rate": 2.940364030395856e-06, "loss": 0.0393, "step": 13922 }, { "epoch": 0.6469795539033457, "grad_norm": 0.3876719843461146, "learning_rate": 2.939678350964398e-06, "loss": 0.0447, "step": 13923 }, { "epoch": 0.6470260223048327, "grad_norm": 0.5511661305157306, "learning_rate": 2.9389927182013933e-06, "loss": 0.0339, "step": 13924 }, { "epoch": 0.6470724907063197, "grad_norm": 0.4114283994819524, "learning_rate": 2.938307132122376e-06, "loss": 0.0531, "step": 13925 }, { "epoch": 0.6471189591078067, "grad_norm": 0.4822180663504569, "learning_rate": 2.9376215927428687e-06, "loss": 0.0487, "step": 13926 }, { "epoch": 0.6471654275092937, "grad_norm": 0.546290408270542, "learning_rate": 2.9369361000784046e-06, "loss": 0.0482, "step": 13927 }, { "epoch": 0.6472118959107807, "grad_norm": 0.4582236994597986, "learning_rate": 2.936250654144508e-06, "loss": 0.0469, "step": 13928 }, { "epoch": 0.6472583643122677, "grad_norm": 0.446914251210146, "learning_rate": 2.9355652549567054e-06, "loss": 0.0562, "step": 13929 }, { "epoch": 0.6473048327137546, "grad_norm": 0.6472896634616353, "learning_rate": 2.934879902530524e-06, "loss": 0.0524, "step": 13930 }, { "epoch": 0.6473513011152416, "grad_norm": 0.6154957666303666, "learning_rate": 2.934194596881488e-06, "loss": 0.0729, "step": 13931 }, { "epoch": 0.6473977695167287, "grad_norm": 0.7566116593499675, "learning_rate": 2.933509338025117e-06, "loss": 0.0507, "step": 13932 }, { "epoch": 0.6474442379182156, "grad_norm": 0.6334651035422206, "learning_rate": 2.932824125976933e-06, "loss": 0.0557, "step": 13933 }, { "epoch": 0.6474907063197026, "grad_norm": 0.6297419230249272, "learning_rate": 2.9321389607524597e-06, "loss": 0.0559, "step": 13934 }, { "epoch": 0.6475371747211895, "grad_norm": 0.8587664562089928, "learning_rate": 2.931453842367214e-06, "loss": 0.0477, "step": 13935 }, { "epoch": 0.6475836431226766, "grad_norm": 0.4680087447979913, "learning_rate": 2.930768770836718e-06, "loss": 0.0496, "step": 13936 }, { "epoch": 0.6476301115241636, "grad_norm": 0.714054216399558, "learning_rate": 2.9300837461764863e-06, "loss": 0.0465, "step": 13937 }, { "epoch": 0.6476765799256505, "grad_norm": 0.5422299467631958, "learning_rate": 2.9293987684020353e-06, "loss": 0.0556, "step": 13938 }, { "epoch": 0.6477230483271376, "grad_norm": 1.0684212535975486, "learning_rate": 2.9287138375288824e-06, "loss": 0.0426, "step": 13939 }, { "epoch": 0.6477695167286245, "grad_norm": 0.40014395761275284, "learning_rate": 2.9280289535725405e-06, "loss": 0.0497, "step": 13940 }, { "epoch": 0.6478159851301115, "grad_norm": 0.5408083773932139, "learning_rate": 2.9273441165485227e-06, "loss": 0.0674, "step": 13941 }, { "epoch": 0.6478624535315985, "grad_norm": 0.8219503084777368, "learning_rate": 2.9266593264723453e-06, "loss": 0.0491, "step": 13942 }, { "epoch": 0.6479089219330855, "grad_norm": 0.8119231380128479, "learning_rate": 2.9259745833595132e-06, "loss": 0.0323, "step": 13943 }, { "epoch": 0.6479553903345725, "grad_norm": 0.75123814926997, "learning_rate": 2.9252898872255416e-06, "loss": 0.0559, "step": 13944 }, { "epoch": 0.6480018587360595, "grad_norm": 0.623999913201275, "learning_rate": 2.924605238085936e-06, "loss": 0.0379, "step": 13945 }, { "epoch": 0.6480483271375465, "grad_norm": 0.40278349558081133, "learning_rate": 2.9239206359562054e-06, "loss": 0.0353, "step": 13946 }, { "epoch": 0.6480947955390335, "grad_norm": 0.9820124444845785, "learning_rate": 2.9232360808518596e-06, "loss": 0.059, "step": 13947 }, { "epoch": 0.6481412639405204, "grad_norm": 0.6823627786626907, "learning_rate": 2.9225515727884025e-06, "loss": 0.0523, "step": 13948 }, { "epoch": 0.6481877323420074, "grad_norm": 0.7856720700389809, "learning_rate": 2.921867111781338e-06, "loss": 0.0713, "step": 13949 }, { "epoch": 0.6482342007434945, "grad_norm": 0.7856720700389809, "learning_rate": 2.921867111781338e-06, "loss": 0.0576, "step": 13950 }, { "epoch": 0.6482806691449814, "grad_norm": 0.591962925183246, "learning_rate": 2.9211826978461697e-06, "loss": 0.069, "step": 13951 }, { "epoch": 0.6483271375464684, "grad_norm": 3.392267766241915, "learning_rate": 2.9204983309984015e-06, "loss": 0.0884, "step": 13952 }, { "epoch": 0.6483736059479553, "grad_norm": 0.5599471983623929, "learning_rate": 2.9198140112535357e-06, "loss": 0.0537, "step": 13953 }, { "epoch": 0.6484200743494424, "grad_norm": 1.2093600355355847, "learning_rate": 2.919129738627071e-06, "loss": 0.0729, "step": 13954 }, { "epoch": 0.6484665427509294, "grad_norm": 0.8616989544937078, "learning_rate": 2.918445513134509e-06, "loss": 0.074, "step": 13955 }, { "epoch": 0.6485130111524163, "grad_norm": 1.4927495804677624, "learning_rate": 2.917761334791346e-06, "loss": 0.0761, "step": 13956 }, { "epoch": 0.6485594795539034, "grad_norm": 0.9593708773651876, "learning_rate": 2.917077203613081e-06, "loss": 0.0652, "step": 13957 }, { "epoch": 0.6486059479553903, "grad_norm": 0.5032041825781083, "learning_rate": 2.9163931196152075e-06, "loss": 0.0436, "step": 13958 }, { "epoch": 0.6486524163568773, "grad_norm": 0.8313539005415995, "learning_rate": 2.915709082813224e-06, "loss": 0.0797, "step": 13959 }, { "epoch": 0.6486988847583643, "grad_norm": 0.7553018086968769, "learning_rate": 2.9150250932226243e-06, "loss": 0.0615, "step": 13960 }, { "epoch": 0.6487453531598513, "grad_norm": 0.5783021524339782, "learning_rate": 2.914341150858902e-06, "loss": 0.0561, "step": 13961 }, { "epoch": 0.6487918215613383, "grad_norm": 0.6496501443192308, "learning_rate": 2.9136572557375444e-06, "loss": 0.0477, "step": 13962 }, { "epoch": 0.6488382899628253, "grad_norm": 0.484081257184577, "learning_rate": 2.9129734078740466e-06, "loss": 0.0427, "step": 13963 }, { "epoch": 0.6488847583643123, "grad_norm": 0.9824159001638304, "learning_rate": 2.912289607283899e-06, "loss": 0.0714, "step": 13964 }, { "epoch": 0.6489312267657993, "grad_norm": 0.7163492932278334, "learning_rate": 2.9116058539825863e-06, "loss": 0.0666, "step": 13965 }, { "epoch": 0.6489776951672862, "grad_norm": 0.590357658459946, "learning_rate": 2.9109221479856015e-06, "loss": 0.072, "step": 13966 }, { "epoch": 0.6490241635687732, "grad_norm": 0.6470469397702355, "learning_rate": 2.910238489308431e-06, "loss": 0.0392, "step": 13967 }, { "epoch": 0.6490706319702603, "grad_norm": 0.6228637385977575, "learning_rate": 2.9095548779665533e-06, "loss": 0.0553, "step": 13968 }, { "epoch": 0.6491171003717472, "grad_norm": 0.8420930210466193, "learning_rate": 2.90887131397546e-06, "loss": 0.0519, "step": 13969 }, { "epoch": 0.6491635687732342, "grad_norm": 0.3676292969110137, "learning_rate": 2.9081877973506323e-06, "loss": 0.0204, "step": 13970 }, { "epoch": 0.6492100371747211, "grad_norm": 0.505454357175063, "learning_rate": 2.9075043281075534e-06, "loss": 0.0429, "step": 13971 }, { "epoch": 0.6492565055762082, "grad_norm": 0.5136899831766795, "learning_rate": 2.906820906261701e-06, "loss": 0.0638, "step": 13972 }, { "epoch": 0.6493029739776952, "grad_norm": 1.857844135913284, "learning_rate": 2.9061375318285624e-06, "loss": 0.0708, "step": 13973 }, { "epoch": 0.6493494423791821, "grad_norm": 0.8662005925452871, "learning_rate": 2.9054542048236102e-06, "loss": 0.0444, "step": 13974 }, { "epoch": 0.6493959107806692, "grad_norm": 0.5310009232320501, "learning_rate": 2.9047709252623223e-06, "loss": 0.0652, "step": 13975 }, { "epoch": 0.6494423791821561, "grad_norm": 2.389260700727436, "learning_rate": 2.9040876931601807e-06, "loss": 0.1156, "step": 13976 }, { "epoch": 0.6494888475836431, "grad_norm": 0.8340454209936958, "learning_rate": 2.903404508532659e-06, "loss": 0.0574, "step": 13977 }, { "epoch": 0.6495353159851301, "grad_norm": 0.5551000083839872, "learning_rate": 2.902721371395232e-06, "loss": 0.0548, "step": 13978 }, { "epoch": 0.6495817843866171, "grad_norm": 0.3736244397607356, "learning_rate": 2.9020382817633717e-06, "loss": 0.0353, "step": 13979 }, { "epoch": 0.6496282527881041, "grad_norm": 0.462172540097184, "learning_rate": 2.901355239652554e-06, "loss": 0.0292, "step": 13980 }, { "epoch": 0.6496747211895911, "grad_norm": 0.5843917661741811, "learning_rate": 2.900672245078249e-06, "loss": 0.0484, "step": 13981 }, { "epoch": 0.649721189591078, "grad_norm": 0.454284053714717, "learning_rate": 2.899989298055924e-06, "loss": 0.0597, "step": 13982 }, { "epoch": 0.6497676579925651, "grad_norm": 1.045224037123475, "learning_rate": 2.8993063986010538e-06, "loss": 0.0487, "step": 13983 }, { "epoch": 0.649814126394052, "grad_norm": 0.8372411742974448, "learning_rate": 2.8986235467291058e-06, "loss": 0.0257, "step": 13984 }, { "epoch": 0.649860594795539, "grad_norm": 0.4915463740281924, "learning_rate": 2.897940742455545e-06, "loss": 0.0429, "step": 13985 }, { "epoch": 0.6499070631970261, "grad_norm": 0.9949884702235425, "learning_rate": 2.897257985795839e-06, "loss": 0.0581, "step": 13986 }, { "epoch": 0.649953531598513, "grad_norm": 0.5209646660282526, "learning_rate": 2.8965752767654527e-06, "loss": 0.0729, "step": 13987 }, { "epoch": 0.65, "grad_norm": 0.6162293101645784, "learning_rate": 2.8958926153798507e-06, "loss": 0.0707, "step": 13988 }, { "epoch": 0.6500464684014869, "grad_norm": 0.837082578097609, "learning_rate": 2.8952100016544926e-06, "loss": 0.0844, "step": 13989 }, { "epoch": 0.650092936802974, "grad_norm": 0.9025332156791207, "learning_rate": 2.894527435604846e-06, "loss": 0.0626, "step": 13990 }, { "epoch": 0.650139405204461, "grad_norm": 1.4605786788168473, "learning_rate": 2.89384491724637e-06, "loss": 0.0482, "step": 13991 }, { "epoch": 0.6501858736059479, "grad_norm": 0.8592820824316637, "learning_rate": 2.8931624465945206e-06, "loss": 0.067, "step": 13992 }, { "epoch": 0.650232342007435, "grad_norm": 0.5663872159131944, "learning_rate": 2.89248002366476e-06, "loss": 0.0709, "step": 13993 }, { "epoch": 0.6502788104089219, "grad_norm": 0.2809065339276609, "learning_rate": 2.8917976484725456e-06, "loss": 0.0386, "step": 13994 }, { "epoch": 0.6503252788104089, "grad_norm": 0.47596377062857825, "learning_rate": 2.891115321033333e-06, "loss": 0.0338, "step": 13995 }, { "epoch": 0.650371747211896, "grad_norm": 0.5755348691595266, "learning_rate": 2.8904330413625786e-06, "loss": 0.066, "step": 13996 }, { "epoch": 0.6504182156133829, "grad_norm": 0.5243690588431477, "learning_rate": 2.8897508094757352e-06, "loss": 0.054, "step": 13997 }, { "epoch": 0.6504646840148699, "grad_norm": 0.46267806702229447, "learning_rate": 2.8890686253882583e-06, "loss": 0.0317, "step": 13998 }, { "epoch": 0.6505111524163569, "grad_norm": 0.5309653584255895, "learning_rate": 2.8883864891155956e-06, "loss": 0.0536, "step": 13999 }, { "epoch": 0.6505576208178439, "grad_norm": 0.6042059274529017, "learning_rate": 2.8877044006732038e-06, "loss": 0.0544, "step": 14000 }, { "epoch": 0.6506040892193309, "grad_norm": 0.5208725092894388, "learning_rate": 2.88702236007653e-06, "loss": 0.0551, "step": 14001 }, { "epoch": 0.6506505576208178, "grad_norm": 1.0425548115494274, "learning_rate": 2.886340367341024e-06, "loss": 0.0551, "step": 14002 }, { "epoch": 0.6506970260223048, "grad_norm": 0.5215469522582291, "learning_rate": 2.885658422482133e-06, "loss": 0.0539, "step": 14003 }, { "epoch": 0.6507434944237919, "grad_norm": 0.45678831412161097, "learning_rate": 2.8849765255153045e-06, "loss": 0.0642, "step": 14004 }, { "epoch": 0.6507899628252788, "grad_norm": 0.526048009395035, "learning_rate": 2.884294676455983e-06, "loss": 0.0498, "step": 14005 }, { "epoch": 0.6508364312267658, "grad_norm": 0.36568656650610104, "learning_rate": 2.8836128753196122e-06, "loss": 0.0365, "step": 14006 }, { "epoch": 0.6508828996282527, "grad_norm": 0.5380812121859583, "learning_rate": 2.8829311221216394e-06, "loss": 0.0483, "step": 14007 }, { "epoch": 0.6509293680297398, "grad_norm": 1.0687863525953478, "learning_rate": 2.8822494168775073e-06, "loss": 0.0543, "step": 14008 }, { "epoch": 0.6509758364312268, "grad_norm": 0.9424302799011203, "learning_rate": 2.8815677596026505e-06, "loss": 0.0564, "step": 14009 }, { "epoch": 0.6510223048327137, "grad_norm": 0.6547045975583798, "learning_rate": 2.8808861503125147e-06, "loss": 0.0538, "step": 14010 }, { "epoch": 0.6510687732342008, "grad_norm": 0.4653202938880638, "learning_rate": 2.8802045890225383e-06, "loss": 0.045, "step": 14011 }, { "epoch": 0.6511152416356877, "grad_norm": 1.0535979547158492, "learning_rate": 2.8795230757481597e-06, "loss": 0.0527, "step": 14012 }, { "epoch": 0.6511617100371747, "grad_norm": 0.5330777828023315, "learning_rate": 2.8788416105048124e-06, "loss": 0.0559, "step": 14013 }, { "epoch": 0.6512081784386617, "grad_norm": 0.4427646838648988, "learning_rate": 2.8781601933079394e-06, "loss": 0.0394, "step": 14014 }, { "epoch": 0.6512546468401487, "grad_norm": 0.5922465801931229, "learning_rate": 2.8774788241729695e-06, "loss": 0.0597, "step": 14015 }, { "epoch": 0.6513011152416357, "grad_norm": 0.5694485966697211, "learning_rate": 2.8767975031153363e-06, "loss": 0.0748, "step": 14016 }, { "epoch": 0.6513475836431227, "grad_norm": 0.6372683296220106, "learning_rate": 2.876116230150475e-06, "loss": 0.0381, "step": 14017 }, { "epoch": 0.6513940520446097, "grad_norm": 0.5342337214719929, "learning_rate": 2.8754350052938174e-06, "loss": 0.0549, "step": 14018 }, { "epoch": 0.6514405204460967, "grad_norm": 0.6364700088065086, "learning_rate": 2.8747538285607936e-06, "loss": 0.0652, "step": 14019 }, { "epoch": 0.6514869888475836, "grad_norm": 0.8433357000227396, "learning_rate": 2.8740726999668313e-06, "loss": 0.0639, "step": 14020 }, { "epoch": 0.6515334572490706, "grad_norm": 0.5535914386658455, "learning_rate": 2.8733916195273603e-06, "loss": 0.0479, "step": 14021 }, { "epoch": 0.6515799256505577, "grad_norm": 0.6766727283560061, "learning_rate": 2.872710587257807e-06, "loss": 0.0587, "step": 14022 }, { "epoch": 0.6516263940520446, "grad_norm": 0.7737920914837872, "learning_rate": 2.872029603173596e-06, "loss": 0.051, "step": 14023 }, { "epoch": 0.6516728624535316, "grad_norm": 0.40842655709751674, "learning_rate": 2.871348667290156e-06, "loss": 0.0551, "step": 14024 }, { "epoch": 0.6517193308550185, "grad_norm": 0.6576542697095653, "learning_rate": 2.8706677796229097e-06, "loss": 0.0513, "step": 14025 }, { "epoch": 0.6517657992565056, "grad_norm": 1.080459507235864, "learning_rate": 2.8699869401872787e-06, "loss": 0.0499, "step": 14026 }, { "epoch": 0.6518122676579926, "grad_norm": 0.8843693773721634, "learning_rate": 2.869306148998685e-06, "loss": 0.0431, "step": 14027 }, { "epoch": 0.6518587360594795, "grad_norm": 1.4940685516323116, "learning_rate": 2.8686254060725505e-06, "loss": 0.0732, "step": 14028 }, { "epoch": 0.6519052044609666, "grad_norm": 1.6064247538089822, "learning_rate": 2.867944711424293e-06, "loss": 0.0667, "step": 14029 }, { "epoch": 0.6519516728624535, "grad_norm": 0.5515355451944488, "learning_rate": 2.86726406506933e-06, "loss": 0.0612, "step": 14030 }, { "epoch": 0.6519981412639405, "grad_norm": 0.38156060334340886, "learning_rate": 2.866583467023083e-06, "loss": 0.0499, "step": 14031 }, { "epoch": 0.6520446096654275, "grad_norm": 0.4044502551939574, "learning_rate": 2.865902917300967e-06, "loss": 0.0425, "step": 14032 }, { "epoch": 0.6520910780669145, "grad_norm": 0.8257723272993908, "learning_rate": 2.8652224159183927e-06, "loss": 0.0623, "step": 14033 }, { "epoch": 0.6521375464684015, "grad_norm": 0.3630056766251782, "learning_rate": 2.864541962890779e-06, "loss": 0.0347, "step": 14034 }, { "epoch": 0.6521840148698885, "grad_norm": 0.853019398704108, "learning_rate": 2.863861558233538e-06, "loss": 0.0739, "step": 14035 }, { "epoch": 0.6522304832713754, "grad_norm": 0.6190099374299475, "learning_rate": 2.86318120196208e-06, "loss": 0.0384, "step": 14036 }, { "epoch": 0.6522769516728625, "grad_norm": 1.4027460600267192, "learning_rate": 2.862500894091815e-06, "loss": 0.0802, "step": 14037 }, { "epoch": 0.6523234200743494, "grad_norm": 0.3121697539737637, "learning_rate": 2.8618206346381592e-06, "loss": 0.0509, "step": 14038 }, { "epoch": 0.6523698884758364, "grad_norm": 0.36097309474477046, "learning_rate": 2.861140423616514e-06, "loss": 0.0395, "step": 14039 }, { "epoch": 0.6524163568773235, "grad_norm": 0.5024396873098567, "learning_rate": 2.860460261042287e-06, "loss": 0.0561, "step": 14040 }, { "epoch": 0.6524628252788104, "grad_norm": 0.342400212903046, "learning_rate": 2.8597801469308895e-06, "loss": 0.0283, "step": 14041 }, { "epoch": 0.6525092936802974, "grad_norm": 0.47620184182554975, "learning_rate": 2.8591000812977245e-06, "loss": 0.0587, "step": 14042 }, { "epoch": 0.6525557620817843, "grad_norm": 0.5661782125130037, "learning_rate": 2.858420064158195e-06, "loss": 0.0595, "step": 14043 }, { "epoch": 0.6526022304832714, "grad_norm": 2.6901384676721642, "learning_rate": 2.857740095527705e-06, "loss": 0.0763, "step": 14044 }, { "epoch": 0.6526486988847584, "grad_norm": 0.6326289841332828, "learning_rate": 2.857060175421657e-06, "loss": 0.073, "step": 14045 }, { "epoch": 0.6526951672862453, "grad_norm": 0.7071931249362284, "learning_rate": 2.8563803038554517e-06, "loss": 0.0471, "step": 14046 }, { "epoch": 0.6527416356877324, "grad_norm": 0.4252315142556644, "learning_rate": 2.8557004808444856e-06, "loss": 0.0401, "step": 14047 }, { "epoch": 0.6527881040892193, "grad_norm": 0.5425864704814795, "learning_rate": 2.8550207064041636e-06, "loss": 0.0522, "step": 14048 }, { "epoch": 0.6528345724907063, "grad_norm": 0.642502189106918, "learning_rate": 2.854340980549879e-06, "loss": 0.0485, "step": 14049 }, { "epoch": 0.6528810408921933, "grad_norm": 1.1129911332572726, "learning_rate": 2.8536613032970305e-06, "loss": 0.0528, "step": 14050 }, { "epoch": 0.6529275092936803, "grad_norm": 0.5316932577437511, "learning_rate": 2.8529816746610117e-06, "loss": 0.0511, "step": 14051 }, { "epoch": 0.6529739776951673, "grad_norm": 0.5775514502172174, "learning_rate": 2.852302094657218e-06, "loss": 0.0619, "step": 14052 }, { "epoch": 0.6530204460966543, "grad_norm": 1.154960346838816, "learning_rate": 2.851622563301042e-06, "loss": 0.0508, "step": 14053 }, { "epoch": 0.6530669144981412, "grad_norm": 0.5468609396343831, "learning_rate": 2.8509430806078746e-06, "loss": 0.0531, "step": 14054 }, { "epoch": 0.6531133828996283, "grad_norm": 0.5851331011258593, "learning_rate": 2.850263646593112e-06, "loss": 0.0608, "step": 14055 }, { "epoch": 0.6531598513011152, "grad_norm": 0.8538991048494795, "learning_rate": 2.849584261272138e-06, "loss": 0.0609, "step": 14056 }, { "epoch": 0.6532063197026022, "grad_norm": 0.8122649826378969, "learning_rate": 2.8489049246603422e-06, "loss": 0.0715, "step": 14057 }, { "epoch": 0.6532527881040893, "grad_norm": 0.8457384517427257, "learning_rate": 2.848225636773116e-06, "loss": 0.0566, "step": 14058 }, { "epoch": 0.6532992565055762, "grad_norm": 0.8083485944055515, "learning_rate": 2.8475463976258433e-06, "loss": 0.072, "step": 14059 }, { "epoch": 0.6533457249070632, "grad_norm": 0.5158738276854484, "learning_rate": 2.846867207233911e-06, "loss": 0.0397, "step": 14060 }, { "epoch": 0.6533921933085501, "grad_norm": 0.45984409310174373, "learning_rate": 2.8461880656127015e-06, "loss": 0.0404, "step": 14061 }, { "epoch": 0.6534386617100372, "grad_norm": 0.6537149899832285, "learning_rate": 2.8455089727775996e-06, "loss": 0.0559, "step": 14062 }, { "epoch": 0.6534851301115242, "grad_norm": 0.4583658871333303, "learning_rate": 2.844829928743987e-06, "loss": 0.036, "step": 14063 }, { "epoch": 0.6535315985130111, "grad_norm": 0.37297128079189823, "learning_rate": 2.8441509335272433e-06, "loss": 0.0465, "step": 14064 }, { "epoch": 0.6535780669144982, "grad_norm": 0.6699751714015866, "learning_rate": 2.8434719871427514e-06, "loss": 0.047, "step": 14065 }, { "epoch": 0.6536245353159851, "grad_norm": 0.6765410117087108, "learning_rate": 2.8427930896058885e-06, "loss": 0.0613, "step": 14066 }, { "epoch": 0.6536710037174721, "grad_norm": 0.40298835073376393, "learning_rate": 2.842114240932033e-06, "loss": 0.0544, "step": 14067 }, { "epoch": 0.6537174721189591, "grad_norm": 0.6814562113397075, "learning_rate": 2.841435441136561e-06, "loss": 0.0585, "step": 14068 }, { "epoch": 0.6537639405204461, "grad_norm": 0.6072487645522434, "learning_rate": 2.8407566902348477e-06, "loss": 0.0486, "step": 14069 }, { "epoch": 0.6538104089219331, "grad_norm": 0.655127612373621, "learning_rate": 2.840077988242268e-06, "loss": 0.082, "step": 14070 }, { "epoch": 0.6538568773234201, "grad_norm": 0.4239490152932181, "learning_rate": 2.8393993351741933e-06, "loss": 0.0359, "step": 14071 }, { "epoch": 0.653903345724907, "grad_norm": 1.342639517080899, "learning_rate": 2.838720731046e-06, "loss": 0.0711, "step": 14072 }, { "epoch": 0.6539498141263941, "grad_norm": 0.5714898915335379, "learning_rate": 2.8380421758730584e-06, "loss": 0.0606, "step": 14073 }, { "epoch": 0.653996282527881, "grad_norm": 0.9681589037294882, "learning_rate": 2.8373636696707336e-06, "loss": 0.045, "step": 14074 }, { "epoch": 0.654042750929368, "grad_norm": 0.5178514491219595, "learning_rate": 2.8366852124543993e-06, "loss": 0.0396, "step": 14075 }, { "epoch": 0.6540892193308551, "grad_norm": 0.6632097276092901, "learning_rate": 2.8360068042394216e-06, "loss": 0.0532, "step": 14076 }, { "epoch": 0.654135687732342, "grad_norm": 0.5734521552850851, "learning_rate": 2.835328445041168e-06, "loss": 0.0718, "step": 14077 }, { "epoch": 0.654182156133829, "grad_norm": 0.5327536632858384, "learning_rate": 2.8346501348750006e-06, "loss": 0.0392, "step": 14078 }, { "epoch": 0.6542286245353159, "grad_norm": 1.2066379947017056, "learning_rate": 2.833971873756291e-06, "loss": 0.0538, "step": 14079 }, { "epoch": 0.654275092936803, "grad_norm": 0.6211972241470455, "learning_rate": 2.8332936617003957e-06, "loss": 0.0387, "step": 14080 }, { "epoch": 0.65432156133829, "grad_norm": 0.4672456731119442, "learning_rate": 2.8326154987226782e-06, "loss": 0.0455, "step": 14081 }, { "epoch": 0.6543680297397769, "grad_norm": 0.3984474437855652, "learning_rate": 2.8319373848385025e-06, "loss": 0.0494, "step": 14082 }, { "epoch": 0.654414498141264, "grad_norm": 0.6821236207897451, "learning_rate": 2.8312593200632265e-06, "loss": 0.0475, "step": 14083 }, { "epoch": 0.6544609665427509, "grad_norm": 0.8556391589962884, "learning_rate": 2.83058130441221e-06, "loss": 0.0495, "step": 14084 }, { "epoch": 0.6545074349442379, "grad_norm": 0.7467887594690609, "learning_rate": 2.82990333790081e-06, "loss": 0.0661, "step": 14085 }, { "epoch": 0.6545539033457249, "grad_norm": 0.4354809593469837, "learning_rate": 2.8292254205443837e-06, "loss": 0.0388, "step": 14086 }, { "epoch": 0.6546003717472119, "grad_norm": 0.6965020588101007, "learning_rate": 2.8285475523582854e-06, "loss": 0.0634, "step": 14087 }, { "epoch": 0.6546468401486989, "grad_norm": 0.6454541374635522, "learning_rate": 2.827869733357869e-06, "loss": 0.067, "step": 14088 }, { "epoch": 0.6546933085501859, "grad_norm": 0.7596792349498479, "learning_rate": 2.8271919635584916e-06, "loss": 0.0722, "step": 14089 }, { "epoch": 0.6547397769516728, "grad_norm": 0.6146282458954544, "learning_rate": 2.8265142429755026e-06, "loss": 0.0448, "step": 14090 }, { "epoch": 0.6547862453531599, "grad_norm": 1.3283673107807925, "learning_rate": 2.8258365716242543e-06, "loss": 0.0695, "step": 14091 }, { "epoch": 0.6548327137546468, "grad_norm": 0.6944128912092393, "learning_rate": 2.825158949520095e-06, "loss": 0.0523, "step": 14092 }, { "epoch": 0.6548791821561338, "grad_norm": 0.6389473763865622, "learning_rate": 2.8244813766783747e-06, "loss": 0.0735, "step": 14093 }, { "epoch": 0.6549256505576209, "grad_norm": 0.5718759502071463, "learning_rate": 2.823803853114441e-06, "loss": 0.0494, "step": 14094 }, { "epoch": 0.6549721189591078, "grad_norm": 0.5063894563741895, "learning_rate": 2.823126378843638e-06, "loss": 0.0409, "step": 14095 }, { "epoch": 0.6550185873605948, "grad_norm": 0.39152531118126965, "learning_rate": 2.8224489538813165e-06, "loss": 0.0362, "step": 14096 }, { "epoch": 0.6550650557620817, "grad_norm": 0.8634142836421773, "learning_rate": 2.8217715782428196e-06, "loss": 0.0765, "step": 14097 }, { "epoch": 0.6551115241635688, "grad_norm": 0.6937222653993242, "learning_rate": 2.8210942519434846e-06, "loss": 0.0607, "step": 14098 }, { "epoch": 0.6551579925650558, "grad_norm": 0.6110228088068539, "learning_rate": 2.820416974998661e-06, "loss": 0.0538, "step": 14099 }, { "epoch": 0.6552044609665427, "grad_norm": 0.4806959103294524, "learning_rate": 2.8197397474236865e-06, "loss": 0.0556, "step": 14100 }, { "epoch": 0.6552509293680298, "grad_norm": 0.3869968311587145, "learning_rate": 2.8190625692339013e-06, "loss": 0.0414, "step": 14101 }, { "epoch": 0.6552973977695167, "grad_norm": 0.45242998900327897, "learning_rate": 2.818385440444642e-06, "loss": 0.0461, "step": 14102 }, { "epoch": 0.6553438661710037, "grad_norm": 0.563222009469884, "learning_rate": 2.817708361071253e-06, "loss": 0.062, "step": 14103 }, { "epoch": 0.6553903345724907, "grad_norm": 0.5319671164124452, "learning_rate": 2.8170313311290642e-06, "loss": 0.067, "step": 14104 }, { "epoch": 0.6554368029739777, "grad_norm": 1.661220078453464, "learning_rate": 2.8163543506334113e-06, "loss": 0.0613, "step": 14105 }, { "epoch": 0.6554832713754647, "grad_norm": 0.6939529324831971, "learning_rate": 2.8156774195996328e-06, "loss": 0.0752, "step": 14106 }, { "epoch": 0.6555297397769517, "grad_norm": 0.5830801217022189, "learning_rate": 2.815000538043059e-06, "loss": 0.0614, "step": 14107 }, { "epoch": 0.6555762081784386, "grad_norm": 0.7323612426003441, "learning_rate": 2.8143237059790233e-06, "loss": 0.0662, "step": 14108 }, { "epoch": 0.6556226765799257, "grad_norm": 0.4976562771156821, "learning_rate": 2.813646923422856e-06, "loss": 0.0506, "step": 14109 }, { "epoch": 0.6556691449814126, "grad_norm": 0.5984659927805427, "learning_rate": 2.8129701903898872e-06, "loss": 0.058, "step": 14110 }, { "epoch": 0.6557156133828996, "grad_norm": 0.5417018541880215, "learning_rate": 2.812293506895445e-06, "loss": 0.0504, "step": 14111 }, { "epoch": 0.6557620817843867, "grad_norm": 0.4034836698491638, "learning_rate": 2.8116168729548553e-06, "loss": 0.0472, "step": 14112 }, { "epoch": 0.6558085501858736, "grad_norm": 0.6280244755638362, "learning_rate": 2.810940288583449e-06, "loss": 0.0441, "step": 14113 }, { "epoch": 0.6558550185873606, "grad_norm": 0.4443180684004409, "learning_rate": 2.8102637537965516e-06, "loss": 0.0384, "step": 14114 }, { "epoch": 0.6559014869888475, "grad_norm": 0.5557935834691188, "learning_rate": 2.8095872686094797e-06, "loss": 0.0589, "step": 14115 }, { "epoch": 0.6559479553903346, "grad_norm": 0.6820316335718714, "learning_rate": 2.8089108330375644e-06, "loss": 0.07, "step": 14116 }, { "epoch": 0.6559944237918216, "grad_norm": 0.5524037637702133, "learning_rate": 2.8082344470961245e-06, "loss": 0.0674, "step": 14117 }, { "epoch": 0.6560408921933085, "grad_norm": 0.5390999323473921, "learning_rate": 2.807558110800481e-06, "loss": 0.0511, "step": 14118 }, { "epoch": 0.6560873605947956, "grad_norm": 0.37382854605925514, "learning_rate": 2.806881824165952e-06, "loss": 0.0484, "step": 14119 }, { "epoch": 0.6561338289962825, "grad_norm": 0.40007048076725266, "learning_rate": 2.806205587207862e-06, "loss": 0.0356, "step": 14120 }, { "epoch": 0.6561802973977695, "grad_norm": 0.3828732893863176, "learning_rate": 2.805529399941522e-06, "loss": 0.06, "step": 14121 }, { "epoch": 0.6562267657992565, "grad_norm": 0.42099723996639243, "learning_rate": 2.8048532623822487e-06, "loss": 0.0555, "step": 14122 }, { "epoch": 0.6562732342007435, "grad_norm": 1.3671945641864844, "learning_rate": 2.8041771745453615e-06, "loss": 0.0587, "step": 14123 }, { "epoch": 0.6563197026022305, "grad_norm": 0.34699859309823483, "learning_rate": 2.8035011364461727e-06, "loss": 0.0291, "step": 14124 }, { "epoch": 0.6563661710037175, "grad_norm": 0.3133039453355814, "learning_rate": 2.8028251480999946e-06, "loss": 0.0217, "step": 14125 }, { "epoch": 0.6564126394052044, "grad_norm": 0.4318841954114753, "learning_rate": 2.80214920952214e-06, "loss": 0.0303, "step": 14126 }, { "epoch": 0.6564591078066915, "grad_norm": 0.3341037529179908, "learning_rate": 2.801473320727918e-06, "loss": 0.0438, "step": 14127 }, { "epoch": 0.6565055762081784, "grad_norm": 0.36133484474828687, "learning_rate": 2.80079748173264e-06, "loss": 0.0272, "step": 14128 }, { "epoch": 0.6565520446096654, "grad_norm": 0.4783020762627447, "learning_rate": 2.800121692551611e-06, "loss": 0.0578, "step": 14129 }, { "epoch": 0.6565985130111525, "grad_norm": 0.5348832966236993, "learning_rate": 2.7994459532001437e-06, "loss": 0.0484, "step": 14130 }, { "epoch": 0.6566449814126394, "grad_norm": 0.3168145895100185, "learning_rate": 2.798770263693541e-06, "loss": 0.0467, "step": 14131 }, { "epoch": 0.6566914498141264, "grad_norm": 0.36802067160506524, "learning_rate": 2.7980946240471085e-06, "loss": 0.0334, "step": 14132 }, { "epoch": 0.6567379182156133, "grad_norm": 1.1412233932480174, "learning_rate": 2.797419034276151e-06, "loss": 0.0817, "step": 14133 }, { "epoch": 0.6567843866171004, "grad_norm": 0.6523286750224556, "learning_rate": 2.79674349439597e-06, "loss": 0.0353, "step": 14134 }, { "epoch": 0.6568308550185874, "grad_norm": 0.42176005747950196, "learning_rate": 2.796068004421868e-06, "loss": 0.0525, "step": 14135 }, { "epoch": 0.6568773234200743, "grad_norm": 0.7060456634734885, "learning_rate": 2.7953925643691425e-06, "loss": 0.0727, "step": 14136 }, { "epoch": 0.6569237918215614, "grad_norm": 0.46406316380566265, "learning_rate": 2.794717174253099e-06, "loss": 0.0652, "step": 14137 }, { "epoch": 0.6569702602230483, "grad_norm": 0.3543487358112784, "learning_rate": 2.7940418340890335e-06, "loss": 0.0441, "step": 14138 }, { "epoch": 0.6570167286245353, "grad_norm": 0.8198796883431259, "learning_rate": 2.7933665438922388e-06, "loss": 0.0695, "step": 14139 }, { "epoch": 0.6570631970260223, "grad_norm": 0.4647789335869784, "learning_rate": 2.792691303678016e-06, "loss": 0.0352, "step": 14140 }, { "epoch": 0.6571096654275093, "grad_norm": 0.5026159774129666, "learning_rate": 2.7920161134616584e-06, "loss": 0.0578, "step": 14141 }, { "epoch": 0.6571561338289963, "grad_norm": 0.7442677586629408, "learning_rate": 2.7913409732584595e-06, "loss": 0.0697, "step": 14142 }, { "epoch": 0.6572026022304833, "grad_norm": 0.6648967479613976, "learning_rate": 2.7906658830837104e-06, "loss": 0.0522, "step": 14143 }, { "epoch": 0.6572490706319702, "grad_norm": 1.3312189543648423, "learning_rate": 2.789990842952708e-06, "loss": 0.0624, "step": 14144 }, { "epoch": 0.6572955390334573, "grad_norm": 0.5470579735861818, "learning_rate": 2.7893158528807373e-06, "loss": 0.0619, "step": 14145 }, { "epoch": 0.6573420074349442, "grad_norm": 0.8474139365703522, "learning_rate": 2.788640912883087e-06, "loss": 0.0675, "step": 14146 }, { "epoch": 0.6573884758364312, "grad_norm": 0.6315184757744731, "learning_rate": 2.7879660229750497e-06, "loss": 0.0712, "step": 14147 }, { "epoch": 0.6574349442379183, "grad_norm": 0.37864437519133604, "learning_rate": 2.7872911831719103e-06, "loss": 0.0409, "step": 14148 }, { "epoch": 0.6574814126394052, "grad_norm": 0.417009537758481, "learning_rate": 2.7866163934889538e-06, "loss": 0.0459, "step": 14149 }, { "epoch": 0.6575278810408922, "grad_norm": 1.1651995105129276, "learning_rate": 2.785941653941466e-06, "loss": 0.07, "step": 14150 }, { "epoch": 0.6575743494423791, "grad_norm": 0.41951787569063564, "learning_rate": 2.78526696454473e-06, "loss": 0.0505, "step": 14151 }, { "epoch": 0.6576208178438662, "grad_norm": 0.5166631941534018, "learning_rate": 2.784592325314028e-06, "loss": 0.0447, "step": 14152 }, { "epoch": 0.6576672862453532, "grad_norm": 0.62881309968059, "learning_rate": 2.78391773626464e-06, "loss": 0.0671, "step": 14153 }, { "epoch": 0.6577137546468401, "grad_norm": 0.5273902061728507, "learning_rate": 2.7832431974118492e-06, "loss": 0.0368, "step": 14154 }, { "epoch": 0.6577602230483272, "grad_norm": 0.5681438740063168, "learning_rate": 2.782568708770933e-06, "loss": 0.0576, "step": 14155 }, { "epoch": 0.6578066914498141, "grad_norm": 0.5804691187974895, "learning_rate": 2.7818942703571694e-06, "loss": 0.0475, "step": 14156 }, { "epoch": 0.6578531598513011, "grad_norm": 0.5784670245790243, "learning_rate": 2.781219882185835e-06, "loss": 0.0594, "step": 14157 }, { "epoch": 0.6578996282527881, "grad_norm": 0.6984157474909473, "learning_rate": 2.780545544272206e-06, "loss": 0.0761, "step": 14158 }, { "epoch": 0.6579460966542751, "grad_norm": 1.154870960476426, "learning_rate": 2.7798712566315556e-06, "loss": 0.0676, "step": 14159 }, { "epoch": 0.6579925650557621, "grad_norm": 0.8052300197045379, "learning_rate": 2.779197019279156e-06, "loss": 0.0521, "step": 14160 }, { "epoch": 0.6580390334572491, "grad_norm": 0.5712893111176832, "learning_rate": 2.7785228322302837e-06, "loss": 0.0486, "step": 14161 }, { "epoch": 0.658085501858736, "grad_norm": 0.5277430715639272, "learning_rate": 2.7778486955002094e-06, "loss": 0.0373, "step": 14162 }, { "epoch": 0.6581319702602231, "grad_norm": 0.6705848525325672, "learning_rate": 2.777174609104197e-06, "loss": 0.0647, "step": 14163 }, { "epoch": 0.65817843866171, "grad_norm": 0.49997343387587334, "learning_rate": 2.7765005730575206e-06, "loss": 0.0649, "step": 14164 }, { "epoch": 0.658224907063197, "grad_norm": 0.6365741395959419, "learning_rate": 2.7758265873754462e-06, "loss": 0.0488, "step": 14165 }, { "epoch": 0.6582713754646841, "grad_norm": 0.5003465298738654, "learning_rate": 2.775152652073242e-06, "loss": 0.0395, "step": 14166 }, { "epoch": 0.658317843866171, "grad_norm": 0.6205164898285535, "learning_rate": 2.7744787671661684e-06, "loss": 0.0459, "step": 14167 }, { "epoch": 0.658364312267658, "grad_norm": 0.4604312281330296, "learning_rate": 2.773804932669497e-06, "loss": 0.049, "step": 14168 }, { "epoch": 0.6584107806691449, "grad_norm": 0.40730775402228525, "learning_rate": 2.7731311485984856e-06, "loss": 0.0464, "step": 14169 }, { "epoch": 0.658457249070632, "grad_norm": 0.4130114056831352, "learning_rate": 2.7724574149683954e-06, "loss": 0.0396, "step": 14170 }, { "epoch": 0.658503717472119, "grad_norm": 0.4853378290098831, "learning_rate": 2.7717837317944905e-06, "loss": 0.0484, "step": 14171 }, { "epoch": 0.6585501858736059, "grad_norm": 0.23113360531404867, "learning_rate": 2.7711100990920305e-06, "loss": 0.0215, "step": 14172 }, { "epoch": 0.658596654275093, "grad_norm": 0.5635762582509576, "learning_rate": 2.7704365168762713e-06, "loss": 0.0392, "step": 14173 }, { "epoch": 0.6586431226765799, "grad_norm": 0.934022686787605, "learning_rate": 2.769762985162472e-06, "loss": 0.0737, "step": 14174 }, { "epoch": 0.6586895910780669, "grad_norm": 0.5578202032150364, "learning_rate": 2.7690895039658886e-06, "loss": 0.0489, "step": 14175 }, { "epoch": 0.6587360594795539, "grad_norm": 0.41493313011510774, "learning_rate": 2.768416073301776e-06, "loss": 0.0525, "step": 14176 }, { "epoch": 0.6587825278810409, "grad_norm": 0.47482869941215916, "learning_rate": 2.767742693185386e-06, "loss": 0.0371, "step": 14177 }, { "epoch": 0.6588289962825279, "grad_norm": 0.37753449627818203, "learning_rate": 2.767069363631975e-06, "loss": 0.0565, "step": 14178 }, { "epoch": 0.6588754646840149, "grad_norm": 0.6927158824353905, "learning_rate": 2.766396084656795e-06, "loss": 0.0596, "step": 14179 }, { "epoch": 0.6589219330855018, "grad_norm": 0.6358743158592116, "learning_rate": 2.765722856275091e-06, "loss": 0.0533, "step": 14180 }, { "epoch": 0.6589684014869889, "grad_norm": 0.46371125269592467, "learning_rate": 2.7650496785021175e-06, "loss": 0.0575, "step": 14181 }, { "epoch": 0.6590148698884758, "grad_norm": 0.5187510404674102, "learning_rate": 2.764376551353121e-06, "loss": 0.0545, "step": 14182 }, { "epoch": 0.6590613382899628, "grad_norm": 0.5715209551027558, "learning_rate": 2.7637034748433488e-06, "loss": 0.0467, "step": 14183 }, { "epoch": 0.6591078066914499, "grad_norm": 0.9042448668078873, "learning_rate": 2.7630304489880444e-06, "loss": 0.0405, "step": 14184 }, { "epoch": 0.6591542750929368, "grad_norm": 0.6570588277343438, "learning_rate": 2.7623574738024593e-06, "loss": 0.064, "step": 14185 }, { "epoch": 0.6592007434944238, "grad_norm": 0.6304957708845961, "learning_rate": 2.7616845493018295e-06, "loss": 0.0579, "step": 14186 }, { "epoch": 0.6592472118959107, "grad_norm": 0.7756177739619109, "learning_rate": 2.7610116755013995e-06, "loss": 0.0543, "step": 14187 }, { "epoch": 0.6592936802973978, "grad_norm": 0.6473432401835735, "learning_rate": 2.760338852416413e-06, "loss": 0.0644, "step": 14188 }, { "epoch": 0.6593401486988848, "grad_norm": 0.4849590216114211, "learning_rate": 2.7596660800621076e-06, "loss": 0.0369, "step": 14189 }, { "epoch": 0.6593866171003717, "grad_norm": 0.3778613767483939, "learning_rate": 2.7589933584537246e-06, "loss": 0.0382, "step": 14190 }, { "epoch": 0.6594330855018588, "grad_norm": 1.4310300110412777, "learning_rate": 2.7583206876065007e-06, "loss": 0.0508, "step": 14191 }, { "epoch": 0.6594795539033457, "grad_norm": 0.8134925220917066, "learning_rate": 2.7576480675356726e-06, "loss": 0.0334, "step": 14192 }, { "epoch": 0.6595260223048327, "grad_norm": 0.7056298061595759, "learning_rate": 2.7569754982564755e-06, "loss": 0.0627, "step": 14193 }, { "epoch": 0.6595724907063197, "grad_norm": 0.5786248992723158, "learning_rate": 2.756302979784142e-06, "loss": 0.0552, "step": 14194 }, { "epoch": 0.6596189591078067, "grad_norm": 0.4970786136195275, "learning_rate": 2.75563051213391e-06, "loss": 0.0514, "step": 14195 }, { "epoch": 0.6596654275092937, "grad_norm": 0.47994654398120357, "learning_rate": 2.7549580953210086e-06, "loss": 0.0589, "step": 14196 }, { "epoch": 0.6597118959107807, "grad_norm": 0.43451651004555203, "learning_rate": 2.754285729360669e-06, "loss": 0.0536, "step": 14197 }, { "epoch": 0.6597583643122676, "grad_norm": 0.43223600830313497, "learning_rate": 2.7536134142681214e-06, "loss": 0.0342, "step": 14198 }, { "epoch": 0.6598048327137547, "grad_norm": 0.5259901108755832, "learning_rate": 2.7529411500585947e-06, "loss": 0.0557, "step": 14199 }, { "epoch": 0.6598513011152416, "grad_norm": 0.7276461021619611, "learning_rate": 2.7522689367473154e-06, "loss": 0.0537, "step": 14200 }, { "epoch": 0.6598977695167286, "grad_norm": 0.4416018757475021, "learning_rate": 2.7515967743495082e-06, "loss": 0.0478, "step": 14201 }, { "epoch": 0.6599442379182157, "grad_norm": 0.583545632847983, "learning_rate": 2.750924662880403e-06, "loss": 0.0421, "step": 14202 }, { "epoch": 0.6599907063197026, "grad_norm": 0.26905484792712103, "learning_rate": 2.7502526023552227e-06, "loss": 0.0292, "step": 14203 }, { "epoch": 0.6600371747211896, "grad_norm": 0.7774722935659675, "learning_rate": 2.749580592789185e-06, "loss": 0.0509, "step": 14204 }, { "epoch": 0.6600836431226765, "grad_norm": 0.7341201346655533, "learning_rate": 2.748908634197518e-06, "loss": 0.0565, "step": 14205 }, { "epoch": 0.6601301115241636, "grad_norm": 0.390998225050674, "learning_rate": 2.748236726595439e-06, "loss": 0.0518, "step": 14206 }, { "epoch": 0.6601765799256506, "grad_norm": 0.5784933384985904, "learning_rate": 2.7475648699981683e-06, "loss": 0.0646, "step": 14207 }, { "epoch": 0.6602230483271375, "grad_norm": 0.6318134657724904, "learning_rate": 2.7468930644209217e-06, "loss": 0.0502, "step": 14208 }, { "epoch": 0.6602695167286246, "grad_norm": 0.6079829360363043, "learning_rate": 2.746221309878923e-06, "loss": 0.0666, "step": 14209 }, { "epoch": 0.6603159851301115, "grad_norm": 0.7347854249825765, "learning_rate": 2.7455496063873816e-06, "loss": 0.0659, "step": 14210 }, { "epoch": 0.6603624535315985, "grad_norm": 0.6354857384516378, "learning_rate": 2.744877953961512e-06, "loss": 0.057, "step": 14211 }, { "epoch": 0.6604089219330855, "grad_norm": 0.3976259376412462, "learning_rate": 2.7442063526165325e-06, "loss": 0.0482, "step": 14212 }, { "epoch": 0.6604553903345725, "grad_norm": 0.48785661271846364, "learning_rate": 2.7435348023676533e-06, "loss": 0.0359, "step": 14213 }, { "epoch": 0.6605018587360595, "grad_norm": 0.6098913108718376, "learning_rate": 2.7428633032300855e-06, "loss": 0.0548, "step": 14214 }, { "epoch": 0.6605483271375465, "grad_norm": 0.43672369704842945, "learning_rate": 2.7421918552190397e-06, "loss": 0.052, "step": 14215 }, { "epoch": 0.6605947955390334, "grad_norm": 0.7239706314999167, "learning_rate": 2.7415204583497245e-06, "loss": 0.0671, "step": 14216 }, { "epoch": 0.6606412639405205, "grad_norm": 0.8421571915320054, "learning_rate": 2.740849112637348e-06, "loss": 0.0609, "step": 14217 }, { "epoch": 0.6606877323420074, "grad_norm": 0.5459672589054914, "learning_rate": 2.740177818097115e-06, "loss": 0.0635, "step": 14218 }, { "epoch": 0.6607342007434944, "grad_norm": 0.5828770404073103, "learning_rate": 2.739506574744235e-06, "loss": 0.0402, "step": 14219 }, { "epoch": 0.6607806691449815, "grad_norm": 0.5623494839320382, "learning_rate": 2.7388353825939096e-06, "loss": 0.0387, "step": 14220 }, { "epoch": 0.6608271375464684, "grad_norm": 0.6663452881693206, "learning_rate": 2.738164241661343e-06, "loss": 0.0697, "step": 14221 }, { "epoch": 0.6608736059479554, "grad_norm": 0.4296242755457827, "learning_rate": 2.7374931519617363e-06, "loss": 0.0515, "step": 14222 }, { "epoch": 0.6609200743494423, "grad_norm": 0.8747514932627974, "learning_rate": 2.736822113510292e-06, "loss": 0.0605, "step": 14223 }, { "epoch": 0.6609665427509294, "grad_norm": 0.5122852268677928, "learning_rate": 2.7361511263222085e-06, "loss": 0.05, "step": 14224 }, { "epoch": 0.6610130111524164, "grad_norm": 0.5452072888372115, "learning_rate": 2.735480190412682e-06, "loss": 0.0606, "step": 14225 }, { "epoch": 0.6610594795539033, "grad_norm": 0.4043174258891237, "learning_rate": 2.734809305796915e-06, "loss": 0.0506, "step": 14226 }, { "epoch": 0.6611059479553903, "grad_norm": 0.4738245342904037, "learning_rate": 2.734138472490103e-06, "loss": 0.0312, "step": 14227 }, { "epoch": 0.6611524163568773, "grad_norm": 0.6012692062158811, "learning_rate": 2.7334676905074355e-06, "loss": 0.0619, "step": 14228 }, { "epoch": 0.6611988847583643, "grad_norm": 0.4610140983700432, "learning_rate": 2.7327969598641125e-06, "loss": 0.0285, "step": 14229 }, { "epoch": 0.6612453531598513, "grad_norm": 0.4183928840173354, "learning_rate": 2.732126280575325e-06, "loss": 0.0543, "step": 14230 }, { "epoch": 0.6612918215613383, "grad_norm": 0.4165391369466843, "learning_rate": 2.731455652656263e-06, "loss": 0.0287, "step": 14231 }, { "epoch": 0.6613382899628253, "grad_norm": 0.4811662453797971, "learning_rate": 2.7307850761221187e-06, "loss": 0.0731, "step": 14232 }, { "epoch": 0.6613847583643123, "grad_norm": 0.5154308488324577, "learning_rate": 2.7301145509880804e-06, "loss": 0.0648, "step": 14233 }, { "epoch": 0.6614312267657992, "grad_norm": 0.6275769722516289, "learning_rate": 2.729444077269337e-06, "loss": 0.0458, "step": 14234 }, { "epoch": 0.6614776951672863, "grad_norm": 0.651827266907606, "learning_rate": 2.728773654981073e-06, "loss": 0.0546, "step": 14235 }, { "epoch": 0.6615241635687732, "grad_norm": 0.5668254133837412, "learning_rate": 2.728103284138478e-06, "loss": 0.0431, "step": 14236 }, { "epoch": 0.6615706319702602, "grad_norm": 0.3523014232820062, "learning_rate": 2.7274329647567342e-06, "loss": 0.0502, "step": 14237 }, { "epoch": 0.6616171003717473, "grad_norm": 0.7360619147943193, "learning_rate": 2.7267626968510264e-06, "loss": 0.0615, "step": 14238 }, { "epoch": 0.6616635687732342, "grad_norm": 0.7641453320281077, "learning_rate": 2.7260924804365352e-06, "loss": 0.0661, "step": 14239 }, { "epoch": 0.6617100371747212, "grad_norm": 0.575398840884278, "learning_rate": 2.725422315528443e-06, "loss": 0.0443, "step": 14240 }, { "epoch": 0.6617565055762081, "grad_norm": 0.6943197656537196, "learning_rate": 2.7247522021419295e-06, "loss": 0.0637, "step": 14241 }, { "epoch": 0.6618029739776952, "grad_norm": 0.6997578746618164, "learning_rate": 2.724082140292171e-06, "loss": 0.0383, "step": 14242 }, { "epoch": 0.6618494423791822, "grad_norm": 0.4236397329822338, "learning_rate": 2.7234121299943497e-06, "loss": 0.0586, "step": 14243 }, { "epoch": 0.6618959107806691, "grad_norm": 0.45255161884151757, "learning_rate": 2.7227421712636415e-06, "loss": 0.0433, "step": 14244 }, { "epoch": 0.6619423791821561, "grad_norm": 0.449869336318779, "learning_rate": 2.722072264115216e-06, "loss": 0.0391, "step": 14245 }, { "epoch": 0.6619888475836431, "grad_norm": 0.805642380593027, "learning_rate": 2.7214024085642533e-06, "loss": 0.0737, "step": 14246 }, { "epoch": 0.6620353159851301, "grad_norm": 0.6709173994610975, "learning_rate": 2.7207326046259236e-06, "loss": 0.066, "step": 14247 }, { "epoch": 0.6620817843866171, "grad_norm": 1.4064532461077857, "learning_rate": 2.7200628523154e-06, "loss": 0.0641, "step": 14248 }, { "epoch": 0.662128252788104, "grad_norm": 0.6713830934958305, "learning_rate": 2.71939315164785e-06, "loss": 0.0561, "step": 14249 }, { "epoch": 0.6621747211895911, "grad_norm": 0.3517890523200586, "learning_rate": 2.7187235026384496e-06, "loss": 0.0312, "step": 14250 }, { "epoch": 0.6622211895910781, "grad_norm": 0.48809293663118714, "learning_rate": 2.71805390530236e-06, "loss": 0.0456, "step": 14251 }, { "epoch": 0.662267657992565, "grad_norm": 0.6631676054541547, "learning_rate": 2.7173843596547507e-06, "loss": 0.0649, "step": 14252 }, { "epoch": 0.6623141263940521, "grad_norm": 0.4720118103353028, "learning_rate": 2.7167148657107888e-06, "loss": 0.0455, "step": 14253 }, { "epoch": 0.662360594795539, "grad_norm": 0.6854621323782837, "learning_rate": 2.716045423485639e-06, "loss": 0.0618, "step": 14254 }, { "epoch": 0.662407063197026, "grad_norm": 0.4456539073235221, "learning_rate": 2.715376032994464e-06, "loss": 0.0257, "step": 14255 }, { "epoch": 0.6624535315985131, "grad_norm": 0.36999723487000236, "learning_rate": 2.7147066942524264e-06, "loss": 0.041, "step": 14256 }, { "epoch": 0.6625, "grad_norm": 0.43102626955627077, "learning_rate": 2.7140374072746877e-06, "loss": 0.0423, "step": 14257 }, { "epoch": 0.662546468401487, "grad_norm": 0.9780520533543989, "learning_rate": 2.713368172076407e-06, "loss": 0.07, "step": 14258 }, { "epoch": 0.6625929368029739, "grad_norm": 0.4741430479832639, "learning_rate": 2.7126989886727424e-06, "loss": 0.048, "step": 14259 }, { "epoch": 0.662639405204461, "grad_norm": 0.5029938217203991, "learning_rate": 2.7120298570788544e-06, "loss": 0.0466, "step": 14260 }, { "epoch": 0.662685873605948, "grad_norm": 0.45809785408094106, "learning_rate": 2.7113607773098988e-06, "loss": 0.0608, "step": 14261 }, { "epoch": 0.6627323420074349, "grad_norm": 0.6596629012765546, "learning_rate": 2.7106917493810303e-06, "loss": 0.0581, "step": 14262 }, { "epoch": 0.662778810408922, "grad_norm": 0.4661299359139475, "learning_rate": 2.7100227733074025e-06, "loss": 0.0551, "step": 14263 }, { "epoch": 0.6628252788104089, "grad_norm": 0.3588430635456308, "learning_rate": 2.7093538491041694e-06, "loss": 0.0504, "step": 14264 }, { "epoch": 0.6628717472118959, "grad_norm": 0.4600739561863425, "learning_rate": 2.708684976786483e-06, "loss": 0.0488, "step": 14265 }, { "epoch": 0.6629182156133829, "grad_norm": 0.8183273994785755, "learning_rate": 2.7080161563694905e-06, "loss": 0.0554, "step": 14266 }, { "epoch": 0.6629646840148699, "grad_norm": 0.5767128713541252, "learning_rate": 2.7073473878683476e-06, "loss": 0.0605, "step": 14267 }, { "epoch": 0.6630111524163569, "grad_norm": 0.6278406552869923, "learning_rate": 2.7066786712982e-06, "loss": 0.0553, "step": 14268 }, { "epoch": 0.6630576208178439, "grad_norm": 0.36341017708279644, "learning_rate": 2.7060100066741913e-06, "loss": 0.0404, "step": 14269 }, { "epoch": 0.6631040892193308, "grad_norm": 0.4083982017504708, "learning_rate": 2.7053413940114716e-06, "loss": 0.036, "step": 14270 }, { "epoch": 0.6631505576208179, "grad_norm": 0.5669710268940227, "learning_rate": 2.704672833325184e-06, "loss": 0.0545, "step": 14271 }, { "epoch": 0.6631970260223048, "grad_norm": 0.7643661886281015, "learning_rate": 2.7040043246304736e-06, "loss": 0.064, "step": 14272 }, { "epoch": 0.6632434944237918, "grad_norm": 0.44667453633180926, "learning_rate": 2.7033358679424794e-06, "loss": 0.0346, "step": 14273 }, { "epoch": 0.6632899628252789, "grad_norm": 0.9495337350280906, "learning_rate": 2.7026674632763485e-06, "loss": 0.0405, "step": 14274 }, { "epoch": 0.6633364312267658, "grad_norm": 0.6379588502298663, "learning_rate": 2.7019991106472154e-06, "loss": 0.0506, "step": 14275 }, { "epoch": 0.6633828996282528, "grad_norm": 0.8060422414634073, "learning_rate": 2.7013308100702195e-06, "loss": 0.0611, "step": 14276 }, { "epoch": 0.6634293680297397, "grad_norm": 0.7066858622301622, "learning_rate": 2.700662561560502e-06, "loss": 0.0528, "step": 14277 }, { "epoch": 0.6634758364312268, "grad_norm": 0.7129823387566266, "learning_rate": 2.699994365133198e-06, "loss": 0.0635, "step": 14278 }, { "epoch": 0.6635223048327138, "grad_norm": 0.512199878726687, "learning_rate": 2.6993262208034416e-06, "loss": 0.049, "step": 14279 }, { "epoch": 0.6635687732342007, "grad_norm": 0.45887665224505914, "learning_rate": 2.6986581285863677e-06, "loss": 0.0416, "step": 14280 }, { "epoch": 0.6636152416356877, "grad_norm": 0.6811633672804508, "learning_rate": 2.69799008849711e-06, "loss": 0.055, "step": 14281 }, { "epoch": 0.6636617100371747, "grad_norm": 0.8833355756553603, "learning_rate": 2.697322100550799e-06, "loss": 0.0632, "step": 14282 }, { "epoch": 0.6637081784386617, "grad_norm": 0.7653318194068243, "learning_rate": 2.6966541647625644e-06, "loss": 0.0636, "step": 14283 }, { "epoch": 0.6637546468401487, "grad_norm": 0.4303999797095672, "learning_rate": 2.6959862811475385e-06, "loss": 0.0535, "step": 14284 }, { "epoch": 0.6638011152416357, "grad_norm": 0.6725561145518186, "learning_rate": 2.6953184497208485e-06, "loss": 0.0593, "step": 14285 }, { "epoch": 0.6638475836431227, "grad_norm": 0.5862388075957791, "learning_rate": 2.6946506704976215e-06, "loss": 0.048, "step": 14286 }, { "epoch": 0.6638940520446097, "grad_norm": 0.569580916907808, "learning_rate": 2.6939829434929834e-06, "loss": 0.0602, "step": 14287 }, { "epoch": 0.6639405204460966, "grad_norm": 0.5100435135632894, "learning_rate": 2.6933152687220583e-06, "loss": 0.0532, "step": 14288 }, { "epoch": 0.6639869888475837, "grad_norm": 0.7843278833761874, "learning_rate": 2.6926476461999693e-06, "loss": 0.052, "step": 14289 }, { "epoch": 0.6640334572490706, "grad_norm": 0.5195323517837439, "learning_rate": 2.6919800759418393e-06, "loss": 0.0498, "step": 14290 }, { "epoch": 0.6640799256505576, "grad_norm": 0.4254061360402809, "learning_rate": 2.6913125579627924e-06, "loss": 0.0551, "step": 14291 }, { "epoch": 0.6641263940520447, "grad_norm": 0.37803932748457975, "learning_rate": 2.6906450922779448e-06, "loss": 0.0423, "step": 14292 }, { "epoch": 0.6641728624535316, "grad_norm": 0.6170359187471667, "learning_rate": 2.6899776789024134e-06, "loss": 0.0493, "step": 14293 }, { "epoch": 0.6642193308550186, "grad_norm": 0.758855802067942, "learning_rate": 2.6893103178513213e-06, "loss": 0.0563, "step": 14294 }, { "epoch": 0.6642657992565055, "grad_norm": 1.2228583302413116, "learning_rate": 2.688643009139783e-06, "loss": 0.0872, "step": 14295 }, { "epoch": 0.6643122676579926, "grad_norm": 0.4996573535416577, "learning_rate": 2.6879757527829126e-06, "loss": 0.0282, "step": 14296 }, { "epoch": 0.6643587360594796, "grad_norm": 0.5922285764169615, "learning_rate": 2.687308548795825e-06, "loss": 0.0366, "step": 14297 }, { "epoch": 0.6644052044609665, "grad_norm": 0.4820273917409333, "learning_rate": 2.686641397193632e-06, "loss": 0.0527, "step": 14298 }, { "epoch": 0.6644516728624535, "grad_norm": 0.5754395717533316, "learning_rate": 2.685974297991447e-06, "loss": 0.0524, "step": 14299 }, { "epoch": 0.6644981412639405, "grad_norm": 0.564783451941539, "learning_rate": 2.6853072512043777e-06, "loss": 0.0518, "step": 14300 }, { "epoch": 0.6645446096654275, "grad_norm": 0.619108991618755, "learning_rate": 2.684640256847537e-06, "loss": 0.0707, "step": 14301 }, { "epoch": 0.6645910780669145, "grad_norm": 0.7075297658658559, "learning_rate": 2.6839733149360316e-06, "loss": 0.0689, "step": 14302 }, { "epoch": 0.6646375464684015, "grad_norm": 0.5805067655572036, "learning_rate": 2.6833064254849684e-06, "loss": 0.0543, "step": 14303 }, { "epoch": 0.6646840148698885, "grad_norm": 0.7168906586822436, "learning_rate": 2.6826395885094526e-06, "loss": 0.0626, "step": 14304 }, { "epoch": 0.6647304832713755, "grad_norm": 0.7005162878673148, "learning_rate": 2.6819728040245884e-06, "loss": 0.0497, "step": 14305 }, { "epoch": 0.6647769516728624, "grad_norm": 1.0501906195547643, "learning_rate": 2.681306072045481e-06, "loss": 0.0453, "step": 14306 }, { "epoch": 0.6648234200743495, "grad_norm": 0.3432775877170876, "learning_rate": 2.680639392587229e-06, "loss": 0.0557, "step": 14307 }, { "epoch": 0.6648698884758364, "grad_norm": 0.44785758561481065, "learning_rate": 2.6799727656649378e-06, "loss": 0.0449, "step": 14308 }, { "epoch": 0.6649163568773234, "grad_norm": 0.6192991458735193, "learning_rate": 2.679306191293707e-06, "loss": 0.0548, "step": 14309 }, { "epoch": 0.6649628252788105, "grad_norm": 1.479271732836431, "learning_rate": 2.6786396694886295e-06, "loss": 0.068, "step": 14310 }, { "epoch": 0.6650092936802974, "grad_norm": 0.5457185661999167, "learning_rate": 2.6779732002648094e-06, "loss": 0.0599, "step": 14311 }, { "epoch": 0.6650557620817844, "grad_norm": 0.47278994160533755, "learning_rate": 2.6773067836373397e-06, "loss": 0.0489, "step": 14312 }, { "epoch": 0.6651022304832713, "grad_norm": 0.44439499702040147, "learning_rate": 2.6766404196213157e-06, "loss": 0.0455, "step": 14313 }, { "epoch": 0.6651486988847584, "grad_norm": 0.4693445392617165, "learning_rate": 2.6759741082318295e-06, "loss": 0.0362, "step": 14314 }, { "epoch": 0.6651951672862454, "grad_norm": 0.5853581102113503, "learning_rate": 2.6753078494839803e-06, "loss": 0.0562, "step": 14315 }, { "epoch": 0.6652416356877323, "grad_norm": 1.1636001763101596, "learning_rate": 2.6746416433928524e-06, "loss": 0.0741, "step": 14316 }, { "epoch": 0.6652881040892193, "grad_norm": 0.5637821616977272, "learning_rate": 2.673975489973537e-06, "loss": 0.0542, "step": 14317 }, { "epoch": 0.6653345724907063, "grad_norm": 0.512529635198778, "learning_rate": 2.6733093892411265e-06, "loss": 0.0458, "step": 14318 }, { "epoch": 0.6653810408921933, "grad_norm": 0.7195068366254971, "learning_rate": 2.672643341210708e-06, "loss": 0.0662, "step": 14319 }, { "epoch": 0.6654275092936803, "grad_norm": 0.58218342133411, "learning_rate": 2.6719773458973665e-06, "loss": 0.0417, "step": 14320 }, { "epoch": 0.6654739776951673, "grad_norm": 0.637912198115672, "learning_rate": 2.6713114033161886e-06, "loss": 0.0613, "step": 14321 }, { "epoch": 0.6655204460966543, "grad_norm": 0.54577866465102, "learning_rate": 2.6706455134822583e-06, "loss": 0.0431, "step": 14322 }, { "epoch": 0.6655669144981413, "grad_norm": 0.5808158128391283, "learning_rate": 2.6699796764106588e-06, "loss": 0.0482, "step": 14323 }, { "epoch": 0.6656133828996282, "grad_norm": 0.6656728817272909, "learning_rate": 2.6693138921164696e-06, "loss": 0.0668, "step": 14324 }, { "epoch": 0.6656598513011153, "grad_norm": 0.8579475355595158, "learning_rate": 2.6686481606147765e-06, "loss": 0.0727, "step": 14325 }, { "epoch": 0.6657063197026022, "grad_norm": 0.6859880872089439, "learning_rate": 2.6679824819206558e-06, "loss": 0.0634, "step": 14326 }, { "epoch": 0.6657527881040892, "grad_norm": 0.7346508759274366, "learning_rate": 2.667316856049186e-06, "loss": 0.0585, "step": 14327 }, { "epoch": 0.6657992565055763, "grad_norm": 0.7346508759274366, "learning_rate": 2.667316856049186e-06, "loss": 0.052, "step": 14328 }, { "epoch": 0.6658457249070632, "grad_norm": 0.5362379186550315, "learning_rate": 2.6666512830154446e-06, "loss": 0.0468, "step": 14329 }, { "epoch": 0.6658921933085502, "grad_norm": 0.4640073303203558, "learning_rate": 2.6659857628345077e-06, "loss": 0.0388, "step": 14330 }, { "epoch": 0.6659386617100371, "grad_norm": 0.7061498574528495, "learning_rate": 2.6653202955214497e-06, "loss": 0.0398, "step": 14331 }, { "epoch": 0.6659851301115242, "grad_norm": 0.4023173540384849, "learning_rate": 2.664654881091342e-06, "loss": 0.046, "step": 14332 }, { "epoch": 0.6660315985130112, "grad_norm": 0.8523325887998261, "learning_rate": 2.663989519559261e-06, "loss": 0.0607, "step": 14333 }, { "epoch": 0.6660780669144981, "grad_norm": 0.5495078053463529, "learning_rate": 2.6633242109402785e-06, "loss": 0.0568, "step": 14334 }, { "epoch": 0.6661245353159851, "grad_norm": 0.4826019819720168, "learning_rate": 2.6626589552494574e-06, "loss": 0.0481, "step": 14335 }, { "epoch": 0.6661710037174721, "grad_norm": 0.585060749616693, "learning_rate": 2.6619937525018737e-06, "loss": 0.049, "step": 14336 }, { "epoch": 0.6662174721189591, "grad_norm": 0.38083738738938044, "learning_rate": 2.661328602712592e-06, "loss": 0.0475, "step": 14337 }, { "epoch": 0.6662639405204461, "grad_norm": 0.2676876642494389, "learning_rate": 2.6606635058966782e-06, "loss": 0.0315, "step": 14338 }, { "epoch": 0.666310408921933, "grad_norm": 0.5426391409450374, "learning_rate": 2.6599984620691967e-06, "loss": 0.0454, "step": 14339 }, { "epoch": 0.6663568773234201, "grad_norm": 1.080691943019154, "learning_rate": 2.659333471245217e-06, "loss": 0.0681, "step": 14340 }, { "epoch": 0.6664033457249071, "grad_norm": 0.37173227594280017, "learning_rate": 2.6586685334397953e-06, "loss": 0.0491, "step": 14341 }, { "epoch": 0.666449814126394, "grad_norm": 0.7621191403457956, "learning_rate": 2.658003648667994e-06, "loss": 0.0586, "step": 14342 }, { "epoch": 0.6664962825278811, "grad_norm": 0.5491419838030384, "learning_rate": 2.657338816944876e-06, "loss": 0.0587, "step": 14343 }, { "epoch": 0.666542750929368, "grad_norm": 2.3392439835920213, "learning_rate": 2.6566740382855005e-06, "loss": 0.0659, "step": 14344 }, { "epoch": 0.666589219330855, "grad_norm": 0.5196934133400274, "learning_rate": 2.6560093127049237e-06, "loss": 0.0499, "step": 14345 }, { "epoch": 0.666635687732342, "grad_norm": 1.4945921763323855, "learning_rate": 2.6553446402182024e-06, "loss": 0.0647, "step": 14346 }, { "epoch": 0.666682156133829, "grad_norm": 0.9676256046468648, "learning_rate": 2.654680020840394e-06, "loss": 0.0384, "step": 14347 }, { "epoch": 0.666728624535316, "grad_norm": 0.5940286437856748, "learning_rate": 2.6540154545865503e-06, "loss": 0.0562, "step": 14348 }, { "epoch": 0.6667750929368029, "grad_norm": 0.5714996004150521, "learning_rate": 2.653350941471725e-06, "loss": 0.0477, "step": 14349 }, { "epoch": 0.66682156133829, "grad_norm": 0.6433822004706587, "learning_rate": 2.6526864815109722e-06, "loss": 0.0485, "step": 14350 }, { "epoch": 0.666868029739777, "grad_norm": 0.5673174750197995, "learning_rate": 2.6520220747193427e-06, "loss": 0.0523, "step": 14351 }, { "epoch": 0.6669144981412639, "grad_norm": 0.4950257011332031, "learning_rate": 2.6513577211118803e-06, "loss": 0.0435, "step": 14352 }, { "epoch": 0.6669609665427509, "grad_norm": 0.7058686459281895, "learning_rate": 2.6506934207036393e-06, "loss": 0.0472, "step": 14353 }, { "epoch": 0.6670074349442379, "grad_norm": 0.6844730367081174, "learning_rate": 2.650029173509666e-06, "loss": 0.0464, "step": 14354 }, { "epoch": 0.6670539033457249, "grad_norm": 0.5683399768967109, "learning_rate": 2.6493649795450044e-06, "loss": 0.0653, "step": 14355 }, { "epoch": 0.6671003717472119, "grad_norm": 0.6848558854543855, "learning_rate": 2.6487008388246984e-06, "loss": 0.0569, "step": 14356 }, { "epoch": 0.6671468401486988, "grad_norm": 0.5367323154890764, "learning_rate": 2.6480367513637963e-06, "loss": 0.0675, "step": 14357 }, { "epoch": 0.6671933085501859, "grad_norm": 1.0818406976311725, "learning_rate": 2.6473727171773355e-06, "loss": 0.0617, "step": 14358 }, { "epoch": 0.6672397769516729, "grad_norm": 0.666446359337323, "learning_rate": 2.6467087362803572e-06, "loss": 0.0609, "step": 14359 }, { "epoch": 0.6672862453531598, "grad_norm": 0.8158672049452801, "learning_rate": 2.6460448086879048e-06, "loss": 0.0707, "step": 14360 }, { "epoch": 0.6673327137546469, "grad_norm": 0.30650859543333364, "learning_rate": 2.645380934415015e-06, "loss": 0.0447, "step": 14361 }, { "epoch": 0.6673791821561338, "grad_norm": 0.4275895507026151, "learning_rate": 2.6447171134767245e-06, "loss": 0.0375, "step": 14362 }, { "epoch": 0.6674256505576208, "grad_norm": 0.558760982695593, "learning_rate": 2.644053345888071e-06, "loss": 0.0497, "step": 14363 }, { "epoch": 0.6674721189591079, "grad_norm": 0.5736044998473209, "learning_rate": 2.6433896316640884e-06, "loss": 0.0666, "step": 14364 }, { "epoch": 0.6675185873605948, "grad_norm": 0.4360176020899142, "learning_rate": 2.6427259708198115e-06, "loss": 0.0535, "step": 14365 }, { "epoch": 0.6675650557620818, "grad_norm": 0.5306421248275682, "learning_rate": 2.6420623633702695e-06, "loss": 0.0489, "step": 14366 }, { "epoch": 0.6676115241635687, "grad_norm": 1.1000149311772798, "learning_rate": 2.641398809330499e-06, "loss": 0.0556, "step": 14367 }, { "epoch": 0.6676579925650558, "grad_norm": 0.49372965271724933, "learning_rate": 2.6407353087155287e-06, "loss": 0.058, "step": 14368 }, { "epoch": 0.6677044609665428, "grad_norm": 0.4602597038608839, "learning_rate": 2.6400718615403852e-06, "loss": 0.0565, "step": 14369 }, { "epoch": 0.6677509293680297, "grad_norm": 0.5613821868350046, "learning_rate": 2.6394084678200994e-06, "loss": 0.0509, "step": 14370 }, { "epoch": 0.6677973977695167, "grad_norm": 0.45934421485528637, "learning_rate": 2.638745127569695e-06, "loss": 0.0461, "step": 14371 }, { "epoch": 0.6678438661710037, "grad_norm": 0.6859135560687653, "learning_rate": 2.6380818408041993e-06, "loss": 0.0479, "step": 14372 }, { "epoch": 0.6678903345724907, "grad_norm": 0.5319659455843639, "learning_rate": 2.637418607538633e-06, "loss": 0.0637, "step": 14373 }, { "epoch": 0.6679368029739777, "grad_norm": 0.7012625544536731, "learning_rate": 2.636755427788025e-06, "loss": 0.0628, "step": 14374 }, { "epoch": 0.6679832713754646, "grad_norm": 0.4191689296356549, "learning_rate": 2.6360923015673956e-06, "loss": 0.0545, "step": 14375 }, { "epoch": 0.6680297397769517, "grad_norm": 0.930793677665034, "learning_rate": 2.6354292288917588e-06, "loss": 0.0487, "step": 14376 }, { "epoch": 0.6680762081784387, "grad_norm": 0.41255379568373757, "learning_rate": 2.6347662097761413e-06, "loss": 0.0319, "step": 14377 }, { "epoch": 0.6681226765799256, "grad_norm": 0.741984266500006, "learning_rate": 2.634103244235558e-06, "loss": 0.0657, "step": 14378 }, { "epoch": 0.6681691449814127, "grad_norm": 0.6261485364579538, "learning_rate": 2.6334403322850264e-06, "loss": 0.0652, "step": 14379 }, { "epoch": 0.6682156133828996, "grad_norm": 0.4894686213027806, "learning_rate": 2.6327774739395606e-06, "loss": 0.0525, "step": 14380 }, { "epoch": 0.6682620817843866, "grad_norm": 0.4686675070125856, "learning_rate": 2.63211466921418e-06, "loss": 0.0459, "step": 14381 }, { "epoch": 0.6683085501858737, "grad_norm": 0.7357124402073272, "learning_rate": 2.6314519181238917e-06, "loss": 0.0554, "step": 14382 }, { "epoch": 0.6683550185873606, "grad_norm": 0.4969483686228378, "learning_rate": 2.6307892206837095e-06, "loss": 0.0582, "step": 14383 }, { "epoch": 0.6684014869888476, "grad_norm": 0.9784286037107728, "learning_rate": 2.630126576908646e-06, "loss": 0.0742, "step": 14384 }, { "epoch": 0.6684479553903345, "grad_norm": 0.3975749915784916, "learning_rate": 2.6294639868137107e-06, "loss": 0.0355, "step": 14385 }, { "epoch": 0.6684944237918216, "grad_norm": 0.8071497062105633, "learning_rate": 2.6288014504139104e-06, "loss": 0.0538, "step": 14386 }, { "epoch": 0.6685408921933086, "grad_norm": 0.8673553733678137, "learning_rate": 2.628138967724253e-06, "loss": 0.0472, "step": 14387 }, { "epoch": 0.6685873605947955, "grad_norm": 0.4509586311897086, "learning_rate": 2.627476538759744e-06, "loss": 0.046, "step": 14388 }, { "epoch": 0.6686338289962825, "grad_norm": 0.25082541074876297, "learning_rate": 2.626814163535389e-06, "loss": 0.023, "step": 14389 }, { "epoch": 0.6686802973977695, "grad_norm": 0.4651777054469426, "learning_rate": 2.6261518420661884e-06, "loss": 0.0335, "step": 14390 }, { "epoch": 0.6687267657992565, "grad_norm": 1.1764452973996096, "learning_rate": 2.625489574367149e-06, "loss": 0.0502, "step": 14391 }, { "epoch": 0.6687732342007435, "grad_norm": 0.9651483906158759, "learning_rate": 2.62482736045327e-06, "loss": 0.0634, "step": 14392 }, { "epoch": 0.6688197026022304, "grad_norm": 0.5365114959445358, "learning_rate": 2.6241652003395524e-06, "loss": 0.0439, "step": 14393 }, { "epoch": 0.6688661710037175, "grad_norm": 0.4885933833365137, "learning_rate": 2.623503094040992e-06, "loss": 0.0561, "step": 14394 }, { "epoch": 0.6689126394052045, "grad_norm": 0.44277030021825653, "learning_rate": 2.622841041572588e-06, "loss": 0.0506, "step": 14395 }, { "epoch": 0.6689591078066914, "grad_norm": 0.3927909691634037, "learning_rate": 2.6221790429493364e-06, "loss": 0.0368, "step": 14396 }, { "epoch": 0.6690055762081785, "grad_norm": 0.5709487850349203, "learning_rate": 2.62151709818623e-06, "loss": 0.0596, "step": 14397 }, { "epoch": 0.6690520446096654, "grad_norm": 0.6065340028444437, "learning_rate": 2.6208552072982674e-06, "loss": 0.055, "step": 14398 }, { "epoch": 0.6690985130111524, "grad_norm": 0.6043509085486553, "learning_rate": 2.6201933703004394e-06, "loss": 0.0766, "step": 14399 }, { "epoch": 0.6691449814126395, "grad_norm": 0.5639795939245118, "learning_rate": 2.619531587207733e-06, "loss": 0.0409, "step": 14400 }, { "epoch": 0.6691914498141264, "grad_norm": 0.7464672918698465, "learning_rate": 2.6188698580351427e-06, "loss": 0.0606, "step": 14401 }, { "epoch": 0.6692379182156134, "grad_norm": 0.3984779579317229, "learning_rate": 2.618208182797657e-06, "loss": 0.044, "step": 14402 }, { "epoch": 0.6692843866171003, "grad_norm": 0.7377087671271931, "learning_rate": 2.61754656151026e-06, "loss": 0.0543, "step": 14403 }, { "epoch": 0.6693308550185874, "grad_norm": 0.502690590852053, "learning_rate": 2.6168849941879438e-06, "loss": 0.0565, "step": 14404 }, { "epoch": 0.6693773234200744, "grad_norm": 0.9764883368986563, "learning_rate": 2.616223480845692e-06, "loss": 0.0765, "step": 14405 }, { "epoch": 0.6694237918215613, "grad_norm": 0.5073246374852823, "learning_rate": 2.6155620214984854e-06, "loss": 0.0556, "step": 14406 }, { "epoch": 0.6694702602230483, "grad_norm": 0.5415061383516578, "learning_rate": 2.6149006161613067e-06, "loss": 0.0654, "step": 14407 }, { "epoch": 0.6695167286245353, "grad_norm": 0.5806414505486717, "learning_rate": 2.6142392648491412e-06, "loss": 0.0628, "step": 14408 }, { "epoch": 0.6695631970260223, "grad_norm": 0.5501959316015113, "learning_rate": 2.613577967576967e-06, "loss": 0.0573, "step": 14409 }, { "epoch": 0.6696096654275093, "grad_norm": 0.7348340604873206, "learning_rate": 2.6129167243597642e-06, "loss": 0.0676, "step": 14410 }, { "epoch": 0.6696561338289962, "grad_norm": 0.6121008797822053, "learning_rate": 2.6122555352125097e-06, "loss": 0.0759, "step": 14411 }, { "epoch": 0.6697026022304833, "grad_norm": 0.36928777867667784, "learning_rate": 2.6115944001501803e-06, "loss": 0.0282, "step": 14412 }, { "epoch": 0.6697490706319703, "grad_norm": 1.1120545804372777, "learning_rate": 2.6109333191877513e-06, "loss": 0.0807, "step": 14413 }, { "epoch": 0.6697955390334572, "grad_norm": 0.7205763905757236, "learning_rate": 2.610272292340195e-06, "loss": 0.0535, "step": 14414 }, { "epoch": 0.6698420074349443, "grad_norm": 0.7312690625883632, "learning_rate": 2.6096113196224892e-06, "loss": 0.0589, "step": 14415 }, { "epoch": 0.6698884758364312, "grad_norm": 0.6210258636778194, "learning_rate": 2.6089504010496034e-06, "loss": 0.053, "step": 14416 }, { "epoch": 0.6699349442379182, "grad_norm": 0.5109544712045477, "learning_rate": 2.608289536636505e-06, "loss": 0.055, "step": 14417 }, { "epoch": 0.6699814126394052, "grad_norm": 0.5535053775047905, "learning_rate": 2.607628726398167e-06, "loss": 0.0385, "step": 14418 }, { "epoch": 0.6700278810408922, "grad_norm": 0.5354196668347123, "learning_rate": 2.6069679703495564e-06, "loss": 0.0423, "step": 14419 }, { "epoch": 0.6700743494423792, "grad_norm": 0.6128683646898329, "learning_rate": 2.6063072685056385e-06, "loss": 0.0405, "step": 14420 }, { "epoch": 0.6701208178438661, "grad_norm": 0.5402642782736932, "learning_rate": 2.605646620881382e-06, "loss": 0.049, "step": 14421 }, { "epoch": 0.6701672862453532, "grad_norm": 0.5228649025892985, "learning_rate": 2.604986027491751e-06, "loss": 0.0544, "step": 14422 }, { "epoch": 0.6702137546468402, "grad_norm": 0.6002258463381948, "learning_rate": 2.6043254883517054e-06, "loss": 0.0468, "step": 14423 }, { "epoch": 0.6702602230483271, "grad_norm": 0.6797726417117992, "learning_rate": 2.6036650034762074e-06, "loss": 0.0435, "step": 14424 }, { "epoch": 0.6703066914498141, "grad_norm": 0.39283257749641304, "learning_rate": 2.603004572880221e-06, "loss": 0.0406, "step": 14425 }, { "epoch": 0.6703531598513011, "grad_norm": 0.35274892023287935, "learning_rate": 2.6023441965787034e-06, "loss": 0.0502, "step": 14426 }, { "epoch": 0.6703996282527881, "grad_norm": 0.45974167276013933, "learning_rate": 2.601683874586611e-06, "loss": 0.057, "step": 14427 }, { "epoch": 0.6704460966542751, "grad_norm": 0.5259925096750733, "learning_rate": 2.6010236069189075e-06, "loss": 0.0314, "step": 14428 }, { "epoch": 0.670492565055762, "grad_norm": 0.4390089616260128, "learning_rate": 2.6003633935905427e-06, "loss": 0.0601, "step": 14429 }, { "epoch": 0.6705390334572491, "grad_norm": 1.002860558252765, "learning_rate": 2.599703234616472e-06, "loss": 0.0832, "step": 14430 }, { "epoch": 0.6705855018587361, "grad_norm": 0.35300328454880425, "learning_rate": 2.599043130011647e-06, "loss": 0.0392, "step": 14431 }, { "epoch": 0.670631970260223, "grad_norm": 0.44935514369606605, "learning_rate": 2.5983830797910247e-06, "loss": 0.0526, "step": 14432 }, { "epoch": 0.6706784386617101, "grad_norm": 0.6850128911323824, "learning_rate": 2.5977230839695534e-06, "loss": 0.0643, "step": 14433 }, { "epoch": 0.670724907063197, "grad_norm": 0.4192246103051991, "learning_rate": 2.5970631425621817e-06, "loss": 0.0432, "step": 14434 }, { "epoch": 0.670771375464684, "grad_norm": 0.8047709529314755, "learning_rate": 2.5964032555838593e-06, "loss": 0.0687, "step": 14435 }, { "epoch": 0.670817843866171, "grad_norm": 0.5505026857699437, "learning_rate": 2.5957434230495333e-06, "loss": 0.0293, "step": 14436 }, { "epoch": 0.670864312267658, "grad_norm": 0.6950769264619342, "learning_rate": 2.5950836449741467e-06, "loss": 0.0616, "step": 14437 }, { "epoch": 0.670910780669145, "grad_norm": 0.5495730961167294, "learning_rate": 2.594423921372649e-06, "loss": 0.042, "step": 14438 }, { "epoch": 0.6709572490706319, "grad_norm": 0.4210219997908567, "learning_rate": 2.5937642522599815e-06, "loss": 0.0429, "step": 14439 }, { "epoch": 0.671003717472119, "grad_norm": 0.7701272521928219, "learning_rate": 2.5931046376510875e-06, "loss": 0.0425, "step": 14440 }, { "epoch": 0.671050185873606, "grad_norm": 0.6494261404852225, "learning_rate": 2.5924450775609034e-06, "loss": 0.048, "step": 14441 }, { "epoch": 0.6710966542750929, "grad_norm": 0.5237569866672793, "learning_rate": 2.5917855720043737e-06, "loss": 0.0674, "step": 14442 }, { "epoch": 0.6711431226765799, "grad_norm": 1.0324376258332972, "learning_rate": 2.5911261209964355e-06, "loss": 0.0891, "step": 14443 }, { "epoch": 0.6711895910780669, "grad_norm": 0.5986425656048093, "learning_rate": 2.5904667245520245e-06, "loss": 0.0468, "step": 14444 }, { "epoch": 0.6712360594795539, "grad_norm": 0.5168415147907492, "learning_rate": 2.5898073826860794e-06, "loss": 0.0683, "step": 14445 }, { "epoch": 0.6712825278810409, "grad_norm": 0.5132808010379372, "learning_rate": 2.5891480954135355e-06, "loss": 0.0634, "step": 14446 }, { "epoch": 0.6713289962825278, "grad_norm": 0.48336749071953955, "learning_rate": 2.588488862749321e-06, "loss": 0.0559, "step": 14447 }, { "epoch": 0.6713754646840149, "grad_norm": 0.570055521579685, "learning_rate": 2.5878296847083746e-06, "loss": 0.0464, "step": 14448 }, { "epoch": 0.6714219330855019, "grad_norm": 0.4338175253644332, "learning_rate": 2.5871705613056237e-06, "loss": 0.0336, "step": 14449 }, { "epoch": 0.6714684014869888, "grad_norm": 0.7494010342746031, "learning_rate": 2.586511492555999e-06, "loss": 0.0539, "step": 14450 }, { "epoch": 0.6715148698884759, "grad_norm": 0.8174149689584291, "learning_rate": 2.5858524784744277e-06, "loss": 0.0708, "step": 14451 }, { "epoch": 0.6715613382899628, "grad_norm": 0.63537970590465, "learning_rate": 2.5851935190758426e-06, "loss": 0.0617, "step": 14452 }, { "epoch": 0.6716078066914498, "grad_norm": 0.6285134615459594, "learning_rate": 2.5845346143751636e-06, "loss": 0.0414, "step": 14453 }, { "epoch": 0.6716542750929368, "grad_norm": 0.5311575085144937, "learning_rate": 2.5838757643873157e-06, "loss": 0.061, "step": 14454 }, { "epoch": 0.6717007434944238, "grad_norm": 0.42507494303229365, "learning_rate": 2.583216969127227e-06, "loss": 0.0453, "step": 14455 }, { "epoch": 0.6717472118959108, "grad_norm": 0.7800577592301523, "learning_rate": 2.5825582286098173e-06, "loss": 0.0602, "step": 14456 }, { "epoch": 0.6717936802973977, "grad_norm": 0.5898493129507187, "learning_rate": 2.581899542850008e-06, "loss": 0.0446, "step": 14457 }, { "epoch": 0.6718401486988848, "grad_norm": 0.7305684451215836, "learning_rate": 2.581240911862719e-06, "loss": 0.0584, "step": 14458 }, { "epoch": 0.6718866171003718, "grad_norm": 0.5061918321452391, "learning_rate": 2.580582335662869e-06, "loss": 0.0289, "step": 14459 }, { "epoch": 0.6719330855018587, "grad_norm": 0.8589682273095386, "learning_rate": 2.579923814265376e-06, "loss": 0.0672, "step": 14460 }, { "epoch": 0.6719795539033457, "grad_norm": 0.4160728517551424, "learning_rate": 2.5792653476851537e-06, "loss": 0.0499, "step": 14461 }, { "epoch": 0.6720260223048327, "grad_norm": 0.4435055583830975, "learning_rate": 2.57860693593712e-06, "loss": 0.036, "step": 14462 }, { "epoch": 0.6720724907063197, "grad_norm": 0.4970633684487472, "learning_rate": 2.577948579036189e-06, "loss": 0.0621, "step": 14463 }, { "epoch": 0.6721189591078067, "grad_norm": 0.3880992958490707, "learning_rate": 2.577290276997271e-06, "loss": 0.0522, "step": 14464 }, { "epoch": 0.6721654275092936, "grad_norm": 0.41707074460078825, "learning_rate": 2.5766320298352783e-06, "loss": 0.0585, "step": 14465 }, { "epoch": 0.6722118959107807, "grad_norm": 0.6403492394662464, "learning_rate": 2.5759738375651207e-06, "loss": 0.0465, "step": 14466 }, { "epoch": 0.6722583643122677, "grad_norm": 1.4567431582941064, "learning_rate": 2.5753157002017073e-06, "loss": 0.0488, "step": 14467 }, { "epoch": 0.6723048327137546, "grad_norm": 0.379590330706935, "learning_rate": 2.5746576177599427e-06, "loss": 0.0476, "step": 14468 }, { "epoch": 0.6723513011152417, "grad_norm": 0.8406164451516861, "learning_rate": 2.57399959025474e-06, "loss": 0.0643, "step": 14469 }, { "epoch": 0.6723977695167286, "grad_norm": 0.7501719982738957, "learning_rate": 2.5733416177009973e-06, "loss": 0.0389, "step": 14470 }, { "epoch": 0.6724442379182156, "grad_norm": 0.6428610648112327, "learning_rate": 2.5726837001136194e-06, "loss": 0.05, "step": 14471 }, { "epoch": 0.6724907063197026, "grad_norm": 0.8014119172101837, "learning_rate": 2.5720258375075126e-06, "loss": 0.0707, "step": 14472 }, { "epoch": 0.6725371747211896, "grad_norm": 0.3754545042348177, "learning_rate": 2.571368029897575e-06, "loss": 0.0455, "step": 14473 }, { "epoch": 0.6725836431226766, "grad_norm": 0.5182812271358671, "learning_rate": 2.5707102772987093e-06, "loss": 0.0333, "step": 14474 }, { "epoch": 0.6726301115241635, "grad_norm": 0.8222526104607684, "learning_rate": 2.5700525797258113e-06, "loss": 0.0569, "step": 14475 }, { "epoch": 0.6726765799256506, "grad_norm": 0.6954453498105062, "learning_rate": 2.56939493719378e-06, "loss": 0.038, "step": 14476 }, { "epoch": 0.6727230483271376, "grad_norm": 0.6096081798004575, "learning_rate": 2.568737349717512e-06, "loss": 0.05, "step": 14477 }, { "epoch": 0.6727695167286245, "grad_norm": 0.2389488595883944, "learning_rate": 2.568079817311899e-06, "loss": 0.0379, "step": 14478 }, { "epoch": 0.6728159851301115, "grad_norm": 0.34797424941857147, "learning_rate": 2.567422339991841e-06, "loss": 0.0461, "step": 14479 }, { "epoch": 0.6728624535315985, "grad_norm": 0.7321895028793662, "learning_rate": 2.5667649177722266e-06, "loss": 0.0553, "step": 14480 }, { "epoch": 0.6729089219330855, "grad_norm": 0.5735889937211787, "learning_rate": 2.5661075506679477e-06, "loss": 0.0527, "step": 14481 }, { "epoch": 0.6729553903345725, "grad_norm": 0.42138158877320736, "learning_rate": 2.565450238693894e-06, "loss": 0.0444, "step": 14482 }, { "epoch": 0.6730018587360594, "grad_norm": 0.45946529973937755, "learning_rate": 2.564792981864955e-06, "loss": 0.0626, "step": 14483 }, { "epoch": 0.6730483271375465, "grad_norm": 0.8206184464363935, "learning_rate": 2.5641357801960186e-06, "loss": 0.0526, "step": 14484 }, { "epoch": 0.6730947955390335, "grad_norm": 0.6136546857740287, "learning_rate": 2.563478633701968e-06, "loss": 0.0593, "step": 14485 }, { "epoch": 0.6731412639405204, "grad_norm": 0.726319876056279, "learning_rate": 2.562821542397693e-06, "loss": 0.0637, "step": 14486 }, { "epoch": 0.6731877323420075, "grad_norm": 0.29000058069737056, "learning_rate": 2.562164506298076e-06, "loss": 0.0253, "step": 14487 }, { "epoch": 0.6732342007434944, "grad_norm": 0.5501413592401783, "learning_rate": 2.561507525417996e-06, "loss": 0.036, "step": 14488 }, { "epoch": 0.6732806691449814, "grad_norm": 0.5978233783646696, "learning_rate": 2.5608505997723386e-06, "loss": 0.0581, "step": 14489 }, { "epoch": 0.6733271375464684, "grad_norm": 0.4585647251617716, "learning_rate": 2.5601937293759828e-06, "loss": 0.0397, "step": 14490 }, { "epoch": 0.6733736059479554, "grad_norm": 0.616036537771694, "learning_rate": 2.559536914243806e-06, "loss": 0.0566, "step": 14491 }, { "epoch": 0.6734200743494424, "grad_norm": 0.5550415029297172, "learning_rate": 2.558880154390685e-06, "loss": 0.0363, "step": 14492 }, { "epoch": 0.6734665427509293, "grad_norm": 0.5195106864174971, "learning_rate": 2.558223449831502e-06, "loss": 0.0493, "step": 14493 }, { "epoch": 0.6735130111524164, "grad_norm": 0.4022998514198923, "learning_rate": 2.557566800581125e-06, "loss": 0.0412, "step": 14494 }, { "epoch": 0.6735594795539034, "grad_norm": 0.5051667752380095, "learning_rate": 2.5569102066544295e-06, "loss": 0.0593, "step": 14495 }, { "epoch": 0.6736059479553903, "grad_norm": 0.5358565264648425, "learning_rate": 2.55625366806629e-06, "loss": 0.0465, "step": 14496 }, { "epoch": 0.6736524163568773, "grad_norm": 1.8636377607478158, "learning_rate": 2.5555971848315777e-06, "loss": 0.0647, "step": 14497 }, { "epoch": 0.6736988847583643, "grad_norm": 0.49005663406631766, "learning_rate": 2.5549407569651618e-06, "loss": 0.0485, "step": 14498 }, { "epoch": 0.6737453531598513, "grad_norm": 0.691643636666332, "learning_rate": 2.554284384481911e-06, "loss": 0.0488, "step": 14499 }, { "epoch": 0.6737918215613383, "grad_norm": 0.5503106368075129, "learning_rate": 2.553628067396693e-06, "loss": 0.0607, "step": 14500 }, { "epoch": 0.6738382899628252, "grad_norm": 0.3784501519355981, "learning_rate": 2.5529718057243746e-06, "loss": 0.0488, "step": 14501 }, { "epoch": 0.6738847583643123, "grad_norm": 0.5053048198261896, "learning_rate": 2.5523155994798175e-06, "loss": 0.0392, "step": 14502 }, { "epoch": 0.6739312267657993, "grad_norm": 0.6820387611473409, "learning_rate": 2.5516594486778906e-06, "loss": 0.0662, "step": 14503 }, { "epoch": 0.6739776951672862, "grad_norm": 0.9095604194521282, "learning_rate": 2.551003353333454e-06, "loss": 0.0456, "step": 14504 }, { "epoch": 0.6740241635687733, "grad_norm": 0.5625984647128232, "learning_rate": 2.5503473134613686e-06, "loss": 0.0468, "step": 14505 }, { "epoch": 0.6740706319702602, "grad_norm": 4.6608939615337786, "learning_rate": 2.549691329076495e-06, "loss": 0.0782, "step": 14506 }, { "epoch": 0.6741171003717472, "grad_norm": 0.5305416036578835, "learning_rate": 2.5490354001936916e-06, "loss": 0.0282, "step": 14507 }, { "epoch": 0.6741635687732342, "grad_norm": 0.43773591092697445, "learning_rate": 2.548379526827816e-06, "loss": 0.0507, "step": 14508 }, { "epoch": 0.6742100371747212, "grad_norm": 0.3672434164300139, "learning_rate": 2.5477237089937223e-06, "loss": 0.0285, "step": 14509 }, { "epoch": 0.6742565055762082, "grad_norm": 0.49936986728000726, "learning_rate": 2.5470679467062697e-06, "loss": 0.0668, "step": 14510 }, { "epoch": 0.6743029739776951, "grad_norm": 0.6001702387202743, "learning_rate": 2.5464122399803126e-06, "loss": 0.0438, "step": 14511 }, { "epoch": 0.6743494423791822, "grad_norm": 0.5312299365730437, "learning_rate": 2.5457565888306957e-06, "loss": 0.0465, "step": 14512 }, { "epoch": 0.6743959107806692, "grad_norm": 0.8092719340216282, "learning_rate": 2.545100993272277e-06, "loss": 0.0716, "step": 14513 }, { "epoch": 0.6744423791821561, "grad_norm": 0.5357978557653986, "learning_rate": 2.5444454533199046e-06, "loss": 0.0515, "step": 14514 }, { "epoch": 0.6744888475836431, "grad_norm": 0.6929737166218474, "learning_rate": 2.543789968988428e-06, "loss": 0.0476, "step": 14515 }, { "epoch": 0.6745353159851301, "grad_norm": 0.3936451307806299, "learning_rate": 2.5431345402926903e-06, "loss": 0.0531, "step": 14516 }, { "epoch": 0.6745817843866171, "grad_norm": 0.436675855343386, "learning_rate": 2.542479167247546e-06, "loss": 0.0387, "step": 14517 }, { "epoch": 0.6746282527881041, "grad_norm": 0.7395092554643798, "learning_rate": 2.5418238498678327e-06, "loss": 0.0546, "step": 14518 }, { "epoch": 0.674674721189591, "grad_norm": 0.37042596229922053, "learning_rate": 2.5411685881683947e-06, "loss": 0.0495, "step": 14519 }, { "epoch": 0.6747211895910781, "grad_norm": 0.6707152403575428, "learning_rate": 2.540513382164078e-06, "loss": 0.0596, "step": 14520 }, { "epoch": 0.6747676579925651, "grad_norm": 0.5640636774372565, "learning_rate": 2.539858231869722e-06, "loss": 0.0379, "step": 14521 }, { "epoch": 0.674814126394052, "grad_norm": 0.5380734853727995, "learning_rate": 2.5392031373001656e-06, "loss": 0.0431, "step": 14522 }, { "epoch": 0.6748605947955391, "grad_norm": 0.5729652209786917, "learning_rate": 2.5385480984702482e-06, "loss": 0.0526, "step": 14523 }, { "epoch": 0.674907063197026, "grad_norm": 0.5016654264215663, "learning_rate": 2.537893115394807e-06, "loss": 0.0686, "step": 14524 }, { "epoch": 0.674953531598513, "grad_norm": 0.7964395537687612, "learning_rate": 2.537238188088679e-06, "loss": 0.0753, "step": 14525 }, { "epoch": 0.675, "grad_norm": 0.3804571803716073, "learning_rate": 2.5365833165666947e-06, "loss": 0.0474, "step": 14526 }, { "epoch": 0.675046468401487, "grad_norm": 0.40052443384338327, "learning_rate": 2.535928500843694e-06, "loss": 0.0423, "step": 14527 }, { "epoch": 0.675092936802974, "grad_norm": 0.5575540836828542, "learning_rate": 2.5352737409345074e-06, "loss": 0.0648, "step": 14528 }, { "epoch": 0.6751394052044609, "grad_norm": 2.3496125345770484, "learning_rate": 2.5346190368539615e-06, "loss": 0.0555, "step": 14529 }, { "epoch": 0.675185873605948, "grad_norm": 0.468964130152161, "learning_rate": 2.53396438861689e-06, "loss": 0.0377, "step": 14530 }, { "epoch": 0.675232342007435, "grad_norm": 3.0773434493441374, "learning_rate": 2.5333097962381216e-06, "loss": 0.0732, "step": 14531 }, { "epoch": 0.6752788104089219, "grad_norm": 0.3826794397716985, "learning_rate": 2.5326552597324814e-06, "loss": 0.0458, "step": 14532 }, { "epoch": 0.6753252788104089, "grad_norm": 0.39949852006656905, "learning_rate": 2.5320007791147954e-06, "loss": 0.0267, "step": 14533 }, { "epoch": 0.6753717472118959, "grad_norm": 0.700464712638563, "learning_rate": 2.5313463543998928e-06, "loss": 0.0472, "step": 14534 }, { "epoch": 0.6754182156133829, "grad_norm": 0.3833921778719601, "learning_rate": 2.5306919856025917e-06, "loss": 0.0455, "step": 14535 }, { "epoch": 0.6754646840148699, "grad_norm": 0.6635352066002712, "learning_rate": 2.5300376727377142e-06, "loss": 0.0503, "step": 14536 }, { "epoch": 0.6755111524163568, "grad_norm": 0.3263878931612692, "learning_rate": 2.5293834158200845e-06, "loss": 0.0477, "step": 14537 }, { "epoch": 0.6755576208178439, "grad_norm": 0.2834791717928691, "learning_rate": 2.5287292148645216e-06, "loss": 0.026, "step": 14538 }, { "epoch": 0.6756040892193309, "grad_norm": 0.5188528087476464, "learning_rate": 2.528075069885842e-06, "loss": 0.0434, "step": 14539 }, { "epoch": 0.6756505576208178, "grad_norm": 0.6588518034662214, "learning_rate": 2.5274209808988638e-06, "loss": 0.069, "step": 14540 }, { "epoch": 0.6756970260223049, "grad_norm": 0.9013005033229418, "learning_rate": 2.5267669479184033e-06, "loss": 0.0535, "step": 14541 }, { "epoch": 0.6757434944237918, "grad_norm": 0.596041637115755, "learning_rate": 2.5261129709592734e-06, "loss": 0.0466, "step": 14542 }, { "epoch": 0.6757899628252788, "grad_norm": 0.927276538238178, "learning_rate": 2.5254590500362874e-06, "loss": 0.0411, "step": 14543 }, { "epoch": 0.6758364312267658, "grad_norm": 0.4500379902475317, "learning_rate": 2.524805185164261e-06, "loss": 0.0396, "step": 14544 }, { "epoch": 0.6758828996282528, "grad_norm": 0.42495902828342774, "learning_rate": 2.5241513763580007e-06, "loss": 0.0566, "step": 14545 }, { "epoch": 0.6759293680297398, "grad_norm": 0.5207731826526044, "learning_rate": 2.523497623632318e-06, "loss": 0.034, "step": 14546 }, { "epoch": 0.6759758364312267, "grad_norm": 1.277369339767713, "learning_rate": 2.522843927002021e-06, "loss": 0.046, "step": 14547 }, { "epoch": 0.6760223048327137, "grad_norm": 0.6721043840904121, "learning_rate": 2.522190286481916e-06, "loss": 0.0635, "step": 14548 }, { "epoch": 0.6760687732342008, "grad_norm": 0.8456928090815152, "learning_rate": 2.521536702086809e-06, "loss": 0.0578, "step": 14549 }, { "epoch": 0.6761152416356877, "grad_norm": 0.3610177462841776, "learning_rate": 2.520883173831502e-06, "loss": 0.0353, "step": 14550 }, { "epoch": 0.6761617100371747, "grad_norm": 0.4105248075897244, "learning_rate": 2.520229701730802e-06, "loss": 0.0508, "step": 14551 }, { "epoch": 0.6762081784386617, "grad_norm": 1.3955424879029565, "learning_rate": 2.5195762857995117e-06, "loss": 0.0812, "step": 14552 }, { "epoch": 0.6762546468401487, "grad_norm": 0.5353350863427356, "learning_rate": 2.518922926052425e-06, "loss": 0.0445, "step": 14553 }, { "epoch": 0.6763011152416357, "grad_norm": 0.46987325826334303, "learning_rate": 2.518269622504347e-06, "loss": 0.0592, "step": 14554 }, { "epoch": 0.6763475836431226, "grad_norm": 0.9661172089242783, "learning_rate": 2.5176163751700745e-06, "loss": 0.0367, "step": 14555 }, { "epoch": 0.6763940520446097, "grad_norm": 1.1594628569555798, "learning_rate": 2.516963184064404e-06, "loss": 0.0531, "step": 14556 }, { "epoch": 0.6764405204460967, "grad_norm": 0.4832526839665996, "learning_rate": 2.5163100492021273e-06, "loss": 0.0465, "step": 14557 }, { "epoch": 0.6764869888475836, "grad_norm": 0.814240823194865, "learning_rate": 2.515656970598047e-06, "loss": 0.0497, "step": 14558 }, { "epoch": 0.6765334572490707, "grad_norm": 0.46159868578697333, "learning_rate": 2.5150039482669496e-06, "loss": 0.0378, "step": 14559 }, { "epoch": 0.6765799256505576, "grad_norm": 0.8167970631063702, "learning_rate": 2.5143509822236255e-06, "loss": 0.0454, "step": 14560 }, { "epoch": 0.6766263940520446, "grad_norm": 0.4801922696058146, "learning_rate": 2.5136980724828697e-06, "loss": 0.0439, "step": 14561 }, { "epoch": 0.6766728624535316, "grad_norm": 0.5317652009547126, "learning_rate": 2.51304521905947e-06, "loss": 0.0513, "step": 14562 }, { "epoch": 0.6767193308550186, "grad_norm": 0.574278469739652, "learning_rate": 2.5123924219682127e-06, "loss": 0.0454, "step": 14563 }, { "epoch": 0.6767657992565056, "grad_norm": 1.185673576858011, "learning_rate": 2.511739681223886e-06, "loss": 0.0762, "step": 14564 }, { "epoch": 0.6768122676579925, "grad_norm": 0.5927889138714432, "learning_rate": 2.5110869968412734e-06, "loss": 0.0663, "step": 14565 }, { "epoch": 0.6768587360594795, "grad_norm": 0.4943204722383084, "learning_rate": 2.510434368835161e-06, "loss": 0.0612, "step": 14566 }, { "epoch": 0.6769052044609666, "grad_norm": 0.5856540196281335, "learning_rate": 2.5097817972203277e-06, "loss": 0.0533, "step": 14567 }, { "epoch": 0.6769516728624535, "grad_norm": 0.3498631191580138, "learning_rate": 2.50912928201156e-06, "loss": 0.0435, "step": 14568 }, { "epoch": 0.6769981412639405, "grad_norm": 0.5688527880876129, "learning_rate": 2.5084768232236346e-06, "loss": 0.0474, "step": 14569 }, { "epoch": 0.6770446096654275, "grad_norm": 0.46211054659883055, "learning_rate": 2.5078244208713322e-06, "loss": 0.0545, "step": 14570 }, { "epoch": 0.6770910780669145, "grad_norm": 0.5176091865243202, "learning_rate": 2.5071720749694297e-06, "loss": 0.0592, "step": 14571 }, { "epoch": 0.6771375464684015, "grad_norm": 0.7455319294890371, "learning_rate": 2.5065197855327034e-06, "loss": 0.0552, "step": 14572 }, { "epoch": 0.6771840148698884, "grad_norm": 0.7507993689957486, "learning_rate": 2.505867552575928e-06, "loss": 0.0512, "step": 14573 }, { "epoch": 0.6772304832713755, "grad_norm": 0.4050268233193984, "learning_rate": 2.505215376113876e-06, "loss": 0.0455, "step": 14574 }, { "epoch": 0.6772769516728625, "grad_norm": 0.4326280707275221, "learning_rate": 2.504563256161323e-06, "loss": 0.052, "step": 14575 }, { "epoch": 0.6773234200743494, "grad_norm": 1.882574594407324, "learning_rate": 2.5039111927330408e-06, "loss": 0.0781, "step": 14576 }, { "epoch": 0.6773698884758365, "grad_norm": 0.6445608596044143, "learning_rate": 2.5032591858437934e-06, "loss": 0.0685, "step": 14577 }, { "epoch": 0.6774163568773234, "grad_norm": 0.40463523633180415, "learning_rate": 2.5026072355083545e-06, "loss": 0.0547, "step": 14578 }, { "epoch": 0.6774628252788104, "grad_norm": 0.544781396663953, "learning_rate": 2.501955341741491e-06, "loss": 0.0604, "step": 14579 }, { "epoch": 0.6775092936802974, "grad_norm": 0.5799581533162804, "learning_rate": 2.5013035045579682e-06, "loss": 0.0597, "step": 14580 }, { "epoch": 0.6775557620817844, "grad_norm": 0.5414607979306292, "learning_rate": 2.500651723972548e-06, "loss": 0.0524, "step": 14581 }, { "epoch": 0.6776022304832714, "grad_norm": 0.47039850661530214, "learning_rate": 2.5000000000000015e-06, "loss": 0.044, "step": 14582 }, { "epoch": 0.6776486988847583, "grad_norm": 1.3823199091745035, "learning_rate": 2.4993483326550838e-06, "loss": 0.075, "step": 14583 }, { "epoch": 0.6776951672862453, "grad_norm": 0.6134495024112335, "learning_rate": 2.4986967219525565e-06, "loss": 0.0638, "step": 14584 }, { "epoch": 0.6777416356877324, "grad_norm": 0.786151743742205, "learning_rate": 2.4980451679071826e-06, "loss": 0.0466, "step": 14585 }, { "epoch": 0.6777881040892193, "grad_norm": 1.2828925940276126, "learning_rate": 2.4973936705337194e-06, "loss": 0.0779, "step": 14586 }, { "epoch": 0.6778345724907063, "grad_norm": 0.7679376540020612, "learning_rate": 2.4967422298469236e-06, "loss": 0.0784, "step": 14587 }, { "epoch": 0.6778810408921933, "grad_norm": 0.4577898835915454, "learning_rate": 2.49609084586155e-06, "loss": 0.0517, "step": 14588 }, { "epoch": 0.6779275092936803, "grad_norm": 0.5265397230334503, "learning_rate": 2.4954395185923554e-06, "loss": 0.0629, "step": 14589 }, { "epoch": 0.6779739776951673, "grad_norm": 0.9652583790047559, "learning_rate": 2.4947882480540906e-06, "loss": 0.0577, "step": 14590 }, { "epoch": 0.6780204460966542, "grad_norm": 0.3821995944740958, "learning_rate": 2.4941370342615073e-06, "loss": 0.0617, "step": 14591 }, { "epoch": 0.6780669144981413, "grad_norm": 0.5342730040828213, "learning_rate": 2.493485877229359e-06, "loss": 0.0464, "step": 14592 }, { "epoch": 0.6781133828996283, "grad_norm": 0.6407089986437969, "learning_rate": 2.492834776972396e-06, "loss": 0.0657, "step": 14593 }, { "epoch": 0.6781598513011152, "grad_norm": 0.43604647241656896, "learning_rate": 2.49218373350536e-06, "loss": 0.0542, "step": 14594 }, { "epoch": 0.6782063197026023, "grad_norm": 1.3612594967289573, "learning_rate": 2.491532746843004e-06, "loss": 0.0587, "step": 14595 }, { "epoch": 0.6782527881040892, "grad_norm": 0.4529086162555498, "learning_rate": 2.4908818170000713e-06, "loss": 0.0586, "step": 14596 }, { "epoch": 0.6782992565055762, "grad_norm": 0.40294450151459577, "learning_rate": 2.4902309439913064e-06, "loss": 0.0481, "step": 14597 }, { "epoch": 0.6783457249070632, "grad_norm": 0.6637188746744939, "learning_rate": 2.4895801278314495e-06, "loss": 0.0561, "step": 14598 }, { "epoch": 0.6783921933085502, "grad_norm": 0.3760663968150155, "learning_rate": 2.4889293685352497e-06, "loss": 0.0319, "step": 14599 }, { "epoch": 0.6784386617100372, "grad_norm": 0.6975861344596556, "learning_rate": 2.48827866611744e-06, "loss": 0.0494, "step": 14600 }, { "epoch": 0.6784851301115241, "grad_norm": 0.7139255510654477, "learning_rate": 2.4876280205927606e-06, "loss": 0.0575, "step": 14601 }, { "epoch": 0.6785315985130111, "grad_norm": 0.49834796394555403, "learning_rate": 2.4869774319759523e-06, "loss": 0.0669, "step": 14602 }, { "epoch": 0.6785780669144982, "grad_norm": 0.507030242531403, "learning_rate": 2.48632690028175e-06, "loss": 0.033, "step": 14603 }, { "epoch": 0.6786245353159851, "grad_norm": 0.668805269655133, "learning_rate": 2.485676425524889e-06, "loss": 0.0568, "step": 14604 }, { "epoch": 0.6786710037174721, "grad_norm": 0.4879623350537982, "learning_rate": 2.4850260077201033e-06, "loss": 0.0584, "step": 14605 }, { "epoch": 0.678717472118959, "grad_norm": 0.42429526593962996, "learning_rate": 2.4843756468821258e-06, "loss": 0.0421, "step": 14606 }, { "epoch": 0.6787639405204461, "grad_norm": 0.4884307579545338, "learning_rate": 2.4837253430256873e-06, "loss": 0.0406, "step": 14607 }, { "epoch": 0.6788104089219331, "grad_norm": 0.6800921491152572, "learning_rate": 2.483075096165516e-06, "loss": 0.0566, "step": 14608 }, { "epoch": 0.67885687732342, "grad_norm": 0.49639361925912956, "learning_rate": 2.4824249063163453e-06, "loss": 0.0422, "step": 14609 }, { "epoch": 0.6789033457249071, "grad_norm": 0.6084840058912926, "learning_rate": 2.4817747734928995e-06, "loss": 0.0547, "step": 14610 }, { "epoch": 0.6789498141263941, "grad_norm": 0.6929145722223702, "learning_rate": 2.481124697709906e-06, "loss": 0.0685, "step": 14611 }, { "epoch": 0.678996282527881, "grad_norm": 0.5008230961950432, "learning_rate": 2.480474678982089e-06, "loss": 0.0387, "step": 14612 }, { "epoch": 0.679042750929368, "grad_norm": 0.6778532217830725, "learning_rate": 2.479824717324172e-06, "loss": 0.0687, "step": 14613 }, { "epoch": 0.679089219330855, "grad_norm": 0.540239284148743, "learning_rate": 2.4791748127508774e-06, "loss": 0.0483, "step": 14614 }, { "epoch": 0.679135687732342, "grad_norm": 0.6982563982008901, "learning_rate": 2.478524965276925e-06, "loss": 0.0397, "step": 14615 }, { "epoch": 0.679182156133829, "grad_norm": 0.9064500234505738, "learning_rate": 2.4778751749170376e-06, "loss": 0.0501, "step": 14616 }, { "epoch": 0.679228624535316, "grad_norm": 0.2811733524365298, "learning_rate": 2.477225441685934e-06, "loss": 0.0244, "step": 14617 }, { "epoch": 0.679275092936803, "grad_norm": 0.7165776137595414, "learning_rate": 2.4765757655983257e-06, "loss": 0.0578, "step": 14618 }, { "epoch": 0.6793215613382899, "grad_norm": 0.6387945143862757, "learning_rate": 2.4759261466689337e-06, "loss": 0.0511, "step": 14619 }, { "epoch": 0.679368029739777, "grad_norm": 0.4387351102321174, "learning_rate": 2.4752765849124718e-06, "loss": 0.0305, "step": 14620 }, { "epoch": 0.679414498141264, "grad_norm": 2.5376324603619795, "learning_rate": 2.474627080343651e-06, "loss": 0.0482, "step": 14621 }, { "epoch": 0.6794609665427509, "grad_norm": 0.2879269423374396, "learning_rate": 2.4739776329771843e-06, "loss": 0.0372, "step": 14622 }, { "epoch": 0.6795074349442379, "grad_norm": 0.7185980435858855, "learning_rate": 2.473328242827786e-06, "loss": 0.0444, "step": 14623 }, { "epoch": 0.6795539033457249, "grad_norm": 0.5173830249037065, "learning_rate": 2.47267890991016e-06, "loss": 0.0663, "step": 14624 }, { "epoch": 0.6796003717472119, "grad_norm": 0.6243628848683116, "learning_rate": 2.4720296342390153e-06, "loss": 0.0468, "step": 14625 }, { "epoch": 0.6796468401486989, "grad_norm": 0.7577742445215443, "learning_rate": 2.471380415829062e-06, "loss": 0.0471, "step": 14626 }, { "epoch": 0.6796933085501858, "grad_norm": 0.5873638930739945, "learning_rate": 2.470731254695003e-06, "loss": 0.0564, "step": 14627 }, { "epoch": 0.6797397769516729, "grad_norm": 0.34127996574357594, "learning_rate": 2.470082150851544e-06, "loss": 0.0485, "step": 14628 }, { "epoch": 0.6797862453531599, "grad_norm": 0.365331879382147, "learning_rate": 2.4694331043133863e-06, "loss": 0.0406, "step": 14629 }, { "epoch": 0.6798327137546468, "grad_norm": 0.7474297893249798, "learning_rate": 2.468784115095232e-06, "loss": 0.063, "step": 14630 }, { "epoch": 0.6798791821561339, "grad_norm": 0.6378800422509682, "learning_rate": 2.4681351832117815e-06, "loss": 0.0457, "step": 14631 }, { "epoch": 0.6799256505576208, "grad_norm": 0.7013245790698831, "learning_rate": 2.467486308677733e-06, "loss": 0.0347, "step": 14632 }, { "epoch": 0.6799721189591078, "grad_norm": 0.7334838403804298, "learning_rate": 2.4668374915077856e-06, "loss": 0.062, "step": 14633 }, { "epoch": 0.6800185873605948, "grad_norm": 0.6412506972411928, "learning_rate": 2.466188731716636e-06, "loss": 0.0596, "step": 14634 }, { "epoch": 0.6800650557620818, "grad_norm": 0.6081591712512115, "learning_rate": 2.465540029318978e-06, "loss": 0.052, "step": 14635 }, { "epoch": 0.6801115241635688, "grad_norm": 0.6142850874204274, "learning_rate": 2.4648913843295064e-06, "loss": 0.0614, "step": 14636 }, { "epoch": 0.6801579925650557, "grad_norm": 0.7257573483640866, "learning_rate": 2.4642427967629123e-06, "loss": 0.0626, "step": 14637 }, { "epoch": 0.6802044609665427, "grad_norm": 0.4981137807518718, "learning_rate": 2.463594266633888e-06, "loss": 0.0535, "step": 14638 }, { "epoch": 0.6802509293680298, "grad_norm": 0.470422591174103, "learning_rate": 2.462945793957121e-06, "loss": 0.0469, "step": 14639 }, { "epoch": 0.6802973977695167, "grad_norm": 0.2960249598263634, "learning_rate": 2.4622973787473043e-06, "loss": 0.0346, "step": 14640 }, { "epoch": 0.6803438661710037, "grad_norm": 0.2695496905248611, "learning_rate": 2.4616490210191244e-06, "loss": 0.0283, "step": 14641 }, { "epoch": 0.6803903345724907, "grad_norm": 0.5529458063876165, "learning_rate": 2.461000720787262e-06, "loss": 0.0587, "step": 14642 }, { "epoch": 0.6804368029739777, "grad_norm": 0.5169454870149279, "learning_rate": 2.4603524780664078e-06, "loss": 0.0609, "step": 14643 }, { "epoch": 0.6804832713754647, "grad_norm": 0.4591859067654898, "learning_rate": 2.4597042928712427e-06, "loss": 0.0365, "step": 14644 }, { "epoch": 0.6805297397769516, "grad_norm": 0.6188212058773633, "learning_rate": 2.459056165216449e-06, "loss": 0.0698, "step": 14645 }, { "epoch": 0.6805762081784387, "grad_norm": 0.5560496414423943, "learning_rate": 2.458408095116707e-06, "loss": 0.0416, "step": 14646 }, { "epoch": 0.6806226765799257, "grad_norm": 0.7167563669441199, "learning_rate": 2.457760082586697e-06, "loss": 0.0615, "step": 14647 }, { "epoch": 0.6806691449814126, "grad_norm": 0.6018556316136308, "learning_rate": 2.457112127641097e-06, "loss": 0.0585, "step": 14648 }, { "epoch": 0.6807156133828997, "grad_norm": 0.6777477246509936, "learning_rate": 2.4564642302945813e-06, "loss": 0.0635, "step": 14649 }, { "epoch": 0.6807620817843866, "grad_norm": 0.3555483750284505, "learning_rate": 2.45581639056183e-06, "loss": 0.0365, "step": 14650 }, { "epoch": 0.6808085501858736, "grad_norm": 0.535434944659742, "learning_rate": 2.455168608457515e-06, "loss": 0.0433, "step": 14651 }, { "epoch": 0.6808550185873606, "grad_norm": 0.9170618072603657, "learning_rate": 2.45452088399631e-06, "loss": 0.0705, "step": 14652 }, { "epoch": 0.6809014869888476, "grad_norm": 0.7185350503096027, "learning_rate": 2.4538732171928847e-06, "loss": 0.0546, "step": 14653 }, { "epoch": 0.6809479553903346, "grad_norm": 0.4062316191961527, "learning_rate": 2.453225608061912e-06, "loss": 0.0426, "step": 14654 }, { "epoch": 0.6809944237918215, "grad_norm": 0.5794145659136374, "learning_rate": 2.4525780566180592e-06, "loss": 0.0595, "step": 14655 }, { "epoch": 0.6810408921933085, "grad_norm": 0.31962835595827765, "learning_rate": 2.4519305628759925e-06, "loss": 0.03, "step": 14656 }, { "epoch": 0.6810873605947956, "grad_norm": 0.39423162005681067, "learning_rate": 2.4512831268503827e-06, "loss": 0.0399, "step": 14657 }, { "epoch": 0.6811338289962825, "grad_norm": 0.6259958168229909, "learning_rate": 2.450635748555894e-06, "loss": 0.0615, "step": 14658 }, { "epoch": 0.6811802973977695, "grad_norm": 0.6613406152334963, "learning_rate": 2.4499884280071856e-06, "loss": 0.0456, "step": 14659 }, { "epoch": 0.6812267657992565, "grad_norm": 0.4118285907356125, "learning_rate": 2.4493411652189243e-06, "loss": 0.0367, "step": 14660 }, { "epoch": 0.6812732342007435, "grad_norm": 0.47978798051805704, "learning_rate": 2.4486939602057706e-06, "loss": 0.0498, "step": 14661 }, { "epoch": 0.6813197026022305, "grad_norm": 0.6169977201657155, "learning_rate": 2.448046812982384e-06, "loss": 0.0612, "step": 14662 }, { "epoch": 0.6813661710037174, "grad_norm": 0.5203840582585262, "learning_rate": 2.4473997235634205e-06, "loss": 0.0551, "step": 14663 }, { "epoch": 0.6814126394052045, "grad_norm": 0.6763888355496541, "learning_rate": 2.446752691963544e-06, "loss": 0.0466, "step": 14664 }, { "epoch": 0.6814591078066915, "grad_norm": 0.6847044603362588, "learning_rate": 2.446105718197405e-06, "loss": 0.0483, "step": 14665 }, { "epoch": 0.6815055762081784, "grad_norm": 0.5316175581045316, "learning_rate": 2.445458802279656e-06, "loss": 0.0439, "step": 14666 }, { "epoch": 0.6815520446096655, "grad_norm": 0.5771567700583317, "learning_rate": 2.4448119442249567e-06, "loss": 0.0587, "step": 14667 }, { "epoch": 0.6815985130111524, "grad_norm": 0.35106023131827896, "learning_rate": 2.4441651440479556e-06, "loss": 0.056, "step": 14668 }, { "epoch": 0.6816449814126394, "grad_norm": 0.46981985271229953, "learning_rate": 2.4435184017633047e-06, "loss": 0.046, "step": 14669 }, { "epoch": 0.6816914498141264, "grad_norm": 0.6253510669280128, "learning_rate": 2.442871717385652e-06, "loss": 0.0589, "step": 14670 }, { "epoch": 0.6817379182156134, "grad_norm": 0.5295291608712066, "learning_rate": 2.4422250909296465e-06, "loss": 0.0454, "step": 14671 }, { "epoch": 0.6817843866171004, "grad_norm": 0.6671000631981827, "learning_rate": 2.4415785224099346e-06, "loss": 0.0685, "step": 14672 }, { "epoch": 0.6818308550185873, "grad_norm": 0.46931158567442544, "learning_rate": 2.44093201184116e-06, "loss": 0.0514, "step": 14673 }, { "epoch": 0.6818773234200743, "grad_norm": 0.6335949309115209, "learning_rate": 2.4402855592379705e-06, "loss": 0.0592, "step": 14674 }, { "epoch": 0.6819237918215614, "grad_norm": 0.40532901161412344, "learning_rate": 2.4396391646150074e-06, "loss": 0.0331, "step": 14675 }, { "epoch": 0.6819702602230483, "grad_norm": 0.6769845949416217, "learning_rate": 2.4389928279869117e-06, "loss": 0.0538, "step": 14676 }, { "epoch": 0.6820167286245353, "grad_norm": 0.5697580664811621, "learning_rate": 2.438346549368324e-06, "loss": 0.0465, "step": 14677 }, { "epoch": 0.6820631970260222, "grad_norm": 0.8362782530599717, "learning_rate": 2.437700328773883e-06, "loss": 0.0578, "step": 14678 }, { "epoch": 0.6821096654275093, "grad_norm": 0.3707665857368243, "learning_rate": 2.437054166218226e-06, "loss": 0.0269, "step": 14679 }, { "epoch": 0.6821561338289963, "grad_norm": 0.4521816418305058, "learning_rate": 2.4364080617159885e-06, "loss": 0.0384, "step": 14680 }, { "epoch": 0.6822026022304832, "grad_norm": 0.45348061180153576, "learning_rate": 2.435762015281808e-06, "loss": 0.0256, "step": 14681 }, { "epoch": 0.6822490706319703, "grad_norm": 0.5331468278748832, "learning_rate": 2.435116026930318e-06, "loss": 0.0415, "step": 14682 }, { "epoch": 0.6822955390334573, "grad_norm": 0.4653629781412631, "learning_rate": 2.4344700966761463e-06, "loss": 0.0345, "step": 14683 }, { "epoch": 0.6823420074349442, "grad_norm": 0.5437988264571131, "learning_rate": 2.433824224533929e-06, "loss": 0.0499, "step": 14684 }, { "epoch": 0.6823884758364313, "grad_norm": 0.7094113197859839, "learning_rate": 2.4331784105182936e-06, "loss": 0.0481, "step": 14685 }, { "epoch": 0.6824349442379182, "grad_norm": 0.38876872725029776, "learning_rate": 2.432532654643869e-06, "loss": 0.0543, "step": 14686 }, { "epoch": 0.6824814126394052, "grad_norm": 0.5195917702426951, "learning_rate": 2.4318869569252802e-06, "loss": 0.0485, "step": 14687 }, { "epoch": 0.6825278810408922, "grad_norm": 0.4802127850080023, "learning_rate": 2.4312413173771583e-06, "loss": 0.0618, "step": 14688 }, { "epoch": 0.6825743494423792, "grad_norm": 0.35873174943620806, "learning_rate": 2.430595736014122e-06, "loss": 0.0433, "step": 14689 }, { "epoch": 0.6826208178438662, "grad_norm": 0.528515570270883, "learning_rate": 2.4299502128507945e-06, "loss": 0.0521, "step": 14690 }, { "epoch": 0.6826672862453531, "grad_norm": 0.4733766027954954, "learning_rate": 2.4293047479018022e-06, "loss": 0.0527, "step": 14691 }, { "epoch": 0.6827137546468401, "grad_norm": 0.59613365717048, "learning_rate": 2.4286593411817626e-06, "loss": 0.0782, "step": 14692 }, { "epoch": 0.6827602230483272, "grad_norm": 0.5218279547257243, "learning_rate": 2.4280139927052953e-06, "loss": 0.0544, "step": 14693 }, { "epoch": 0.6828066914498141, "grad_norm": 1.545283525279111, "learning_rate": 2.427368702487018e-06, "loss": 0.0789, "step": 14694 }, { "epoch": 0.6828531598513011, "grad_norm": 0.5561759007429954, "learning_rate": 2.426723470541547e-06, "loss": 0.0548, "step": 14695 }, { "epoch": 0.682899628252788, "grad_norm": 0.7691146694683197, "learning_rate": 2.4260782968834977e-06, "loss": 0.0551, "step": 14696 }, { "epoch": 0.6829460966542751, "grad_norm": 0.895611626484661, "learning_rate": 2.4254331815274822e-06, "loss": 0.0728, "step": 14697 }, { "epoch": 0.6829925650557621, "grad_norm": 0.2853574273393758, "learning_rate": 2.4247881244881173e-06, "loss": 0.0392, "step": 14698 }, { "epoch": 0.683039033457249, "grad_norm": 0.48930147488732845, "learning_rate": 2.424143125780011e-06, "loss": 0.0419, "step": 14699 }, { "epoch": 0.6830855018587361, "grad_norm": 0.5963626483225756, "learning_rate": 2.4234981854177743e-06, "loss": 0.0786, "step": 14700 }, { "epoch": 0.6831319702602231, "grad_norm": 0.41848645298472714, "learning_rate": 2.422853303416015e-06, "loss": 0.044, "step": 14701 }, { "epoch": 0.68317843866171, "grad_norm": 0.5538869490736381, "learning_rate": 2.4222084797893413e-06, "loss": 0.0667, "step": 14702 }, { "epoch": 0.683224907063197, "grad_norm": 0.3060638321459228, "learning_rate": 2.421563714552358e-06, "loss": 0.0415, "step": 14703 }, { "epoch": 0.683271375464684, "grad_norm": 0.3463483001165844, "learning_rate": 2.4209190077196693e-06, "loss": 0.0487, "step": 14704 }, { "epoch": 0.683317843866171, "grad_norm": 0.6271222994036104, "learning_rate": 2.4202743593058824e-06, "loss": 0.0689, "step": 14705 }, { "epoch": 0.683364312267658, "grad_norm": 0.47984289649455214, "learning_rate": 2.4196297693255948e-06, "loss": 0.0566, "step": 14706 }, { "epoch": 0.683410780669145, "grad_norm": 0.42185558650962524, "learning_rate": 2.418985237793407e-06, "loss": 0.0415, "step": 14707 }, { "epoch": 0.683457249070632, "grad_norm": 0.4549506961752218, "learning_rate": 2.418340764723922e-06, "loss": 0.0427, "step": 14708 }, { "epoch": 0.6835037174721189, "grad_norm": 0.3969208697406001, "learning_rate": 2.417696350131736e-06, "loss": 0.0563, "step": 14709 }, { "epoch": 0.6835501858736059, "grad_norm": 1.0427130991863949, "learning_rate": 2.4170519940314447e-06, "loss": 0.0771, "step": 14710 }, { "epoch": 0.683596654275093, "grad_norm": 0.540449751413873, "learning_rate": 2.416407696437645e-06, "loss": 0.0617, "step": 14711 }, { "epoch": 0.6836431226765799, "grad_norm": 0.37971146098568853, "learning_rate": 2.41576345736493e-06, "loss": 0.0298, "step": 14712 }, { "epoch": 0.6836895910780669, "grad_norm": 0.47413696533091804, "learning_rate": 2.415119276827892e-06, "loss": 0.0506, "step": 14713 }, { "epoch": 0.6837360594795538, "grad_norm": 0.6963955228997873, "learning_rate": 2.4144751548411217e-06, "loss": 0.056, "step": 14714 }, { "epoch": 0.6837825278810409, "grad_norm": 0.5494994134392962, "learning_rate": 2.4138310914192113e-06, "loss": 0.0399, "step": 14715 }, { "epoch": 0.6838289962825279, "grad_norm": 0.4587082129703811, "learning_rate": 2.41318708657675e-06, "loss": 0.0456, "step": 14716 }, { "epoch": 0.6838754646840148, "grad_norm": 0.5768596818146668, "learning_rate": 2.4125431403283227e-06, "loss": 0.0617, "step": 14717 }, { "epoch": 0.6839219330855019, "grad_norm": 0.4270420017658902, "learning_rate": 2.4118992526885166e-06, "loss": 0.0431, "step": 14718 }, { "epoch": 0.6839684014869889, "grad_norm": 0.4498009237574582, "learning_rate": 2.4112554236719164e-06, "loss": 0.0374, "step": 14719 }, { "epoch": 0.6840148698884758, "grad_norm": 0.727577006763298, "learning_rate": 2.4106116532931057e-06, "loss": 0.0631, "step": 14720 }, { "epoch": 0.6840613382899629, "grad_norm": 0.5278400041656863, "learning_rate": 2.4099679415666645e-06, "loss": 0.0504, "step": 14721 }, { "epoch": 0.6841078066914498, "grad_norm": 0.5354209958561819, "learning_rate": 2.409324288507177e-06, "loss": 0.032, "step": 14722 }, { "epoch": 0.6841542750929368, "grad_norm": 0.29690509712907104, "learning_rate": 2.4086806941292233e-06, "loss": 0.0306, "step": 14723 }, { "epoch": 0.6842007434944238, "grad_norm": 0.8482989490990824, "learning_rate": 2.408037158447375e-06, "loss": 0.0644, "step": 14724 }, { "epoch": 0.6842472118959108, "grad_norm": 0.5345061000013416, "learning_rate": 2.407393681476215e-06, "loss": 0.0564, "step": 14725 }, { "epoch": 0.6842936802973978, "grad_norm": 0.5228773899937091, "learning_rate": 2.406750263230317e-06, "loss": 0.0439, "step": 14726 }, { "epoch": 0.6843401486988847, "grad_norm": 0.4399407020569403, "learning_rate": 2.4061069037242545e-06, "loss": 0.0606, "step": 14727 }, { "epoch": 0.6843866171003717, "grad_norm": 0.28868256335972275, "learning_rate": 2.4054636029725992e-06, "loss": 0.0341, "step": 14728 }, { "epoch": 0.6844330855018588, "grad_norm": 0.7900427777066283, "learning_rate": 2.404820360989928e-06, "loss": 0.0565, "step": 14729 }, { "epoch": 0.6844795539033457, "grad_norm": 0.5932616178924031, "learning_rate": 2.4041771777908047e-06, "loss": 0.0595, "step": 14730 }, { "epoch": 0.6845260223048327, "grad_norm": 0.2934590511523104, "learning_rate": 2.4035340533897983e-06, "loss": 0.0321, "step": 14731 }, { "epoch": 0.6845724907063196, "grad_norm": 0.6709265841995848, "learning_rate": 2.4028909878014807e-06, "loss": 0.0518, "step": 14732 }, { "epoch": 0.6846189591078067, "grad_norm": 2.1244198814766357, "learning_rate": 2.402247981040415e-06, "loss": 0.0552, "step": 14733 }, { "epoch": 0.6846654275092937, "grad_norm": 0.6839052867909102, "learning_rate": 2.401605033121167e-06, "loss": 0.0567, "step": 14734 }, { "epoch": 0.6847118959107806, "grad_norm": 0.6725872329044487, "learning_rate": 2.4009621440583e-06, "loss": 0.0468, "step": 14735 }, { "epoch": 0.6847583643122677, "grad_norm": 0.548700048802015, "learning_rate": 2.4003193138663756e-06, "loss": 0.0527, "step": 14736 }, { "epoch": 0.6848048327137547, "grad_norm": 0.528718317137198, "learning_rate": 2.3996765425599543e-06, "loss": 0.0432, "step": 14737 }, { "epoch": 0.6848513011152416, "grad_norm": 0.9279661308313926, "learning_rate": 2.399033830153595e-06, "loss": 0.0708, "step": 14738 }, { "epoch": 0.6848977695167286, "grad_norm": 0.6895491009303488, "learning_rate": 2.3983911766618584e-06, "loss": 0.0767, "step": 14739 }, { "epoch": 0.6849442379182156, "grad_norm": 1.088717216385838, "learning_rate": 2.3977485820993e-06, "loss": 0.0712, "step": 14740 }, { "epoch": 0.6849907063197026, "grad_norm": 0.40679464634205087, "learning_rate": 2.397106046480475e-06, "loss": 0.0371, "step": 14741 }, { "epoch": 0.6850371747211896, "grad_norm": 1.0180931912092834, "learning_rate": 2.3964635698199378e-06, "loss": 0.0683, "step": 14742 }, { "epoch": 0.6850836431226766, "grad_norm": 0.5194695075579264, "learning_rate": 2.3958211521322405e-06, "loss": 0.0481, "step": 14743 }, { "epoch": 0.6851301115241636, "grad_norm": 0.4935713420793171, "learning_rate": 2.3951787934319356e-06, "loss": 0.0591, "step": 14744 }, { "epoch": 0.6851765799256505, "grad_norm": 0.5900892889789926, "learning_rate": 2.3945364937335697e-06, "loss": 0.0594, "step": 14745 }, { "epoch": 0.6852230483271375, "grad_norm": 0.434711084432746, "learning_rate": 2.393894253051697e-06, "loss": 0.0596, "step": 14746 }, { "epoch": 0.6852695167286246, "grad_norm": 0.5056263200859481, "learning_rate": 2.3932520714008638e-06, "loss": 0.0474, "step": 14747 }, { "epoch": 0.6853159851301115, "grad_norm": 0.7090489823739854, "learning_rate": 2.392609948795611e-06, "loss": 0.0791, "step": 14748 }, { "epoch": 0.6853624535315985, "grad_norm": 0.6046584045797774, "learning_rate": 2.3919678852504892e-06, "loss": 0.0665, "step": 14749 }, { "epoch": 0.6854089219330854, "grad_norm": 0.39574819402598427, "learning_rate": 2.39132588078004e-06, "loss": 0.0338, "step": 14750 }, { "epoch": 0.6854553903345725, "grad_norm": 0.7974522861625782, "learning_rate": 2.390683935398805e-06, "loss": 0.0824, "step": 14751 }, { "epoch": 0.6855018587360595, "grad_norm": 0.968496696643874, "learning_rate": 2.3900420491213226e-06, "loss": 0.0458, "step": 14752 }, { "epoch": 0.6855483271375464, "grad_norm": 0.661207511326847, "learning_rate": 2.3894002219621386e-06, "loss": 0.0441, "step": 14753 }, { "epoch": 0.6855947955390335, "grad_norm": 0.7467234802619646, "learning_rate": 2.388758453935786e-06, "loss": 0.063, "step": 14754 }, { "epoch": 0.6856412639405205, "grad_norm": 0.6474202323891805, "learning_rate": 2.3881167450568005e-06, "loss": 0.059, "step": 14755 }, { "epoch": 0.6856877323420074, "grad_norm": 0.5299976275923597, "learning_rate": 2.387475095339721e-06, "loss": 0.0556, "step": 14756 }, { "epoch": 0.6857342007434944, "grad_norm": 0.521856921311475, "learning_rate": 2.3868335047990813e-06, "loss": 0.0439, "step": 14757 }, { "epoch": 0.6857806691449814, "grad_norm": 0.5435264410925458, "learning_rate": 2.3861919734494125e-06, "loss": 0.0617, "step": 14758 }, { "epoch": 0.6858271375464684, "grad_norm": 0.3616645739468873, "learning_rate": 2.3855505013052465e-06, "loss": 0.0323, "step": 14759 }, { "epoch": 0.6858736059479554, "grad_norm": 0.43759224142919906, "learning_rate": 2.3849090883811133e-06, "loss": 0.0282, "step": 14760 }, { "epoch": 0.6859200743494424, "grad_norm": 0.7351436907860209, "learning_rate": 2.3842677346915417e-06, "loss": 0.0537, "step": 14761 }, { "epoch": 0.6859665427509294, "grad_norm": 0.7066471317887916, "learning_rate": 2.383626440251057e-06, "loss": 0.0768, "step": 14762 }, { "epoch": 0.6860130111524163, "grad_norm": 0.4498561445797183, "learning_rate": 2.3829852050741885e-06, "loss": 0.0419, "step": 14763 }, { "epoch": 0.6860594795539033, "grad_norm": 0.608038215086658, "learning_rate": 2.382344029175462e-06, "loss": 0.0485, "step": 14764 }, { "epoch": 0.6861059479553904, "grad_norm": 0.46329658686413466, "learning_rate": 2.381702912569394e-06, "loss": 0.0571, "step": 14765 }, { "epoch": 0.6861524163568773, "grad_norm": 0.6330471364144734, "learning_rate": 2.381061855270513e-06, "loss": 0.0395, "step": 14766 }, { "epoch": 0.6861988847583643, "grad_norm": 0.9810826547893471, "learning_rate": 2.3804208572933367e-06, "loss": 0.0721, "step": 14767 }, { "epoch": 0.6862453531598512, "grad_norm": 0.3758140786320132, "learning_rate": 2.3797799186523855e-06, "loss": 0.0455, "step": 14768 }, { "epoch": 0.6862918215613383, "grad_norm": 0.5392976726164441, "learning_rate": 2.3791390393621757e-06, "loss": 0.0401, "step": 14769 }, { "epoch": 0.6863382899628253, "grad_norm": 0.3951387958634888, "learning_rate": 2.3784982194372276e-06, "loss": 0.04, "step": 14770 }, { "epoch": 0.6863847583643122, "grad_norm": 0.39014496350202815, "learning_rate": 2.377857458892053e-06, "loss": 0.0537, "step": 14771 }, { "epoch": 0.6864312267657993, "grad_norm": 0.5686309585407847, "learning_rate": 2.3772167577411644e-06, "loss": 0.0429, "step": 14772 }, { "epoch": 0.6864776951672863, "grad_norm": 0.5186684458801295, "learning_rate": 2.3765761159990784e-06, "loss": 0.0659, "step": 14773 }, { "epoch": 0.6865241635687732, "grad_norm": 0.5145283037652938, "learning_rate": 2.3759355336803047e-06, "loss": 0.0518, "step": 14774 }, { "epoch": 0.6865706319702602, "grad_norm": 0.3703347712393286, "learning_rate": 2.375295010799353e-06, "loss": 0.0505, "step": 14775 }, { "epoch": 0.6866171003717472, "grad_norm": 0.5118551971373253, "learning_rate": 2.374654547370732e-06, "loss": 0.047, "step": 14776 }, { "epoch": 0.6866635687732342, "grad_norm": 0.4747492289821508, "learning_rate": 2.3740141434089477e-06, "loss": 0.0646, "step": 14777 }, { "epoch": 0.6867100371747212, "grad_norm": 0.34983451085155615, "learning_rate": 2.373373798928507e-06, "loss": 0.0465, "step": 14778 }, { "epoch": 0.6867565055762082, "grad_norm": 0.4580717117904083, "learning_rate": 2.3727335139439133e-06, "loss": 0.063, "step": 14779 }, { "epoch": 0.6868029739776952, "grad_norm": 0.45464304977408027, "learning_rate": 2.3720932884696716e-06, "loss": 0.0425, "step": 14780 }, { "epoch": 0.6868494423791821, "grad_norm": 0.3542638745414098, "learning_rate": 2.3714531225202837e-06, "loss": 0.044, "step": 14781 }, { "epoch": 0.6868959107806691, "grad_norm": 0.5904609860888941, "learning_rate": 2.370813016110248e-06, "loss": 0.0551, "step": 14782 }, { "epoch": 0.6869423791821562, "grad_norm": 0.7377310131147422, "learning_rate": 2.3701729692540653e-06, "loss": 0.0699, "step": 14783 }, { "epoch": 0.6869888475836431, "grad_norm": 0.6823023458160701, "learning_rate": 2.3695329819662325e-06, "loss": 0.0499, "step": 14784 }, { "epoch": 0.6870353159851301, "grad_norm": 0.47240445547310755, "learning_rate": 2.3688930542612464e-06, "loss": 0.0482, "step": 14785 }, { "epoch": 0.687081784386617, "grad_norm": 0.3904515984204652, "learning_rate": 2.3682531861535997e-06, "loss": 0.0427, "step": 14786 }, { "epoch": 0.6871282527881041, "grad_norm": 0.4091029373505135, "learning_rate": 2.3676133776577903e-06, "loss": 0.0533, "step": 14787 }, { "epoch": 0.6871747211895911, "grad_norm": 0.6558765398513778, "learning_rate": 2.36697362878831e-06, "loss": 0.05, "step": 14788 }, { "epoch": 0.687221189591078, "grad_norm": 0.6530963897792922, "learning_rate": 2.3663339395596445e-06, "loss": 0.0531, "step": 14789 }, { "epoch": 0.6872676579925651, "grad_norm": 0.9330583540849327, "learning_rate": 2.3656943099862895e-06, "loss": 0.0687, "step": 14790 }, { "epoch": 0.6873141263940521, "grad_norm": 0.5071460571234391, "learning_rate": 2.3650547400827307e-06, "loss": 0.0545, "step": 14791 }, { "epoch": 0.687360594795539, "grad_norm": 0.988248986054609, "learning_rate": 2.3644152298634546e-06, "loss": 0.05, "step": 14792 }, { "epoch": 0.687407063197026, "grad_norm": 0.572481193107692, "learning_rate": 2.363775779342946e-06, "loss": 0.0693, "step": 14793 }, { "epoch": 0.687453531598513, "grad_norm": 0.5188187346395604, "learning_rate": 2.363136388535695e-06, "loss": 0.0635, "step": 14794 }, { "epoch": 0.6875, "grad_norm": 0.4373557998410099, "learning_rate": 2.3624970574561773e-06, "loss": 0.0505, "step": 14795 }, { "epoch": 0.687546468401487, "grad_norm": 1.3248283967111505, "learning_rate": 2.3618577861188757e-06, "loss": 0.0673, "step": 14796 }, { "epoch": 0.687592936802974, "grad_norm": 0.3700292559064861, "learning_rate": 2.3612185745382735e-06, "loss": 0.0488, "step": 14797 }, { "epoch": 0.687639405204461, "grad_norm": 0.7341819710654707, "learning_rate": 2.360579422728848e-06, "loss": 0.0565, "step": 14798 }, { "epoch": 0.6876858736059479, "grad_norm": 0.3902328153666552, "learning_rate": 2.3599403307050766e-06, "loss": 0.036, "step": 14799 }, { "epoch": 0.6877323420074349, "grad_norm": 0.6629039040158947, "learning_rate": 2.359301298481435e-06, "loss": 0.0529, "step": 14800 }, { "epoch": 0.687778810408922, "grad_norm": 0.6196710161992912, "learning_rate": 2.3586623260723983e-06, "loss": 0.0633, "step": 14801 }, { "epoch": 0.6878252788104089, "grad_norm": 0.42662344311199424, "learning_rate": 2.3580234134924406e-06, "loss": 0.0536, "step": 14802 }, { "epoch": 0.6878717472118959, "grad_norm": 0.6908931813439116, "learning_rate": 2.3573845607560307e-06, "loss": 0.0657, "step": 14803 }, { "epoch": 0.687918215613383, "grad_norm": 0.5767983710940147, "learning_rate": 2.356745767877644e-06, "loss": 0.0602, "step": 14804 }, { "epoch": 0.6879646840148699, "grad_norm": 0.24907833204703145, "learning_rate": 2.356107034871747e-06, "loss": 0.0194, "step": 14805 }, { "epoch": 0.6880111524163569, "grad_norm": 0.3781799411292212, "learning_rate": 2.355468361752809e-06, "loss": 0.06, "step": 14806 }, { "epoch": 0.6880576208178438, "grad_norm": 0.5454115375095779, "learning_rate": 2.3548297485352954e-06, "loss": 0.0362, "step": 14807 }, { "epoch": 0.6881040892193309, "grad_norm": 0.5426324554447921, "learning_rate": 2.354191195233672e-06, "loss": 0.0448, "step": 14808 }, { "epoch": 0.6881505576208179, "grad_norm": 0.4494291957810743, "learning_rate": 2.3535527018624033e-06, "loss": 0.065, "step": 14809 }, { "epoch": 0.6881970260223048, "grad_norm": 0.5771648742660751, "learning_rate": 2.3529142684359493e-06, "loss": 0.0712, "step": 14810 }, { "epoch": 0.6882434944237918, "grad_norm": 0.46486247815267256, "learning_rate": 2.352275894968774e-06, "loss": 0.0658, "step": 14811 }, { "epoch": 0.6882899628252788, "grad_norm": 0.6082228679627005, "learning_rate": 2.351637581475339e-06, "loss": 0.0629, "step": 14812 }, { "epoch": 0.6883364312267658, "grad_norm": 0.9647716418851909, "learning_rate": 2.3509993279700965e-06, "loss": 0.0646, "step": 14813 }, { "epoch": 0.6883828996282528, "grad_norm": 0.5865873783573081, "learning_rate": 2.3503611344675086e-06, "loss": 0.0289, "step": 14814 }, { "epoch": 0.6884293680297398, "grad_norm": 0.8710647740949679, "learning_rate": 2.3497230009820302e-06, "loss": 0.0383, "step": 14815 }, { "epoch": 0.6884758364312268, "grad_norm": 0.6770832043774265, "learning_rate": 2.349084927528115e-06, "loss": 0.0504, "step": 14816 }, { "epoch": 0.6885223048327137, "grad_norm": 0.5879595770778717, "learning_rate": 2.3484469141202143e-06, "loss": 0.0478, "step": 14817 }, { "epoch": 0.6885687732342007, "grad_norm": 0.7004569094730342, "learning_rate": 2.3478089607727855e-06, "loss": 0.0634, "step": 14818 }, { "epoch": 0.6886152416356878, "grad_norm": 0.45634016896989493, "learning_rate": 2.3471710675002732e-06, "loss": 0.0578, "step": 14819 }, { "epoch": 0.6886617100371747, "grad_norm": 0.5841913035837678, "learning_rate": 2.3465332343171266e-06, "loss": 0.0567, "step": 14820 }, { "epoch": 0.6887081784386617, "grad_norm": 0.4983087268422211, "learning_rate": 2.3458954612377967e-06, "loss": 0.0456, "step": 14821 }, { "epoch": 0.6887546468401488, "grad_norm": 0.687747244930213, "learning_rate": 2.345257748276728e-06, "loss": 0.0592, "step": 14822 }, { "epoch": 0.6888011152416357, "grad_norm": 0.6068963374164791, "learning_rate": 2.3446200954483656e-06, "loss": 0.0623, "step": 14823 }, { "epoch": 0.6888475836431227, "grad_norm": 0.8140404613883716, "learning_rate": 2.3439825027671524e-06, "loss": 0.0645, "step": 14824 }, { "epoch": 0.6888940520446096, "grad_norm": 0.5176882267574139, "learning_rate": 2.343344970247531e-06, "loss": 0.048, "step": 14825 }, { "epoch": 0.6889405204460967, "grad_norm": 0.4385389921821123, "learning_rate": 2.342707497903942e-06, "loss": 0.0359, "step": 14826 }, { "epoch": 0.6889869888475837, "grad_norm": 0.8511373920747333, "learning_rate": 2.3420700857508233e-06, "loss": 0.0594, "step": 14827 }, { "epoch": 0.6890334572490706, "grad_norm": 0.4089398133865892, "learning_rate": 2.3414327338026164e-06, "loss": 0.0519, "step": 14828 }, { "epoch": 0.6890799256505576, "grad_norm": 0.652833036450148, "learning_rate": 2.3407954420737582e-06, "loss": 0.0553, "step": 14829 }, { "epoch": 0.6891263940520446, "grad_norm": 0.5767206787084524, "learning_rate": 2.340158210578678e-06, "loss": 0.0399, "step": 14830 }, { "epoch": 0.6891728624535316, "grad_norm": 0.7319989996441565, "learning_rate": 2.339521039331816e-06, "loss": 0.0637, "step": 14831 }, { "epoch": 0.6892193308550186, "grad_norm": 0.6077672458274637, "learning_rate": 2.338883928347602e-06, "loss": 0.0581, "step": 14832 }, { "epoch": 0.6892657992565056, "grad_norm": 0.5971890208143628, "learning_rate": 2.3382468776404683e-06, "loss": 0.0446, "step": 14833 }, { "epoch": 0.6893122676579926, "grad_norm": 0.6042720604600085, "learning_rate": 2.337609887224842e-06, "loss": 0.0561, "step": 14834 }, { "epoch": 0.6893587360594795, "grad_norm": 0.3428164075857847, "learning_rate": 2.336972957115158e-06, "loss": 0.0506, "step": 14835 }, { "epoch": 0.6894052044609665, "grad_norm": 0.4183627466484905, "learning_rate": 2.3363360873258368e-06, "loss": 0.0391, "step": 14836 }, { "epoch": 0.6894516728624536, "grad_norm": 0.4580550964800191, "learning_rate": 2.335699277871305e-06, "loss": 0.052, "step": 14837 }, { "epoch": 0.6894981412639405, "grad_norm": 0.5758855318638227, "learning_rate": 2.3350625287659913e-06, "loss": 0.0367, "step": 14838 }, { "epoch": 0.6895446096654275, "grad_norm": 0.9977158432330148, "learning_rate": 2.334425840024315e-06, "loss": 0.0675, "step": 14839 }, { "epoch": 0.6895910780669146, "grad_norm": 0.5209329747490351, "learning_rate": 2.3337892116606996e-06, "loss": 0.0639, "step": 14840 }, { "epoch": 0.6896375464684015, "grad_norm": 0.382796193993696, "learning_rate": 2.3331526436895647e-06, "loss": 0.0499, "step": 14841 }, { "epoch": 0.6896840148698885, "grad_norm": 0.6381525167251186, "learning_rate": 2.332516136125329e-06, "loss": 0.057, "step": 14842 }, { "epoch": 0.6897304832713754, "grad_norm": 0.3892447268664388, "learning_rate": 2.33187968898241e-06, "loss": 0.0287, "step": 14843 }, { "epoch": 0.6897769516728625, "grad_norm": 0.606517857670226, "learning_rate": 2.331243302275223e-06, "loss": 0.0647, "step": 14844 }, { "epoch": 0.6898234200743495, "grad_norm": 0.7032209130546426, "learning_rate": 2.3306069760181855e-06, "loss": 0.0813, "step": 14845 }, { "epoch": 0.6898698884758364, "grad_norm": 0.5644658060138531, "learning_rate": 2.3299707102257097e-06, "loss": 0.055, "step": 14846 }, { "epoch": 0.6899163568773234, "grad_norm": 0.5335857977420572, "learning_rate": 2.3293345049122076e-06, "loss": 0.042, "step": 14847 }, { "epoch": 0.6899628252788104, "grad_norm": 0.48543174230902375, "learning_rate": 2.3286983600920892e-06, "loss": 0.0338, "step": 14848 }, { "epoch": 0.6900092936802974, "grad_norm": 0.5559490468506071, "learning_rate": 2.328062275779765e-06, "loss": 0.0712, "step": 14849 }, { "epoch": 0.6900557620817844, "grad_norm": 0.5611074156894086, "learning_rate": 2.327426251989642e-06, "loss": 0.0458, "step": 14850 }, { "epoch": 0.6901022304832714, "grad_norm": 0.5054270433288174, "learning_rate": 2.3267902887361252e-06, "loss": 0.0391, "step": 14851 }, { "epoch": 0.6901486988847584, "grad_norm": 0.6180283931342221, "learning_rate": 2.3261543860336244e-06, "loss": 0.0489, "step": 14852 }, { "epoch": 0.6901951672862453, "grad_norm": 0.6228914272150283, "learning_rate": 2.325518543896542e-06, "loss": 0.0458, "step": 14853 }, { "epoch": 0.6902416356877323, "grad_norm": 0.5080979823731622, "learning_rate": 2.324882762339277e-06, "loss": 0.0414, "step": 14854 }, { "epoch": 0.6902881040892194, "grad_norm": 1.6564515142211687, "learning_rate": 2.3242470413762334e-06, "loss": 0.0544, "step": 14855 }, { "epoch": 0.6903345724907063, "grad_norm": 0.30402575776873497, "learning_rate": 2.323611381021811e-06, "loss": 0.0498, "step": 14856 }, { "epoch": 0.6903810408921933, "grad_norm": 0.6658924727477676, "learning_rate": 2.322975781290408e-06, "loss": 0.0636, "step": 14857 }, { "epoch": 0.6904275092936804, "grad_norm": 0.399454327389815, "learning_rate": 2.3223402421964187e-06, "loss": 0.0301, "step": 14858 }, { "epoch": 0.6904739776951673, "grad_norm": 0.6270474924103325, "learning_rate": 2.321704763754245e-06, "loss": 0.0582, "step": 14859 }, { "epoch": 0.6905204460966543, "grad_norm": 0.6136625583940196, "learning_rate": 2.321069345978276e-06, "loss": 0.0636, "step": 14860 }, { "epoch": 0.6905669144981412, "grad_norm": 0.6333584291494959, "learning_rate": 2.320433988882903e-06, "loss": 0.0532, "step": 14861 }, { "epoch": 0.6906133828996283, "grad_norm": 0.4905378459202293, "learning_rate": 2.3197986924825226e-06, "loss": 0.0548, "step": 14862 }, { "epoch": 0.6906598513011153, "grad_norm": 0.5327791216333776, "learning_rate": 2.319163456791523e-06, "loss": 0.061, "step": 14863 }, { "epoch": 0.6907063197026022, "grad_norm": 0.435365360591379, "learning_rate": 2.318528281824292e-06, "loss": 0.0557, "step": 14864 }, { "epoch": 0.6907527881040892, "grad_norm": 0.8898856193934745, "learning_rate": 2.317893167595218e-06, "loss": 0.067, "step": 14865 }, { "epoch": 0.6907992565055762, "grad_norm": 0.67115068967799, "learning_rate": 2.317258114118686e-06, "loss": 0.062, "step": 14866 }, { "epoch": 0.6908457249070632, "grad_norm": 0.6524280593994982, "learning_rate": 2.3166231214090825e-06, "loss": 0.0471, "step": 14867 }, { "epoch": 0.6908921933085502, "grad_norm": 0.5038048767953126, "learning_rate": 2.3159881894807867e-06, "loss": 0.0569, "step": 14868 }, { "epoch": 0.6909386617100371, "grad_norm": 0.7147179986148873, "learning_rate": 2.3153533183481853e-06, "loss": 0.0646, "step": 14869 }, { "epoch": 0.6909851301115242, "grad_norm": 0.4124600870658959, "learning_rate": 2.314718508025657e-06, "loss": 0.0435, "step": 14870 }, { "epoch": 0.6910315985130111, "grad_norm": 0.6346975858164597, "learning_rate": 2.3140837585275806e-06, "loss": 0.0533, "step": 14871 }, { "epoch": 0.6910780669144981, "grad_norm": 0.37381057145382957, "learning_rate": 2.313449069868334e-06, "loss": 0.0449, "step": 14872 }, { "epoch": 0.6911245353159852, "grad_norm": 0.3416467074423332, "learning_rate": 2.312814442062293e-06, "loss": 0.0426, "step": 14873 }, { "epoch": 0.6911710037174721, "grad_norm": 0.593819721608044, "learning_rate": 2.3121798751238337e-06, "loss": 0.0498, "step": 14874 }, { "epoch": 0.6912174721189591, "grad_norm": 0.49559055164981836, "learning_rate": 2.3115453690673272e-06, "loss": 0.0367, "step": 14875 }, { "epoch": 0.6912639405204462, "grad_norm": 0.6135921430147789, "learning_rate": 2.310910923907149e-06, "loss": 0.055, "step": 14876 }, { "epoch": 0.6913104089219331, "grad_norm": 0.22633897152395832, "learning_rate": 2.310276539657671e-06, "loss": 0.0207, "step": 14877 }, { "epoch": 0.6913568773234201, "grad_norm": 0.377069164749551, "learning_rate": 2.3096422163332568e-06, "loss": 0.0329, "step": 14878 }, { "epoch": 0.691403345724907, "grad_norm": 0.6146653815151298, "learning_rate": 2.309007953948279e-06, "loss": 0.0762, "step": 14879 }, { "epoch": 0.6914498141263941, "grad_norm": 0.29538727180571356, "learning_rate": 2.3083737525171033e-06, "loss": 0.0418, "step": 14880 }, { "epoch": 0.6914962825278811, "grad_norm": 0.34716011534796415, "learning_rate": 2.307739612054094e-06, "loss": 0.0317, "step": 14881 }, { "epoch": 0.691542750929368, "grad_norm": 0.8294740776771059, "learning_rate": 2.3071055325736192e-06, "loss": 0.0367, "step": 14882 }, { "epoch": 0.691589219330855, "grad_norm": 0.8002416461503511, "learning_rate": 2.3064715140900363e-06, "loss": 0.0573, "step": 14883 }, { "epoch": 0.691635687732342, "grad_norm": 0.2486098792530412, "learning_rate": 2.305837556617709e-06, "loss": 0.0291, "step": 14884 }, { "epoch": 0.691682156133829, "grad_norm": 0.3738186890138758, "learning_rate": 2.3052036601709945e-06, "loss": 0.047, "step": 14885 }, { "epoch": 0.691728624535316, "grad_norm": 0.6809826788427569, "learning_rate": 2.3045698247642547e-06, "loss": 0.056, "step": 14886 }, { "epoch": 0.691775092936803, "grad_norm": 0.42824375485445765, "learning_rate": 2.303936050411846e-06, "loss": 0.0561, "step": 14887 }, { "epoch": 0.69182156133829, "grad_norm": 0.49541994568502995, "learning_rate": 2.3033023371281226e-06, "loss": 0.0508, "step": 14888 }, { "epoch": 0.6918680297397769, "grad_norm": 0.485784933563391, "learning_rate": 2.30266868492744e-06, "loss": 0.0304, "step": 14889 }, { "epoch": 0.6919144981412639, "grad_norm": 0.6047631937063661, "learning_rate": 2.3020350938241502e-06, "loss": 0.0566, "step": 14890 }, { "epoch": 0.691960966542751, "grad_norm": 0.755683729933129, "learning_rate": 2.3014015638326054e-06, "loss": 0.057, "step": 14891 }, { "epoch": 0.6920074349442379, "grad_norm": 0.3760481039496256, "learning_rate": 2.3007680949671534e-06, "loss": 0.0482, "step": 14892 }, { "epoch": 0.6920539033457249, "grad_norm": 0.456841284024167, "learning_rate": 2.3001346872421463e-06, "loss": 0.0487, "step": 14893 }, { "epoch": 0.692100371747212, "grad_norm": 0.5016236822894742, "learning_rate": 2.299501340671933e-06, "loss": 0.0278, "step": 14894 }, { "epoch": 0.6921468401486989, "grad_norm": 0.5535793804156034, "learning_rate": 2.298868055270852e-06, "loss": 0.0464, "step": 14895 }, { "epoch": 0.6921933085501859, "grad_norm": 0.44397230084071226, "learning_rate": 2.2982348310532548e-06, "loss": 0.0497, "step": 14896 }, { "epoch": 0.6922397769516728, "grad_norm": 0.5165011446878505, "learning_rate": 2.2976016680334823e-06, "loss": 0.0459, "step": 14897 }, { "epoch": 0.6922862453531599, "grad_norm": 0.5487879637099641, "learning_rate": 2.296968566225874e-06, "loss": 0.0345, "step": 14898 }, { "epoch": 0.6923327137546469, "grad_norm": 0.5610860507259632, "learning_rate": 2.2963355256447746e-06, "loss": 0.061, "step": 14899 }, { "epoch": 0.6923791821561338, "grad_norm": 0.8153051047049824, "learning_rate": 2.2957025463045228e-06, "loss": 0.0496, "step": 14900 }, { "epoch": 0.6924256505576208, "grad_norm": 0.4974396603939456, "learning_rate": 2.295069628219453e-06, "loss": 0.0406, "step": 14901 }, { "epoch": 0.6924721189591078, "grad_norm": 0.363757661603722, "learning_rate": 2.2944367714039006e-06, "loss": 0.0433, "step": 14902 }, { "epoch": 0.6925185873605948, "grad_norm": 0.4035573743027868, "learning_rate": 2.293803975872205e-06, "loss": 0.0536, "step": 14903 }, { "epoch": 0.6925650557620818, "grad_norm": 0.6765458443249167, "learning_rate": 2.293171241638698e-06, "loss": 0.056, "step": 14904 }, { "epoch": 0.6926115241635687, "grad_norm": 0.6799254595469707, "learning_rate": 2.2925385687177086e-06, "loss": 0.0544, "step": 14905 }, { "epoch": 0.6926579925650558, "grad_norm": 0.47357798153346226, "learning_rate": 2.291905957123575e-06, "loss": 0.0563, "step": 14906 }, { "epoch": 0.6927044609665427, "grad_norm": 0.9730138942251441, "learning_rate": 2.291273406870619e-06, "loss": 0.048, "step": 14907 }, { "epoch": 0.6927509293680297, "grad_norm": 0.9343651899427733, "learning_rate": 2.290640917973169e-06, "loss": 0.0583, "step": 14908 }, { "epoch": 0.6927973977695168, "grad_norm": 0.4862550846268575, "learning_rate": 2.2900084904455565e-06, "loss": 0.0551, "step": 14909 }, { "epoch": 0.6928438661710037, "grad_norm": 0.5013102975575128, "learning_rate": 2.289376124302104e-06, "loss": 0.0522, "step": 14910 }, { "epoch": 0.6928903345724907, "grad_norm": 0.7654135937608978, "learning_rate": 2.2887438195571343e-06, "loss": 0.0498, "step": 14911 }, { "epoch": 0.6929368029739778, "grad_norm": 0.5835177245658625, "learning_rate": 2.2881115762249713e-06, "loss": 0.0552, "step": 14912 }, { "epoch": 0.6929832713754647, "grad_norm": 0.7649145046119135, "learning_rate": 2.2874793943199346e-06, "loss": 0.0632, "step": 14913 }, { "epoch": 0.6930297397769517, "grad_norm": 0.533538055374319, "learning_rate": 2.2868472738563454e-06, "loss": 0.0671, "step": 14914 }, { "epoch": 0.6930762081784386, "grad_norm": 0.36688366864031613, "learning_rate": 2.2862152148485184e-06, "loss": 0.0268, "step": 14915 }, { "epoch": 0.6931226765799257, "grad_norm": 0.5287628441204741, "learning_rate": 2.2855832173107757e-06, "loss": 0.0478, "step": 14916 }, { "epoch": 0.6931691449814127, "grad_norm": 0.5660712144873576, "learning_rate": 2.2849512812574293e-06, "loss": 0.0422, "step": 14917 }, { "epoch": 0.6932156133828996, "grad_norm": 0.5068020175182794, "learning_rate": 2.2843194067027957e-06, "loss": 0.0542, "step": 14918 }, { "epoch": 0.6932620817843866, "grad_norm": 0.2842351804751829, "learning_rate": 2.2836875936611825e-06, "loss": 0.0269, "step": 14919 }, { "epoch": 0.6933085501858736, "grad_norm": 0.6426907058683657, "learning_rate": 2.2830558421469055e-06, "loss": 0.0478, "step": 14920 }, { "epoch": 0.6933550185873606, "grad_norm": 0.4252723701065078, "learning_rate": 2.2824241521742735e-06, "loss": 0.0364, "step": 14921 }, { "epoch": 0.6934014869888476, "grad_norm": 0.641482426074885, "learning_rate": 2.2817925237575926e-06, "loss": 0.049, "step": 14922 }, { "epoch": 0.6934479553903345, "grad_norm": 1.6858990932204854, "learning_rate": 2.281160956911174e-06, "loss": 0.0704, "step": 14923 }, { "epoch": 0.6934944237918216, "grad_norm": 0.47788556879229904, "learning_rate": 2.280529451649323e-06, "loss": 0.0328, "step": 14924 }, { "epoch": 0.6935408921933085, "grad_norm": 0.740724040347568, "learning_rate": 2.2798980079863386e-06, "loss": 0.0447, "step": 14925 }, { "epoch": 0.6935873605947955, "grad_norm": 0.7575843669181532, "learning_rate": 2.2792666259365286e-06, "loss": 0.0706, "step": 14926 }, { "epoch": 0.6936338289962826, "grad_norm": 0.541050115510296, "learning_rate": 2.2786353055141936e-06, "loss": 0.0513, "step": 14927 }, { "epoch": 0.6936802973977695, "grad_norm": 0.5741138486377212, "learning_rate": 2.2780040467336327e-06, "loss": 0.054, "step": 14928 }, { "epoch": 0.6937267657992565, "grad_norm": 0.336098614057924, "learning_rate": 2.2773728496091434e-06, "loss": 0.0602, "step": 14929 }, { "epoch": 0.6937732342007435, "grad_norm": 0.4650736884071337, "learning_rate": 2.276741714155028e-06, "loss": 0.0434, "step": 14930 }, { "epoch": 0.6938197026022305, "grad_norm": 0.2960797913960355, "learning_rate": 2.276110640385577e-06, "loss": 0.0391, "step": 14931 }, { "epoch": 0.6938661710037175, "grad_norm": 0.7628685075146179, "learning_rate": 2.2754796283150855e-06, "loss": 0.0287, "step": 14932 }, { "epoch": 0.6939126394052044, "grad_norm": 0.46610583235896147, "learning_rate": 2.2748486779578495e-06, "loss": 0.0622, "step": 14933 }, { "epoch": 0.6939591078066915, "grad_norm": 0.5398067629829165, "learning_rate": 2.274217789328159e-06, "loss": 0.0525, "step": 14934 }, { "epoch": 0.6940055762081785, "grad_norm": 0.3447126817502109, "learning_rate": 2.2735869624403047e-06, "loss": 0.0485, "step": 14935 }, { "epoch": 0.6940520446096654, "grad_norm": 0.43766443137965055, "learning_rate": 2.2729561973085753e-06, "loss": 0.0482, "step": 14936 }, { "epoch": 0.6940985130111524, "grad_norm": 0.7112560840370504, "learning_rate": 2.272325493947257e-06, "loss": 0.0507, "step": 14937 }, { "epoch": 0.6941449814126394, "grad_norm": 0.4626804389009102, "learning_rate": 2.2716948523706385e-06, "loss": 0.0606, "step": 14938 }, { "epoch": 0.6941914498141264, "grad_norm": 0.4819553346285539, "learning_rate": 2.2710642725930003e-06, "loss": 0.0366, "step": 14939 }, { "epoch": 0.6942379182156134, "grad_norm": 0.5167629445124706, "learning_rate": 2.2704337546286307e-06, "loss": 0.0408, "step": 14940 }, { "epoch": 0.6942843866171003, "grad_norm": 0.504140849473887, "learning_rate": 2.2698032984918116e-06, "loss": 0.0327, "step": 14941 }, { "epoch": 0.6943308550185874, "grad_norm": 1.8477522404964644, "learning_rate": 2.2691729041968176e-06, "loss": 0.061, "step": 14942 }, { "epoch": 0.6943773234200743, "grad_norm": 0.33563793090264443, "learning_rate": 2.268542571757933e-06, "loss": 0.0482, "step": 14943 }, { "epoch": 0.6944237918215613, "grad_norm": 0.40241835674324233, "learning_rate": 2.267912301189434e-06, "loss": 0.0466, "step": 14944 }, { "epoch": 0.6944702602230484, "grad_norm": 0.6955158204204781, "learning_rate": 2.2672820925055966e-06, "loss": 0.0412, "step": 14945 }, { "epoch": 0.6945167286245353, "grad_norm": 0.5815814582584415, "learning_rate": 2.266651945720694e-06, "loss": 0.0598, "step": 14946 }, { "epoch": 0.6945631970260223, "grad_norm": 0.5565412853794343, "learning_rate": 2.2660218608490054e-06, "loss": 0.0535, "step": 14947 }, { "epoch": 0.6946096654275093, "grad_norm": 0.47985425923173003, "learning_rate": 2.265391837904798e-06, "loss": 0.0519, "step": 14948 }, { "epoch": 0.6946561338289963, "grad_norm": 0.5825225856104573, "learning_rate": 2.264761876902341e-06, "loss": 0.0604, "step": 14949 }, { "epoch": 0.6947026022304833, "grad_norm": 0.5805770480227632, "learning_rate": 2.264131977855908e-06, "loss": 0.053, "step": 14950 }, { "epoch": 0.6947490706319702, "grad_norm": 0.6632832040226049, "learning_rate": 2.263502140779766e-06, "loss": 0.0449, "step": 14951 }, { "epoch": 0.6947955390334573, "grad_norm": 0.7127039453747359, "learning_rate": 2.2628723656881795e-06, "loss": 0.0576, "step": 14952 }, { "epoch": 0.6948420074349443, "grad_norm": 0.4441381619479073, "learning_rate": 2.2622426525954157e-06, "loss": 0.0652, "step": 14953 }, { "epoch": 0.6948884758364312, "grad_norm": 0.6482456928491116, "learning_rate": 2.261613001515737e-06, "loss": 0.053, "step": 14954 }, { "epoch": 0.6949349442379182, "grad_norm": 0.5656043902714352, "learning_rate": 2.260983412463405e-06, "loss": 0.0536, "step": 14955 }, { "epoch": 0.6949814126394052, "grad_norm": 1.0583162869719878, "learning_rate": 2.2603538854526806e-06, "loss": 0.0435, "step": 14956 }, { "epoch": 0.6950278810408922, "grad_norm": 0.8141085652202769, "learning_rate": 2.2597244204978256e-06, "loss": 0.0582, "step": 14957 }, { "epoch": 0.6950743494423792, "grad_norm": 0.4582972640915967, "learning_rate": 2.259095017613097e-06, "loss": 0.0338, "step": 14958 }, { "epoch": 0.6951208178438661, "grad_norm": 0.4432225576406357, "learning_rate": 2.2584656768127506e-06, "loss": 0.0454, "step": 14959 }, { "epoch": 0.6951672862453532, "grad_norm": 0.721270940997787, "learning_rate": 2.2578363981110417e-06, "loss": 0.0598, "step": 14960 }, { "epoch": 0.6952137546468401, "grad_norm": 0.40706518750574905, "learning_rate": 2.2572071815222253e-06, "loss": 0.0418, "step": 14961 }, { "epoch": 0.6952602230483271, "grad_norm": 0.31061088810745446, "learning_rate": 2.2565780270605524e-06, "loss": 0.0274, "step": 14962 }, { "epoch": 0.6953066914498142, "grad_norm": 0.9438677470046609, "learning_rate": 2.255948934740273e-06, "loss": 0.0629, "step": 14963 }, { "epoch": 0.6953531598513011, "grad_norm": 0.4150047569944537, "learning_rate": 2.2553199045756402e-06, "loss": 0.0513, "step": 14964 }, { "epoch": 0.6953996282527881, "grad_norm": 0.5330788486167454, "learning_rate": 2.2546909365809015e-06, "loss": 0.0467, "step": 14965 }, { "epoch": 0.6954460966542751, "grad_norm": 0.8342778561877612, "learning_rate": 2.2540620307702997e-06, "loss": 0.0723, "step": 14966 }, { "epoch": 0.6954925650557621, "grad_norm": 0.6092927622498546, "learning_rate": 2.2534331871580844e-06, "loss": 0.0632, "step": 14967 }, { "epoch": 0.6955390334572491, "grad_norm": 0.786151701170658, "learning_rate": 2.252804405758498e-06, "loss": 0.042, "step": 14968 }, { "epoch": 0.695585501858736, "grad_norm": 0.49336268460880306, "learning_rate": 2.252175686585783e-06, "loss": 0.0634, "step": 14969 }, { "epoch": 0.695631970260223, "grad_norm": 0.5083868803018475, "learning_rate": 2.251547029654179e-06, "loss": 0.0422, "step": 14970 }, { "epoch": 0.6956784386617101, "grad_norm": 1.4542393435818495, "learning_rate": 2.250918434977932e-06, "loss": 0.0687, "step": 14971 }, { "epoch": 0.695724907063197, "grad_norm": 0.48339700990255763, "learning_rate": 2.2502899025712728e-06, "loss": 0.0425, "step": 14972 }, { "epoch": 0.695771375464684, "grad_norm": 0.5440807492277748, "learning_rate": 2.24966143244844e-06, "loss": 0.0392, "step": 14973 }, { "epoch": 0.695817843866171, "grad_norm": 0.35713634669259614, "learning_rate": 2.249033024623672e-06, "loss": 0.0408, "step": 14974 }, { "epoch": 0.695864312267658, "grad_norm": 0.8215236049195067, "learning_rate": 2.248404679111202e-06, "loss": 0.0531, "step": 14975 }, { "epoch": 0.695910780669145, "grad_norm": 0.4265504114841255, "learning_rate": 2.2477763959252614e-06, "loss": 0.0568, "step": 14976 }, { "epoch": 0.695957249070632, "grad_norm": 0.7102524161863942, "learning_rate": 2.2471481750800823e-06, "loss": 0.0739, "step": 14977 }, { "epoch": 0.696003717472119, "grad_norm": 0.452017737135378, "learning_rate": 2.2465200165898943e-06, "loss": 0.0423, "step": 14978 }, { "epoch": 0.6960501858736059, "grad_norm": 0.5586744348522784, "learning_rate": 2.2458919204689254e-06, "loss": 0.0652, "step": 14979 }, { "epoch": 0.6960966542750929, "grad_norm": 0.541707714997182, "learning_rate": 2.245263886731402e-06, "loss": 0.0429, "step": 14980 }, { "epoch": 0.69614312267658, "grad_norm": 2.0535174330492327, "learning_rate": 2.2446359153915524e-06, "loss": 0.0631, "step": 14981 }, { "epoch": 0.6961895910780669, "grad_norm": 0.6762955612971849, "learning_rate": 2.2440080064636e-06, "loss": 0.0584, "step": 14982 }, { "epoch": 0.6962360594795539, "grad_norm": 0.5325064650500438, "learning_rate": 2.2433801599617662e-06, "loss": 0.0483, "step": 14983 }, { "epoch": 0.696282527881041, "grad_norm": 0.4900120920881234, "learning_rate": 2.242752375900273e-06, "loss": 0.0627, "step": 14984 }, { "epoch": 0.6963289962825279, "grad_norm": 0.37014731655953237, "learning_rate": 2.2421246542933406e-06, "loss": 0.0618, "step": 14985 }, { "epoch": 0.6963754646840149, "grad_norm": 0.5135384945752804, "learning_rate": 2.2414969951551873e-06, "loss": 0.0602, "step": 14986 }, { "epoch": 0.6964219330855018, "grad_norm": 0.7831276928186501, "learning_rate": 2.2408693985000283e-06, "loss": 0.0689, "step": 14987 }, { "epoch": 0.6964684014869889, "grad_norm": 0.5909951651798517, "learning_rate": 2.2402418643420844e-06, "loss": 0.0376, "step": 14988 }, { "epoch": 0.6965148698884759, "grad_norm": 0.35337078748187667, "learning_rate": 2.239614392695567e-06, "loss": 0.0377, "step": 14989 }, { "epoch": 0.6965613382899628, "grad_norm": 0.6004466433051961, "learning_rate": 2.2389869835746864e-06, "loss": 0.0659, "step": 14990 }, { "epoch": 0.6966078066914498, "grad_norm": 0.4789562862036705, "learning_rate": 2.238359636993658e-06, "loss": 0.0445, "step": 14991 }, { "epoch": 0.6966542750929368, "grad_norm": 0.4786462905368296, "learning_rate": 2.23773235296669e-06, "loss": 0.0491, "step": 14992 }, { "epoch": 0.6967007434944238, "grad_norm": 0.5046684058379766, "learning_rate": 2.237105131507992e-06, "loss": 0.0464, "step": 14993 }, { "epoch": 0.6967472118959108, "grad_norm": 0.4696554100647362, "learning_rate": 2.2364779726317687e-06, "loss": 0.055, "step": 14994 }, { "epoch": 0.6967936802973977, "grad_norm": 0.5191533691772808, "learning_rate": 2.235850876352231e-06, "loss": 0.0613, "step": 14995 }, { "epoch": 0.6968401486988848, "grad_norm": 0.7912200811173358, "learning_rate": 2.2352238426835794e-06, "loss": 0.0728, "step": 14996 }, { "epoch": 0.6968866171003717, "grad_norm": 0.6415933220411748, "learning_rate": 2.234596871640015e-06, "loss": 0.0718, "step": 14997 }, { "epoch": 0.6969330855018587, "grad_norm": 0.7530972216957325, "learning_rate": 2.233969963235744e-06, "loss": 0.0585, "step": 14998 }, { "epoch": 0.6969795539033458, "grad_norm": 0.41953082368041394, "learning_rate": 2.2333431174849644e-06, "loss": 0.0418, "step": 14999 }, { "epoch": 0.6970260223048327, "grad_norm": 0.4982812121021574, "learning_rate": 2.2327163344018756e-06, "loss": 0.0464, "step": 15000 }, { "epoch": 0.6970724907063197, "grad_norm": 1.6901722986168393, "learning_rate": 2.232089614000674e-06, "loss": 0.0438, "step": 15001 }, { "epoch": 0.6971189591078067, "grad_norm": 0.5308522393415486, "learning_rate": 2.2314629562955552e-06, "loss": 0.0508, "step": 15002 }, { "epoch": 0.6971654275092937, "grad_norm": 0.5176725017502998, "learning_rate": 2.230836361300715e-06, "loss": 0.0493, "step": 15003 }, { "epoch": 0.6972118959107807, "grad_norm": 0.2984645706043438, "learning_rate": 2.2302098290303435e-06, "loss": 0.0331, "step": 15004 }, { "epoch": 0.6972583643122676, "grad_norm": 0.5214455389201175, "learning_rate": 2.2295833594986365e-06, "loss": 0.035, "step": 15005 }, { "epoch": 0.6973048327137547, "grad_norm": 0.5302398875882662, "learning_rate": 2.2289569527197848e-06, "loss": 0.0613, "step": 15006 }, { "epoch": 0.6973513011152417, "grad_norm": 0.45581063720949827, "learning_rate": 2.2283306087079696e-06, "loss": 0.0512, "step": 15007 }, { "epoch": 0.6973977695167286, "grad_norm": 0.44456704748843795, "learning_rate": 2.2277043274773856e-06, "loss": 0.0445, "step": 15008 }, { "epoch": 0.6974442379182156, "grad_norm": 0.6642598575928249, "learning_rate": 2.2270781090422165e-06, "loss": 0.072, "step": 15009 }, { "epoch": 0.6974907063197026, "grad_norm": 0.40437486037902876, "learning_rate": 2.226451953416647e-06, "loss": 0.049, "step": 15010 }, { "epoch": 0.6975371747211896, "grad_norm": 0.5404158241357613, "learning_rate": 2.2258258606148575e-06, "loss": 0.0575, "step": 15011 }, { "epoch": 0.6975836431226766, "grad_norm": 0.6573019510815007, "learning_rate": 2.2251998306510358e-06, "loss": 0.0462, "step": 15012 }, { "epoch": 0.6976301115241635, "grad_norm": 0.6354371812269248, "learning_rate": 2.2245738635393567e-06, "loss": 0.036, "step": 15013 }, { "epoch": 0.6976765799256506, "grad_norm": 0.6824420347592695, "learning_rate": 2.223947959293999e-06, "loss": 0.0583, "step": 15014 }, { "epoch": 0.6977230483271375, "grad_norm": 0.7895790451683088, "learning_rate": 2.223322117929143e-06, "loss": 0.0555, "step": 15015 }, { "epoch": 0.6977695167286245, "grad_norm": 0.5235189447139351, "learning_rate": 2.222696339458964e-06, "loss": 0.057, "step": 15016 }, { "epoch": 0.6978159851301116, "grad_norm": 0.42304604781352345, "learning_rate": 2.2220706238976357e-06, "loss": 0.0394, "step": 15017 }, { "epoch": 0.6978624535315985, "grad_norm": 0.3893864705784689, "learning_rate": 2.2214449712593316e-06, "loss": 0.0412, "step": 15018 }, { "epoch": 0.6979089219330855, "grad_norm": 0.25118985662570353, "learning_rate": 2.220819381558223e-06, "loss": 0.0256, "step": 15019 }, { "epoch": 0.6979553903345725, "grad_norm": 0.3519984388068077, "learning_rate": 2.2201938548084805e-06, "loss": 0.0489, "step": 15020 }, { "epoch": 0.6980018587360595, "grad_norm": 0.47325651324831863, "learning_rate": 2.2195683910242714e-06, "loss": 0.0393, "step": 15021 }, { "epoch": 0.6980483271375465, "grad_norm": 1.6816960187415093, "learning_rate": 2.218942990219766e-06, "loss": 0.0679, "step": 15022 }, { "epoch": 0.6980947955390334, "grad_norm": 0.6705284144451016, "learning_rate": 2.2183176524091293e-06, "loss": 0.0717, "step": 15023 }, { "epoch": 0.6981412639405205, "grad_norm": 0.5222036156804252, "learning_rate": 2.2176923776065253e-06, "loss": 0.0446, "step": 15024 }, { "epoch": 0.6981877323420075, "grad_norm": 1.1341375361079737, "learning_rate": 2.2170671658261173e-06, "loss": 0.0468, "step": 15025 }, { "epoch": 0.6982342007434944, "grad_norm": 0.47432439741731075, "learning_rate": 2.216442017082067e-06, "loss": 0.0534, "step": 15026 }, { "epoch": 0.6982806691449814, "grad_norm": 0.5447457502931389, "learning_rate": 2.215816931388534e-06, "loss": 0.0587, "step": 15027 }, { "epoch": 0.6983271375464684, "grad_norm": 0.6677439092297199, "learning_rate": 2.215191908759677e-06, "loss": 0.0561, "step": 15028 }, { "epoch": 0.6983736059479554, "grad_norm": 0.85584747159232, "learning_rate": 2.214566949209655e-06, "loss": 0.0756, "step": 15029 }, { "epoch": 0.6984200743494424, "grad_norm": 0.4833459309007294, "learning_rate": 2.213942052752626e-06, "loss": 0.0639, "step": 15030 }, { "epoch": 0.6984665427509293, "grad_norm": 0.5078775923354545, "learning_rate": 2.2133172194027374e-06, "loss": 0.0363, "step": 15031 }, { "epoch": 0.6985130111524164, "grad_norm": 0.612571855634593, "learning_rate": 2.212692449174149e-06, "loss": 0.0452, "step": 15032 }, { "epoch": 0.6985594795539033, "grad_norm": 0.6241448094920566, "learning_rate": 2.21206774208101e-06, "loss": 0.0669, "step": 15033 }, { "epoch": 0.6986059479553903, "grad_norm": 0.4412073149348338, "learning_rate": 2.2114430981374713e-06, "loss": 0.0324, "step": 15034 }, { "epoch": 0.6986524163568774, "grad_norm": 0.5017848854218927, "learning_rate": 2.210818517357679e-06, "loss": 0.0492, "step": 15035 }, { "epoch": 0.6986988847583643, "grad_norm": 0.8898232248536664, "learning_rate": 2.2101939997557863e-06, "loss": 0.0753, "step": 15036 }, { "epoch": 0.6987453531598513, "grad_norm": 0.5918844454444875, "learning_rate": 2.209569545345935e-06, "loss": 0.0532, "step": 15037 }, { "epoch": 0.6987918215613383, "grad_norm": 0.49765226400465656, "learning_rate": 2.2089451541422675e-06, "loss": 0.0512, "step": 15038 }, { "epoch": 0.6988382899628253, "grad_norm": 0.7361892152935541, "learning_rate": 2.208320826158932e-06, "loss": 0.0579, "step": 15039 }, { "epoch": 0.6988847583643123, "grad_norm": 0.3606330509679559, "learning_rate": 2.2076965614100675e-06, "loss": 0.0307, "step": 15040 }, { "epoch": 0.6989312267657992, "grad_norm": 0.39151406234062763, "learning_rate": 2.2070723599098153e-06, "loss": 0.0505, "step": 15041 }, { "epoch": 0.6989776951672863, "grad_norm": 0.44371196792079703, "learning_rate": 2.206448221672314e-06, "loss": 0.0458, "step": 15042 }, { "epoch": 0.6990241635687733, "grad_norm": 0.47107715394192584, "learning_rate": 2.2058241467116998e-06, "loss": 0.0604, "step": 15043 }, { "epoch": 0.6990706319702602, "grad_norm": 0.756018622044062, "learning_rate": 2.2052001350421096e-06, "loss": 0.0451, "step": 15044 }, { "epoch": 0.6991171003717472, "grad_norm": 0.38140996588086695, "learning_rate": 2.2045761866776766e-06, "loss": 0.0523, "step": 15045 }, { "epoch": 0.6991635687732342, "grad_norm": 0.5593615751729132, "learning_rate": 2.2039523016325363e-06, "loss": 0.0497, "step": 15046 }, { "epoch": 0.6992100371747212, "grad_norm": 0.7289968510519683, "learning_rate": 2.2033284799208195e-06, "loss": 0.063, "step": 15047 }, { "epoch": 0.6992565055762082, "grad_norm": 0.43315416796544826, "learning_rate": 2.202704721556656e-06, "loss": 0.0483, "step": 15048 }, { "epoch": 0.6993029739776951, "grad_norm": 0.6024052062397912, "learning_rate": 2.2020810265541743e-06, "loss": 0.0342, "step": 15049 }, { "epoch": 0.6993494423791822, "grad_norm": 0.707203344490704, "learning_rate": 2.2014573949275026e-06, "loss": 0.0581, "step": 15050 }, { "epoch": 0.6993959107806691, "grad_norm": 0.439444905222617, "learning_rate": 2.2008338266907662e-06, "loss": 0.0404, "step": 15051 }, { "epoch": 0.6994423791821561, "grad_norm": 0.8264555367277894, "learning_rate": 2.2002103218580878e-06, "loss": 0.0601, "step": 15052 }, { "epoch": 0.6994888475836432, "grad_norm": 0.38211199835928444, "learning_rate": 2.1995868804435945e-06, "loss": 0.0493, "step": 15053 }, { "epoch": 0.6995353159851301, "grad_norm": 0.8460313087802246, "learning_rate": 2.1989635024614068e-06, "loss": 0.0457, "step": 15054 }, { "epoch": 0.6995817843866171, "grad_norm": 0.3534784391131974, "learning_rate": 2.1983401879256405e-06, "loss": 0.0502, "step": 15055 }, { "epoch": 0.6996282527881041, "grad_norm": 0.5296670361018224, "learning_rate": 2.1977169368504196e-06, "loss": 0.0547, "step": 15056 }, { "epoch": 0.6996747211895911, "grad_norm": 0.38340358762728877, "learning_rate": 2.1970937492498597e-06, "loss": 0.0533, "step": 15057 }, { "epoch": 0.6997211895910781, "grad_norm": 0.408442185495782, "learning_rate": 2.196470625138076e-06, "loss": 0.0292, "step": 15058 }, { "epoch": 0.699767657992565, "grad_norm": 0.6732256263546114, "learning_rate": 2.1958475645291834e-06, "loss": 0.054, "step": 15059 }, { "epoch": 0.699814126394052, "grad_norm": 0.510191576820257, "learning_rate": 2.1952245674372948e-06, "loss": 0.054, "step": 15060 }, { "epoch": 0.6998605947955391, "grad_norm": 0.7288023212292596, "learning_rate": 2.1946016338765216e-06, "loss": 0.0619, "step": 15061 }, { "epoch": 0.699907063197026, "grad_norm": 0.6197293010125581, "learning_rate": 2.1939787638609724e-06, "loss": 0.0496, "step": 15062 }, { "epoch": 0.699953531598513, "grad_norm": 0.9912092910334159, "learning_rate": 2.1933559574047588e-06, "loss": 0.078, "step": 15063 }, { "epoch": 0.7, "grad_norm": 0.5874333122028401, "learning_rate": 2.192733214521987e-06, "loss": 0.0456, "step": 15064 }, { "epoch": 0.700046468401487, "grad_norm": 0.6214221741651053, "learning_rate": 2.192110535226763e-06, "loss": 0.05, "step": 15065 }, { "epoch": 0.700092936802974, "grad_norm": 0.7100179128741789, "learning_rate": 2.19148791953319e-06, "loss": 0.0483, "step": 15066 }, { "epoch": 0.7001394052044609, "grad_norm": 0.6730647644491947, "learning_rate": 2.1908653674553716e-06, "loss": 0.0731, "step": 15067 }, { "epoch": 0.700185873605948, "grad_norm": 0.5002568566488022, "learning_rate": 2.1902428790074094e-06, "loss": 0.035, "step": 15068 }, { "epoch": 0.7002323420074349, "grad_norm": 1.0038058348877996, "learning_rate": 2.189620454203401e-06, "loss": 0.0601, "step": 15069 }, { "epoch": 0.7002788104089219, "grad_norm": 0.8592114832605964, "learning_rate": 2.1889980930574495e-06, "loss": 0.0587, "step": 15070 }, { "epoch": 0.700325278810409, "grad_norm": 0.6276587625623811, "learning_rate": 2.1883757955836515e-06, "loss": 0.062, "step": 15071 }, { "epoch": 0.7003717472118959, "grad_norm": 1.3365556362135367, "learning_rate": 2.187753561796097e-06, "loss": 0.0806, "step": 15072 }, { "epoch": 0.7004182156133829, "grad_norm": 0.34877433960739934, "learning_rate": 2.187131391708886e-06, "loss": 0.0357, "step": 15073 }, { "epoch": 0.7004646840148699, "grad_norm": 0.48726950428614135, "learning_rate": 2.1865092853361097e-06, "loss": 0.0419, "step": 15074 }, { "epoch": 0.7005111524163569, "grad_norm": 0.6617993346230746, "learning_rate": 2.1858872426918593e-06, "loss": 0.0703, "step": 15075 }, { "epoch": 0.7005576208178439, "grad_norm": 0.7322143669508347, "learning_rate": 2.1852652637902224e-06, "loss": 0.0573, "step": 15076 }, { "epoch": 0.7006040892193308, "grad_norm": 0.6285321532645342, "learning_rate": 2.1846433486452933e-06, "loss": 0.0655, "step": 15077 }, { "epoch": 0.7006505576208178, "grad_norm": 0.6847450510024262, "learning_rate": 2.1840214972711544e-06, "loss": 0.0514, "step": 15078 }, { "epoch": 0.7006970260223049, "grad_norm": 0.6881789124683362, "learning_rate": 2.1833997096818897e-06, "loss": 0.0553, "step": 15079 }, { "epoch": 0.7007434944237918, "grad_norm": 0.4409457976928214, "learning_rate": 2.1827779858915883e-06, "loss": 0.0536, "step": 15080 }, { "epoch": 0.7007899628252788, "grad_norm": 0.3998743925318735, "learning_rate": 2.18215632591433e-06, "loss": 0.0395, "step": 15081 }, { "epoch": 0.7008364312267658, "grad_norm": 0.47518307469710613, "learning_rate": 2.1815347297641977e-06, "loss": 0.0422, "step": 15082 }, { "epoch": 0.7008828996282528, "grad_norm": 0.6279960088443834, "learning_rate": 2.1809131974552693e-06, "loss": 0.0605, "step": 15083 }, { "epoch": 0.7009293680297398, "grad_norm": 0.5518294483807722, "learning_rate": 2.1802917290016247e-06, "loss": 0.0617, "step": 15084 }, { "epoch": 0.7009758364312267, "grad_norm": 0.4486461764530375, "learning_rate": 2.1796703244173395e-06, "loss": 0.0448, "step": 15085 }, { "epoch": 0.7010223048327138, "grad_norm": 0.5676494570753452, "learning_rate": 2.179048983716488e-06, "loss": 0.0627, "step": 15086 }, { "epoch": 0.7010687732342007, "grad_norm": 0.40003326409193807, "learning_rate": 2.178427706913148e-06, "loss": 0.0428, "step": 15087 }, { "epoch": 0.7011152416356877, "grad_norm": 0.3839458167727589, "learning_rate": 2.17780649402139e-06, "loss": 0.0553, "step": 15088 }, { "epoch": 0.7011617100371748, "grad_norm": 0.5969820765813776, "learning_rate": 2.1771853450552856e-06, "loss": 0.0667, "step": 15089 }, { "epoch": 0.7012081784386617, "grad_norm": 0.5007935605433302, "learning_rate": 2.1765642600289037e-06, "loss": 0.0738, "step": 15090 }, { "epoch": 0.7012546468401487, "grad_norm": 0.9207511810689262, "learning_rate": 2.175943238956313e-06, "loss": 0.0539, "step": 15091 }, { "epoch": 0.7013011152416357, "grad_norm": 0.5061480141565134, "learning_rate": 2.1753222818515802e-06, "loss": 0.0545, "step": 15092 }, { "epoch": 0.7013475836431227, "grad_norm": 0.5258855282301247, "learning_rate": 2.1747013887287686e-06, "loss": 0.0488, "step": 15093 }, { "epoch": 0.7013940520446097, "grad_norm": 0.6482648429326139, "learning_rate": 2.1740805596019465e-06, "loss": 0.0592, "step": 15094 }, { "epoch": 0.7014405204460966, "grad_norm": 0.44905435239523706, "learning_rate": 2.1734597944851755e-06, "loss": 0.0558, "step": 15095 }, { "epoch": 0.7014869888475836, "grad_norm": 0.4201889482468739, "learning_rate": 2.1728390933925113e-06, "loss": 0.0543, "step": 15096 }, { "epoch": 0.7015334572490707, "grad_norm": 0.6221716564521846, "learning_rate": 2.172218456338019e-06, "loss": 0.0678, "step": 15097 }, { "epoch": 0.7015799256505576, "grad_norm": 0.4390499859717949, "learning_rate": 2.1715978833357547e-06, "loss": 0.0358, "step": 15098 }, { "epoch": 0.7016263940520446, "grad_norm": 0.4645053146472429, "learning_rate": 2.1709773743997758e-06, "loss": 0.0512, "step": 15099 }, { "epoch": 0.7016728624535316, "grad_norm": 1.2143336664278015, "learning_rate": 2.1703569295441348e-06, "loss": 0.0646, "step": 15100 }, { "epoch": 0.7017193308550186, "grad_norm": 0.3428953944756961, "learning_rate": 2.1697365487828907e-06, "loss": 0.0412, "step": 15101 }, { "epoch": 0.7017657992565056, "grad_norm": 0.49221818154332847, "learning_rate": 2.169116232130091e-06, "loss": 0.0518, "step": 15102 }, { "epoch": 0.7018122676579925, "grad_norm": 0.7059953406759466, "learning_rate": 2.168495979599786e-06, "loss": 0.0533, "step": 15103 }, { "epoch": 0.7018587360594796, "grad_norm": 0.4487764100919728, "learning_rate": 2.1678757912060293e-06, "loss": 0.0519, "step": 15104 }, { "epoch": 0.7019052044609665, "grad_norm": 0.6084182145126537, "learning_rate": 2.167255666962867e-06, "loss": 0.0631, "step": 15105 }, { "epoch": 0.7019516728624535, "grad_norm": 0.43188389505363195, "learning_rate": 2.1666356068843447e-06, "loss": 0.0384, "step": 15106 }, { "epoch": 0.7019981412639406, "grad_norm": 0.565810903935008, "learning_rate": 2.1660156109845083e-06, "loss": 0.0491, "step": 15107 }, { "epoch": 0.7020446096654275, "grad_norm": 0.650861727341501, "learning_rate": 2.1653956792774005e-06, "loss": 0.0736, "step": 15108 }, { "epoch": 0.7020910780669145, "grad_norm": 0.5616345758452688, "learning_rate": 2.1647758117770647e-06, "loss": 0.0462, "step": 15109 }, { "epoch": 0.7021375464684015, "grad_norm": 0.31583965189947977, "learning_rate": 2.1641560084975383e-06, "loss": 0.0301, "step": 15110 }, { "epoch": 0.7021840148698885, "grad_norm": 0.5662384893851014, "learning_rate": 2.163536269452865e-06, "loss": 0.0496, "step": 15111 }, { "epoch": 0.7022304832713755, "grad_norm": 0.6724574514554641, "learning_rate": 2.1629165946570806e-06, "loss": 0.0565, "step": 15112 }, { "epoch": 0.7022769516728624, "grad_norm": 0.35926137884900455, "learning_rate": 2.1622969841242213e-06, "loss": 0.0459, "step": 15113 }, { "epoch": 0.7023234200743494, "grad_norm": 0.6037889388410465, "learning_rate": 2.161677437868322e-06, "loss": 0.0719, "step": 15114 }, { "epoch": 0.7023698884758365, "grad_norm": 0.48413580158318664, "learning_rate": 2.161057955903416e-06, "loss": 0.0398, "step": 15115 }, { "epoch": 0.7024163568773234, "grad_norm": 0.3771328082100153, "learning_rate": 2.1604385382435355e-06, "loss": 0.0454, "step": 15116 }, { "epoch": 0.7024628252788104, "grad_norm": 0.9786695049906492, "learning_rate": 2.159819184902708e-06, "loss": 0.0627, "step": 15117 }, { "epoch": 0.7025092936802974, "grad_norm": 0.6676971356950584, "learning_rate": 2.15919989589497e-06, "loss": 0.0519, "step": 15118 }, { "epoch": 0.7025557620817844, "grad_norm": 0.6535462246284309, "learning_rate": 2.1585806712343417e-06, "loss": 0.0641, "step": 15119 }, { "epoch": 0.7026022304832714, "grad_norm": 6.722475562860021, "learning_rate": 2.15796151093485e-06, "loss": 0.0461, "step": 15120 }, { "epoch": 0.7026486988847583, "grad_norm": 0.6178044009572526, "learning_rate": 2.1573424150105234e-06, "loss": 0.0626, "step": 15121 }, { "epoch": 0.7026951672862454, "grad_norm": 0.44613706934728625, "learning_rate": 2.1567233834753827e-06, "loss": 0.0284, "step": 15122 }, { "epoch": 0.7027416356877323, "grad_norm": 0.7737776905339787, "learning_rate": 2.1561044163434498e-06, "loss": 0.062, "step": 15123 }, { "epoch": 0.7027881040892193, "grad_norm": 2.1903227865959543, "learning_rate": 2.155485513628745e-06, "loss": 0.0748, "step": 15124 }, { "epoch": 0.7028345724907064, "grad_norm": 0.7255126107935027, "learning_rate": 2.154866675345287e-06, "loss": 0.0346, "step": 15125 }, { "epoch": 0.7028810408921933, "grad_norm": 0.5349253700358102, "learning_rate": 2.1542479015070933e-06, "loss": 0.0488, "step": 15126 }, { "epoch": 0.7029275092936803, "grad_norm": 0.8256719223078494, "learning_rate": 2.153629192128178e-06, "loss": 0.0692, "step": 15127 }, { "epoch": 0.7029739776951673, "grad_norm": 0.265139266944961, "learning_rate": 2.1530105472225594e-06, "loss": 0.0357, "step": 15128 }, { "epoch": 0.7030204460966543, "grad_norm": 0.6633364355794804, "learning_rate": 2.1523919668042474e-06, "loss": 0.0466, "step": 15129 }, { "epoch": 0.7030669144981413, "grad_norm": 0.44216257688245314, "learning_rate": 2.1517734508872555e-06, "loss": 0.0529, "step": 15130 }, { "epoch": 0.7031133828996282, "grad_norm": 0.5753387692649752, "learning_rate": 2.151154999485592e-06, "loss": 0.0508, "step": 15131 }, { "epoch": 0.7031598513011152, "grad_norm": 0.48349568401584114, "learning_rate": 2.1505366126132665e-06, "loss": 0.056, "step": 15132 }, { "epoch": 0.7032063197026023, "grad_norm": 0.6720756247494151, "learning_rate": 2.149918290284285e-06, "loss": 0.0473, "step": 15133 }, { "epoch": 0.7032527881040892, "grad_norm": 0.6367840366373496, "learning_rate": 2.1493000325126527e-06, "loss": 0.0515, "step": 15134 }, { "epoch": 0.7032992565055762, "grad_norm": 0.7549039700366194, "learning_rate": 2.1486818393123766e-06, "loss": 0.0698, "step": 15135 }, { "epoch": 0.7033457249070632, "grad_norm": 0.8639170422717339, "learning_rate": 2.14806371069746e-06, "loss": 0.0581, "step": 15136 }, { "epoch": 0.7033921933085502, "grad_norm": 0.4687314868024322, "learning_rate": 2.1474456466818977e-06, "loss": 0.0623, "step": 15137 }, { "epoch": 0.7034386617100372, "grad_norm": 0.5563154924432204, "learning_rate": 2.146827647279696e-06, "loss": 0.0451, "step": 15138 }, { "epoch": 0.7034851301115241, "grad_norm": 0.5135721269426148, "learning_rate": 2.146209712504851e-06, "loss": 0.0536, "step": 15139 }, { "epoch": 0.7035315985130112, "grad_norm": 0.4635634122771417, "learning_rate": 2.1455918423713597e-06, "loss": 0.0564, "step": 15140 }, { "epoch": 0.7035780669144981, "grad_norm": 0.596692149199329, "learning_rate": 2.144974036893215e-06, "loss": 0.0671, "step": 15141 }, { "epoch": 0.7036245353159851, "grad_norm": 0.6178250796465588, "learning_rate": 2.144356296084417e-06, "loss": 0.0605, "step": 15142 }, { "epoch": 0.7036710037174722, "grad_norm": 0.8154858304722468, "learning_rate": 2.1437386199589526e-06, "loss": 0.0813, "step": 15143 }, { "epoch": 0.7037174721189591, "grad_norm": 2.0182689091962436, "learning_rate": 2.143121008530813e-06, "loss": 0.0429, "step": 15144 }, { "epoch": 0.7037639405204461, "grad_norm": 0.5083356805254975, "learning_rate": 2.142503461813991e-06, "loss": 0.0643, "step": 15145 }, { "epoch": 0.7038104089219331, "grad_norm": 0.48857475210230833, "learning_rate": 2.1418859798224735e-06, "loss": 0.0375, "step": 15146 }, { "epoch": 0.7038568773234201, "grad_norm": 0.41788060289830753, "learning_rate": 2.141268562570246e-06, "loss": 0.0531, "step": 15147 }, { "epoch": 0.7039033457249071, "grad_norm": 0.5716731840235797, "learning_rate": 2.140651210071295e-06, "loss": 0.0464, "step": 15148 }, { "epoch": 0.703949814126394, "grad_norm": 0.7184905673192719, "learning_rate": 2.1400339223396033e-06, "loss": 0.0501, "step": 15149 }, { "epoch": 0.703996282527881, "grad_norm": 0.24873028032294034, "learning_rate": 2.139416699389153e-06, "loss": 0.0241, "step": 15150 }, { "epoch": 0.7040427509293681, "grad_norm": 0.665325306483338, "learning_rate": 2.1387995412339235e-06, "loss": 0.068, "step": 15151 }, { "epoch": 0.704089219330855, "grad_norm": 0.5098657282494402, "learning_rate": 2.1381824478878977e-06, "loss": 0.053, "step": 15152 }, { "epoch": 0.704135687732342, "grad_norm": 1.158027927420723, "learning_rate": 2.1375654193650514e-06, "loss": 0.0403, "step": 15153 }, { "epoch": 0.704182156133829, "grad_norm": 0.46695614549287834, "learning_rate": 2.1369484556793606e-06, "loss": 0.0366, "step": 15154 }, { "epoch": 0.704228624535316, "grad_norm": 0.9860605310747361, "learning_rate": 2.1363315568448013e-06, "loss": 0.0503, "step": 15155 }, { "epoch": 0.704275092936803, "grad_norm": 0.6313817479872508, "learning_rate": 2.1357147228753463e-06, "loss": 0.0635, "step": 15156 }, { "epoch": 0.7043215613382899, "grad_norm": 0.5131924882493033, "learning_rate": 2.1350979537849664e-06, "loss": 0.0469, "step": 15157 }, { "epoch": 0.704368029739777, "grad_norm": 0.3773746405804033, "learning_rate": 2.134481249587632e-06, "loss": 0.0346, "step": 15158 }, { "epoch": 0.7044144981412639, "grad_norm": 0.46735248168989346, "learning_rate": 2.1338646102973147e-06, "loss": 0.0583, "step": 15159 }, { "epoch": 0.7044609665427509, "grad_norm": 0.8096309737297326, "learning_rate": 2.1332480359279827e-06, "loss": 0.0453, "step": 15160 }, { "epoch": 0.704507434944238, "grad_norm": 0.7133325810174141, "learning_rate": 2.132631526493596e-06, "loss": 0.0487, "step": 15161 }, { "epoch": 0.7045539033457249, "grad_norm": 0.8731785868374818, "learning_rate": 2.1320150820081255e-06, "loss": 0.0534, "step": 15162 }, { "epoch": 0.7046003717472119, "grad_norm": 0.491865608445845, "learning_rate": 2.131398702485531e-06, "loss": 0.0477, "step": 15163 }, { "epoch": 0.7046468401486989, "grad_norm": 0.5574649410109106, "learning_rate": 2.1307823879397756e-06, "loss": 0.0647, "step": 15164 }, { "epoch": 0.7046933085501859, "grad_norm": 0.4739479969072752, "learning_rate": 2.1301661383848165e-06, "loss": 0.0484, "step": 15165 }, { "epoch": 0.7047397769516729, "grad_norm": 0.34640486160210343, "learning_rate": 2.129549953834619e-06, "loss": 0.0468, "step": 15166 }, { "epoch": 0.7047862453531598, "grad_norm": 0.49503643289009086, "learning_rate": 2.1289338343031342e-06, "loss": 0.0598, "step": 15167 }, { "epoch": 0.7048327137546468, "grad_norm": 0.621378478722138, "learning_rate": 2.128317779804318e-06, "loss": 0.0436, "step": 15168 }, { "epoch": 0.7048791821561339, "grad_norm": 0.6568191226162895, "learning_rate": 2.1277017903521284e-06, "loss": 0.0692, "step": 15169 }, { "epoch": 0.7049256505576208, "grad_norm": 0.5258276855434959, "learning_rate": 2.127085865960516e-06, "loss": 0.0576, "step": 15170 }, { "epoch": 0.7049721189591078, "grad_norm": 0.5902911076798724, "learning_rate": 2.126470006643432e-06, "loss": 0.0616, "step": 15171 }, { "epoch": 0.7050185873605948, "grad_norm": 0.41773576461366374, "learning_rate": 2.125854212414828e-06, "loss": 0.0502, "step": 15172 }, { "epoch": 0.7050650557620818, "grad_norm": 0.7368646499488738, "learning_rate": 2.12523848328865e-06, "loss": 0.0751, "step": 15173 }, { "epoch": 0.7051115241635688, "grad_norm": 0.3512989301382558, "learning_rate": 2.1246228192788465e-06, "loss": 0.0399, "step": 15174 }, { "epoch": 0.7051579925650557, "grad_norm": 0.4418365267600559, "learning_rate": 2.1240072203993606e-06, "loss": 0.0399, "step": 15175 }, { "epoch": 0.7052044609665428, "grad_norm": 0.6061145514614231, "learning_rate": 2.1233916866641395e-06, "loss": 0.0623, "step": 15176 }, { "epoch": 0.7052509293680297, "grad_norm": 0.43748755558304137, "learning_rate": 2.1227762180871264e-06, "loss": 0.0525, "step": 15177 }, { "epoch": 0.7052973977695167, "grad_norm": 0.4123362504315743, "learning_rate": 2.1221608146822564e-06, "loss": 0.0597, "step": 15178 }, { "epoch": 0.7053438661710038, "grad_norm": 0.5566131861185086, "learning_rate": 2.1215454764634746e-06, "loss": 0.0564, "step": 15179 }, { "epoch": 0.7053903345724907, "grad_norm": 0.45493362852712005, "learning_rate": 2.120930203444717e-06, "loss": 0.0438, "step": 15180 }, { "epoch": 0.7054368029739777, "grad_norm": 0.46237572239351826, "learning_rate": 2.1203149956399212e-06, "loss": 0.0555, "step": 15181 }, { "epoch": 0.7054832713754647, "grad_norm": 0.4875603214458455, "learning_rate": 2.1196998530630193e-06, "loss": 0.0464, "step": 15182 }, { "epoch": 0.7055297397769517, "grad_norm": 0.7307080781597916, "learning_rate": 2.119084775727951e-06, "loss": 0.0499, "step": 15183 }, { "epoch": 0.7055762081784387, "grad_norm": 0.7357821864283195, "learning_rate": 2.118469763648643e-06, "loss": 0.0595, "step": 15184 }, { "epoch": 0.7056226765799256, "grad_norm": 0.6958784067028725, "learning_rate": 2.117854816839025e-06, "loss": 0.0552, "step": 15185 }, { "epoch": 0.7056691449814126, "grad_norm": 0.6503482251661984, "learning_rate": 2.117239935313031e-06, "loss": 0.0695, "step": 15186 }, { "epoch": 0.7057156133828997, "grad_norm": 0.6632922705551948, "learning_rate": 2.116625119084587e-06, "loss": 0.0452, "step": 15187 }, { "epoch": 0.7057620817843866, "grad_norm": 0.7461248686484412, "learning_rate": 2.116010368167618e-06, "loss": 0.0471, "step": 15188 }, { "epoch": 0.7058085501858736, "grad_norm": 0.5695884206745105, "learning_rate": 2.11539568257605e-06, "loss": 0.0558, "step": 15189 }, { "epoch": 0.7058550185873605, "grad_norm": 0.43830832314386553, "learning_rate": 2.1147810623238056e-06, "loss": 0.0544, "step": 15190 }, { "epoch": 0.7059014869888476, "grad_norm": 0.4474846089183392, "learning_rate": 2.1141665074248067e-06, "loss": 0.0257, "step": 15191 }, { "epoch": 0.7059479553903346, "grad_norm": 0.6678286311707883, "learning_rate": 2.1135520178929725e-06, "loss": 0.0524, "step": 15192 }, { "epoch": 0.7059944237918215, "grad_norm": 0.631609991319107, "learning_rate": 2.112937593742224e-06, "loss": 0.0678, "step": 15193 }, { "epoch": 0.7060408921933086, "grad_norm": 0.4025536580283862, "learning_rate": 2.112323234986478e-06, "loss": 0.0478, "step": 15194 }, { "epoch": 0.7060873605947955, "grad_norm": 0.5239887200143852, "learning_rate": 2.1117089416396502e-06, "loss": 0.0537, "step": 15195 }, { "epoch": 0.7061338289962825, "grad_norm": 0.6707901092944848, "learning_rate": 2.1110947137156545e-06, "loss": 0.0724, "step": 15196 }, { "epoch": 0.7061802973977696, "grad_norm": 0.6242727855242144, "learning_rate": 2.110480551228404e-06, "loss": 0.0589, "step": 15197 }, { "epoch": 0.7062267657992565, "grad_norm": 0.9852793910008603, "learning_rate": 2.10986645419181e-06, "loss": 0.0448, "step": 15198 }, { "epoch": 0.7062732342007435, "grad_norm": 0.5872348002252775, "learning_rate": 2.1092524226197806e-06, "loss": 0.0484, "step": 15199 }, { "epoch": 0.7063197026022305, "grad_norm": 0.5160079225718455, "learning_rate": 2.108638456526229e-06, "loss": 0.0466, "step": 15200 }, { "epoch": 0.7063661710037175, "grad_norm": 0.5979250292967359, "learning_rate": 2.1080245559250606e-06, "loss": 0.0527, "step": 15201 }, { "epoch": 0.7064126394052045, "grad_norm": 0.5833363420848051, "learning_rate": 2.1074107208301764e-06, "loss": 0.0441, "step": 15202 }, { "epoch": 0.7064591078066914, "grad_norm": 0.5706639519749538, "learning_rate": 2.1067969512554854e-06, "loss": 0.0478, "step": 15203 }, { "epoch": 0.7065055762081784, "grad_norm": 0.4549863938617455, "learning_rate": 2.1061832472148886e-06, "loss": 0.0537, "step": 15204 }, { "epoch": 0.7065520446096655, "grad_norm": 0.3555018418716161, "learning_rate": 2.105569608722287e-06, "loss": 0.0267, "step": 15205 }, { "epoch": 0.7065985130111524, "grad_norm": 0.4851231566550995, "learning_rate": 2.104956035791578e-06, "loss": 0.0365, "step": 15206 }, { "epoch": 0.7066449814126394, "grad_norm": 0.8103528710764418, "learning_rate": 2.1043425284366654e-06, "loss": 0.0589, "step": 15207 }, { "epoch": 0.7066914498141263, "grad_norm": 0.3444923562065016, "learning_rate": 2.1037290866714406e-06, "loss": 0.0425, "step": 15208 }, { "epoch": 0.7067379182156134, "grad_norm": 0.8391505877936527, "learning_rate": 2.103115710509798e-06, "loss": 0.0615, "step": 15209 }, { "epoch": 0.7067843866171004, "grad_norm": 0.3287195514988062, "learning_rate": 2.1025023999656363e-06, "loss": 0.0455, "step": 15210 }, { "epoch": 0.7068308550185873, "grad_norm": 0.7372725358842018, "learning_rate": 2.101889155052844e-06, "loss": 0.0832, "step": 15211 }, { "epoch": 0.7068773234200744, "grad_norm": 0.560056386723915, "learning_rate": 2.1012759757853125e-06, "loss": 0.0505, "step": 15212 }, { "epoch": 0.7069237918215613, "grad_norm": 0.39681512076801584, "learning_rate": 2.1006628621769314e-06, "loss": 0.0528, "step": 15213 }, { "epoch": 0.7069702602230483, "grad_norm": 0.4342759172457726, "learning_rate": 2.1000498142415876e-06, "loss": 0.0377, "step": 15214 }, { "epoch": 0.7070167286245354, "grad_norm": 0.4348272425258785, "learning_rate": 2.0994368319931678e-06, "loss": 0.0546, "step": 15215 }, { "epoch": 0.7070631970260223, "grad_norm": 0.5274726018473879, "learning_rate": 2.0988239154455547e-06, "loss": 0.043, "step": 15216 }, { "epoch": 0.7071096654275093, "grad_norm": 0.4984367766119705, "learning_rate": 2.098211064612635e-06, "loss": 0.0495, "step": 15217 }, { "epoch": 0.7071561338289963, "grad_norm": 0.3835159626272436, "learning_rate": 2.0975982795082895e-06, "loss": 0.0279, "step": 15218 }, { "epoch": 0.7072026022304833, "grad_norm": 0.4913086097181878, "learning_rate": 2.0969855601463966e-06, "loss": 0.0587, "step": 15219 }, { "epoch": 0.7072490706319703, "grad_norm": 0.6227035432523738, "learning_rate": 2.0963729065408374e-06, "loss": 0.0583, "step": 15220 }, { "epoch": 0.7072955390334572, "grad_norm": 0.41707268171257506, "learning_rate": 2.095760318705487e-06, "loss": 0.048, "step": 15221 }, { "epoch": 0.7073420074349442, "grad_norm": 0.2971867736551299, "learning_rate": 2.0951477966542222e-06, "loss": 0.0252, "step": 15222 }, { "epoch": 0.7073884758364313, "grad_norm": 0.46812474421515016, "learning_rate": 2.094535340400915e-06, "loss": 0.0618, "step": 15223 }, { "epoch": 0.7074349442379182, "grad_norm": 0.4166965905751388, "learning_rate": 2.093922949959443e-06, "loss": 0.0593, "step": 15224 }, { "epoch": 0.7074814126394052, "grad_norm": 0.634087238129226, "learning_rate": 2.0933106253436754e-06, "loss": 0.0655, "step": 15225 }, { "epoch": 0.7075278810408921, "grad_norm": 0.6192896138224298, "learning_rate": 2.0926983665674783e-06, "loss": 0.0405, "step": 15226 }, { "epoch": 0.7075743494423792, "grad_norm": 0.7007532628885234, "learning_rate": 2.092086173644725e-06, "loss": 0.0417, "step": 15227 }, { "epoch": 0.7076208178438662, "grad_norm": 0.5789283409270226, "learning_rate": 2.0914740465892803e-06, "loss": 0.0447, "step": 15228 }, { "epoch": 0.7076672862453531, "grad_norm": 0.5599301162652075, "learning_rate": 2.0908619854150095e-06, "loss": 0.0504, "step": 15229 }, { "epoch": 0.7077137546468402, "grad_norm": 1.0649428172588657, "learning_rate": 2.090249990135775e-06, "loss": 0.0793, "step": 15230 }, { "epoch": 0.7077602230483271, "grad_norm": 0.622648608803546, "learning_rate": 2.0896380607654445e-06, "loss": 0.0617, "step": 15231 }, { "epoch": 0.7078066914498141, "grad_norm": 0.5968897559373594, "learning_rate": 2.0890261973178727e-06, "loss": 0.0552, "step": 15232 }, { "epoch": 0.7078531598513012, "grad_norm": 0.4964770994945894, "learning_rate": 2.0884143998069196e-06, "loss": 0.0601, "step": 15233 }, { "epoch": 0.7078996282527881, "grad_norm": 0.5668829709363605, "learning_rate": 2.0878026682464464e-06, "loss": 0.0535, "step": 15234 }, { "epoch": 0.7079460966542751, "grad_norm": 0.4182452309602715, "learning_rate": 2.0871910026503077e-06, "loss": 0.0537, "step": 15235 }, { "epoch": 0.7079925650557621, "grad_norm": 0.1756170576551342, "learning_rate": 2.0865794030323588e-06, "loss": 0.019, "step": 15236 }, { "epoch": 0.7080390334572491, "grad_norm": 0.44862886014558584, "learning_rate": 2.085967869406453e-06, "loss": 0.0341, "step": 15237 }, { "epoch": 0.7080855018587361, "grad_norm": 0.4236220242230409, "learning_rate": 2.0853564017864414e-06, "loss": 0.0305, "step": 15238 }, { "epoch": 0.708131970260223, "grad_norm": 0.8311483677865386, "learning_rate": 2.084745000186175e-06, "loss": 0.0636, "step": 15239 }, { "epoch": 0.70817843866171, "grad_norm": 0.5752928884610382, "learning_rate": 2.084133664619501e-06, "loss": 0.0637, "step": 15240 }, { "epoch": 0.7082249070631971, "grad_norm": 0.3818635481948142, "learning_rate": 2.08352239510027e-06, "loss": 0.0458, "step": 15241 }, { "epoch": 0.708271375464684, "grad_norm": 0.6118125264077219, "learning_rate": 2.0829111916423284e-06, "loss": 0.0511, "step": 15242 }, { "epoch": 0.708317843866171, "grad_norm": 0.40477511826099094, "learning_rate": 2.0823000542595147e-06, "loss": 0.0442, "step": 15243 }, { "epoch": 0.708364312267658, "grad_norm": 0.5875280334783964, "learning_rate": 2.0816889829656772e-06, "loss": 0.0421, "step": 15244 }, { "epoch": 0.708410780669145, "grad_norm": 0.707645838910715, "learning_rate": 2.081077977774656e-06, "loss": 0.0574, "step": 15245 }, { "epoch": 0.708457249070632, "grad_norm": 0.4471113346316989, "learning_rate": 2.0804670387002907e-06, "loss": 0.0644, "step": 15246 }, { "epoch": 0.7085037174721189, "grad_norm": 0.44434763015637796, "learning_rate": 2.0798561657564177e-06, "loss": 0.0664, "step": 15247 }, { "epoch": 0.708550185873606, "grad_norm": 0.4809406420803257, "learning_rate": 2.0792453589568795e-06, "loss": 0.0361, "step": 15248 }, { "epoch": 0.7085966542750929, "grad_norm": 0.8083946770866965, "learning_rate": 2.0786346183155065e-06, "loss": 0.0602, "step": 15249 }, { "epoch": 0.7086431226765799, "grad_norm": 0.5761794359362874, "learning_rate": 2.078023943846132e-06, "loss": 0.0555, "step": 15250 }, { "epoch": 0.708689591078067, "grad_norm": 0.3801711575784802, "learning_rate": 2.0774133355625925e-06, "loss": 0.0416, "step": 15251 }, { "epoch": 0.7087360594795539, "grad_norm": 0.5723188643779877, "learning_rate": 2.076802793478717e-06, "loss": 0.0407, "step": 15252 }, { "epoch": 0.7087825278810409, "grad_norm": 0.5043718438058493, "learning_rate": 2.076192317608334e-06, "loss": 0.0381, "step": 15253 }, { "epoch": 0.7088289962825279, "grad_norm": 0.6237620807142843, "learning_rate": 2.0755819079652734e-06, "loss": 0.0679, "step": 15254 }, { "epoch": 0.7088754646840149, "grad_norm": 0.45088072689021913, "learning_rate": 2.0749715645633593e-06, "loss": 0.0551, "step": 15255 }, { "epoch": 0.7089219330855019, "grad_norm": 0.7635435299370547, "learning_rate": 2.0743612874164186e-06, "loss": 0.0608, "step": 15256 }, { "epoch": 0.7089684014869888, "grad_norm": 0.739297075406505, "learning_rate": 2.0737510765382724e-06, "loss": 0.0506, "step": 15257 }, { "epoch": 0.7090148698884758, "grad_norm": 1.2469151294011074, "learning_rate": 2.073140931942745e-06, "loss": 0.0521, "step": 15258 }, { "epoch": 0.7090613382899629, "grad_norm": 1.1157874617467967, "learning_rate": 2.0725308536436566e-06, "loss": 0.0714, "step": 15259 }, { "epoch": 0.7091078066914498, "grad_norm": 0.48877927006120436, "learning_rate": 2.071920841654825e-06, "loss": 0.0497, "step": 15260 }, { "epoch": 0.7091542750929368, "grad_norm": 0.7726391025887309, "learning_rate": 2.071310895990069e-06, "loss": 0.0568, "step": 15261 }, { "epoch": 0.7092007434944237, "grad_norm": 0.5664458450403927, "learning_rate": 2.070701016663203e-06, "loss": 0.0623, "step": 15262 }, { "epoch": 0.7092472118959108, "grad_norm": 0.5186306261390085, "learning_rate": 2.070091203688042e-06, "loss": 0.0562, "step": 15263 }, { "epoch": 0.7092936802973978, "grad_norm": 0.5009460952805861, "learning_rate": 2.0694814570783974e-06, "loss": 0.0474, "step": 15264 }, { "epoch": 0.7093401486988847, "grad_norm": 1.6876509056337525, "learning_rate": 2.068871776848083e-06, "loss": 0.0535, "step": 15265 }, { "epoch": 0.7093866171003718, "grad_norm": 0.3408666032643653, "learning_rate": 2.0682621630109104e-06, "loss": 0.0518, "step": 15266 }, { "epoch": 0.7094330855018587, "grad_norm": 0.4917798666207295, "learning_rate": 2.067652615580682e-06, "loss": 0.0493, "step": 15267 }, { "epoch": 0.7094795539033457, "grad_norm": 0.49210455295794503, "learning_rate": 2.0670431345712092e-06, "loss": 0.0438, "step": 15268 }, { "epoch": 0.7095260223048327, "grad_norm": 0.43329230869363733, "learning_rate": 2.066433719996297e-06, "loss": 0.0412, "step": 15269 }, { "epoch": 0.7095724907063197, "grad_norm": 0.6353688937164051, "learning_rate": 2.065824371869748e-06, "loss": 0.0663, "step": 15270 }, { "epoch": 0.7096189591078067, "grad_norm": 0.6465851106736791, "learning_rate": 2.0652150902053636e-06, "loss": 0.0573, "step": 15271 }, { "epoch": 0.7096654275092937, "grad_norm": 0.3882478962541084, "learning_rate": 2.06460587501695e-06, "loss": 0.0341, "step": 15272 }, { "epoch": 0.7097118959107807, "grad_norm": 0.35341880828107375, "learning_rate": 2.063996726318301e-06, "loss": 0.03, "step": 15273 }, { "epoch": 0.7097583643122677, "grad_norm": 0.4592072877872598, "learning_rate": 2.0633876441232137e-06, "loss": 0.0671, "step": 15274 }, { "epoch": 0.7098048327137546, "grad_norm": 0.5234632947597392, "learning_rate": 2.0627786284454893e-06, "loss": 0.0545, "step": 15275 }, { "epoch": 0.7098513011152416, "grad_norm": 0.6875384685489215, "learning_rate": 2.0621696792989206e-06, "loss": 0.0488, "step": 15276 }, { "epoch": 0.7098977695167287, "grad_norm": 0.583697430541576, "learning_rate": 2.0615607966973e-06, "loss": 0.0568, "step": 15277 }, { "epoch": 0.7099442379182156, "grad_norm": 0.5700848853617524, "learning_rate": 2.0609519806544203e-06, "loss": 0.0465, "step": 15278 }, { "epoch": 0.7099907063197026, "grad_norm": 0.41641535598170726, "learning_rate": 2.060343231184072e-06, "loss": 0.0624, "step": 15279 }, { "epoch": 0.7100371747211895, "grad_norm": 0.6223049193807276, "learning_rate": 2.059734548300043e-06, "loss": 0.0523, "step": 15280 }, { "epoch": 0.7100836431226766, "grad_norm": 0.7280207655867523, "learning_rate": 2.0591259320161194e-06, "loss": 0.0601, "step": 15281 }, { "epoch": 0.7101301115241636, "grad_norm": 0.47271285966964266, "learning_rate": 2.05851738234609e-06, "loss": 0.0511, "step": 15282 }, { "epoch": 0.7101765799256505, "grad_norm": 1.2747238208727314, "learning_rate": 2.0579088993037384e-06, "loss": 0.0446, "step": 15283 }, { "epoch": 0.7102230483271376, "grad_norm": 0.42568119725008674, "learning_rate": 2.0573004829028465e-06, "loss": 0.0574, "step": 15284 }, { "epoch": 0.7102695167286245, "grad_norm": 0.7006033194407877, "learning_rate": 2.0566921331571952e-06, "loss": 0.0662, "step": 15285 }, { "epoch": 0.7103159851301115, "grad_norm": 0.6333001135661016, "learning_rate": 2.0560838500805657e-06, "loss": 0.0533, "step": 15286 }, { "epoch": 0.7103624535315985, "grad_norm": 0.59560644327317, "learning_rate": 2.055475633686735e-06, "loss": 0.0408, "step": 15287 }, { "epoch": 0.7104089219330855, "grad_norm": 0.7363018197122142, "learning_rate": 2.054867483989479e-06, "loss": 0.0641, "step": 15288 }, { "epoch": 0.7104553903345725, "grad_norm": 0.6724343360534819, "learning_rate": 2.054259401002575e-06, "loss": 0.058, "step": 15289 }, { "epoch": 0.7105018587360595, "grad_norm": 0.7187961321169165, "learning_rate": 2.0536513847397983e-06, "loss": 0.0613, "step": 15290 }, { "epoch": 0.7105483271375465, "grad_norm": 0.5295563866291768, "learning_rate": 2.0530434352149153e-06, "loss": 0.0766, "step": 15291 }, { "epoch": 0.7105947955390335, "grad_norm": 0.526188032634726, "learning_rate": 2.0524355524417017e-06, "loss": 0.0729, "step": 15292 }, { "epoch": 0.7106412639405204, "grad_norm": 0.5987529870073371, "learning_rate": 2.051827736433925e-06, "loss": 0.0524, "step": 15293 }, { "epoch": 0.7106877323420074, "grad_norm": 0.26717666412536795, "learning_rate": 2.0512199872053527e-06, "loss": 0.034, "step": 15294 }, { "epoch": 0.7107342007434945, "grad_norm": 0.41072659758006336, "learning_rate": 2.050612304769752e-06, "loss": 0.0463, "step": 15295 }, { "epoch": 0.7107806691449814, "grad_norm": 0.6487422212416903, "learning_rate": 2.050004689140886e-06, "loss": 0.0639, "step": 15296 }, { "epoch": 0.7108271375464684, "grad_norm": 0.8062586462733671, "learning_rate": 2.0493971403325193e-06, "loss": 0.065, "step": 15297 }, { "epoch": 0.7108736059479553, "grad_norm": 1.0918842822116088, "learning_rate": 2.0487896583584103e-06, "loss": 0.0657, "step": 15298 }, { "epoch": 0.7109200743494424, "grad_norm": 1.3666023038170443, "learning_rate": 2.048182243232324e-06, "loss": 0.0493, "step": 15299 }, { "epoch": 0.7109665427509294, "grad_norm": 0.288071291043001, "learning_rate": 2.0475748949680162e-06, "loss": 0.043, "step": 15300 }, { "epoch": 0.7110130111524163, "grad_norm": 0.689962580498761, "learning_rate": 2.046967613579245e-06, "loss": 0.0494, "step": 15301 }, { "epoch": 0.7110594795539034, "grad_norm": 0.5205288893336025, "learning_rate": 2.0463603990797647e-06, "loss": 0.0408, "step": 15302 }, { "epoch": 0.7111059479553903, "grad_norm": 0.49322616832072314, "learning_rate": 2.04575325148333e-06, "loss": 0.0495, "step": 15303 }, { "epoch": 0.7111524163568773, "grad_norm": 0.4780783950432201, "learning_rate": 2.045146170803694e-06, "loss": 0.0472, "step": 15304 }, { "epoch": 0.7111988847583643, "grad_norm": 0.6391845053438047, "learning_rate": 2.0445391570546045e-06, "loss": 0.0601, "step": 15305 }, { "epoch": 0.7112453531598513, "grad_norm": 0.3236854628298892, "learning_rate": 2.0439322102498165e-06, "loss": 0.0614, "step": 15306 }, { "epoch": 0.7112918215613383, "grad_norm": 0.7642578539354281, "learning_rate": 2.0433253304030765e-06, "loss": 0.048, "step": 15307 }, { "epoch": 0.7113382899628253, "grad_norm": 0.6613985946560985, "learning_rate": 2.042718517528126e-06, "loss": 0.0628, "step": 15308 }, { "epoch": 0.7113847583643123, "grad_norm": 0.6342387091810046, "learning_rate": 2.0421117716387153e-06, "loss": 0.0667, "step": 15309 }, { "epoch": 0.7114312267657993, "grad_norm": 0.46611965487065493, "learning_rate": 2.0415050927485862e-06, "loss": 0.0401, "step": 15310 }, { "epoch": 0.7114776951672862, "grad_norm": 0.33751439701692754, "learning_rate": 2.0408984808714807e-06, "loss": 0.0494, "step": 15311 }, { "epoch": 0.7115241635687732, "grad_norm": 0.5171893689164156, "learning_rate": 2.0402919360211366e-06, "loss": 0.0313, "step": 15312 }, { "epoch": 0.7115706319702603, "grad_norm": 1.1998942772784742, "learning_rate": 2.039685458211299e-06, "loss": 0.0544, "step": 15313 }, { "epoch": 0.7116171003717472, "grad_norm": 0.6189331985106841, "learning_rate": 2.0390790474557e-06, "loss": 0.0441, "step": 15314 }, { "epoch": 0.7116635687732342, "grad_norm": 0.49810183596200647, "learning_rate": 2.0384727037680745e-06, "loss": 0.0429, "step": 15315 }, { "epoch": 0.7117100371747211, "grad_norm": 0.552858939779164, "learning_rate": 2.037866427162162e-06, "loss": 0.05, "step": 15316 }, { "epoch": 0.7117565055762082, "grad_norm": 0.5876020525559094, "learning_rate": 2.0372602176516914e-06, "loss": 0.0518, "step": 15317 }, { "epoch": 0.7118029739776952, "grad_norm": 0.6214533041515162, "learning_rate": 2.0366540752503955e-06, "loss": 0.077, "step": 15318 }, { "epoch": 0.7118494423791821, "grad_norm": 0.6200966714656034, "learning_rate": 2.036047999972004e-06, "loss": 0.0556, "step": 15319 }, { "epoch": 0.7118959107806692, "grad_norm": 2.039562240253747, "learning_rate": 2.0354419918302443e-06, "loss": 0.0728, "step": 15320 }, { "epoch": 0.7119423791821561, "grad_norm": 0.4405051888005637, "learning_rate": 2.0348360508388444e-06, "loss": 0.0462, "step": 15321 }, { "epoch": 0.7119888475836431, "grad_norm": 0.8042228341484673, "learning_rate": 2.0342301770115265e-06, "loss": 0.0346, "step": 15322 }, { "epoch": 0.7120353159851301, "grad_norm": 0.5607512617291011, "learning_rate": 2.033624370362018e-06, "loss": 0.0628, "step": 15323 }, { "epoch": 0.7120817843866171, "grad_norm": 0.5461290598106993, "learning_rate": 2.0330186309040394e-06, "loss": 0.0509, "step": 15324 }, { "epoch": 0.7121282527881041, "grad_norm": 0.5755758209017442, "learning_rate": 2.032412958651312e-06, "loss": 0.0584, "step": 15325 }, { "epoch": 0.7121747211895911, "grad_norm": 0.4636270188099865, "learning_rate": 2.031807353617554e-06, "loss": 0.0377, "step": 15326 }, { "epoch": 0.712221189591078, "grad_norm": 0.6255140682666823, "learning_rate": 2.0312018158164843e-06, "loss": 0.0449, "step": 15327 }, { "epoch": 0.7122676579925651, "grad_norm": 0.16942969834236, "learning_rate": 2.0305963452618173e-06, "loss": 0.0297, "step": 15328 }, { "epoch": 0.712314126394052, "grad_norm": 0.5235617503859062, "learning_rate": 2.0299909419672664e-06, "loss": 0.0435, "step": 15329 }, { "epoch": 0.712360594795539, "grad_norm": 0.6326070977644443, "learning_rate": 2.029385605946549e-06, "loss": 0.0615, "step": 15330 }, { "epoch": 0.7124070631970261, "grad_norm": 0.7868399372268315, "learning_rate": 2.0287803372133757e-06, "loss": 0.0628, "step": 15331 }, { "epoch": 0.712453531598513, "grad_norm": 0.34899175998928733, "learning_rate": 2.028175135781451e-06, "loss": 0.0607, "step": 15332 }, { "epoch": 0.7125, "grad_norm": 0.6443711277808124, "learning_rate": 2.0275700016644895e-06, "loss": 0.0503, "step": 15333 }, { "epoch": 0.7125464684014869, "grad_norm": 0.4263261347206829, "learning_rate": 2.0269649348761953e-06, "loss": 0.0279, "step": 15334 }, { "epoch": 0.712592936802974, "grad_norm": 0.9583580470027775, "learning_rate": 2.0263599354302748e-06, "loss": 0.0385, "step": 15335 }, { "epoch": 0.712639405204461, "grad_norm": 0.6518195845651967, "learning_rate": 2.0257550033404294e-06, "loss": 0.0494, "step": 15336 }, { "epoch": 0.7126858736059479, "grad_norm": 0.5754625666484033, "learning_rate": 2.0251501386203676e-06, "loss": 0.0504, "step": 15337 }, { "epoch": 0.712732342007435, "grad_norm": 0.5242212414232216, "learning_rate": 2.024545341283784e-06, "loss": 0.0483, "step": 15338 }, { "epoch": 0.7127788104089219, "grad_norm": 0.755893404240507, "learning_rate": 2.0239406113443784e-06, "loss": 0.0662, "step": 15339 }, { "epoch": 0.7128252788104089, "grad_norm": 0.6443841249246444, "learning_rate": 2.0233359488158517e-06, "loss": 0.0464, "step": 15340 }, { "epoch": 0.712871747211896, "grad_norm": 0.5193610226000817, "learning_rate": 2.022731353711899e-06, "loss": 0.0519, "step": 15341 }, { "epoch": 0.7129182156133829, "grad_norm": 0.42129413229281826, "learning_rate": 2.022126826046212e-06, "loss": 0.0451, "step": 15342 }, { "epoch": 0.7129646840148699, "grad_norm": 0.7560320336960393, "learning_rate": 2.0215223658324913e-06, "loss": 0.0788, "step": 15343 }, { "epoch": 0.7130111524163569, "grad_norm": 0.3966506186867559, "learning_rate": 2.0209179730844215e-06, "loss": 0.0375, "step": 15344 }, { "epoch": 0.7130576208178439, "grad_norm": 0.50999454989387, "learning_rate": 2.020313647815695e-06, "loss": 0.067, "step": 15345 }, { "epoch": 0.7131040892193309, "grad_norm": 0.4364434215217686, "learning_rate": 2.019709390039999e-06, "loss": 0.0603, "step": 15346 }, { "epoch": 0.7131505576208178, "grad_norm": 0.4639751567059436, "learning_rate": 2.019105199771023e-06, "loss": 0.0571, "step": 15347 }, { "epoch": 0.7131970260223048, "grad_norm": 0.48482281787616577, "learning_rate": 2.018501077022452e-06, "loss": 0.0555, "step": 15348 }, { "epoch": 0.7132434944237919, "grad_norm": 0.5646643266433139, "learning_rate": 2.01789702180797e-06, "loss": 0.0705, "step": 15349 }, { "epoch": 0.7132899628252788, "grad_norm": 0.6884790109103638, "learning_rate": 2.017293034141259e-06, "loss": 0.0567, "step": 15350 }, { "epoch": 0.7133364312267658, "grad_norm": 0.7202896519370895, "learning_rate": 2.0166891140359995e-06, "loss": 0.0519, "step": 15351 }, { "epoch": 0.7133828996282527, "grad_norm": 0.6614223935716544, "learning_rate": 2.0160852615058716e-06, "loss": 0.0609, "step": 15352 }, { "epoch": 0.7134293680297398, "grad_norm": 0.652512304356068, "learning_rate": 2.0154814765645514e-06, "loss": 0.0505, "step": 15353 }, { "epoch": 0.7134758364312268, "grad_norm": 0.32271885091194924, "learning_rate": 2.01487775922572e-06, "loss": 0.0444, "step": 15354 }, { "epoch": 0.7135223048327137, "grad_norm": 0.3803585732148763, "learning_rate": 2.014274109503048e-06, "loss": 0.0507, "step": 15355 }, { "epoch": 0.7135687732342008, "grad_norm": 0.6316216729127453, "learning_rate": 2.013670527410208e-06, "loss": 0.0653, "step": 15356 }, { "epoch": 0.7136152416356877, "grad_norm": 0.6924957953112881, "learning_rate": 2.0130670129608746e-06, "loss": 0.0445, "step": 15357 }, { "epoch": 0.7136617100371747, "grad_norm": 0.6070729550789802, "learning_rate": 2.012463566168718e-06, "loss": 0.0442, "step": 15358 }, { "epoch": 0.7137081784386617, "grad_norm": 0.5546315321969862, "learning_rate": 2.0118601870474036e-06, "loss": 0.0371, "step": 15359 }, { "epoch": 0.7137546468401487, "grad_norm": 0.7502157416288319, "learning_rate": 2.011256875610605e-06, "loss": 0.0546, "step": 15360 }, { "epoch": 0.7138011152416357, "grad_norm": 0.5836732377930708, "learning_rate": 2.0106536318719816e-06, "loss": 0.0731, "step": 15361 }, { "epoch": 0.7138475836431227, "grad_norm": 0.9186452104441234, "learning_rate": 2.0100504558452e-06, "loss": 0.0571, "step": 15362 }, { "epoch": 0.7138940520446097, "grad_norm": 0.5716248705345507, "learning_rate": 2.00944734754392e-06, "loss": 0.0494, "step": 15363 }, { "epoch": 0.7139405204460967, "grad_norm": 0.3983948989011091, "learning_rate": 2.0088443069818075e-06, "loss": 0.0447, "step": 15364 }, { "epoch": 0.7139869888475836, "grad_norm": 3.466496907220764, "learning_rate": 2.0082413341725193e-06, "loss": 0.0728, "step": 15365 }, { "epoch": 0.7140334572490706, "grad_norm": 0.8671389709772285, "learning_rate": 2.0076384291297134e-06, "loss": 0.073, "step": 15366 }, { "epoch": 0.7140799256505577, "grad_norm": 0.54646511313047, "learning_rate": 2.007035591867047e-06, "loss": 0.0427, "step": 15367 }, { "epoch": 0.7141263940520446, "grad_norm": 0.47465255514386084, "learning_rate": 2.0064328223981743e-06, "loss": 0.0528, "step": 15368 }, { "epoch": 0.7141728624535316, "grad_norm": 0.45815047191984687, "learning_rate": 2.005830120736747e-06, "loss": 0.0601, "step": 15369 }, { "epoch": 0.7142193308550185, "grad_norm": 0.8836404550616326, "learning_rate": 2.0052274868964207e-06, "loss": 0.0463, "step": 15370 }, { "epoch": 0.7142657992565056, "grad_norm": 1.1427502566275967, "learning_rate": 2.0046249208908436e-06, "loss": 0.0582, "step": 15371 }, { "epoch": 0.7143122676579926, "grad_norm": 0.7051162673573875, "learning_rate": 2.0040224227336665e-06, "loss": 0.0588, "step": 15372 }, { "epoch": 0.7143587360594795, "grad_norm": 0.4166841240112399, "learning_rate": 2.0034199924385312e-06, "loss": 0.0333, "step": 15373 }, { "epoch": 0.7144052044609666, "grad_norm": 0.780566314505671, "learning_rate": 2.0028176300190882e-06, "loss": 0.0541, "step": 15374 }, { "epoch": 0.7144516728624535, "grad_norm": 0.5734658285766723, "learning_rate": 2.002215335488981e-06, "loss": 0.037, "step": 15375 }, { "epoch": 0.7144981412639405, "grad_norm": 0.5189084010995804, "learning_rate": 2.0016131088618494e-06, "loss": 0.0411, "step": 15376 }, { "epoch": 0.7145446096654275, "grad_norm": 0.5657738779757734, "learning_rate": 2.0010109501513386e-06, "loss": 0.0508, "step": 15377 }, { "epoch": 0.7145910780669145, "grad_norm": 0.29143103044519725, "learning_rate": 2.000408859371088e-06, "loss": 0.0313, "step": 15378 }, { "epoch": 0.7146375464684015, "grad_norm": 0.21889847958195208, "learning_rate": 1.9998068365347316e-06, "loss": 0.0275, "step": 15379 }, { "epoch": 0.7146840148698885, "grad_norm": 0.6292689613057033, "learning_rate": 1.999204881655907e-06, "loss": 0.0456, "step": 15380 }, { "epoch": 0.7147304832713754, "grad_norm": 0.6070187450112202, "learning_rate": 1.998602994748251e-06, "loss": 0.0439, "step": 15381 }, { "epoch": 0.7147769516728625, "grad_norm": 0.6193106903355255, "learning_rate": 1.9980011758253962e-06, "loss": 0.058, "step": 15382 }, { "epoch": 0.7148234200743494, "grad_norm": 0.4713908371565867, "learning_rate": 1.997399424900973e-06, "loss": 0.0372, "step": 15383 }, { "epoch": 0.7148698884758364, "grad_norm": 0.7997326028513245, "learning_rate": 1.9967977419886165e-06, "loss": 0.0488, "step": 15384 }, { "epoch": 0.7149163568773235, "grad_norm": 0.31948305306308167, "learning_rate": 1.99619612710195e-06, "loss": 0.0289, "step": 15385 }, { "epoch": 0.7149628252788104, "grad_norm": 0.6843460430926109, "learning_rate": 1.995594580254601e-06, "loss": 0.0485, "step": 15386 }, { "epoch": 0.7150092936802974, "grad_norm": 0.523514586443307, "learning_rate": 1.9949931014601987e-06, "loss": 0.0475, "step": 15387 }, { "epoch": 0.7150557620817843, "grad_norm": 1.5474206929618752, "learning_rate": 1.994391690732365e-06, "loss": 0.07, "step": 15388 }, { "epoch": 0.7151022304832714, "grad_norm": 0.8810899799930395, "learning_rate": 1.993790348084723e-06, "loss": 0.0529, "step": 15389 }, { "epoch": 0.7151486988847584, "grad_norm": 0.42172706808474836, "learning_rate": 1.9931890735308935e-06, "loss": 0.0558, "step": 15390 }, { "epoch": 0.7151951672862453, "grad_norm": 0.45106477101044484, "learning_rate": 1.992587867084496e-06, "loss": 0.0402, "step": 15391 }, { "epoch": 0.7152416356877324, "grad_norm": 1.041370784207651, "learning_rate": 1.9919867287591493e-06, "loss": 0.0519, "step": 15392 }, { "epoch": 0.7152881040892193, "grad_norm": 0.5633117433485166, "learning_rate": 1.9913856585684666e-06, "loss": 0.0602, "step": 15393 }, { "epoch": 0.7153345724907063, "grad_norm": 0.5227160445891829, "learning_rate": 1.9907846565260675e-06, "loss": 0.0364, "step": 15394 }, { "epoch": 0.7153810408921933, "grad_norm": 0.6723693591654457, "learning_rate": 1.990183722645563e-06, "loss": 0.0634, "step": 15395 }, { "epoch": 0.7154275092936803, "grad_norm": 0.6947498015206202, "learning_rate": 1.989582856940566e-06, "loss": 0.0645, "step": 15396 }, { "epoch": 0.7154739776951673, "grad_norm": 0.6289811295514204, "learning_rate": 1.988982059424683e-06, "loss": 0.045, "step": 15397 }, { "epoch": 0.7155204460966543, "grad_norm": 0.40101532476903906, "learning_rate": 1.988381330111527e-06, "loss": 0.0442, "step": 15398 }, { "epoch": 0.7155669144981412, "grad_norm": 0.6715249005861665, "learning_rate": 1.9877806690147027e-06, "loss": 0.0558, "step": 15399 }, { "epoch": 0.7156133828996283, "grad_norm": 0.5803445941409128, "learning_rate": 1.9871800761478157e-06, "loss": 0.0575, "step": 15400 }, { "epoch": 0.7156598513011152, "grad_norm": 0.37485665943204205, "learning_rate": 1.9865795515244724e-06, "loss": 0.0349, "step": 15401 }, { "epoch": 0.7157063197026022, "grad_norm": 0.45853246893831895, "learning_rate": 1.985979095158276e-06, "loss": 0.0438, "step": 15402 }, { "epoch": 0.7157527881040893, "grad_norm": 0.5292389840975279, "learning_rate": 1.985378707062821e-06, "loss": 0.035, "step": 15403 }, { "epoch": 0.7157992565055762, "grad_norm": 0.3225644988721467, "learning_rate": 1.9847783872517133e-06, "loss": 0.0305, "step": 15404 }, { "epoch": 0.7158457249070632, "grad_norm": 0.4446347818893568, "learning_rate": 1.9841781357385488e-06, "loss": 0.0656, "step": 15405 }, { "epoch": 0.7158921933085501, "grad_norm": 0.5686123433985549, "learning_rate": 1.983577952536923e-06, "loss": 0.0642, "step": 15406 }, { "epoch": 0.7159386617100372, "grad_norm": 0.32813838746501595, "learning_rate": 1.9829778376604298e-06, "loss": 0.0469, "step": 15407 }, { "epoch": 0.7159851301115242, "grad_norm": 1.0391255577692085, "learning_rate": 1.9823777911226676e-06, "loss": 0.0612, "step": 15408 }, { "epoch": 0.7160315985130111, "grad_norm": 0.551116019456309, "learning_rate": 1.9817778129372227e-06, "loss": 0.039, "step": 15409 }, { "epoch": 0.7160780669144982, "grad_norm": 2.310226243627818, "learning_rate": 1.9811779031176852e-06, "loss": 0.0596, "step": 15410 }, { "epoch": 0.7161245353159851, "grad_norm": 0.3830790097743649, "learning_rate": 1.980578061677647e-06, "loss": 0.0306, "step": 15411 }, { "epoch": 0.7161710037174721, "grad_norm": 0.5213899968141233, "learning_rate": 1.979978288630694e-06, "loss": 0.0595, "step": 15412 }, { "epoch": 0.7162174721189591, "grad_norm": 0.6647198891632942, "learning_rate": 1.9793785839904106e-06, "loss": 0.0552, "step": 15413 }, { "epoch": 0.7162639405204461, "grad_norm": 0.5698458471482467, "learning_rate": 1.9787789477703824e-06, "loss": 0.0448, "step": 15414 }, { "epoch": 0.7163104089219331, "grad_norm": 1.6771071998226332, "learning_rate": 1.97817937998419e-06, "loss": 0.0603, "step": 15415 }, { "epoch": 0.7163568773234201, "grad_norm": 0.32582694631722964, "learning_rate": 1.9775798806454147e-06, "loss": 0.0435, "step": 15416 }, { "epoch": 0.716403345724907, "grad_norm": 0.5823353275934587, "learning_rate": 1.9769804497676355e-06, "loss": 0.0587, "step": 15417 }, { "epoch": 0.7164498141263941, "grad_norm": 0.6209497865042649, "learning_rate": 1.9763810873644316e-06, "loss": 0.0535, "step": 15418 }, { "epoch": 0.716496282527881, "grad_norm": 1.079071913485973, "learning_rate": 1.975781793449381e-06, "loss": 0.0448, "step": 15419 }, { "epoch": 0.716542750929368, "grad_norm": 0.34668604557735094, "learning_rate": 1.9751825680360514e-06, "loss": 0.0356, "step": 15420 }, { "epoch": 0.7165892193308551, "grad_norm": 0.578275659117662, "learning_rate": 1.9745834111380225e-06, "loss": 0.0542, "step": 15421 }, { "epoch": 0.716635687732342, "grad_norm": 0.6336817642298559, "learning_rate": 1.9739843227688634e-06, "loss": 0.0538, "step": 15422 }, { "epoch": 0.716682156133829, "grad_norm": 0.7205354066622752, "learning_rate": 1.973385302942144e-06, "loss": 0.0735, "step": 15423 }, { "epoch": 0.7167286245353159, "grad_norm": 0.6141442002198639, "learning_rate": 1.9727863516714313e-06, "loss": 0.064, "step": 15424 }, { "epoch": 0.716775092936803, "grad_norm": 0.5752511094707682, "learning_rate": 1.9721874689702976e-06, "loss": 0.0671, "step": 15425 }, { "epoch": 0.71682156133829, "grad_norm": 0.6220371578322514, "learning_rate": 1.9715886548523024e-06, "loss": 0.0733, "step": 15426 }, { "epoch": 0.7168680297397769, "grad_norm": 0.7489816484788837, "learning_rate": 1.97098990933101e-06, "loss": 0.0524, "step": 15427 }, { "epoch": 0.716914498141264, "grad_norm": 0.6591499503695664, "learning_rate": 1.9703912324199854e-06, "loss": 0.0502, "step": 15428 }, { "epoch": 0.7169609665427509, "grad_norm": 0.5121912564319145, "learning_rate": 1.969792624132788e-06, "loss": 0.0478, "step": 15429 }, { "epoch": 0.7170074349442379, "grad_norm": 0.5923167922024388, "learning_rate": 1.969194084482977e-06, "loss": 0.0581, "step": 15430 }, { "epoch": 0.7170539033457249, "grad_norm": 0.8615103455321526, "learning_rate": 1.9685956134841104e-06, "loss": 0.0617, "step": 15431 }, { "epoch": 0.7171003717472119, "grad_norm": 0.2914931981211587, "learning_rate": 1.967997211149743e-06, "loss": 0.0224, "step": 15432 }, { "epoch": 0.7171468401486989, "grad_norm": 1.2906799191381253, "learning_rate": 1.9673988774934306e-06, "loss": 0.0707, "step": 15433 }, { "epoch": 0.7171933085501859, "grad_norm": 0.46050484586266627, "learning_rate": 1.966800612528723e-06, "loss": 0.0366, "step": 15434 }, { "epoch": 0.7172397769516728, "grad_norm": 0.4823249668167737, "learning_rate": 1.966202416269176e-06, "loss": 0.0479, "step": 15435 }, { "epoch": 0.7172862453531599, "grad_norm": 0.32317219050318186, "learning_rate": 1.965604288728337e-06, "loss": 0.037, "step": 15436 }, { "epoch": 0.7173327137546468, "grad_norm": 0.6467336649749619, "learning_rate": 1.965006229919755e-06, "loss": 0.0599, "step": 15437 }, { "epoch": 0.7173791821561338, "grad_norm": 0.5948795363500617, "learning_rate": 1.964408239856975e-06, "loss": 0.0624, "step": 15438 }, { "epoch": 0.7174256505576209, "grad_norm": 0.47211350798610474, "learning_rate": 1.9638103185535446e-06, "loss": 0.0332, "step": 15439 }, { "epoch": 0.7174721189591078, "grad_norm": 0.6931914018415566, "learning_rate": 1.9632124660230056e-06, "loss": 0.0623, "step": 15440 }, { "epoch": 0.7175185873605948, "grad_norm": 0.2937636398114949, "learning_rate": 1.9626146822788985e-06, "loss": 0.037, "step": 15441 }, { "epoch": 0.7175650557620817, "grad_norm": 0.6182574580514272, "learning_rate": 1.962016967334767e-06, "loss": 0.0473, "step": 15442 }, { "epoch": 0.7176115241635688, "grad_norm": 0.5079281830431488, "learning_rate": 1.961419321204151e-06, "loss": 0.0345, "step": 15443 }, { "epoch": 0.7176579925650558, "grad_norm": 0.7020639211403843, "learning_rate": 1.9608217439005817e-06, "loss": 0.057, "step": 15444 }, { "epoch": 0.7177044609665427, "grad_norm": 0.8238307037700562, "learning_rate": 1.9602242354376007e-06, "loss": 0.0626, "step": 15445 }, { "epoch": 0.7177509293680298, "grad_norm": 0.511044995178218, "learning_rate": 1.9596267958287397e-06, "loss": 0.0508, "step": 15446 }, { "epoch": 0.7177973977695167, "grad_norm": 0.8594518604406267, "learning_rate": 1.9590294250875312e-06, "loss": 0.0556, "step": 15447 }, { "epoch": 0.7178438661710037, "grad_norm": 0.9421672480798933, "learning_rate": 1.9584321232275053e-06, "loss": 0.044, "step": 15448 }, { "epoch": 0.7178903345724907, "grad_norm": 0.6454972688131649, "learning_rate": 1.9578348902621976e-06, "loss": 0.0773, "step": 15449 }, { "epoch": 0.7179368029739777, "grad_norm": 0.8191956138036667, "learning_rate": 1.9572377262051284e-06, "loss": 0.0417, "step": 15450 }, { "epoch": 0.7179832713754647, "grad_norm": 0.5655448570188587, "learning_rate": 1.956640631069826e-06, "loss": 0.0479, "step": 15451 }, { "epoch": 0.7180297397769517, "grad_norm": 0.6823440585343413, "learning_rate": 1.956043604869818e-06, "loss": 0.0541, "step": 15452 }, { "epoch": 0.7180762081784386, "grad_norm": 0.42362976899272575, "learning_rate": 1.955446647618626e-06, "loss": 0.0386, "step": 15453 }, { "epoch": 0.7181226765799257, "grad_norm": 0.7394563024706664, "learning_rate": 1.9548497593297716e-06, "loss": 0.0511, "step": 15454 }, { "epoch": 0.7181691449814126, "grad_norm": 0.38529483543430565, "learning_rate": 1.9542529400167755e-06, "loss": 0.039, "step": 15455 }, { "epoch": 0.7182156133828996, "grad_norm": 0.9586689603008738, "learning_rate": 1.953656189693156e-06, "loss": 0.0504, "step": 15456 }, { "epoch": 0.7182620817843867, "grad_norm": 0.48216595808051277, "learning_rate": 1.9530595083724294e-06, "loss": 0.0292, "step": 15457 }, { "epoch": 0.7183085501858736, "grad_norm": 0.7439079051056695, "learning_rate": 1.952462896068111e-06, "loss": 0.0765, "step": 15458 }, { "epoch": 0.7183550185873606, "grad_norm": 0.5172344120837684, "learning_rate": 1.951866352793716e-06, "loss": 0.0518, "step": 15459 }, { "epoch": 0.7184014869888475, "grad_norm": 0.47873642550247136, "learning_rate": 1.951269878562757e-06, "loss": 0.0512, "step": 15460 }, { "epoch": 0.7184479553903346, "grad_norm": 0.5425222172999491, "learning_rate": 1.950673473388744e-06, "loss": 0.0553, "step": 15461 }, { "epoch": 0.7184944237918216, "grad_norm": 0.3639393588015583, "learning_rate": 1.9500771372851856e-06, "loss": 0.0401, "step": 15462 }, { "epoch": 0.7185408921933085, "grad_norm": 0.5504720993203241, "learning_rate": 1.9494808702655906e-06, "loss": 0.0459, "step": 15463 }, { "epoch": 0.7185873605947956, "grad_norm": 0.7441056780250186, "learning_rate": 1.9488846723434646e-06, "loss": 0.0596, "step": 15464 }, { "epoch": 0.7186338289962825, "grad_norm": 0.6901981686898632, "learning_rate": 1.9482885435323095e-06, "loss": 0.065, "step": 15465 }, { "epoch": 0.7186802973977695, "grad_norm": 0.6786933713177892, "learning_rate": 1.9476924838456336e-06, "loss": 0.0633, "step": 15466 }, { "epoch": 0.7187267657992565, "grad_norm": 0.38204508422296424, "learning_rate": 1.947096493296936e-06, "loss": 0.048, "step": 15467 }, { "epoch": 0.7187732342007435, "grad_norm": 0.8856616878582241, "learning_rate": 1.9465005718997133e-06, "loss": 0.0669, "step": 15468 }, { "epoch": 0.7188197026022305, "grad_norm": 0.5240577061466049, "learning_rate": 1.945904719667468e-06, "loss": 0.0523, "step": 15469 }, { "epoch": 0.7188661710037175, "grad_norm": 0.6471888138390712, "learning_rate": 1.9453089366136953e-06, "loss": 0.0514, "step": 15470 }, { "epoch": 0.7189126394052044, "grad_norm": 0.38872253665048945, "learning_rate": 1.9447132227518896e-06, "loss": 0.0462, "step": 15471 }, { "epoch": 0.7189591078066915, "grad_norm": 0.7510604284243435, "learning_rate": 1.944117578095545e-06, "loss": 0.067, "step": 15472 }, { "epoch": 0.7190055762081784, "grad_norm": 0.48160652069139354, "learning_rate": 1.943522002658154e-06, "loss": 0.0579, "step": 15473 }, { "epoch": 0.7190520446096654, "grad_norm": 0.5039080161656475, "learning_rate": 1.9429264964532067e-06, "loss": 0.0696, "step": 15474 }, { "epoch": 0.7190985130111525, "grad_norm": 0.8679898440042406, "learning_rate": 1.9423310594941897e-06, "loss": 0.0649, "step": 15475 }, { "epoch": 0.7191449814126394, "grad_norm": 0.6384245412914848, "learning_rate": 1.941735691794594e-06, "loss": 0.0305, "step": 15476 }, { "epoch": 0.7191914498141264, "grad_norm": 0.6344182923333116, "learning_rate": 1.941140393367904e-06, "loss": 0.057, "step": 15477 }, { "epoch": 0.7192379182156133, "grad_norm": 0.41882406582957926, "learning_rate": 1.940545164227603e-06, "loss": 0.0423, "step": 15478 }, { "epoch": 0.7192843866171004, "grad_norm": 0.6532990130457275, "learning_rate": 1.9399500043871745e-06, "loss": 0.0597, "step": 15479 }, { "epoch": 0.7193308550185874, "grad_norm": 0.5457922982222195, "learning_rate": 1.9393549138600987e-06, "loss": 0.0582, "step": 15480 }, { "epoch": 0.7193773234200743, "grad_norm": 0.6533754480936466, "learning_rate": 1.9387598926598554e-06, "loss": 0.0778, "step": 15481 }, { "epoch": 0.7194237918215614, "grad_norm": 0.727987285117722, "learning_rate": 1.9381649407999208e-06, "loss": 0.0663, "step": 15482 }, { "epoch": 0.7194702602230483, "grad_norm": 0.45240257296874925, "learning_rate": 1.9375700582937746e-06, "loss": 0.0363, "step": 15483 }, { "epoch": 0.7195167286245353, "grad_norm": 0.6130264865127724, "learning_rate": 1.936975245154892e-06, "loss": 0.0505, "step": 15484 }, { "epoch": 0.7195631970260223, "grad_norm": 0.5627373837032127, "learning_rate": 1.9363805013967397e-06, "loss": 0.0571, "step": 15485 }, { "epoch": 0.7196096654275093, "grad_norm": 0.6689402520574268, "learning_rate": 1.935785827032795e-06, "loss": 0.0721, "step": 15486 }, { "epoch": 0.7196561338289963, "grad_norm": 0.44908176307197356, "learning_rate": 1.9351912220765267e-06, "loss": 0.0438, "step": 15487 }, { "epoch": 0.7197026022304833, "grad_norm": 0.7850004807284907, "learning_rate": 1.934596686541403e-06, "loss": 0.0571, "step": 15488 }, { "epoch": 0.7197490706319702, "grad_norm": 0.6067197041558413, "learning_rate": 1.9340022204408887e-06, "loss": 0.0518, "step": 15489 }, { "epoch": 0.7197955390334573, "grad_norm": 0.992873263304003, "learning_rate": 1.933407823788454e-06, "loss": 0.0597, "step": 15490 }, { "epoch": 0.7198420074349442, "grad_norm": 0.40358721468305847, "learning_rate": 1.9328134965975586e-06, "loss": 0.0323, "step": 15491 }, { "epoch": 0.7198884758364312, "grad_norm": 0.8195486947540153, "learning_rate": 1.9322192388816635e-06, "loss": 0.0665, "step": 15492 }, { "epoch": 0.7199349442379183, "grad_norm": 0.43903198917133024, "learning_rate": 1.9316250506542333e-06, "loss": 0.0495, "step": 15493 }, { "epoch": 0.7199814126394052, "grad_norm": 0.7659780265471596, "learning_rate": 1.9310309319287258e-06, "loss": 0.0594, "step": 15494 }, { "epoch": 0.7200278810408922, "grad_norm": 1.3337106146466706, "learning_rate": 1.9304368827185964e-06, "loss": 0.0779, "step": 15495 }, { "epoch": 0.7200743494423791, "grad_norm": 0.6417285707104536, "learning_rate": 1.9298429030373034e-06, "loss": 0.0668, "step": 15496 }, { "epoch": 0.7201208178438662, "grad_norm": 0.5738188610275318, "learning_rate": 1.9292489928982996e-06, "loss": 0.063, "step": 15497 }, { "epoch": 0.7201672862453532, "grad_norm": 0.29585950213224266, "learning_rate": 1.9286551523150376e-06, "loss": 0.037, "step": 15498 }, { "epoch": 0.7202137546468401, "grad_norm": 0.4050839479275437, "learning_rate": 1.9280613813009675e-06, "loss": 0.0534, "step": 15499 }, { "epoch": 0.7202602230483272, "grad_norm": 1.6371744217211195, "learning_rate": 1.927467679869542e-06, "loss": 0.0398, "step": 15500 }, { "epoch": 0.7203066914498141, "grad_norm": 0.7452076105101322, "learning_rate": 1.9268740480342074e-06, "loss": 0.0522, "step": 15501 }, { "epoch": 0.7203531598513011, "grad_norm": 0.5165098235479914, "learning_rate": 1.92628048580841e-06, "loss": 0.0527, "step": 15502 }, { "epoch": 0.7203996282527881, "grad_norm": 0.40818126350508854, "learning_rate": 1.9256869932055946e-06, "loss": 0.0436, "step": 15503 }, { "epoch": 0.7204460966542751, "grad_norm": 0.6423851972077063, "learning_rate": 1.9250935702392044e-06, "loss": 0.0692, "step": 15504 }, { "epoch": 0.7204925650557621, "grad_norm": 0.486030090551767, "learning_rate": 1.9245002169226814e-06, "loss": 0.0494, "step": 15505 }, { "epoch": 0.7205390334572491, "grad_norm": 0.3461522586536978, "learning_rate": 1.923906933269463e-06, "loss": 0.0348, "step": 15506 }, { "epoch": 0.720585501858736, "grad_norm": 0.6925890773081651, "learning_rate": 1.9233137192929923e-06, "loss": 0.0575, "step": 15507 }, { "epoch": 0.7206319702602231, "grad_norm": 0.45395724481930655, "learning_rate": 1.9227205750067056e-06, "loss": 0.058, "step": 15508 }, { "epoch": 0.72067843866171, "grad_norm": 0.4862812773243978, "learning_rate": 1.9221275004240336e-06, "loss": 0.0406, "step": 15509 }, { "epoch": 0.720724907063197, "grad_norm": 0.5053731518882902, "learning_rate": 1.921534495558414e-06, "loss": 0.0483, "step": 15510 }, { "epoch": 0.7207713754646841, "grad_norm": 0.5978723112585314, "learning_rate": 1.920941560423279e-06, "loss": 0.0686, "step": 15511 }, { "epoch": 0.720817843866171, "grad_norm": 0.605061516107765, "learning_rate": 1.920348695032058e-06, "loss": 0.0731, "step": 15512 }, { "epoch": 0.720864312267658, "grad_norm": 0.9602568117262893, "learning_rate": 1.9197558993981784e-06, "loss": 0.0617, "step": 15513 }, { "epoch": 0.7209107806691449, "grad_norm": 0.3722086351101482, "learning_rate": 1.919163173535073e-06, "loss": 0.0449, "step": 15514 }, { "epoch": 0.720957249070632, "grad_norm": 0.5703111672466459, "learning_rate": 1.9185705174561626e-06, "loss": 0.052, "step": 15515 }, { "epoch": 0.721003717472119, "grad_norm": 0.5045920415447882, "learning_rate": 1.9179779311748714e-06, "loss": 0.0413, "step": 15516 }, { "epoch": 0.7210501858736059, "grad_norm": 0.5688312740551759, "learning_rate": 1.9173854147046263e-06, "loss": 0.0438, "step": 15517 }, { "epoch": 0.721096654275093, "grad_norm": 0.42784533879702424, "learning_rate": 1.916792968058845e-06, "loss": 0.0415, "step": 15518 }, { "epoch": 0.7211431226765799, "grad_norm": 0.3618527623411183, "learning_rate": 1.916200591250948e-06, "loss": 0.0528, "step": 15519 }, { "epoch": 0.7211895910780669, "grad_norm": 0.3848089444487361, "learning_rate": 1.9156082842943536e-06, "loss": 0.0535, "step": 15520 }, { "epoch": 0.7212360594795539, "grad_norm": 0.7838861448445975, "learning_rate": 1.9150160472024775e-06, "loss": 0.0514, "step": 15521 }, { "epoch": 0.7212825278810409, "grad_norm": 0.5928764896360498, "learning_rate": 1.9144238799887354e-06, "loss": 0.0592, "step": 15522 }, { "epoch": 0.7213289962825279, "grad_norm": 0.5427303232526272, "learning_rate": 1.9138317826665375e-06, "loss": 0.0626, "step": 15523 }, { "epoch": 0.7213754646840149, "grad_norm": 0.38507118033379795, "learning_rate": 1.9132397552493005e-06, "loss": 0.0475, "step": 15524 }, { "epoch": 0.7214219330855018, "grad_norm": 0.36950206390592777, "learning_rate": 1.912647797750431e-06, "loss": 0.0391, "step": 15525 }, { "epoch": 0.7214684014869889, "grad_norm": 0.3791805982176158, "learning_rate": 1.9120559101833384e-06, "loss": 0.0421, "step": 15526 }, { "epoch": 0.7215148698884758, "grad_norm": 0.6280377175808304, "learning_rate": 1.9114640925614295e-06, "loss": 0.0399, "step": 15527 }, { "epoch": 0.7215613382899628, "grad_norm": 0.46193397097701133, "learning_rate": 1.9108723448981097e-06, "loss": 0.0605, "step": 15528 }, { "epoch": 0.7216078066914499, "grad_norm": 0.5302651926379913, "learning_rate": 1.9102806672067824e-06, "loss": 0.0407, "step": 15529 }, { "epoch": 0.7216542750929368, "grad_norm": 0.5007018847841563, "learning_rate": 1.9096890595008484e-06, "loss": 0.0566, "step": 15530 }, { "epoch": 0.7217007434944238, "grad_norm": 0.39343880913314616, "learning_rate": 1.9090975217937125e-06, "loss": 0.0554, "step": 15531 }, { "epoch": 0.7217472118959107, "grad_norm": 0.3389656543545744, "learning_rate": 1.9085060540987693e-06, "loss": 0.0333, "step": 15532 }, { "epoch": 0.7217936802973978, "grad_norm": 0.5595352048676494, "learning_rate": 1.9079146564294167e-06, "loss": 0.0481, "step": 15533 }, { "epoch": 0.7218401486988848, "grad_norm": 0.6682581905381855, "learning_rate": 1.9073233287990522e-06, "loss": 0.0503, "step": 15534 }, { "epoch": 0.7218866171003717, "grad_norm": 0.6299385364091552, "learning_rate": 1.9067320712210702e-06, "loss": 0.0617, "step": 15535 }, { "epoch": 0.7219330855018588, "grad_norm": 0.707223375114058, "learning_rate": 1.9061408837088618e-06, "loss": 0.0618, "step": 15536 }, { "epoch": 0.7219795539033457, "grad_norm": 0.4001770577202398, "learning_rate": 1.9055497662758188e-06, "loss": 0.0447, "step": 15537 }, { "epoch": 0.7220260223048327, "grad_norm": 0.41434749449955244, "learning_rate": 1.9049587189353301e-06, "loss": 0.0314, "step": 15538 }, { "epoch": 0.7220724907063197, "grad_norm": 0.5272243174611043, "learning_rate": 1.9043677417007844e-06, "loss": 0.0472, "step": 15539 }, { "epoch": 0.7221189591078067, "grad_norm": 0.619939725348665, "learning_rate": 1.9037768345855657e-06, "loss": 0.0498, "step": 15540 }, { "epoch": 0.7221654275092937, "grad_norm": 0.3105514219322804, "learning_rate": 1.903185997603062e-06, "loss": 0.0343, "step": 15541 }, { "epoch": 0.7222118959107807, "grad_norm": 0.626415688864254, "learning_rate": 1.902595230766655e-06, "loss": 0.0555, "step": 15542 }, { "epoch": 0.7222583643122676, "grad_norm": 0.6109603646563305, "learning_rate": 1.9020045340897259e-06, "loss": 0.033, "step": 15543 }, { "epoch": 0.7223048327137547, "grad_norm": 0.5202903495948894, "learning_rate": 1.9014139075856541e-06, "loss": 0.0336, "step": 15544 }, { "epoch": 0.7223513011152416, "grad_norm": 0.5414673496250088, "learning_rate": 1.9008233512678193e-06, "loss": 0.0553, "step": 15545 }, { "epoch": 0.7223977695167286, "grad_norm": 0.4450665461337601, "learning_rate": 1.9002328651495965e-06, "loss": 0.0484, "step": 15546 }, { "epoch": 0.7224442379182157, "grad_norm": 0.5792305083734902, "learning_rate": 1.899642449244361e-06, "loss": 0.0528, "step": 15547 }, { "epoch": 0.7224907063197026, "grad_norm": 0.8744187530084457, "learning_rate": 1.8990521035654885e-06, "loss": 0.0637, "step": 15548 }, { "epoch": 0.7225371747211896, "grad_norm": 0.696218256165517, "learning_rate": 1.8984618281263512e-06, "loss": 0.0438, "step": 15549 }, { "epoch": 0.7225836431226765, "grad_norm": 0.523288021834298, "learning_rate": 1.8978716229403144e-06, "loss": 0.0538, "step": 15550 }, { "epoch": 0.7226301115241636, "grad_norm": 0.3851723437998182, "learning_rate": 1.8972814880207519e-06, "loss": 0.0373, "step": 15551 }, { "epoch": 0.7226765799256506, "grad_norm": 0.4243472206329502, "learning_rate": 1.896691423381029e-06, "loss": 0.0505, "step": 15552 }, { "epoch": 0.7227230483271375, "grad_norm": 0.3917700411742722, "learning_rate": 1.896101429034512e-06, "loss": 0.0254, "step": 15553 }, { "epoch": 0.7227695167286246, "grad_norm": 0.46324141362184407, "learning_rate": 1.895511504994562e-06, "loss": 0.036, "step": 15554 }, { "epoch": 0.7228159851301115, "grad_norm": 0.563533399559064, "learning_rate": 1.8949216512745483e-06, "loss": 0.0541, "step": 15555 }, { "epoch": 0.7228624535315985, "grad_norm": 0.6648307761160028, "learning_rate": 1.8943318678878247e-06, "loss": 0.0652, "step": 15556 }, { "epoch": 0.7229089219330855, "grad_norm": 0.3377439719422184, "learning_rate": 1.8937421548477514e-06, "loss": 0.049, "step": 15557 }, { "epoch": 0.7229553903345725, "grad_norm": 0.3968702055042363, "learning_rate": 1.8931525121676897e-06, "loss": 0.0555, "step": 15558 }, { "epoch": 0.7230018587360595, "grad_norm": 0.5929591776325475, "learning_rate": 1.892562939860993e-06, "loss": 0.0638, "step": 15559 }, { "epoch": 0.7230483271375465, "grad_norm": 0.4566959257729955, "learning_rate": 1.8919734379410165e-06, "loss": 0.0459, "step": 15560 }, { "epoch": 0.7230947955390334, "grad_norm": 0.9162951979718186, "learning_rate": 1.8913840064211126e-06, "loss": 0.035, "step": 15561 }, { "epoch": 0.7231412639405205, "grad_norm": 0.6652778491732041, "learning_rate": 1.890794645314633e-06, "loss": 0.0623, "step": 15562 }, { "epoch": 0.7231877323420074, "grad_norm": 0.38541254535342695, "learning_rate": 1.8902053546349269e-06, "loss": 0.0424, "step": 15563 }, { "epoch": 0.7232342007434944, "grad_norm": 0.46589677380384026, "learning_rate": 1.8896161343953407e-06, "loss": 0.0339, "step": 15564 }, { "epoch": 0.7232806691449815, "grad_norm": 0.4677481744083193, "learning_rate": 1.8890269846092247e-06, "loss": 0.042, "step": 15565 }, { "epoch": 0.7233271375464684, "grad_norm": 0.4523129908177851, "learning_rate": 1.888437905289921e-06, "loss": 0.0584, "step": 15566 }, { "epoch": 0.7233736059479554, "grad_norm": 0.6063418997744134, "learning_rate": 1.887848896450774e-06, "loss": 0.0368, "step": 15567 }, { "epoch": 0.7234200743494423, "grad_norm": 0.8756539208612173, "learning_rate": 1.8872599581051254e-06, "loss": 0.0535, "step": 15568 }, { "epoch": 0.7234665427509294, "grad_norm": 0.5744773089017342, "learning_rate": 1.8866710902663144e-06, "loss": 0.0541, "step": 15569 }, { "epoch": 0.7235130111524164, "grad_norm": 0.435479455827783, "learning_rate": 1.8860822929476801e-06, "loss": 0.0455, "step": 15570 }, { "epoch": 0.7235594795539033, "grad_norm": 0.41115584420736645, "learning_rate": 1.8854935661625572e-06, "loss": 0.0539, "step": 15571 }, { "epoch": 0.7236059479553903, "grad_norm": 0.39548851300157745, "learning_rate": 1.8849049099242856e-06, "loss": 0.0393, "step": 15572 }, { "epoch": 0.7236524163568773, "grad_norm": 0.33535076940645375, "learning_rate": 1.884316324246197e-06, "loss": 0.0513, "step": 15573 }, { "epoch": 0.7236988847583643, "grad_norm": 0.6602236824864683, "learning_rate": 1.88372780914162e-06, "loss": 0.0523, "step": 15574 }, { "epoch": 0.7237453531598513, "grad_norm": 0.7865643744276477, "learning_rate": 1.8831393646238894e-06, "loss": 0.0592, "step": 15575 }, { "epoch": 0.7237918215613383, "grad_norm": 0.5233850317179733, "learning_rate": 1.8825509907063328e-06, "loss": 0.045, "step": 15576 }, { "epoch": 0.7238382899628253, "grad_norm": 0.44015939431928097, "learning_rate": 1.8819626874022773e-06, "loss": 0.044, "step": 15577 }, { "epoch": 0.7238847583643123, "grad_norm": 0.7956723990952379, "learning_rate": 1.8813744547250468e-06, "loss": 0.0563, "step": 15578 }, { "epoch": 0.7239312267657992, "grad_norm": 0.3710684729364228, "learning_rate": 1.8807862926879706e-06, "loss": 0.0289, "step": 15579 }, { "epoch": 0.7239776951672863, "grad_norm": 0.493707267331483, "learning_rate": 1.8801982013043662e-06, "loss": 0.0586, "step": 15580 }, { "epoch": 0.7240241635687732, "grad_norm": 0.4274269981353747, "learning_rate": 1.8796101805875544e-06, "loss": 0.0638, "step": 15581 }, { "epoch": 0.7240706319702602, "grad_norm": 0.5291218734616316, "learning_rate": 1.8790222305508576e-06, "loss": 0.0483, "step": 15582 }, { "epoch": 0.7241171003717473, "grad_norm": 0.6017837945935409, "learning_rate": 1.8784343512075926e-06, "loss": 0.0478, "step": 15583 }, { "epoch": 0.7241635687732342, "grad_norm": 0.5667862561865395, "learning_rate": 1.877846542571074e-06, "loss": 0.0544, "step": 15584 }, { "epoch": 0.7242100371747212, "grad_norm": 0.8490223708747409, "learning_rate": 1.877258804654618e-06, "loss": 0.048, "step": 15585 }, { "epoch": 0.7242565055762081, "grad_norm": 0.4657254985084965, "learning_rate": 1.8766711374715362e-06, "loss": 0.0415, "step": 15586 }, { "epoch": 0.7243029739776952, "grad_norm": 1.0191780566457405, "learning_rate": 1.8760835410351409e-06, "loss": 0.0811, "step": 15587 }, { "epoch": 0.7243494423791822, "grad_norm": 0.5166290111768657, "learning_rate": 1.8754960153587393e-06, "loss": 0.0445, "step": 15588 }, { "epoch": 0.7243959107806691, "grad_norm": 0.26109974172932565, "learning_rate": 1.8749085604556433e-06, "loss": 0.0466, "step": 15589 }, { "epoch": 0.7244423791821561, "grad_norm": 1.3223835659223855, "learning_rate": 1.874321176339159e-06, "loss": 0.0782, "step": 15590 }, { "epoch": 0.7244888475836431, "grad_norm": 0.60267526352451, "learning_rate": 1.8737338630225866e-06, "loss": 0.0622, "step": 15591 }, { "epoch": 0.7245353159851301, "grad_norm": 0.7044152416125914, "learning_rate": 1.873146620519234e-06, "loss": 0.0623, "step": 15592 }, { "epoch": 0.7245817843866171, "grad_norm": 0.6436474188921607, "learning_rate": 1.8725594488424015e-06, "loss": 0.0449, "step": 15593 }, { "epoch": 0.724628252788104, "grad_norm": 0.5663798294373662, "learning_rate": 1.8719723480053887e-06, "loss": 0.0824, "step": 15594 }, { "epoch": 0.7246747211895911, "grad_norm": 0.7831081571639721, "learning_rate": 1.8713853180214931e-06, "loss": 0.0737, "step": 15595 }, { "epoch": 0.7247211895910781, "grad_norm": 0.7135856658148648, "learning_rate": 1.8707983589040164e-06, "loss": 0.0436, "step": 15596 }, { "epoch": 0.724767657992565, "grad_norm": 0.4319023490206124, "learning_rate": 1.8702114706662484e-06, "loss": 0.0278, "step": 15597 }, { "epoch": 0.7248141263940521, "grad_norm": 0.4575933538640859, "learning_rate": 1.8696246533214823e-06, "loss": 0.0477, "step": 15598 }, { "epoch": 0.724860594795539, "grad_norm": 0.288449815176077, "learning_rate": 1.869037906883015e-06, "loss": 0.0259, "step": 15599 }, { "epoch": 0.724907063197026, "grad_norm": 0.37472440735640267, "learning_rate": 1.868451231364134e-06, "loss": 0.03, "step": 15600 }, { "epoch": 0.7249535315985131, "grad_norm": 0.91529800254819, "learning_rate": 1.8678646267781287e-06, "loss": 0.0628, "step": 15601 }, { "epoch": 0.725, "grad_norm": 0.4650468644588464, "learning_rate": 1.8672780931382861e-06, "loss": 0.0437, "step": 15602 }, { "epoch": 0.725046468401487, "grad_norm": 0.4811911070846, "learning_rate": 1.866691630457892e-06, "loss": 0.0597, "step": 15603 }, { "epoch": 0.7250929368029739, "grad_norm": 0.4403799214024754, "learning_rate": 1.86610523875023e-06, "loss": 0.0538, "step": 15604 }, { "epoch": 0.725139405204461, "grad_norm": 0.5117587439467238, "learning_rate": 1.865518918028581e-06, "loss": 0.0506, "step": 15605 }, { "epoch": 0.725185873605948, "grad_norm": 0.4835638364955626, "learning_rate": 1.8649326683062291e-06, "loss": 0.0667, "step": 15606 }, { "epoch": 0.7252323420074349, "grad_norm": 0.42555394495558435, "learning_rate": 1.8643464895964519e-06, "loss": 0.0416, "step": 15607 }, { "epoch": 0.725278810408922, "grad_norm": 0.5853610489678059, "learning_rate": 1.8637603819125272e-06, "loss": 0.0412, "step": 15608 }, { "epoch": 0.7253252788104089, "grad_norm": 0.7120421390218447, "learning_rate": 1.8631743452677304e-06, "loss": 0.0698, "step": 15609 }, { "epoch": 0.7253717472118959, "grad_norm": 0.6349528132814742, "learning_rate": 1.8625883796753362e-06, "loss": 0.0648, "step": 15610 }, { "epoch": 0.7254182156133829, "grad_norm": 0.5350248515094751, "learning_rate": 1.8620024851486174e-06, "loss": 0.0562, "step": 15611 }, { "epoch": 0.7254646840148699, "grad_norm": 0.3159304976931502, "learning_rate": 1.8614166617008427e-06, "loss": 0.0391, "step": 15612 }, { "epoch": 0.7255111524163569, "grad_norm": 0.5987375657105326, "learning_rate": 1.8608309093452865e-06, "loss": 0.0607, "step": 15613 }, { "epoch": 0.7255576208178439, "grad_norm": 0.7295307867958607, "learning_rate": 1.8602452280952156e-06, "loss": 0.0678, "step": 15614 }, { "epoch": 0.7256040892193308, "grad_norm": 0.5176807175889493, "learning_rate": 1.859659617963891e-06, "loss": 0.0594, "step": 15615 }, { "epoch": 0.7256505576208179, "grad_norm": 0.4199048275291962, "learning_rate": 1.8590740789645828e-06, "loss": 0.0338, "step": 15616 }, { "epoch": 0.7256970260223048, "grad_norm": 0.45220622015029066, "learning_rate": 1.8584886111105526e-06, "loss": 0.0317, "step": 15617 }, { "epoch": 0.7257434944237918, "grad_norm": 0.46801204961929355, "learning_rate": 1.857903214415062e-06, "loss": 0.04, "step": 15618 }, { "epoch": 0.7257899628252789, "grad_norm": 0.4592651376678821, "learning_rate": 1.857317888891369e-06, "loss": 0.0423, "step": 15619 }, { "epoch": 0.7258364312267658, "grad_norm": 0.3950693454726813, "learning_rate": 1.8567326345527364e-06, "loss": 0.0485, "step": 15620 }, { "epoch": 0.7258828996282528, "grad_norm": 0.3762530219723978, "learning_rate": 1.8561474514124167e-06, "loss": 0.0518, "step": 15621 }, { "epoch": 0.7259293680297397, "grad_norm": 0.6703091632187393, "learning_rate": 1.8555623394836642e-06, "loss": 0.0576, "step": 15622 }, { "epoch": 0.7259758364312268, "grad_norm": 0.7688155357057661, "learning_rate": 1.8549772987797354e-06, "loss": 0.0574, "step": 15623 }, { "epoch": 0.7260223048327138, "grad_norm": 0.6092758464362796, "learning_rate": 1.8543923293138816e-06, "loss": 0.0515, "step": 15624 }, { "epoch": 0.7260687732342007, "grad_norm": 0.7440082007891701, "learning_rate": 1.8538074310993516e-06, "loss": 0.0761, "step": 15625 }, { "epoch": 0.7261152416356877, "grad_norm": 0.7652812091282315, "learning_rate": 1.853222604149395e-06, "loss": 0.0448, "step": 15626 }, { "epoch": 0.7261617100371747, "grad_norm": 0.5471437627888912, "learning_rate": 1.8526378484772579e-06, "loss": 0.0282, "step": 15627 }, { "epoch": 0.7262081784386617, "grad_norm": 0.6847455801388893, "learning_rate": 1.8520531640961865e-06, "loss": 0.0469, "step": 15628 }, { "epoch": 0.7262546468401487, "grad_norm": 0.47423039179769716, "learning_rate": 1.851468551019422e-06, "loss": 0.04, "step": 15629 }, { "epoch": 0.7263011152416357, "grad_norm": 0.4169908824646945, "learning_rate": 1.8508840092602105e-06, "loss": 0.0376, "step": 15630 }, { "epoch": 0.7263475836431227, "grad_norm": 0.9483851964484924, "learning_rate": 1.8502995388317907e-06, "loss": 0.0606, "step": 15631 }, { "epoch": 0.7263940520446097, "grad_norm": 0.61941989342857, "learning_rate": 1.8497151397474011e-06, "loss": 0.0647, "step": 15632 }, { "epoch": 0.7264405204460966, "grad_norm": 1.383178649273554, "learning_rate": 1.8491308120202794e-06, "loss": 0.0606, "step": 15633 }, { "epoch": 0.7264869888475837, "grad_norm": 0.3777230011282118, "learning_rate": 1.848546555663661e-06, "loss": 0.0308, "step": 15634 }, { "epoch": 0.7265334572490706, "grad_norm": 0.6628376543957362, "learning_rate": 1.8479623706907802e-06, "loss": 0.0528, "step": 15635 }, { "epoch": 0.7265799256505576, "grad_norm": 1.1000649109164684, "learning_rate": 1.8473782571148669e-06, "loss": 0.0633, "step": 15636 }, { "epoch": 0.7266263940520447, "grad_norm": 0.4956640306507253, "learning_rate": 1.8467942149491563e-06, "loss": 0.0361, "step": 15637 }, { "epoch": 0.7266728624535316, "grad_norm": 0.6037408162917586, "learning_rate": 1.8462102442068769e-06, "loss": 0.054, "step": 15638 }, { "epoch": 0.7267193308550186, "grad_norm": 0.351962613030786, "learning_rate": 1.8456263449012513e-06, "loss": 0.0322, "step": 15639 }, { "epoch": 0.7267657992565055, "grad_norm": 1.5473604129547731, "learning_rate": 1.8450425170455112e-06, "loss": 0.0682, "step": 15640 }, { "epoch": 0.7268122676579926, "grad_norm": 0.3582106886962206, "learning_rate": 1.844458760652878e-06, "loss": 0.0378, "step": 15641 }, { "epoch": 0.7268587360594796, "grad_norm": 0.47922032161760214, "learning_rate": 1.8438750757365747e-06, "loss": 0.0495, "step": 15642 }, { "epoch": 0.7269052044609665, "grad_norm": 0.5630401460627658, "learning_rate": 1.843291462309822e-06, "loss": 0.0595, "step": 15643 }, { "epoch": 0.7269516728624535, "grad_norm": 0.6236045177681457, "learning_rate": 1.8427079203858429e-06, "loss": 0.0614, "step": 15644 }, { "epoch": 0.7269981412639405, "grad_norm": 0.42309459783042724, "learning_rate": 1.8421244499778506e-06, "loss": 0.0515, "step": 15645 }, { "epoch": 0.7270446096654275, "grad_norm": 0.516353229656852, "learning_rate": 1.8415410510990612e-06, "loss": 0.0423, "step": 15646 }, { "epoch": 0.7270910780669145, "grad_norm": 0.9240945749505632, "learning_rate": 1.8409577237626935e-06, "loss": 0.0811, "step": 15647 }, { "epoch": 0.7271375464684015, "grad_norm": 0.46205522512077196, "learning_rate": 1.840374467981958e-06, "loss": 0.0445, "step": 15648 }, { "epoch": 0.7271840148698885, "grad_norm": 0.523027034389977, "learning_rate": 1.839791283770066e-06, "loss": 0.0574, "step": 15649 }, { "epoch": 0.7272304832713755, "grad_norm": 0.4260170040153085, "learning_rate": 1.8392081711402276e-06, "loss": 0.0488, "step": 15650 }, { "epoch": 0.7272769516728624, "grad_norm": 2.4819700746470597, "learning_rate": 1.8386251301056506e-06, "loss": 0.0687, "step": 15651 }, { "epoch": 0.7273234200743495, "grad_norm": 0.5220198352089777, "learning_rate": 1.8380421606795424e-06, "loss": 0.0373, "step": 15652 }, { "epoch": 0.7273698884758364, "grad_norm": 0.45369400879569866, "learning_rate": 1.8374592628751047e-06, "loss": 0.0572, "step": 15653 }, { "epoch": 0.7274163568773234, "grad_norm": 0.443669519150824, "learning_rate": 1.8368764367055452e-06, "loss": 0.0574, "step": 15654 }, { "epoch": 0.7274628252788105, "grad_norm": 0.6344271912216246, "learning_rate": 1.8362936821840654e-06, "loss": 0.0475, "step": 15655 }, { "epoch": 0.7275092936802974, "grad_norm": 0.5236595543934673, "learning_rate": 1.8357109993238597e-06, "loss": 0.0573, "step": 15656 }, { "epoch": 0.7275557620817844, "grad_norm": 0.5812318468176492, "learning_rate": 1.835128388138132e-06, "loss": 0.0467, "step": 15657 }, { "epoch": 0.7276022304832713, "grad_norm": 0.4025586471700167, "learning_rate": 1.8345458486400774e-06, "loss": 0.0532, "step": 15658 }, { "epoch": 0.7276486988847584, "grad_norm": 0.7965694200523595, "learning_rate": 1.8339633808428915e-06, "loss": 0.0722, "step": 15659 }, { "epoch": 0.7276951672862454, "grad_norm": 0.6491625416396539, "learning_rate": 1.8333809847597644e-06, "loss": 0.0512, "step": 15660 }, { "epoch": 0.7277416356877323, "grad_norm": 0.6108262776608486, "learning_rate": 1.832798660403895e-06, "loss": 0.0519, "step": 15661 }, { "epoch": 0.7277881040892193, "grad_norm": 0.36479999529925417, "learning_rate": 1.8322164077884675e-06, "loss": 0.0672, "step": 15662 }, { "epoch": 0.7278345724907063, "grad_norm": 0.744425118798712, "learning_rate": 1.8316342269266702e-06, "loss": 0.0619, "step": 15663 }, { "epoch": 0.7278810408921933, "grad_norm": 0.5582132791239397, "learning_rate": 1.8310521178316949e-06, "loss": 0.0533, "step": 15664 }, { "epoch": 0.7279275092936803, "grad_norm": 0.9501913237890207, "learning_rate": 1.8304700805167235e-06, "loss": 0.0493, "step": 15665 }, { "epoch": 0.7279739776951673, "grad_norm": 0.41069449895493076, "learning_rate": 1.829888114994941e-06, "loss": 0.039, "step": 15666 }, { "epoch": 0.7280204460966543, "grad_norm": 0.6441074616457112, "learning_rate": 1.8293062212795294e-06, "loss": 0.0695, "step": 15667 }, { "epoch": 0.7280669144981413, "grad_norm": 0.6583864638383871, "learning_rate": 1.8287243993836685e-06, "loss": 0.0598, "step": 15668 }, { "epoch": 0.7281133828996282, "grad_norm": 0.5732222933308531, "learning_rate": 1.828142649320538e-06, "loss": 0.0608, "step": 15669 }, { "epoch": 0.7281598513011153, "grad_norm": 0.33393635162634794, "learning_rate": 1.8275609711033133e-06, "loss": 0.031, "step": 15670 }, { "epoch": 0.7282063197026022, "grad_norm": 0.4443561439852686, "learning_rate": 1.8269793647451727e-06, "loss": 0.0458, "step": 15671 }, { "epoch": 0.7282527881040892, "grad_norm": 0.4221680846007002, "learning_rate": 1.8263978302592895e-06, "loss": 0.0401, "step": 15672 }, { "epoch": 0.7282992565055763, "grad_norm": 0.8885408363473009, "learning_rate": 1.8258163676588353e-06, "loss": 0.0473, "step": 15673 }, { "epoch": 0.7283457249070632, "grad_norm": 0.6815808783062943, "learning_rate": 1.825234976956981e-06, "loss": 0.0606, "step": 15674 }, { "epoch": 0.7283921933085502, "grad_norm": 0.6094272910417353, "learning_rate": 1.824653658166896e-06, "loss": 0.0488, "step": 15675 }, { "epoch": 0.7284386617100371, "grad_norm": 0.6175172599360208, "learning_rate": 1.8240724113017477e-06, "loss": 0.0639, "step": 15676 }, { "epoch": 0.7284851301115242, "grad_norm": 0.7189109145133834, "learning_rate": 1.8234912363747004e-06, "loss": 0.0617, "step": 15677 }, { "epoch": 0.7285315985130112, "grad_norm": 0.905992564460873, "learning_rate": 1.8229101333989214e-06, "loss": 0.0807, "step": 15678 }, { "epoch": 0.7285780669144981, "grad_norm": 0.6382296752613162, "learning_rate": 1.8223291023875733e-06, "loss": 0.0409, "step": 15679 }, { "epoch": 0.7286245353159851, "grad_norm": 0.7769824768441573, "learning_rate": 1.821748143353812e-06, "loss": 0.0756, "step": 15680 }, { "epoch": 0.7286710037174721, "grad_norm": 0.5761869361760324, "learning_rate": 1.8211672563108025e-06, "loss": 0.0509, "step": 15681 }, { "epoch": 0.7287174721189591, "grad_norm": 0.4552075961759763, "learning_rate": 1.8205864412716994e-06, "loss": 0.0621, "step": 15682 }, { "epoch": 0.7287639405204461, "grad_norm": 0.4573736556220224, "learning_rate": 1.8200056982496605e-06, "loss": 0.0384, "step": 15683 }, { "epoch": 0.728810408921933, "grad_norm": 0.57270412747966, "learning_rate": 1.8194250272578374e-06, "loss": 0.0582, "step": 15684 }, { "epoch": 0.7288568773234201, "grad_norm": 0.46421438894229256, "learning_rate": 1.8188444283093882e-06, "loss": 0.053, "step": 15685 }, { "epoch": 0.7289033457249071, "grad_norm": 0.4930921287138586, "learning_rate": 1.8182639014174592e-06, "loss": 0.0461, "step": 15686 }, { "epoch": 0.728949814126394, "grad_norm": 0.6487469921048333, "learning_rate": 1.8176834465951997e-06, "loss": 0.0598, "step": 15687 }, { "epoch": 0.7289962825278811, "grad_norm": 0.4673609765126591, "learning_rate": 1.8171030638557618e-06, "loss": 0.0546, "step": 15688 }, { "epoch": 0.729042750929368, "grad_norm": 0.6774074018169697, "learning_rate": 1.8165227532122886e-06, "loss": 0.0621, "step": 15689 }, { "epoch": 0.729089219330855, "grad_norm": 0.7110611742296163, "learning_rate": 1.8159425146779263e-06, "loss": 0.0617, "step": 15690 }, { "epoch": 0.729135687732342, "grad_norm": 0.9058540495661197, "learning_rate": 1.8153623482658167e-06, "loss": 0.0252, "step": 15691 }, { "epoch": 0.729182156133829, "grad_norm": 0.37042652579253155, "learning_rate": 1.8147822539891018e-06, "loss": 0.0572, "step": 15692 }, { "epoch": 0.729228624535316, "grad_norm": 0.5605868619954009, "learning_rate": 1.8142022318609214e-06, "loss": 0.0614, "step": 15693 }, { "epoch": 0.7292750929368029, "grad_norm": 0.53288642939356, "learning_rate": 1.8136222818944117e-06, "loss": 0.0683, "step": 15694 }, { "epoch": 0.72932156133829, "grad_norm": 0.5188116842940207, "learning_rate": 1.8130424041027128e-06, "loss": 0.0556, "step": 15695 }, { "epoch": 0.729368029739777, "grad_norm": 0.4436255179341636, "learning_rate": 1.812462598498957e-06, "loss": 0.0588, "step": 15696 }, { "epoch": 0.7294144981412639, "grad_norm": 0.48281931899702357, "learning_rate": 1.8118828650962788e-06, "loss": 0.034, "step": 15697 }, { "epoch": 0.7294609665427509, "grad_norm": 0.3751482131837771, "learning_rate": 1.8113032039078088e-06, "loss": 0.044, "step": 15698 }, { "epoch": 0.7295074349442379, "grad_norm": 0.6750423195099158, "learning_rate": 1.8107236149466773e-06, "loss": 0.0674, "step": 15699 }, { "epoch": 0.7295539033457249, "grad_norm": 0.6459046246850999, "learning_rate": 1.8101440982260127e-06, "loss": 0.0483, "step": 15700 }, { "epoch": 0.7296003717472119, "grad_norm": 0.6559483109753266, "learning_rate": 1.80956465375894e-06, "loss": 0.0414, "step": 15701 }, { "epoch": 0.7296468401486988, "grad_norm": 0.7973453023383134, "learning_rate": 1.8089852815585873e-06, "loss": 0.0497, "step": 15702 }, { "epoch": 0.7296933085501859, "grad_norm": 0.7668905030282004, "learning_rate": 1.8084059816380779e-06, "loss": 0.0624, "step": 15703 }, { "epoch": 0.7297397769516729, "grad_norm": 0.4480069908468192, "learning_rate": 1.8078267540105288e-06, "loss": 0.037, "step": 15704 }, { "epoch": 0.7297862453531598, "grad_norm": 0.6449411673989504, "learning_rate": 1.8072475986890652e-06, "loss": 0.0504, "step": 15705 }, { "epoch": 0.7298327137546469, "grad_norm": 0.6844433270022404, "learning_rate": 1.8066685156868042e-06, "loss": 0.0552, "step": 15706 }, { "epoch": 0.7298791821561338, "grad_norm": 3.5648087567447417, "learning_rate": 1.806089505016862e-06, "loss": 0.0249, "step": 15707 }, { "epoch": 0.7299256505576208, "grad_norm": 0.9742957289238069, "learning_rate": 1.8055105666923538e-06, "loss": 0.0687, "step": 15708 }, { "epoch": 0.7299721189591079, "grad_norm": 0.35049239557515066, "learning_rate": 1.8049317007263933e-06, "loss": 0.0532, "step": 15709 }, { "epoch": 0.7300185873605948, "grad_norm": 0.6532303064621321, "learning_rate": 1.804352907132093e-06, "loss": 0.0352, "step": 15710 }, { "epoch": 0.7300650557620818, "grad_norm": 0.8989904445693964, "learning_rate": 1.8037741859225604e-06, "loss": 0.0699, "step": 15711 }, { "epoch": 0.7301115241635687, "grad_norm": 0.6578584596617861, "learning_rate": 1.803195537110909e-06, "loss": 0.0678, "step": 15712 }, { "epoch": 0.7301579925650558, "grad_norm": 0.46964091780938205, "learning_rate": 1.8026169607102429e-06, "loss": 0.0509, "step": 15713 }, { "epoch": 0.7302044609665428, "grad_norm": 0.780222358068589, "learning_rate": 1.8020384567336673e-06, "loss": 0.0614, "step": 15714 }, { "epoch": 0.7302509293680297, "grad_norm": 0.5930939991263594, "learning_rate": 1.801460025194287e-06, "loss": 0.0418, "step": 15715 }, { "epoch": 0.7302973977695167, "grad_norm": 0.4455027293912029, "learning_rate": 1.8008816661052032e-06, "loss": 0.0721, "step": 15716 }, { "epoch": 0.7303438661710037, "grad_norm": 0.5611189532059396, "learning_rate": 1.8003033794795172e-06, "loss": 0.0494, "step": 15717 }, { "epoch": 0.7303903345724907, "grad_norm": 0.5142470257256541, "learning_rate": 1.7997251653303249e-06, "loss": 0.0463, "step": 15718 }, { "epoch": 0.7304368029739777, "grad_norm": 0.7334817531512535, "learning_rate": 1.7991470236707282e-06, "loss": 0.0481, "step": 15719 }, { "epoch": 0.7304832713754646, "grad_norm": 0.5847497470034128, "learning_rate": 1.7985689545138213e-06, "loss": 0.0456, "step": 15720 }, { "epoch": 0.7305297397769517, "grad_norm": 0.5328791898133342, "learning_rate": 1.797990957872694e-06, "loss": 0.0644, "step": 15721 }, { "epoch": 0.7305762081784387, "grad_norm": 0.4861798312804651, "learning_rate": 1.7974130337604434e-06, "loss": 0.0378, "step": 15722 }, { "epoch": 0.7306226765799256, "grad_norm": 0.5924053160308889, "learning_rate": 1.796835182190158e-06, "loss": 0.0493, "step": 15723 }, { "epoch": 0.7306691449814127, "grad_norm": 0.9347325378121715, "learning_rate": 1.796257403174927e-06, "loss": 0.0472, "step": 15724 }, { "epoch": 0.7307156133828996, "grad_norm": 0.4328394598429992, "learning_rate": 1.7956796967278362e-06, "loss": 0.0578, "step": 15725 }, { "epoch": 0.7307620817843866, "grad_norm": 0.34178755556757767, "learning_rate": 1.795102062861976e-06, "loss": 0.0505, "step": 15726 }, { "epoch": 0.7308085501858737, "grad_norm": 0.7350811274764522, "learning_rate": 1.7945245015904256e-06, "loss": 0.0465, "step": 15727 }, { "epoch": 0.7308550185873606, "grad_norm": 0.31981018143865775, "learning_rate": 1.7939470129262671e-06, "loss": 0.0246, "step": 15728 }, { "epoch": 0.7309014869888476, "grad_norm": 0.5568214098011344, "learning_rate": 1.7933695968825843e-06, "loss": 0.044, "step": 15729 }, { "epoch": 0.7309479553903345, "grad_norm": 0.8543703162472948, "learning_rate": 1.7927922534724557e-06, "loss": 0.0653, "step": 15730 }, { "epoch": 0.7309944237918216, "grad_norm": 0.4209235305239519, "learning_rate": 1.792214982708958e-06, "loss": 0.0619, "step": 15731 }, { "epoch": 0.7310408921933086, "grad_norm": 0.6254557097031962, "learning_rate": 1.791637784605167e-06, "loss": 0.06, "step": 15732 }, { "epoch": 0.7310873605947955, "grad_norm": 0.35442822498609716, "learning_rate": 1.791060659174157e-06, "loss": 0.049, "step": 15733 }, { "epoch": 0.7311338289962825, "grad_norm": 0.6396846927382912, "learning_rate": 1.7904836064289998e-06, "loss": 0.0537, "step": 15734 }, { "epoch": 0.7311802973977695, "grad_norm": 0.5071381647329575, "learning_rate": 1.7899066263827658e-06, "loss": 0.0643, "step": 15735 }, { "epoch": 0.7312267657992565, "grad_norm": 0.37353189447136037, "learning_rate": 1.789329719048527e-06, "loss": 0.0468, "step": 15736 }, { "epoch": 0.7312732342007435, "grad_norm": 0.4583993618020426, "learning_rate": 1.7887528844393493e-06, "loss": 0.0407, "step": 15737 }, { "epoch": 0.7313197026022304, "grad_norm": 0.6289712917859762, "learning_rate": 1.788176122568298e-06, "loss": 0.0781, "step": 15738 }, { "epoch": 0.7313661710037175, "grad_norm": 0.6603123563203385, "learning_rate": 1.7875994334484387e-06, "loss": 0.0613, "step": 15739 }, { "epoch": 0.7314126394052045, "grad_norm": 0.3317308193789701, "learning_rate": 1.7870228170928328e-06, "loss": 0.0281, "step": 15740 }, { "epoch": 0.7314591078066914, "grad_norm": 0.44643542224446375, "learning_rate": 1.7864462735145416e-06, "loss": 0.0595, "step": 15741 }, { "epoch": 0.7315055762081785, "grad_norm": 0.39746662238661123, "learning_rate": 1.7858698027266224e-06, "loss": 0.0524, "step": 15742 }, { "epoch": 0.7315520446096654, "grad_norm": 0.495541811413492, "learning_rate": 1.7852934047421371e-06, "loss": 0.0646, "step": 15743 }, { "epoch": 0.7315985130111524, "grad_norm": 0.7224942982938254, "learning_rate": 1.7847170795741414e-06, "loss": 0.0639, "step": 15744 }, { "epoch": 0.7316449814126395, "grad_norm": 1.5434027073131598, "learning_rate": 1.7841408272356848e-06, "loss": 0.0836, "step": 15745 }, { "epoch": 0.7316914498141264, "grad_norm": 0.8973843279474254, "learning_rate": 1.7835646477398243e-06, "loss": 0.0558, "step": 15746 }, { "epoch": 0.7317379182156134, "grad_norm": 0.7485645957682008, "learning_rate": 1.78298854109961e-06, "loss": 0.0649, "step": 15747 }, { "epoch": 0.7317843866171003, "grad_norm": 0.6741506743412287, "learning_rate": 1.7824125073280906e-06, "loss": 0.0523, "step": 15748 }, { "epoch": 0.7318308550185874, "grad_norm": 1.2045766952596753, "learning_rate": 1.7818365464383131e-06, "loss": 0.0435, "step": 15749 }, { "epoch": 0.7318773234200744, "grad_norm": 0.547517526286641, "learning_rate": 1.7812606584433284e-06, "loss": 0.0348, "step": 15750 }, { "epoch": 0.7319237918215613, "grad_norm": 0.42997324162730444, "learning_rate": 1.7806848433561752e-06, "loss": 0.0429, "step": 15751 }, { "epoch": 0.7319702602230483, "grad_norm": 0.466501860795194, "learning_rate": 1.7801091011898974e-06, "loss": 0.0586, "step": 15752 }, { "epoch": 0.7320167286245353, "grad_norm": 1.307186712327159, "learning_rate": 1.779533431957539e-06, "loss": 0.0266, "step": 15753 }, { "epoch": 0.7320631970260223, "grad_norm": 0.5514329918762756, "learning_rate": 1.7789578356721377e-06, "loss": 0.0676, "step": 15754 }, { "epoch": 0.7321096654275093, "grad_norm": 0.6243844307664854, "learning_rate": 1.7783823123467314e-06, "loss": 0.0728, "step": 15755 }, { "epoch": 0.7321561338289962, "grad_norm": 0.4127765600388969, "learning_rate": 1.7778068619943567e-06, "loss": 0.0451, "step": 15756 }, { "epoch": 0.7322026022304833, "grad_norm": 0.46376037995622044, "learning_rate": 1.7772314846280476e-06, "loss": 0.0476, "step": 15757 }, { "epoch": 0.7322490706319703, "grad_norm": 0.6179944523084024, "learning_rate": 1.7766561802608374e-06, "loss": 0.0519, "step": 15758 }, { "epoch": 0.7322955390334572, "grad_norm": 0.447897522465227, "learning_rate": 1.7760809489057552e-06, "loss": 0.0469, "step": 15759 }, { "epoch": 0.7323420074349443, "grad_norm": 0.4740522027472916, "learning_rate": 1.775505790575835e-06, "loss": 0.0739, "step": 15760 }, { "epoch": 0.7323884758364312, "grad_norm": 1.485331397840519, "learning_rate": 1.7749307052841019e-06, "loss": 0.0521, "step": 15761 }, { "epoch": 0.7324349442379182, "grad_norm": 0.7302781385875882, "learning_rate": 1.7743556930435824e-06, "loss": 0.0435, "step": 15762 }, { "epoch": 0.7324814126394052, "grad_norm": 0.36059685494518656, "learning_rate": 1.7737807538673018e-06, "loss": 0.0449, "step": 15763 }, { "epoch": 0.7325278810408922, "grad_norm": 0.3661011706046044, "learning_rate": 1.773205887768283e-06, "loss": 0.0359, "step": 15764 }, { "epoch": 0.7325743494423792, "grad_norm": 0.7310439678557221, "learning_rate": 1.772631094759546e-06, "loss": 0.0632, "step": 15765 }, { "epoch": 0.7326208178438661, "grad_norm": 0.728070695476442, "learning_rate": 1.7720563748541102e-06, "loss": 0.0561, "step": 15766 }, { "epoch": 0.7326672862453532, "grad_norm": 0.36142477817666246, "learning_rate": 1.7714817280649987e-06, "loss": 0.0464, "step": 15767 }, { "epoch": 0.7327137546468402, "grad_norm": 0.6394091663140364, "learning_rate": 1.770907154405222e-06, "loss": 0.054, "step": 15768 }, { "epoch": 0.7327602230483271, "grad_norm": 0.45917422258941526, "learning_rate": 1.7703326538877951e-06, "loss": 0.0334, "step": 15769 }, { "epoch": 0.7328066914498141, "grad_norm": 0.48294657864356993, "learning_rate": 1.769758226525734e-06, "loss": 0.046, "step": 15770 }, { "epoch": 0.7328531598513011, "grad_norm": 0.8119429585699406, "learning_rate": 1.7691838723320493e-06, "loss": 0.0588, "step": 15771 }, { "epoch": 0.7328996282527881, "grad_norm": 0.6469968234233361, "learning_rate": 1.7686095913197503e-06, "loss": 0.0452, "step": 15772 }, { "epoch": 0.7329460966542751, "grad_norm": 0.4566247406521596, "learning_rate": 1.768035383501845e-06, "loss": 0.0413, "step": 15773 }, { "epoch": 0.732992565055762, "grad_norm": 0.6780895413677291, "learning_rate": 1.76746124889134e-06, "loss": 0.0394, "step": 15774 }, { "epoch": 0.7330390334572491, "grad_norm": 0.521480909992371, "learning_rate": 1.7668871875012393e-06, "loss": 0.049, "step": 15775 }, { "epoch": 0.7330855018587361, "grad_norm": 0.68049060560789, "learning_rate": 1.7663131993445453e-06, "loss": 0.0666, "step": 15776 }, { "epoch": 0.733131970260223, "grad_norm": 0.38284332803498894, "learning_rate": 1.7657392844342624e-06, "loss": 0.0709, "step": 15777 }, { "epoch": 0.7331784386617101, "grad_norm": 0.48966588394537613, "learning_rate": 1.7651654427833893e-06, "loss": 0.0368, "step": 15778 }, { "epoch": 0.733224907063197, "grad_norm": 0.4718323964603238, "learning_rate": 1.7645916744049225e-06, "loss": 0.0416, "step": 15779 }, { "epoch": 0.733271375464684, "grad_norm": 0.34755168945950654, "learning_rate": 1.76401797931186e-06, "loss": 0.0381, "step": 15780 }, { "epoch": 0.733317843866171, "grad_norm": 0.3865654891001642, "learning_rate": 1.7634443575171962e-06, "loss": 0.0533, "step": 15781 }, { "epoch": 0.733364312267658, "grad_norm": 0.5737780552421043, "learning_rate": 1.7628708090339236e-06, "loss": 0.0466, "step": 15782 }, { "epoch": 0.733410780669145, "grad_norm": 1.261494272186241, "learning_rate": 1.7622973338750331e-06, "loss": 0.0529, "step": 15783 }, { "epoch": 0.7334572490706319, "grad_norm": 0.6552293762500456, "learning_rate": 1.7617239320535168e-06, "loss": 0.0348, "step": 15784 }, { "epoch": 0.733503717472119, "grad_norm": 0.5833589423694194, "learning_rate": 1.761150603582364e-06, "loss": 0.0586, "step": 15785 }, { "epoch": 0.733550185873606, "grad_norm": 1.0803611614808397, "learning_rate": 1.7605773484745548e-06, "loss": 0.0518, "step": 15786 }, { "epoch": 0.7335966542750929, "grad_norm": 0.6426940940356686, "learning_rate": 1.7600041667430801e-06, "loss": 0.0695, "step": 15787 }, { "epoch": 0.7336431226765799, "grad_norm": 0.47126406423273937, "learning_rate": 1.759431058400921e-06, "loss": 0.06, "step": 15788 }, { "epoch": 0.7336895910780669, "grad_norm": 0.8971077711970696, "learning_rate": 1.7588580234610592e-06, "loss": 0.0473, "step": 15789 }, { "epoch": 0.7337360594795539, "grad_norm": 0.6177943338433686, "learning_rate": 1.758285061936472e-06, "loss": 0.0547, "step": 15790 }, { "epoch": 0.7337825278810409, "grad_norm": 0.6261404495068161, "learning_rate": 1.7577121738401442e-06, "loss": 0.0601, "step": 15791 }, { "epoch": 0.7338289962825278, "grad_norm": 0.6320680012112364, "learning_rate": 1.757139359185046e-06, "loss": 0.0577, "step": 15792 }, { "epoch": 0.7338754646840149, "grad_norm": 0.5390306520527574, "learning_rate": 1.756566617984153e-06, "loss": 0.0518, "step": 15793 }, { "epoch": 0.7339219330855019, "grad_norm": 0.42406365361771514, "learning_rate": 1.7559939502504415e-06, "loss": 0.037, "step": 15794 }, { "epoch": 0.7339684014869888, "grad_norm": 0.5001917328872103, "learning_rate": 1.7554213559968807e-06, "loss": 0.0553, "step": 15795 }, { "epoch": 0.7340148698884759, "grad_norm": 0.6327926498323434, "learning_rate": 1.7548488352364418e-06, "loss": 0.0657, "step": 15796 }, { "epoch": 0.7340613382899628, "grad_norm": 0.47573007863183175, "learning_rate": 1.7542763879820918e-06, "loss": 0.0522, "step": 15797 }, { "epoch": 0.7341078066914498, "grad_norm": 0.5698932792951503, "learning_rate": 1.7537040142467976e-06, "loss": 0.0524, "step": 15798 }, { "epoch": 0.7341542750929368, "grad_norm": 0.44170828574328386, "learning_rate": 1.7531317140435244e-06, "loss": 0.051, "step": 15799 }, { "epoch": 0.7342007434944238, "grad_norm": 0.6010462319061006, "learning_rate": 1.752559487385233e-06, "loss": 0.0571, "step": 15800 }, { "epoch": 0.7342472118959108, "grad_norm": 0.38511688160476326, "learning_rate": 1.7519873342848893e-06, "loss": 0.0617, "step": 15801 }, { "epoch": 0.7342936802973977, "grad_norm": 0.4588305514464603, "learning_rate": 1.7514152547554508e-06, "loss": 0.0529, "step": 15802 }, { "epoch": 0.7343401486988848, "grad_norm": 0.6188304966438553, "learning_rate": 1.7508432488098753e-06, "loss": 0.0599, "step": 15803 }, { "epoch": 0.7343866171003718, "grad_norm": 0.6538893303346374, "learning_rate": 1.7502713164611206e-06, "loss": 0.0506, "step": 15804 }, { "epoch": 0.7344330855018587, "grad_norm": 0.35535627770687206, "learning_rate": 1.7496994577221405e-06, "loss": 0.0242, "step": 15805 }, { "epoch": 0.7344795539033457, "grad_norm": 0.6269072709815612, "learning_rate": 1.7491276726058887e-06, "loss": 0.0535, "step": 15806 }, { "epoch": 0.7345260223048327, "grad_norm": 1.050682099761406, "learning_rate": 1.748555961125315e-06, "loss": 0.0665, "step": 15807 }, { "epoch": 0.7345724907063197, "grad_norm": 0.36979396863785924, "learning_rate": 1.7479843232933724e-06, "loss": 0.0477, "step": 15808 }, { "epoch": 0.7346189591078067, "grad_norm": 0.32239858660020027, "learning_rate": 1.7474127591230095e-06, "loss": 0.0408, "step": 15809 }, { "epoch": 0.7346654275092936, "grad_norm": 0.4977608194526255, "learning_rate": 1.7468412686271675e-06, "loss": 0.0542, "step": 15810 }, { "epoch": 0.7347118959107807, "grad_norm": 0.5134986897053166, "learning_rate": 1.746269851818796e-06, "loss": 0.0425, "step": 15811 }, { "epoch": 0.7347583643122677, "grad_norm": 0.5134986897053166, "learning_rate": 1.746269851818796e-06, "loss": 0.0742, "step": 15812 }, { "epoch": 0.7348048327137546, "grad_norm": 0.6272366278882612, "learning_rate": 1.745698508710837e-06, "loss": 0.0683, "step": 15813 }, { "epoch": 0.7348513011152417, "grad_norm": 0.7377250101909377, "learning_rate": 1.7451272393162315e-06, "loss": 0.079, "step": 15814 }, { "epoch": 0.7348977695167286, "grad_norm": 0.5137593696371647, "learning_rate": 1.7445560436479187e-06, "loss": 0.0472, "step": 15815 }, { "epoch": 0.7349442379182156, "grad_norm": 0.5482365969981046, "learning_rate": 1.7439849217188409e-06, "loss": 0.0416, "step": 15816 }, { "epoch": 0.7349907063197026, "grad_norm": 0.5736729919678613, "learning_rate": 1.74341387354193e-06, "loss": 0.0482, "step": 15817 }, { "epoch": 0.7350371747211896, "grad_norm": 0.5985270955758539, "learning_rate": 1.7428428991301205e-06, "loss": 0.0589, "step": 15818 }, { "epoch": 0.7350836431226766, "grad_norm": 0.5329426713635865, "learning_rate": 1.7422719984963487e-06, "loss": 0.0723, "step": 15819 }, { "epoch": 0.7351301115241635, "grad_norm": 0.43291965173300406, "learning_rate": 1.7417011716535453e-06, "loss": 0.0401, "step": 15820 }, { "epoch": 0.7351765799256506, "grad_norm": 0.4827792058767393, "learning_rate": 1.7411304186146382e-06, "loss": 0.047, "step": 15821 }, { "epoch": 0.7352230483271376, "grad_norm": 0.8973782237154184, "learning_rate": 1.74055973939256e-06, "loss": 0.0517, "step": 15822 }, { "epoch": 0.7352695167286245, "grad_norm": 0.414391104921279, "learning_rate": 1.7399891340002328e-06, "loss": 0.046, "step": 15823 }, { "epoch": 0.7353159851301115, "grad_norm": 0.5488393884383078, "learning_rate": 1.739418602450582e-06, "loss": 0.0597, "step": 15824 }, { "epoch": 0.7353624535315985, "grad_norm": 0.4658391862363363, "learning_rate": 1.7388481447565308e-06, "loss": 0.0434, "step": 15825 }, { "epoch": 0.7354089219330855, "grad_norm": 0.45217043976560706, "learning_rate": 1.738277760931003e-06, "loss": 0.0341, "step": 15826 }, { "epoch": 0.7354553903345725, "grad_norm": 0.4029659586167582, "learning_rate": 1.737707450986918e-06, "loss": 0.0333, "step": 15827 }, { "epoch": 0.7355018587360594, "grad_norm": 0.4554008449909981, "learning_rate": 1.73713721493719e-06, "loss": 0.0463, "step": 15828 }, { "epoch": 0.7355483271375465, "grad_norm": 0.49307792107261705, "learning_rate": 1.7365670527947392e-06, "loss": 0.0404, "step": 15829 }, { "epoch": 0.7355947955390335, "grad_norm": 0.4781885267935692, "learning_rate": 1.7359969645724801e-06, "loss": 0.0549, "step": 15830 }, { "epoch": 0.7356412639405204, "grad_norm": 0.41062272314299497, "learning_rate": 1.7354269502833243e-06, "loss": 0.0538, "step": 15831 }, { "epoch": 0.7356877323420075, "grad_norm": 0.4923925322189053, "learning_rate": 1.734857009940183e-06, "loss": 0.0587, "step": 15832 }, { "epoch": 0.7357342007434944, "grad_norm": 0.4452509039596551, "learning_rate": 1.7342871435559699e-06, "loss": 0.043, "step": 15833 }, { "epoch": 0.7357806691449814, "grad_norm": 0.29805092900840785, "learning_rate": 1.7337173511435884e-06, "loss": 0.053, "step": 15834 }, { "epoch": 0.7358271375464684, "grad_norm": 0.7794220168891492, "learning_rate": 1.7331476327159446e-06, "loss": 0.059, "step": 15835 }, { "epoch": 0.7358736059479554, "grad_norm": 0.7493891857928484, "learning_rate": 1.7325779882859472e-06, "loss": 0.0627, "step": 15836 }, { "epoch": 0.7359200743494424, "grad_norm": 0.5555433983435873, "learning_rate": 1.7320084178664976e-06, "loss": 0.0556, "step": 15837 }, { "epoch": 0.7359665427509293, "grad_norm": 0.7695120244818794, "learning_rate": 1.7314389214704947e-06, "loss": 0.0555, "step": 15838 }, { "epoch": 0.7360130111524164, "grad_norm": 0.42974275263878137, "learning_rate": 1.7308694991108438e-06, "loss": 0.0408, "step": 15839 }, { "epoch": 0.7360594795539034, "grad_norm": 0.583851251593206, "learning_rate": 1.730300150800438e-06, "loss": 0.0493, "step": 15840 }, { "epoch": 0.7361059479553903, "grad_norm": 0.751955095495446, "learning_rate": 1.7297308765521748e-06, "loss": 0.0683, "step": 15841 }, { "epoch": 0.7361524163568773, "grad_norm": 0.3405009334462801, "learning_rate": 1.7291616763789475e-06, "loss": 0.0334, "step": 15842 }, { "epoch": 0.7361988847583643, "grad_norm": 0.5492736029906093, "learning_rate": 1.728592550293653e-06, "loss": 0.0503, "step": 15843 }, { "epoch": 0.7362453531598513, "grad_norm": 0.23460972611613837, "learning_rate": 1.7280234983091792e-06, "loss": 0.0236, "step": 15844 }, { "epoch": 0.7362918215613383, "grad_norm": 0.3715551693302363, "learning_rate": 1.727454520438418e-06, "loss": 0.04, "step": 15845 }, { "epoch": 0.7363382899628252, "grad_norm": 0.45246831603435195, "learning_rate": 1.7268856166942556e-06, "loss": 0.035, "step": 15846 }, { "epoch": 0.7363847583643123, "grad_norm": 0.6576266529117716, "learning_rate": 1.7263167870895787e-06, "loss": 0.0499, "step": 15847 }, { "epoch": 0.7364312267657993, "grad_norm": 0.4618739266357079, "learning_rate": 1.7257480316372704e-06, "loss": 0.0448, "step": 15848 }, { "epoch": 0.7364776951672862, "grad_norm": 0.48156418711241505, "learning_rate": 1.7251793503502173e-06, "loss": 0.0362, "step": 15849 }, { "epoch": 0.7365241635687733, "grad_norm": 0.47272556970808255, "learning_rate": 1.7246107432412984e-06, "loss": 0.0416, "step": 15850 }, { "epoch": 0.7365706319702602, "grad_norm": 0.5799155332408701, "learning_rate": 1.7240422103233956e-06, "loss": 0.0518, "step": 15851 }, { "epoch": 0.7366171003717472, "grad_norm": 0.4881015156933674, "learning_rate": 1.7234737516093808e-06, "loss": 0.0504, "step": 15852 }, { "epoch": 0.7366635687732342, "grad_norm": 0.6819079366781046, "learning_rate": 1.7229053671121355e-06, "loss": 0.0577, "step": 15853 }, { "epoch": 0.7367100371747212, "grad_norm": 0.36699542440975036, "learning_rate": 1.722337056844533e-06, "loss": 0.0297, "step": 15854 }, { "epoch": 0.7367565055762082, "grad_norm": 0.4751556450124848, "learning_rate": 1.7217688208194444e-06, "loss": 0.0565, "step": 15855 }, { "epoch": 0.7368029739776951, "grad_norm": 1.6953247758797783, "learning_rate": 1.7212006590497438e-06, "loss": 0.0382, "step": 15856 }, { "epoch": 0.7368494423791822, "grad_norm": 0.3228166504779879, "learning_rate": 1.7206325715483003e-06, "loss": 0.0432, "step": 15857 }, { "epoch": 0.7368959107806692, "grad_norm": 0.5565979943578283, "learning_rate": 1.7200645583279797e-06, "loss": 0.0532, "step": 15858 }, { "epoch": 0.7369423791821561, "grad_norm": 0.46302530534231057, "learning_rate": 1.7194966194016466e-06, "loss": 0.065, "step": 15859 }, { "epoch": 0.7369888475836431, "grad_norm": 0.5493015289026167, "learning_rate": 1.7189287547821698e-06, "loss": 0.0407, "step": 15860 }, { "epoch": 0.7370353159851301, "grad_norm": 0.5390110947245493, "learning_rate": 1.7183609644824096e-06, "loss": 0.0422, "step": 15861 }, { "epoch": 0.7370817843866171, "grad_norm": 1.0694839598617376, "learning_rate": 1.717793248515226e-06, "loss": 0.0575, "step": 15862 }, { "epoch": 0.7371282527881041, "grad_norm": 0.6193556618843917, "learning_rate": 1.717225606893483e-06, "loss": 0.0425, "step": 15863 }, { "epoch": 0.737174721189591, "grad_norm": 0.42905911757838283, "learning_rate": 1.7166580396300336e-06, "loss": 0.0544, "step": 15864 }, { "epoch": 0.7372211895910781, "grad_norm": 0.6615838899985736, "learning_rate": 1.716090546737733e-06, "loss": 0.0757, "step": 15865 }, { "epoch": 0.7372676579925651, "grad_norm": 0.8741801653158804, "learning_rate": 1.7155231282294398e-06, "loss": 0.0827, "step": 15866 }, { "epoch": 0.737314126394052, "grad_norm": 0.6145872949248449, "learning_rate": 1.7149557841180048e-06, "loss": 0.0501, "step": 15867 }, { "epoch": 0.7373605947955391, "grad_norm": 0.5147493596107563, "learning_rate": 1.7143885144162787e-06, "loss": 0.0519, "step": 15868 }, { "epoch": 0.737407063197026, "grad_norm": 0.5142097749562923, "learning_rate": 1.7138213191371106e-06, "loss": 0.0452, "step": 15869 }, { "epoch": 0.737453531598513, "grad_norm": 0.353503838616227, "learning_rate": 1.7132541982933487e-06, "loss": 0.055, "step": 15870 }, { "epoch": 0.7375, "grad_norm": 0.6990180653900078, "learning_rate": 1.7126871518978378e-06, "loss": 0.0654, "step": 15871 }, { "epoch": 0.737546468401487, "grad_norm": 0.5531652599989781, "learning_rate": 1.7121201799634213e-06, "loss": 0.0358, "step": 15872 }, { "epoch": 0.737592936802974, "grad_norm": 0.42452228593118946, "learning_rate": 1.711553282502945e-06, "loss": 0.0508, "step": 15873 }, { "epoch": 0.7376394052044609, "grad_norm": 0.4147084661920779, "learning_rate": 1.7109864595292481e-06, "loss": 0.0434, "step": 15874 }, { "epoch": 0.737685873605948, "grad_norm": 0.40167619299443125, "learning_rate": 1.710419711055169e-06, "loss": 0.0377, "step": 15875 }, { "epoch": 0.737732342007435, "grad_norm": 0.6834841228025194, "learning_rate": 1.7098530370935456e-06, "loss": 0.0466, "step": 15876 }, { "epoch": 0.7377788104089219, "grad_norm": 0.6699950479498965, "learning_rate": 1.7092864376572143e-06, "loss": 0.0541, "step": 15877 }, { "epoch": 0.7378252788104089, "grad_norm": 0.4423692703237585, "learning_rate": 1.708719912759008e-06, "loss": 0.0546, "step": 15878 }, { "epoch": 0.7378717472118959, "grad_norm": 0.552486650320877, "learning_rate": 1.7081534624117586e-06, "loss": 0.0544, "step": 15879 }, { "epoch": 0.7379182156133829, "grad_norm": 0.6330200082895223, "learning_rate": 1.7075870866282989e-06, "loss": 0.0551, "step": 15880 }, { "epoch": 0.7379646840148699, "grad_norm": 0.5049364096196733, "learning_rate": 1.7070207854214593e-06, "loss": 0.0593, "step": 15881 }, { "epoch": 0.7380111524163568, "grad_norm": 0.5310189520713993, "learning_rate": 1.706454558804061e-06, "loss": 0.0501, "step": 15882 }, { "epoch": 0.7380576208178439, "grad_norm": 0.2910548657884315, "learning_rate": 1.7058884067889347e-06, "loss": 0.0275, "step": 15883 }, { "epoch": 0.7381040892193309, "grad_norm": 0.37498396802339734, "learning_rate": 1.7053223293889031e-06, "loss": 0.0445, "step": 15884 }, { "epoch": 0.7381505576208178, "grad_norm": 0.4621321470561838, "learning_rate": 1.7047563266167888e-06, "loss": 0.0416, "step": 15885 }, { "epoch": 0.7381970260223049, "grad_norm": 0.4559710585952469, "learning_rate": 1.7041903984854113e-06, "loss": 0.0529, "step": 15886 }, { "epoch": 0.7382434944237918, "grad_norm": 0.8084185298678851, "learning_rate": 1.7036245450075906e-06, "loss": 0.0489, "step": 15887 }, { "epoch": 0.7382899628252788, "grad_norm": 0.5442280223962769, "learning_rate": 1.703058766196143e-06, "loss": 0.0562, "step": 15888 }, { "epoch": 0.7383364312267658, "grad_norm": 0.5187054008813139, "learning_rate": 1.702493062063883e-06, "loss": 0.0424, "step": 15889 }, { "epoch": 0.7383828996282528, "grad_norm": 0.37075440406871857, "learning_rate": 1.7019274326236269e-06, "loss": 0.0539, "step": 15890 }, { "epoch": 0.7384293680297398, "grad_norm": 0.43022708856182235, "learning_rate": 1.7013618778881858e-06, "loss": 0.0584, "step": 15891 }, { "epoch": 0.7384758364312267, "grad_norm": 0.604313035497074, "learning_rate": 1.7007963978703696e-06, "loss": 0.0409, "step": 15892 }, { "epoch": 0.7385223048327137, "grad_norm": 0.474559593219758, "learning_rate": 1.700230992582988e-06, "loss": 0.0409, "step": 15893 }, { "epoch": 0.7385687732342008, "grad_norm": 0.4025596386702212, "learning_rate": 1.6996656620388468e-06, "loss": 0.0441, "step": 15894 }, { "epoch": 0.7386152416356877, "grad_norm": 0.7316790424108307, "learning_rate": 1.6991004062507522e-06, "loss": 0.0839, "step": 15895 }, { "epoch": 0.7386617100371747, "grad_norm": 0.516907189866041, "learning_rate": 1.6985352252315056e-06, "loss": 0.0587, "step": 15896 }, { "epoch": 0.7387081784386617, "grad_norm": 0.5343188549981625, "learning_rate": 1.697970118993913e-06, "loss": 0.0357, "step": 15897 }, { "epoch": 0.7387546468401487, "grad_norm": 0.6047752139074678, "learning_rate": 1.6974050875507736e-06, "loss": 0.0537, "step": 15898 }, { "epoch": 0.7388011152416357, "grad_norm": 3.219972936712992, "learning_rate": 1.6968401309148819e-06, "loss": 0.0697, "step": 15899 }, { "epoch": 0.7388475836431226, "grad_norm": 0.5750369728571609, "learning_rate": 1.6962752490990392e-06, "loss": 0.0517, "step": 15900 }, { "epoch": 0.7388940520446097, "grad_norm": 0.5640566086214281, "learning_rate": 1.6957104421160398e-06, "loss": 0.0532, "step": 15901 }, { "epoch": 0.7389405204460967, "grad_norm": 0.4888034340953507, "learning_rate": 1.695145709978676e-06, "loss": 0.0503, "step": 15902 }, { "epoch": 0.7389869888475836, "grad_norm": 0.6551342953946568, "learning_rate": 1.6945810526997385e-06, "loss": 0.0617, "step": 15903 }, { "epoch": 0.7390334572490707, "grad_norm": 0.5095807018750488, "learning_rate": 1.6940164702920225e-06, "loss": 0.0601, "step": 15904 }, { "epoch": 0.7390799256505576, "grad_norm": 0.3620165860499036, "learning_rate": 1.6934519627683116e-06, "loss": 0.0393, "step": 15905 }, { "epoch": 0.7391263940520446, "grad_norm": 0.41092057597159937, "learning_rate": 1.6928875301413916e-06, "loss": 0.0564, "step": 15906 }, { "epoch": 0.7391728624535316, "grad_norm": 0.8798372980737911, "learning_rate": 1.6923231724240519e-06, "loss": 0.0596, "step": 15907 }, { "epoch": 0.7392193308550186, "grad_norm": 0.48970164551465256, "learning_rate": 1.6917588896290733e-06, "loss": 0.0467, "step": 15908 }, { "epoch": 0.7392657992565056, "grad_norm": 0.5104746368090171, "learning_rate": 1.6911946817692377e-06, "loss": 0.0581, "step": 15909 }, { "epoch": 0.7393122676579925, "grad_norm": 0.41397099223464223, "learning_rate": 1.6906305488573248e-06, "loss": 0.0285, "step": 15910 }, { "epoch": 0.7393587360594795, "grad_norm": 0.8285945764161685, "learning_rate": 1.6900664909061138e-06, "loss": 0.0676, "step": 15911 }, { "epoch": 0.7394052044609666, "grad_norm": 0.5825804295301458, "learning_rate": 1.6895025079283794e-06, "loss": 0.0502, "step": 15912 }, { "epoch": 0.7394516728624535, "grad_norm": 0.6575215293354156, "learning_rate": 1.6889385999368968e-06, "loss": 0.0543, "step": 15913 }, { "epoch": 0.7394981412639405, "grad_norm": 0.5158867390457091, "learning_rate": 1.6883747669444406e-06, "loss": 0.0434, "step": 15914 }, { "epoch": 0.7395446096654275, "grad_norm": 0.8304948419977757, "learning_rate": 1.6878110089637817e-06, "loss": 0.0522, "step": 15915 }, { "epoch": 0.7395910780669145, "grad_norm": 0.7415910742244406, "learning_rate": 1.6872473260076894e-06, "loss": 0.0533, "step": 15916 }, { "epoch": 0.7396375464684015, "grad_norm": 0.5561906942988198, "learning_rate": 1.6866837180889318e-06, "loss": 0.0415, "step": 15917 }, { "epoch": 0.7396840148698884, "grad_norm": 0.7676024554315509, "learning_rate": 1.6861201852202753e-06, "loss": 0.064, "step": 15918 }, { "epoch": 0.7397304832713755, "grad_norm": 0.511157914966784, "learning_rate": 1.685556727414484e-06, "loss": 0.0365, "step": 15919 }, { "epoch": 0.7397769516728625, "grad_norm": 0.38453602917368884, "learning_rate": 1.6849933446843198e-06, "loss": 0.0363, "step": 15920 }, { "epoch": 0.7398234200743494, "grad_norm": 0.4299337207597403, "learning_rate": 1.684430037042547e-06, "loss": 0.0348, "step": 15921 }, { "epoch": 0.7398698884758365, "grad_norm": 0.4653920897937453, "learning_rate": 1.6838668045019252e-06, "loss": 0.0367, "step": 15922 }, { "epoch": 0.7399163568773234, "grad_norm": 0.6168809245207797, "learning_rate": 1.6833036470752072e-06, "loss": 0.0639, "step": 15923 }, { "epoch": 0.7399628252788104, "grad_norm": 0.4016179515464488, "learning_rate": 1.6827405647751537e-06, "loss": 0.0601, "step": 15924 }, { "epoch": 0.7400092936802974, "grad_norm": 0.46887321362132195, "learning_rate": 1.6821775576145177e-06, "loss": 0.0642, "step": 15925 }, { "epoch": 0.7400557620817844, "grad_norm": 0.581716545697992, "learning_rate": 1.6816146256060522e-06, "loss": 0.0405, "step": 15926 }, { "epoch": 0.7401022304832714, "grad_norm": 0.5175314555966729, "learning_rate": 1.6810517687625067e-06, "loss": 0.0503, "step": 15927 }, { "epoch": 0.7401486988847583, "grad_norm": 0.5275305445165311, "learning_rate": 1.6804889870966346e-06, "loss": 0.0626, "step": 15928 }, { "epoch": 0.7401951672862453, "grad_norm": 0.4445433437229295, "learning_rate": 1.6799262806211786e-06, "loss": 0.0428, "step": 15929 }, { "epoch": 0.7402416356877324, "grad_norm": 0.5616657026504631, "learning_rate": 1.6793636493488858e-06, "loss": 0.0369, "step": 15930 }, { "epoch": 0.7402881040892193, "grad_norm": 0.8280398594991603, "learning_rate": 1.6788010932925025e-06, "loss": 0.0619, "step": 15931 }, { "epoch": 0.7403345724907063, "grad_norm": 2.2105183952182625, "learning_rate": 1.67823861246477e-06, "loss": 0.0663, "step": 15932 }, { "epoch": 0.7403810408921933, "grad_norm": 0.5807685109263451, "learning_rate": 1.6776762068784292e-06, "loss": 0.0428, "step": 15933 }, { "epoch": 0.7404275092936803, "grad_norm": 0.7508992261386833, "learning_rate": 1.6771138765462187e-06, "loss": 0.0622, "step": 15934 }, { "epoch": 0.7404739776951673, "grad_norm": 0.4123713072745001, "learning_rate": 1.6765516214808764e-06, "loss": 0.0387, "step": 15935 }, { "epoch": 0.7405204460966542, "grad_norm": 0.4258104991223086, "learning_rate": 1.6759894416951371e-06, "loss": 0.0678, "step": 15936 }, { "epoch": 0.7405669144981413, "grad_norm": 0.6978319030640713, "learning_rate": 1.6754273372017343e-06, "loss": 0.0568, "step": 15937 }, { "epoch": 0.7406133828996283, "grad_norm": 0.7956577145996188, "learning_rate": 1.674865308013403e-06, "loss": 0.0568, "step": 15938 }, { "epoch": 0.7406598513011152, "grad_norm": 0.5210733009179882, "learning_rate": 1.6743033541428721e-06, "loss": 0.0516, "step": 15939 }, { "epoch": 0.7407063197026023, "grad_norm": 0.38188547330463113, "learning_rate": 1.6737414756028708e-06, "loss": 0.0455, "step": 15940 }, { "epoch": 0.7407527881040892, "grad_norm": 0.5623123756059687, "learning_rate": 1.6731796724061255e-06, "loss": 0.055, "step": 15941 }, { "epoch": 0.7407992565055762, "grad_norm": 0.44423074016411357, "learning_rate": 1.6726179445653623e-06, "loss": 0.0488, "step": 15942 }, { "epoch": 0.7408457249070632, "grad_norm": 1.6607111921299453, "learning_rate": 1.6720562920933042e-06, "loss": 0.0548, "step": 15943 }, { "epoch": 0.7408921933085502, "grad_norm": 0.7482903872633078, "learning_rate": 1.6714947150026722e-06, "loss": 0.0571, "step": 15944 }, { "epoch": 0.7409386617100372, "grad_norm": 0.37437010730722964, "learning_rate": 1.6709332133061917e-06, "loss": 0.052, "step": 15945 }, { "epoch": 0.7409851301115241, "grad_norm": 0.38107509423072533, "learning_rate": 1.670371787016576e-06, "loss": 0.0417, "step": 15946 }, { "epoch": 0.7410315985130111, "grad_norm": 0.7574437085286968, "learning_rate": 1.6698104361465416e-06, "loss": 0.0594, "step": 15947 }, { "epoch": 0.7410780669144982, "grad_norm": 0.5964621801249915, "learning_rate": 1.669249160708808e-06, "loss": 0.0585, "step": 15948 }, { "epoch": 0.7411245353159851, "grad_norm": 0.8327922207284812, "learning_rate": 1.668687960716086e-06, "loss": 0.0593, "step": 15949 }, { "epoch": 0.7411710037174721, "grad_norm": 0.47949080689545026, "learning_rate": 1.6681268361810881e-06, "loss": 0.0415, "step": 15950 }, { "epoch": 0.741217472118959, "grad_norm": 0.43075738522101564, "learning_rate": 1.667565787116524e-06, "loss": 0.0472, "step": 15951 }, { "epoch": 0.7412639405204461, "grad_norm": 0.8036030285384843, "learning_rate": 1.667004813535102e-06, "loss": 0.0518, "step": 15952 }, { "epoch": 0.7413104089219331, "grad_norm": 0.5202334561973512, "learning_rate": 1.6664439154495289e-06, "loss": 0.0287, "step": 15953 }, { "epoch": 0.74135687732342, "grad_norm": 1.0762834529466219, "learning_rate": 1.6658830928725084e-06, "loss": 0.0814, "step": 15954 }, { "epoch": 0.7414033457249071, "grad_norm": 0.4405623007839207, "learning_rate": 1.665322345816746e-06, "loss": 0.0507, "step": 15955 }, { "epoch": 0.7414498141263941, "grad_norm": 1.5871595611819933, "learning_rate": 1.6647616742949424e-06, "loss": 0.072, "step": 15956 }, { "epoch": 0.741496282527881, "grad_norm": 2.2237092067591138, "learning_rate": 1.6642010783197977e-06, "loss": 0.0593, "step": 15957 }, { "epoch": 0.741542750929368, "grad_norm": 0.575044462785303, "learning_rate": 1.6636405579040087e-06, "loss": 0.0599, "step": 15958 }, { "epoch": 0.741589219330855, "grad_norm": 0.4384330609043411, "learning_rate": 1.6630801130602725e-06, "loss": 0.0467, "step": 15959 }, { "epoch": 0.741635687732342, "grad_norm": 0.3475286324560644, "learning_rate": 1.6625197438012842e-06, "loss": 0.0393, "step": 15960 }, { "epoch": 0.741682156133829, "grad_norm": 0.613773534963243, "learning_rate": 1.6619594501397346e-06, "loss": 0.0448, "step": 15961 }, { "epoch": 0.741728624535316, "grad_norm": 0.40925992358486574, "learning_rate": 1.6613992320883183e-06, "loss": 0.043, "step": 15962 }, { "epoch": 0.741775092936803, "grad_norm": 2.6171944296108953, "learning_rate": 1.6608390896597248e-06, "loss": 0.0582, "step": 15963 }, { "epoch": 0.7418215613382899, "grad_norm": 0.7726505281577887, "learning_rate": 1.660279022866637e-06, "loss": 0.0681, "step": 15964 }, { "epoch": 0.741868029739777, "grad_norm": 0.6948255737555162, "learning_rate": 1.659719031721746e-06, "loss": 0.0554, "step": 15965 }, { "epoch": 0.741914498141264, "grad_norm": 0.4656243339035077, "learning_rate": 1.6591591162377347e-06, "loss": 0.0455, "step": 15966 }, { "epoch": 0.7419609665427509, "grad_norm": 0.690033725913158, "learning_rate": 1.6585992764272851e-06, "loss": 0.0627, "step": 15967 }, { "epoch": 0.7420074349442379, "grad_norm": 0.6317093270724077, "learning_rate": 1.6580395123030773e-06, "loss": 0.0478, "step": 15968 }, { "epoch": 0.7420539033457249, "grad_norm": 0.4519111047672996, "learning_rate": 1.657479823877795e-06, "loss": 0.0544, "step": 15969 }, { "epoch": 0.7421003717472119, "grad_norm": 0.24737328562620656, "learning_rate": 1.6569202111641115e-06, "loss": 0.0379, "step": 15970 }, { "epoch": 0.7421468401486989, "grad_norm": 0.5510529663029324, "learning_rate": 1.6563606741747018e-06, "loss": 0.056, "step": 15971 }, { "epoch": 0.7421933085501858, "grad_norm": 0.526971829726559, "learning_rate": 1.6558012129222434e-06, "loss": 0.0498, "step": 15972 }, { "epoch": 0.7422397769516729, "grad_norm": 0.5567534531070535, "learning_rate": 1.6552418274194076e-06, "loss": 0.0503, "step": 15973 }, { "epoch": 0.7422862453531599, "grad_norm": 0.35832438047158816, "learning_rate": 1.6546825176788645e-06, "loss": 0.0299, "step": 15974 }, { "epoch": 0.7423327137546468, "grad_norm": 0.48529799582628197, "learning_rate": 1.654123283713283e-06, "loss": 0.0461, "step": 15975 }, { "epoch": 0.7423791821561339, "grad_norm": 0.3598382867003932, "learning_rate": 1.6535641255353313e-06, "loss": 0.0611, "step": 15976 }, { "epoch": 0.7424256505576208, "grad_norm": 0.5410626895958391, "learning_rate": 1.6530050431576734e-06, "loss": 0.0486, "step": 15977 }, { "epoch": 0.7424721189591078, "grad_norm": 0.441116286036957, "learning_rate": 1.6524460365929728e-06, "loss": 0.0681, "step": 15978 }, { "epoch": 0.7425185873605948, "grad_norm": 0.4372195148582648, "learning_rate": 1.6518871058538943e-06, "loss": 0.0456, "step": 15979 }, { "epoch": 0.7425650557620818, "grad_norm": 0.5971776685507614, "learning_rate": 1.6513282509530965e-06, "loss": 0.0638, "step": 15980 }, { "epoch": 0.7426115241635688, "grad_norm": 0.5821806775650857, "learning_rate": 1.650769471903238e-06, "loss": 0.0538, "step": 15981 }, { "epoch": 0.7426579925650557, "grad_norm": 1.234578652598568, "learning_rate": 1.6502107687169765e-06, "loss": 0.0736, "step": 15982 }, { "epoch": 0.7427044609665427, "grad_norm": 0.5141803088236432, "learning_rate": 1.6496521414069667e-06, "loss": 0.0394, "step": 15983 }, { "epoch": 0.7427509293680298, "grad_norm": 0.42519628913817886, "learning_rate": 1.6490935899858618e-06, "loss": 0.0371, "step": 15984 }, { "epoch": 0.7427973977695167, "grad_norm": 0.6318875471045238, "learning_rate": 1.648535114466312e-06, "loss": 0.0508, "step": 15985 }, { "epoch": 0.7428438661710037, "grad_norm": 0.8251222364244918, "learning_rate": 1.647976714860971e-06, "loss": 0.0815, "step": 15986 }, { "epoch": 0.7428903345724907, "grad_norm": 0.6167139995065949, "learning_rate": 1.6474183911824871e-06, "loss": 0.0653, "step": 15987 }, { "epoch": 0.7429368029739777, "grad_norm": 0.5329718425518968, "learning_rate": 1.6468601434435022e-06, "loss": 0.0318, "step": 15988 }, { "epoch": 0.7429832713754647, "grad_norm": 1.6637664849354146, "learning_rate": 1.6463019716566653e-06, "loss": 0.038, "step": 15989 }, { "epoch": 0.7430297397769516, "grad_norm": 0.44796625707552745, "learning_rate": 1.6457438758346184e-06, "loss": 0.0675, "step": 15990 }, { "epoch": 0.7430762081784387, "grad_norm": 0.7349533173466555, "learning_rate": 1.6451858559900036e-06, "loss": 0.0446, "step": 15991 }, { "epoch": 0.7431226765799257, "grad_norm": 0.6575233196476507, "learning_rate": 1.644627912135458e-06, "loss": 0.0554, "step": 15992 }, { "epoch": 0.7431691449814126, "grad_norm": 0.42226238478180894, "learning_rate": 1.644070044283625e-06, "loss": 0.0489, "step": 15993 }, { "epoch": 0.7432156133828997, "grad_norm": 0.617369689759597, "learning_rate": 1.6435122524471358e-06, "loss": 0.0665, "step": 15994 }, { "epoch": 0.7432620817843866, "grad_norm": 0.666371146321629, "learning_rate": 1.6429545366386252e-06, "loss": 0.0341, "step": 15995 }, { "epoch": 0.7433085501858736, "grad_norm": 0.5445607098352053, "learning_rate": 1.6423968968707288e-06, "loss": 0.0469, "step": 15996 }, { "epoch": 0.7433550185873606, "grad_norm": 0.6450379653556102, "learning_rate": 1.6418393331560772e-06, "loss": 0.0538, "step": 15997 }, { "epoch": 0.7434014869888476, "grad_norm": 0.5263690831412287, "learning_rate": 1.6412818455072988e-06, "loss": 0.0438, "step": 15998 }, { "epoch": 0.7434479553903346, "grad_norm": 0.39122669955450606, "learning_rate": 1.6407244339370215e-06, "loss": 0.0376, "step": 15999 }, { "epoch": 0.7434944237918215, "grad_norm": 0.7371863790107129, "learning_rate": 1.6401670984578711e-06, "loss": 0.0808, "step": 16000 }, { "epoch": 0.7435408921933085, "grad_norm": 0.9285740196628574, "learning_rate": 1.6396098390824728e-06, "loss": 0.0513, "step": 16001 }, { "epoch": 0.7435873605947956, "grad_norm": 0.7537685947838654, "learning_rate": 1.639052655823446e-06, "loss": 0.0393, "step": 16002 }, { "epoch": 0.7436338289962825, "grad_norm": 0.5622386493452448, "learning_rate": 1.6384955486934157e-06, "loss": 0.0577, "step": 16003 }, { "epoch": 0.7436802973977695, "grad_norm": 0.71763049451674, "learning_rate": 1.637938517705001e-06, "loss": 0.0716, "step": 16004 }, { "epoch": 0.7437267657992565, "grad_norm": 0.5602315208312871, "learning_rate": 1.637381562870814e-06, "loss": 0.0467, "step": 16005 }, { "epoch": 0.7437732342007435, "grad_norm": 0.6834628473725503, "learning_rate": 1.6368246842034752e-06, "loss": 0.0541, "step": 16006 }, { "epoch": 0.7438197026022305, "grad_norm": 0.5620641256353699, "learning_rate": 1.6362678817155975e-06, "loss": 0.0468, "step": 16007 }, { "epoch": 0.7438661710037174, "grad_norm": 0.7375172997371934, "learning_rate": 1.6357111554197918e-06, "loss": 0.0517, "step": 16008 }, { "epoch": 0.7439126394052045, "grad_norm": 0.49238105414576655, "learning_rate": 1.6351545053286683e-06, "loss": 0.0441, "step": 16009 }, { "epoch": 0.7439591078066915, "grad_norm": 0.5261282107689187, "learning_rate": 1.6345979314548393e-06, "loss": 0.0446, "step": 16010 }, { "epoch": 0.7440055762081784, "grad_norm": 0.48304137574376566, "learning_rate": 1.6340414338109085e-06, "loss": 0.0626, "step": 16011 }, { "epoch": 0.7440520446096655, "grad_norm": 0.6507618154907008, "learning_rate": 1.6334850124094792e-06, "loss": 0.0835, "step": 16012 }, { "epoch": 0.7440985130111524, "grad_norm": 0.8859357151678268, "learning_rate": 1.63292866726316e-06, "loss": 0.0544, "step": 16013 }, { "epoch": 0.7441449814126394, "grad_norm": 0.442008579044386, "learning_rate": 1.6323723983845496e-06, "loss": 0.0376, "step": 16014 }, { "epoch": 0.7441914498141264, "grad_norm": 0.44891137065873604, "learning_rate": 1.631816205786249e-06, "loss": 0.0417, "step": 16015 }, { "epoch": 0.7442379182156134, "grad_norm": 0.4259149793702367, "learning_rate": 1.631260089480856e-06, "loss": 0.0527, "step": 16016 }, { "epoch": 0.7442843866171004, "grad_norm": 0.4753259111458143, "learning_rate": 1.6307040494809685e-06, "loss": 0.054, "step": 16017 }, { "epoch": 0.7443308550185873, "grad_norm": 0.556275321146172, "learning_rate": 1.630148085799179e-06, "loss": 0.0458, "step": 16018 }, { "epoch": 0.7443773234200743, "grad_norm": 0.4029549062856106, "learning_rate": 1.6295921984480816e-06, "loss": 0.0548, "step": 16019 }, { "epoch": 0.7444237918215614, "grad_norm": 0.46334310073165724, "learning_rate": 1.6290363874402693e-06, "loss": 0.0469, "step": 16020 }, { "epoch": 0.7444702602230483, "grad_norm": 0.679588381261459, "learning_rate": 1.6284806527883307e-06, "loss": 0.0609, "step": 16021 }, { "epoch": 0.7445167286245353, "grad_norm": 0.3953637025990724, "learning_rate": 1.6279249945048542e-06, "loss": 0.0413, "step": 16022 }, { "epoch": 0.7445631970260222, "grad_norm": 0.5247118958137849, "learning_rate": 1.6273694126024247e-06, "loss": 0.0613, "step": 16023 }, { "epoch": 0.7446096654275093, "grad_norm": 0.5884316942819305, "learning_rate": 1.6268139070936285e-06, "loss": 0.0595, "step": 16024 }, { "epoch": 0.7446561338289963, "grad_norm": 0.4997529203284479, "learning_rate": 1.6262584779910472e-06, "loss": 0.0407, "step": 16025 }, { "epoch": 0.7447026022304832, "grad_norm": 0.698989562303672, "learning_rate": 1.6257031253072603e-06, "loss": 0.0497, "step": 16026 }, { "epoch": 0.7447490706319703, "grad_norm": 0.5710865704009327, "learning_rate": 1.625147849054851e-06, "loss": 0.06, "step": 16027 }, { "epoch": 0.7447955390334573, "grad_norm": 0.5405028793287074, "learning_rate": 1.6245926492463966e-06, "loss": 0.0551, "step": 16028 }, { "epoch": 0.7448420074349442, "grad_norm": 0.5770306957391659, "learning_rate": 1.624037525894468e-06, "loss": 0.0579, "step": 16029 }, { "epoch": 0.7448884758364313, "grad_norm": 0.8620270679670703, "learning_rate": 1.6234824790116438e-06, "loss": 0.0776, "step": 16030 }, { "epoch": 0.7449349442379182, "grad_norm": 0.5743794432796471, "learning_rate": 1.6229275086104952e-06, "loss": 0.0508, "step": 16031 }, { "epoch": 0.7449814126394052, "grad_norm": 0.6625617685321282, "learning_rate": 1.6223726147035929e-06, "loss": 0.0467, "step": 16032 }, { "epoch": 0.7450278810408922, "grad_norm": 0.6093406054558302, "learning_rate": 1.6218177973035043e-06, "loss": 0.0524, "step": 16033 }, { "epoch": 0.7450743494423792, "grad_norm": 0.6287126231000804, "learning_rate": 1.6212630564228017e-06, "loss": 0.0431, "step": 16034 }, { "epoch": 0.7451208178438662, "grad_norm": 0.7150796933731063, "learning_rate": 1.6207083920740446e-06, "loss": 0.0666, "step": 16035 }, { "epoch": 0.7451672862453531, "grad_norm": 0.506325620547876, "learning_rate": 1.6201538042697972e-06, "loss": 0.0657, "step": 16036 }, { "epoch": 0.7452137546468401, "grad_norm": 0.4518805259894368, "learning_rate": 1.6195992930226257e-06, "loss": 0.0292, "step": 16037 }, { "epoch": 0.7452602230483272, "grad_norm": 0.64105860991759, "learning_rate": 1.619044858345088e-06, "loss": 0.0726, "step": 16038 }, { "epoch": 0.7453066914498141, "grad_norm": 0.6892667168024751, "learning_rate": 1.6184905002497425e-06, "loss": 0.053, "step": 16039 }, { "epoch": 0.7453531598513011, "grad_norm": 0.945026704613146, "learning_rate": 1.6179362187491459e-06, "loss": 0.0741, "step": 16040 }, { "epoch": 0.745399628252788, "grad_norm": 0.4375422877666304, "learning_rate": 1.6173820138558538e-06, "loss": 0.0442, "step": 16041 }, { "epoch": 0.7454460966542751, "grad_norm": 0.4033301080877435, "learning_rate": 1.6168278855824188e-06, "loss": 0.0445, "step": 16042 }, { "epoch": 0.7454925650557621, "grad_norm": 0.8760902991737936, "learning_rate": 1.616273833941392e-06, "loss": 0.0437, "step": 16043 }, { "epoch": 0.745539033457249, "grad_norm": 0.6160892739470546, "learning_rate": 1.6157198589453249e-06, "loss": 0.0527, "step": 16044 }, { "epoch": 0.7455855018587361, "grad_norm": 0.46357116674261406, "learning_rate": 1.6151659606067655e-06, "loss": 0.0459, "step": 16045 }, { "epoch": 0.7456319702602231, "grad_norm": 0.4723565973052813, "learning_rate": 1.6146121389382596e-06, "loss": 0.0377, "step": 16046 }, { "epoch": 0.74567843866171, "grad_norm": 0.5148522690912332, "learning_rate": 1.6140583939523518e-06, "loss": 0.0608, "step": 16047 }, { "epoch": 0.745724907063197, "grad_norm": 0.5861101847249687, "learning_rate": 1.6135047256615848e-06, "loss": 0.0638, "step": 16048 }, { "epoch": 0.745771375464684, "grad_norm": 0.5229673910844582, "learning_rate": 1.6129511340784999e-06, "loss": 0.0475, "step": 16049 }, { "epoch": 0.745817843866171, "grad_norm": 0.6211348635572868, "learning_rate": 1.6123976192156355e-06, "loss": 0.0506, "step": 16050 }, { "epoch": 0.745864312267658, "grad_norm": 0.5113955279864097, "learning_rate": 1.611844181085531e-06, "loss": 0.0457, "step": 16051 }, { "epoch": 0.745910780669145, "grad_norm": 0.4237429749950227, "learning_rate": 1.6112908197007243e-06, "loss": 0.0444, "step": 16052 }, { "epoch": 0.745957249070632, "grad_norm": 0.43147285119499157, "learning_rate": 1.6107375350737437e-06, "loss": 0.0492, "step": 16053 }, { "epoch": 0.7460037174721189, "grad_norm": 0.3617204711398739, "learning_rate": 1.610184327217127e-06, "loss": 0.0475, "step": 16054 }, { "epoch": 0.7460501858736059, "grad_norm": 1.08724572300939, "learning_rate": 1.6096311961434026e-06, "loss": 0.0624, "step": 16055 }, { "epoch": 0.746096654275093, "grad_norm": 0.4841736867656586, "learning_rate": 1.6090781418650997e-06, "loss": 0.0562, "step": 16056 }, { "epoch": 0.7461431226765799, "grad_norm": 0.8043010310795408, "learning_rate": 1.6085251643947442e-06, "loss": 0.0629, "step": 16057 }, { "epoch": 0.7461895910780669, "grad_norm": 1.0805833210913671, "learning_rate": 1.6079722637448663e-06, "loss": 0.0497, "step": 16058 }, { "epoch": 0.7462360594795538, "grad_norm": 0.5309480572303538, "learning_rate": 1.6074194399279857e-06, "loss": 0.0616, "step": 16059 }, { "epoch": 0.7462825278810409, "grad_norm": 0.4129575857885877, "learning_rate": 1.6068666929566235e-06, "loss": 0.0658, "step": 16060 }, { "epoch": 0.7463289962825279, "grad_norm": 0.6008078908685769, "learning_rate": 1.6063140228433033e-06, "loss": 0.0421, "step": 16061 }, { "epoch": 0.7463754646840148, "grad_norm": 0.38380199048125974, "learning_rate": 1.6057614296005424e-06, "loss": 0.0286, "step": 16062 }, { "epoch": 0.7464219330855019, "grad_norm": 0.5886297213445099, "learning_rate": 1.6052089132408578e-06, "loss": 0.0636, "step": 16063 }, { "epoch": 0.7464684014869889, "grad_norm": 0.5317497662927586, "learning_rate": 1.604656473776764e-06, "loss": 0.0399, "step": 16064 }, { "epoch": 0.7465148698884758, "grad_norm": 0.5416861963943692, "learning_rate": 1.6041041112207745e-06, "loss": 0.064, "step": 16065 }, { "epoch": 0.7465613382899629, "grad_norm": 0.37433644796665805, "learning_rate": 1.603551825585401e-06, "loss": 0.0359, "step": 16066 }, { "epoch": 0.7466078066914498, "grad_norm": 0.5104714761598503, "learning_rate": 1.602999616883152e-06, "loss": 0.0335, "step": 16067 }, { "epoch": 0.7466542750929368, "grad_norm": 0.49616035563822714, "learning_rate": 1.6024474851265381e-06, "loss": 0.0551, "step": 16068 }, { "epoch": 0.7467007434944238, "grad_norm": 0.6526234881777081, "learning_rate": 1.6018954303280665e-06, "loss": 0.0446, "step": 16069 }, { "epoch": 0.7467472118959108, "grad_norm": 0.6738838522110708, "learning_rate": 1.6013434525002364e-06, "loss": 0.0542, "step": 16070 }, { "epoch": 0.7467936802973978, "grad_norm": 0.4634548947406894, "learning_rate": 1.600791551655556e-06, "loss": 0.0343, "step": 16071 }, { "epoch": 0.7468401486988847, "grad_norm": 0.576722866263219, "learning_rate": 1.6002397278065246e-06, "loss": 0.0536, "step": 16072 }, { "epoch": 0.7468866171003717, "grad_norm": 0.5101128695933296, "learning_rate": 1.5996879809656412e-06, "loss": 0.0403, "step": 16073 }, { "epoch": 0.7469330855018588, "grad_norm": 0.4135195832179969, "learning_rate": 1.5991363111454023e-06, "loss": 0.0435, "step": 16074 }, { "epoch": 0.7469795539033457, "grad_norm": 0.38353493160907537, "learning_rate": 1.5985847183583086e-06, "loss": 0.0403, "step": 16075 }, { "epoch": 0.7470260223048327, "grad_norm": 0.4698832147701705, "learning_rate": 1.5980332026168494e-06, "loss": 0.0448, "step": 16076 }, { "epoch": 0.7470724907063196, "grad_norm": 0.4237641661527016, "learning_rate": 1.597481763933517e-06, "loss": 0.0471, "step": 16077 }, { "epoch": 0.7471189591078067, "grad_norm": 0.7804632504439566, "learning_rate": 1.5969304023208053e-06, "loss": 0.0705, "step": 16078 }, { "epoch": 0.7471654275092937, "grad_norm": 0.5304068425875683, "learning_rate": 1.5963791177912019e-06, "loss": 0.0639, "step": 16079 }, { "epoch": 0.7472118959107806, "grad_norm": 0.8124271305250945, "learning_rate": 1.5958279103571932e-06, "loss": 0.0621, "step": 16080 }, { "epoch": 0.7472583643122677, "grad_norm": 0.5947199582513643, "learning_rate": 1.5952767800312657e-06, "loss": 0.046, "step": 16081 }, { "epoch": 0.7473048327137547, "grad_norm": 0.863956504907093, "learning_rate": 1.594725726825902e-06, "loss": 0.0672, "step": 16082 }, { "epoch": 0.7473513011152416, "grad_norm": 0.5967489726407121, "learning_rate": 1.594174750753585e-06, "loss": 0.0599, "step": 16083 }, { "epoch": 0.7473977695167286, "grad_norm": 0.3038166864645604, "learning_rate": 1.5936238518267928e-06, "loss": 0.0445, "step": 16084 }, { "epoch": 0.7474442379182156, "grad_norm": 0.6110111362280335, "learning_rate": 1.5930730300580072e-06, "loss": 0.0472, "step": 16085 }, { "epoch": 0.7474907063197026, "grad_norm": 0.5557844767897617, "learning_rate": 1.5925222854597028e-06, "loss": 0.0655, "step": 16086 }, { "epoch": 0.7475371747211896, "grad_norm": 0.6877091361504578, "learning_rate": 1.591971618044355e-06, "loss": 0.0597, "step": 16087 }, { "epoch": 0.7475836431226766, "grad_norm": 0.3585775943600933, "learning_rate": 1.5914210278244373e-06, "loss": 0.0389, "step": 16088 }, { "epoch": 0.7476301115241636, "grad_norm": 0.5948765512770797, "learning_rate": 1.590870514812421e-06, "loss": 0.0551, "step": 16089 }, { "epoch": 0.7476765799256505, "grad_norm": 0.591995995343381, "learning_rate": 1.5903200790207751e-06, "loss": 0.0685, "step": 16090 }, { "epoch": 0.7477230483271375, "grad_norm": 0.5422948598486554, "learning_rate": 1.5897697204619661e-06, "loss": 0.0596, "step": 16091 }, { "epoch": 0.7477695167286246, "grad_norm": 0.5055621574649232, "learning_rate": 1.5892194391484644e-06, "loss": 0.0561, "step": 16092 }, { "epoch": 0.7478159851301115, "grad_norm": 0.5902626646351845, "learning_rate": 1.5886692350927336e-06, "loss": 0.0478, "step": 16093 }, { "epoch": 0.7478624535315985, "grad_norm": 1.749459909759422, "learning_rate": 1.5881191083072317e-06, "loss": 0.0509, "step": 16094 }, { "epoch": 0.7479089219330854, "grad_norm": 0.4622771042723493, "learning_rate": 1.5875690588044245e-06, "loss": 0.0343, "step": 16095 }, { "epoch": 0.7479553903345725, "grad_norm": 0.6465564768000238, "learning_rate": 1.5870190865967694e-06, "loss": 0.0584, "step": 16096 }, { "epoch": 0.7480018587360595, "grad_norm": 0.50715891429474, "learning_rate": 1.5864691916967245e-06, "loss": 0.0496, "step": 16097 }, { "epoch": 0.7480483271375464, "grad_norm": 0.5560440601913929, "learning_rate": 1.5859193741167433e-06, "loss": 0.0504, "step": 16098 }, { "epoch": 0.7480947955390335, "grad_norm": 0.41289460241944176, "learning_rate": 1.5853696338692849e-06, "loss": 0.0537, "step": 16099 }, { "epoch": 0.7481412639405205, "grad_norm": 0.46359543535935627, "learning_rate": 1.5848199709667967e-06, "loss": 0.0595, "step": 16100 }, { "epoch": 0.7481877323420074, "grad_norm": 0.6210091248586395, "learning_rate": 1.5842703854217284e-06, "loss": 0.0582, "step": 16101 }, { "epoch": 0.7482342007434944, "grad_norm": 0.53168810946311, "learning_rate": 1.583720877246533e-06, "loss": 0.0412, "step": 16102 }, { "epoch": 0.7482806691449814, "grad_norm": 0.4864948577003357, "learning_rate": 1.5831714464536547e-06, "loss": 0.0778, "step": 16103 }, { "epoch": 0.7483271375464684, "grad_norm": 0.36605659474018887, "learning_rate": 1.5826220930555396e-06, "loss": 0.0294, "step": 16104 }, { "epoch": 0.7483736059479554, "grad_norm": 0.5887785057732401, "learning_rate": 1.582072817064631e-06, "loss": 0.0509, "step": 16105 }, { "epoch": 0.7484200743494424, "grad_norm": 0.5131949525966825, "learning_rate": 1.5815236184933702e-06, "loss": 0.0559, "step": 16106 }, { "epoch": 0.7484665427509294, "grad_norm": 0.6188080557767104, "learning_rate": 1.5809744973541973e-06, "loss": 0.0517, "step": 16107 }, { "epoch": 0.7485130111524163, "grad_norm": 0.6188080557767104, "learning_rate": 1.5809744973541973e-06, "loss": 0.0535, "step": 16108 }, { "epoch": 0.7485594795539033, "grad_norm": 0.29307997710837863, "learning_rate": 1.5804254536595492e-06, "loss": 0.0319, "step": 16109 }, { "epoch": 0.7486059479553904, "grad_norm": 0.9562892427823632, "learning_rate": 1.5798764874218648e-06, "loss": 0.0638, "step": 16110 }, { "epoch": 0.7486524163568773, "grad_norm": 0.6734689958221024, "learning_rate": 1.5793275986535778e-06, "loss": 0.0529, "step": 16111 }, { "epoch": 0.7486988847583643, "grad_norm": 0.5407006088017415, "learning_rate": 1.5787787873671206e-06, "loss": 0.0422, "step": 16112 }, { "epoch": 0.7487453531598512, "grad_norm": 0.8191823629517166, "learning_rate": 1.5782300535749251e-06, "loss": 0.0718, "step": 16113 }, { "epoch": 0.7487918215613383, "grad_norm": 0.7590314409482946, "learning_rate": 1.57768139728942e-06, "loss": 0.042, "step": 16114 }, { "epoch": 0.7488382899628253, "grad_norm": 0.36970178434347506, "learning_rate": 1.5771328185230333e-06, "loss": 0.0471, "step": 16115 }, { "epoch": 0.7488847583643122, "grad_norm": 0.49280322742115595, "learning_rate": 1.576584317288189e-06, "loss": 0.0675, "step": 16116 }, { "epoch": 0.7489312267657993, "grad_norm": 0.48947837514935494, "learning_rate": 1.5760358935973146e-06, "loss": 0.0542, "step": 16117 }, { "epoch": 0.7489776951672863, "grad_norm": 0.6406289289494131, "learning_rate": 1.5754875474628322e-06, "loss": 0.041, "step": 16118 }, { "epoch": 0.7490241635687732, "grad_norm": 0.5993582460563592, "learning_rate": 1.5749392788971586e-06, "loss": 0.0409, "step": 16119 }, { "epoch": 0.7490706319702602, "grad_norm": 0.631499546979741, "learning_rate": 1.5743910879127162e-06, "loss": 0.0537, "step": 16120 }, { "epoch": 0.7491171003717472, "grad_norm": 0.4480786281158345, "learning_rate": 1.5738429745219208e-06, "loss": 0.0549, "step": 16121 }, { "epoch": 0.7491635687732342, "grad_norm": 0.7761920049879442, "learning_rate": 1.5732949387371888e-06, "loss": 0.074, "step": 16122 }, { "epoch": 0.7492100371747212, "grad_norm": 0.3749186903460359, "learning_rate": 1.5727469805709321e-06, "loss": 0.0567, "step": 16123 }, { "epoch": 0.7492565055762082, "grad_norm": 0.5677688475865192, "learning_rate": 1.572199100035564e-06, "loss": 0.058, "step": 16124 }, { "epoch": 0.7493029739776952, "grad_norm": 0.43137096112557305, "learning_rate": 1.571651297143494e-06, "loss": 0.0548, "step": 16125 }, { "epoch": 0.7493494423791821, "grad_norm": 0.544023656212212, "learning_rate": 1.5711035719071283e-06, "loss": 0.0601, "step": 16126 }, { "epoch": 0.7493959107806691, "grad_norm": 1.3072108549520185, "learning_rate": 1.570555924338878e-06, "loss": 0.0561, "step": 16127 }, { "epoch": 0.7494423791821562, "grad_norm": 0.9594225751053062, "learning_rate": 1.5700083544511446e-06, "loss": 0.0773, "step": 16128 }, { "epoch": 0.7494888475836431, "grad_norm": 0.5277625661632953, "learning_rate": 1.5694608622563324e-06, "loss": 0.0596, "step": 16129 }, { "epoch": 0.7495353159851301, "grad_norm": 0.7220580229102542, "learning_rate": 1.5689134477668421e-06, "loss": 0.0635, "step": 16130 }, { "epoch": 0.749581784386617, "grad_norm": 0.41559136221507875, "learning_rate": 1.5683661109950731e-06, "loss": 0.0461, "step": 16131 }, { "epoch": 0.7496282527881041, "grad_norm": 0.561622857625577, "learning_rate": 1.5678188519534238e-06, "loss": 0.0572, "step": 16132 }, { "epoch": 0.7496747211895911, "grad_norm": 0.3738029772333769, "learning_rate": 1.5672716706542884e-06, "loss": 0.0424, "step": 16133 }, { "epoch": 0.749721189591078, "grad_norm": 0.4277122675253673, "learning_rate": 1.5667245671100633e-06, "loss": 0.0509, "step": 16134 }, { "epoch": 0.7497676579925651, "grad_norm": 0.7815851287853471, "learning_rate": 1.5661775413331426e-06, "loss": 0.0615, "step": 16135 }, { "epoch": 0.7498141263940521, "grad_norm": 0.626284910713428, "learning_rate": 1.5656305933359112e-06, "loss": 0.074, "step": 16136 }, { "epoch": 0.749860594795539, "grad_norm": 0.40507872816497503, "learning_rate": 1.565083723130763e-06, "loss": 0.0502, "step": 16137 }, { "epoch": 0.749907063197026, "grad_norm": 0.4684826802739196, "learning_rate": 1.5645369307300839e-06, "loss": 0.0497, "step": 16138 }, { "epoch": 0.749953531598513, "grad_norm": 0.6167474817632507, "learning_rate": 1.563990216146259e-06, "loss": 0.0639, "step": 16139 }, { "epoch": 0.75, "grad_norm": 0.45181618582279204, "learning_rate": 1.5634435793916703e-06, "loss": 0.0453, "step": 16140 }, { "epoch": 0.750046468401487, "grad_norm": 0.49169364062208887, "learning_rate": 1.5628970204787048e-06, "loss": 0.05, "step": 16141 }, { "epoch": 0.750092936802974, "grad_norm": 1.5263333905825527, "learning_rate": 1.5623505394197375e-06, "loss": 0.0479, "step": 16142 }, { "epoch": 0.750139405204461, "grad_norm": 0.4686088075382632, "learning_rate": 1.561804136227147e-06, "loss": 0.0584, "step": 16143 }, { "epoch": 0.7501858736059479, "grad_norm": 0.39032752078092225, "learning_rate": 1.5612578109133125e-06, "loss": 0.0602, "step": 16144 }, { "epoch": 0.7502323420074349, "grad_norm": 0.2653648660584622, "learning_rate": 1.5607115634906083e-06, "loss": 0.04, "step": 16145 }, { "epoch": 0.750278810408922, "grad_norm": 0.5317407548582505, "learning_rate": 1.5601653939714073e-06, "loss": 0.0372, "step": 16146 }, { "epoch": 0.7503252788104089, "grad_norm": 0.580634801472049, "learning_rate": 1.55961930236808e-06, "loss": 0.0487, "step": 16147 }, { "epoch": 0.7503717472118959, "grad_norm": 0.37365049871754663, "learning_rate": 1.5590732886929965e-06, "loss": 0.0301, "step": 16148 }, { "epoch": 0.750418215613383, "grad_norm": 0.36878035154892197, "learning_rate": 1.5585273529585248e-06, "loss": 0.053, "step": 16149 }, { "epoch": 0.7504646840148699, "grad_norm": 0.8942189316350924, "learning_rate": 1.5579814951770289e-06, "loss": 0.0513, "step": 16150 }, { "epoch": 0.7505111524163569, "grad_norm": 0.41635702333039065, "learning_rate": 1.5574357153608767e-06, "loss": 0.0528, "step": 16151 }, { "epoch": 0.7505576208178438, "grad_norm": 0.6284729267380447, "learning_rate": 1.556890013522428e-06, "loss": 0.0389, "step": 16152 }, { "epoch": 0.7506040892193309, "grad_norm": 0.4437380278412108, "learning_rate": 1.556344389674045e-06, "loss": 0.057, "step": 16153 }, { "epoch": 0.7506505576208179, "grad_norm": 0.7345675887119293, "learning_rate": 1.5557988438280858e-06, "loss": 0.0541, "step": 16154 }, { "epoch": 0.7506970260223048, "grad_norm": 0.9953318183225541, "learning_rate": 1.5552533759969073e-06, "loss": 0.0821, "step": 16155 }, { "epoch": 0.7507434944237918, "grad_norm": 0.4969314659709548, "learning_rate": 1.554707986192866e-06, "loss": 0.0471, "step": 16156 }, { "epoch": 0.7507899628252788, "grad_norm": 0.5205816329312887, "learning_rate": 1.5541626744283128e-06, "loss": 0.0538, "step": 16157 }, { "epoch": 0.7508364312267658, "grad_norm": 0.32106046416948647, "learning_rate": 1.5536174407156035e-06, "loss": 0.0466, "step": 16158 }, { "epoch": 0.7508828996282528, "grad_norm": 0.6990614947802305, "learning_rate": 1.5530722850670881e-06, "loss": 0.0371, "step": 16159 }, { "epoch": 0.7509293680297398, "grad_norm": 0.6692282127316618, "learning_rate": 1.5525272074951097e-06, "loss": 0.0608, "step": 16160 }, { "epoch": 0.7509758364312268, "grad_norm": 0.5990725691263178, "learning_rate": 1.5519822080120206e-06, "loss": 0.0563, "step": 16161 }, { "epoch": 0.7510223048327137, "grad_norm": 0.3131948622389985, "learning_rate": 1.5514372866301635e-06, "loss": 0.03, "step": 16162 }, { "epoch": 0.7510687732342007, "grad_norm": 0.5372072774230371, "learning_rate": 1.5508924433618811e-06, "loss": 0.0575, "step": 16163 }, { "epoch": 0.7511152416356878, "grad_norm": 0.6118891497586254, "learning_rate": 1.5503476782195132e-06, "loss": 0.0495, "step": 16164 }, { "epoch": 0.7511617100371747, "grad_norm": 0.6330729305481774, "learning_rate": 1.5498029912154044e-06, "loss": 0.0467, "step": 16165 }, { "epoch": 0.7512081784386617, "grad_norm": 0.7658341778030253, "learning_rate": 1.5492583823618878e-06, "loss": 0.0644, "step": 16166 }, { "epoch": 0.7512546468401488, "grad_norm": 0.31952988757630724, "learning_rate": 1.548713851671299e-06, "loss": 0.056, "step": 16167 }, { "epoch": 0.7513011152416357, "grad_norm": 1.1527573876071397, "learning_rate": 1.5481693991559755e-06, "loss": 0.0603, "step": 16168 }, { "epoch": 0.7513475836431227, "grad_norm": 0.4067644300686703, "learning_rate": 1.5476250248282476e-06, "loss": 0.0523, "step": 16169 }, { "epoch": 0.7513940520446096, "grad_norm": 0.29293087190501826, "learning_rate": 1.5470807287004469e-06, "loss": 0.0441, "step": 16170 }, { "epoch": 0.7514405204460967, "grad_norm": 0.8112438855070855, "learning_rate": 1.5465365107849018e-06, "loss": 0.0594, "step": 16171 }, { "epoch": 0.7514869888475837, "grad_norm": 0.38724436640542015, "learning_rate": 1.545992371093939e-06, "loss": 0.0498, "step": 16172 }, { "epoch": 0.7515334572490706, "grad_norm": 0.5566755025875118, "learning_rate": 1.5454483096398848e-06, "loss": 0.059, "step": 16173 }, { "epoch": 0.7515799256505576, "grad_norm": 0.31972463740061874, "learning_rate": 1.5449043264350605e-06, "loss": 0.0349, "step": 16174 }, { "epoch": 0.7516263940520446, "grad_norm": 0.41343389449046636, "learning_rate": 1.5443604214917912e-06, "loss": 0.06, "step": 16175 }, { "epoch": 0.7516728624535316, "grad_norm": 0.6950661841018532, "learning_rate": 1.5438165948223955e-06, "loss": 0.0642, "step": 16176 }, { "epoch": 0.7517193308550186, "grad_norm": 0.7228355797320372, "learning_rate": 1.5432728464391911e-06, "loss": 0.0756, "step": 16177 }, { "epoch": 0.7517657992565056, "grad_norm": 0.3586300709610097, "learning_rate": 1.5427291763544954e-06, "loss": 0.0344, "step": 16178 }, { "epoch": 0.7518122676579926, "grad_norm": 0.5176877736744141, "learning_rate": 1.542185584580622e-06, "loss": 0.0704, "step": 16179 }, { "epoch": 0.7518587360594795, "grad_norm": 0.8366487868816023, "learning_rate": 1.5416420711298847e-06, "loss": 0.057, "step": 16180 }, { "epoch": 0.7519052044609665, "grad_norm": 0.5717121138035408, "learning_rate": 1.541098636014593e-06, "loss": 0.0627, "step": 16181 }, { "epoch": 0.7519516728624536, "grad_norm": 0.5398613096743721, "learning_rate": 1.5405552792470602e-06, "loss": 0.0727, "step": 16182 }, { "epoch": 0.7519981412639405, "grad_norm": 0.5616870503066183, "learning_rate": 1.5400120008395903e-06, "loss": 0.0608, "step": 16183 }, { "epoch": 0.7520446096654275, "grad_norm": 0.48513437267346243, "learning_rate": 1.5394688008044883e-06, "loss": 0.0524, "step": 16184 }, { "epoch": 0.7520910780669146, "grad_norm": 0.7383488461258887, "learning_rate": 1.5389256791540614e-06, "loss": 0.0601, "step": 16185 }, { "epoch": 0.7521375464684015, "grad_norm": 0.5690686179131121, "learning_rate": 1.538382635900611e-06, "loss": 0.0663, "step": 16186 }, { "epoch": 0.7521840148698885, "grad_norm": 0.6344365133518224, "learning_rate": 1.5378396710564369e-06, "loss": 0.053, "step": 16187 }, { "epoch": 0.7522304832713754, "grad_norm": 0.5447929824534926, "learning_rate": 1.5372967846338388e-06, "loss": 0.0413, "step": 16188 }, { "epoch": 0.7522769516728625, "grad_norm": 0.5525990084865838, "learning_rate": 1.5367539766451123e-06, "loss": 0.0536, "step": 16189 }, { "epoch": 0.7523234200743495, "grad_norm": 0.709464064782684, "learning_rate": 1.536211247102553e-06, "loss": 0.0678, "step": 16190 }, { "epoch": 0.7523698884758364, "grad_norm": 0.5748481697250671, "learning_rate": 1.5356685960184536e-06, "loss": 0.0376, "step": 16191 }, { "epoch": 0.7524163568773234, "grad_norm": 0.8718534555034695, "learning_rate": 1.5351260234051079e-06, "loss": 0.0729, "step": 16192 }, { "epoch": 0.7524628252788104, "grad_norm": 0.4607022453119628, "learning_rate": 1.5345835292748046e-06, "loss": 0.0533, "step": 16193 }, { "epoch": 0.7525092936802974, "grad_norm": 0.4522715844784212, "learning_rate": 1.5340411136398315e-06, "loss": 0.0271, "step": 16194 }, { "epoch": 0.7525557620817844, "grad_norm": 0.6535017859020184, "learning_rate": 1.5334987765124753e-06, "loss": 0.0504, "step": 16195 }, { "epoch": 0.7526022304832714, "grad_norm": 0.4102370169297926, "learning_rate": 1.5329565179050199e-06, "loss": 0.0475, "step": 16196 }, { "epoch": 0.7526486988847584, "grad_norm": 0.4438609140658562, "learning_rate": 1.5324143378297486e-06, "loss": 0.0499, "step": 16197 }, { "epoch": 0.7526951672862453, "grad_norm": 0.3438667946212317, "learning_rate": 1.5318722362989401e-06, "loss": 0.035, "step": 16198 }, { "epoch": 0.7527416356877323, "grad_norm": 0.6481455189734125, "learning_rate": 1.5313302133248774e-06, "loss": 0.0522, "step": 16199 }, { "epoch": 0.7527881040892194, "grad_norm": 0.4775106659393384, "learning_rate": 1.5307882689198384e-06, "loss": 0.0488, "step": 16200 }, { "epoch": 0.7528345724907063, "grad_norm": 0.5403662368977252, "learning_rate": 1.5302464030960924e-06, "loss": 0.0542, "step": 16201 }, { "epoch": 0.7528810408921933, "grad_norm": 0.4432781143092202, "learning_rate": 1.529704615865919e-06, "loss": 0.0574, "step": 16202 }, { "epoch": 0.7529275092936804, "grad_norm": 0.4825918035469579, "learning_rate": 1.5291629072415881e-06, "loss": 0.0335, "step": 16203 }, { "epoch": 0.7529739776951673, "grad_norm": 0.7663305969827412, "learning_rate": 1.5286212772353708e-06, "loss": 0.0607, "step": 16204 }, { "epoch": 0.7530204460966543, "grad_norm": 0.6244736398358474, "learning_rate": 1.5280797258595337e-06, "loss": 0.0734, "step": 16205 }, { "epoch": 0.7530669144981412, "grad_norm": 0.3640735130083483, "learning_rate": 1.5275382531263478e-06, "loss": 0.0496, "step": 16206 }, { "epoch": 0.7531133828996283, "grad_norm": 0.3925218800140544, "learning_rate": 1.5269968590480739e-06, "loss": 0.0326, "step": 16207 }, { "epoch": 0.7531598513011153, "grad_norm": 0.4897675129275846, "learning_rate": 1.5264555436369744e-06, "loss": 0.0638, "step": 16208 }, { "epoch": 0.7532063197026022, "grad_norm": 0.42990896668545897, "learning_rate": 1.5259143069053144e-06, "loss": 0.0426, "step": 16209 }, { "epoch": 0.7532527881040892, "grad_norm": 0.5505571476087627, "learning_rate": 1.5253731488653518e-06, "loss": 0.0733, "step": 16210 }, { "epoch": 0.7532992565055762, "grad_norm": 0.5727470279128207, "learning_rate": 1.5248320695293439e-06, "loss": 0.0538, "step": 16211 }, { "epoch": 0.7533457249070632, "grad_norm": 0.6023596829741018, "learning_rate": 1.524291068909548e-06, "loss": 0.049, "step": 16212 }, { "epoch": 0.7533921933085502, "grad_norm": 0.6139433615746995, "learning_rate": 1.523750147018217e-06, "loss": 0.04, "step": 16213 }, { "epoch": 0.7534386617100371, "grad_norm": 0.573490214144939, "learning_rate": 1.5232093038676038e-06, "loss": 0.0594, "step": 16214 }, { "epoch": 0.7534851301115242, "grad_norm": 0.49875850966361257, "learning_rate": 1.522668539469958e-06, "loss": 0.0409, "step": 16215 }, { "epoch": 0.7535315985130111, "grad_norm": 0.6674061031176636, "learning_rate": 1.5221278538375306e-06, "loss": 0.06, "step": 16216 }, { "epoch": 0.7535780669144981, "grad_norm": 0.3796844613864373, "learning_rate": 1.5215872469825682e-06, "loss": 0.0553, "step": 16217 }, { "epoch": 0.7536245353159852, "grad_norm": 0.78621528347286, "learning_rate": 1.5210467189173155e-06, "loss": 0.0546, "step": 16218 }, { "epoch": 0.7536710037174721, "grad_norm": 0.4342412239112802, "learning_rate": 1.5205062696540162e-06, "loss": 0.0316, "step": 16219 }, { "epoch": 0.7537174721189591, "grad_norm": 0.3775824125037302, "learning_rate": 1.5199658992049116e-06, "loss": 0.037, "step": 16220 }, { "epoch": 0.7537639405204462, "grad_norm": 0.5633596255844452, "learning_rate": 1.5194256075822422e-06, "loss": 0.0477, "step": 16221 }, { "epoch": 0.7538104089219331, "grad_norm": 0.4908563261255269, "learning_rate": 1.5188853947982446e-06, "loss": 0.0331, "step": 16222 }, { "epoch": 0.7538568773234201, "grad_norm": 0.4995578903057257, "learning_rate": 1.5183452608651584e-06, "loss": 0.0379, "step": 16223 }, { "epoch": 0.753903345724907, "grad_norm": 0.5827404844030405, "learning_rate": 1.5178052057952174e-06, "loss": 0.0749, "step": 16224 }, { "epoch": 0.7539498141263941, "grad_norm": 0.5771703497807349, "learning_rate": 1.5172652296006508e-06, "loss": 0.0504, "step": 16225 }, { "epoch": 0.7539962825278811, "grad_norm": 0.5684274655681018, "learning_rate": 1.5167253322936937e-06, "loss": 0.0699, "step": 16226 }, { "epoch": 0.754042750929368, "grad_norm": 0.6218992145326911, "learning_rate": 1.5161855138865738e-06, "loss": 0.0654, "step": 16227 }, { "epoch": 0.754089219330855, "grad_norm": 0.4087824724993943, "learning_rate": 1.5156457743915187e-06, "loss": 0.0502, "step": 16228 }, { "epoch": 0.754135687732342, "grad_norm": 0.3900535992225743, "learning_rate": 1.5151061138207529e-06, "loss": 0.0523, "step": 16229 }, { "epoch": 0.754182156133829, "grad_norm": 0.5687265662677505, "learning_rate": 1.5145665321865037e-06, "loss": 0.0495, "step": 16230 }, { "epoch": 0.754228624535316, "grad_norm": 0.4306347259370799, "learning_rate": 1.51402702950099e-06, "loss": 0.0561, "step": 16231 }, { "epoch": 0.754275092936803, "grad_norm": 0.4361233677747226, "learning_rate": 1.5134876057764313e-06, "loss": 0.0538, "step": 16232 }, { "epoch": 0.75432156133829, "grad_norm": 0.3563210928870965, "learning_rate": 1.512948261025049e-06, "loss": 0.0471, "step": 16233 }, { "epoch": 0.7543680297397769, "grad_norm": 0.35489694669793087, "learning_rate": 1.5124089952590593e-06, "loss": 0.0535, "step": 16234 }, { "epoch": 0.7544144981412639, "grad_norm": 0.47731548293880743, "learning_rate": 1.5118698084906759e-06, "loss": 0.0558, "step": 16235 }, { "epoch": 0.754460966542751, "grad_norm": 0.6593962054788179, "learning_rate": 1.5113307007321126e-06, "loss": 0.04, "step": 16236 }, { "epoch": 0.7545074349442379, "grad_norm": 0.31698308732236297, "learning_rate": 1.510791671995581e-06, "loss": 0.0243, "step": 16237 }, { "epoch": 0.7545539033457249, "grad_norm": 0.40490914429000346, "learning_rate": 1.51025272229329e-06, "loss": 0.0502, "step": 16238 }, { "epoch": 0.754600371747212, "grad_norm": 0.6530542993895911, "learning_rate": 1.5097138516374466e-06, "loss": 0.0418, "step": 16239 }, { "epoch": 0.7546468401486989, "grad_norm": 0.412254971086756, "learning_rate": 1.5091750600402593e-06, "loss": 0.0526, "step": 16240 }, { "epoch": 0.7546933085501859, "grad_norm": 0.681691878555108, "learning_rate": 1.508636347513932e-06, "loss": 0.0495, "step": 16241 }, { "epoch": 0.7547397769516728, "grad_norm": 0.4000416508971757, "learning_rate": 1.5080977140706637e-06, "loss": 0.0329, "step": 16242 }, { "epoch": 0.7547862453531599, "grad_norm": 0.5788308426765159, "learning_rate": 1.5075591597226586e-06, "loss": 0.0616, "step": 16243 }, { "epoch": 0.7548327137546469, "grad_norm": 0.5524971257745903, "learning_rate": 1.507020684482114e-06, "loss": 0.0524, "step": 16244 }, { "epoch": 0.7548791821561338, "grad_norm": 0.421656277998696, "learning_rate": 1.5064822883612274e-06, "loss": 0.0452, "step": 16245 }, { "epoch": 0.7549256505576208, "grad_norm": 0.7276982582161645, "learning_rate": 1.5059439713721924e-06, "loss": 0.0384, "step": 16246 }, { "epoch": 0.7549721189591078, "grad_norm": 0.7490135036057274, "learning_rate": 1.5054057335272066e-06, "loss": 0.0626, "step": 16247 }, { "epoch": 0.7550185873605948, "grad_norm": 2.0718195002549953, "learning_rate": 1.5048675748384579e-06, "loss": 0.0459, "step": 16248 }, { "epoch": 0.7550650557620818, "grad_norm": 0.31275105085920984, "learning_rate": 1.5043294953181349e-06, "loss": 0.0357, "step": 16249 }, { "epoch": 0.7551115241635687, "grad_norm": 0.474693434016671, "learning_rate": 1.50379149497843e-06, "loss": 0.041, "step": 16250 }, { "epoch": 0.7551579925650558, "grad_norm": 0.653875442114993, "learning_rate": 1.5032535738315273e-06, "loss": 0.062, "step": 16251 }, { "epoch": 0.7552044609665427, "grad_norm": 0.5281806211664617, "learning_rate": 1.5027157318896112e-06, "loss": 0.048, "step": 16252 }, { "epoch": 0.7552509293680297, "grad_norm": 0.7918655859344622, "learning_rate": 1.5021779691648652e-06, "loss": 0.0697, "step": 16253 }, { "epoch": 0.7552973977695168, "grad_norm": 0.5465926022119042, "learning_rate": 1.5016402856694695e-06, "loss": 0.0486, "step": 16254 }, { "epoch": 0.7553438661710037, "grad_norm": 0.4397329303412333, "learning_rate": 1.5011026814156026e-06, "loss": 0.0444, "step": 16255 }, { "epoch": 0.7553903345724907, "grad_norm": 0.5217020627472885, "learning_rate": 1.5005651564154416e-06, "loss": 0.0545, "step": 16256 }, { "epoch": 0.7554368029739778, "grad_norm": 0.5484029687065793, "learning_rate": 1.5000277106811644e-06, "loss": 0.0641, "step": 16257 }, { "epoch": 0.7554832713754647, "grad_norm": 0.5642604203581896, "learning_rate": 1.499490344224943e-06, "loss": 0.0501, "step": 16258 }, { "epoch": 0.7555297397769517, "grad_norm": 0.726294570679349, "learning_rate": 1.4989530570589495e-06, "loss": 0.0395, "step": 16259 }, { "epoch": 0.7555762081784386, "grad_norm": 0.6022398852137911, "learning_rate": 1.4984158491953543e-06, "loss": 0.0643, "step": 16260 }, { "epoch": 0.7556226765799257, "grad_norm": 0.5458104797575662, "learning_rate": 1.4978787206463253e-06, "loss": 0.0597, "step": 16261 }, { "epoch": 0.7556691449814127, "grad_norm": 0.483796380148216, "learning_rate": 1.497341671424029e-06, "loss": 0.0611, "step": 16262 }, { "epoch": 0.7557156133828996, "grad_norm": 0.5960949205932664, "learning_rate": 1.496804701540629e-06, "loss": 0.0643, "step": 16263 }, { "epoch": 0.7557620817843866, "grad_norm": 0.4339079101872252, "learning_rate": 1.4962678110082907e-06, "loss": 0.0635, "step": 16264 }, { "epoch": 0.7558085501858736, "grad_norm": 0.6379953665149865, "learning_rate": 1.4957309998391761e-06, "loss": 0.0536, "step": 16265 }, { "epoch": 0.7558550185873606, "grad_norm": 0.41466580986257867, "learning_rate": 1.4951942680454396e-06, "loss": 0.0625, "step": 16266 }, { "epoch": 0.7559014869888476, "grad_norm": 0.476946825118332, "learning_rate": 1.4946576156392428e-06, "loss": 0.057, "step": 16267 }, { "epoch": 0.7559479553903345, "grad_norm": 0.6069362202495612, "learning_rate": 1.49412104263274e-06, "loss": 0.0408, "step": 16268 }, { "epoch": 0.7559944237918216, "grad_norm": 0.41481064901136583, "learning_rate": 1.4935845490380864e-06, "loss": 0.0403, "step": 16269 }, { "epoch": 0.7560408921933085, "grad_norm": 0.5360334598320414, "learning_rate": 1.4930481348674308e-06, "loss": 0.0341, "step": 16270 }, { "epoch": 0.7560873605947955, "grad_norm": 0.4445955313630442, "learning_rate": 1.4925118001329297e-06, "loss": 0.052, "step": 16271 }, { "epoch": 0.7561338289962826, "grad_norm": 0.4278908982376236, "learning_rate": 1.491975544846726e-06, "loss": 0.0357, "step": 16272 }, { "epoch": 0.7561802973977695, "grad_norm": 0.5596297458910503, "learning_rate": 1.4914393690209666e-06, "loss": 0.06, "step": 16273 }, { "epoch": 0.7562267657992565, "grad_norm": 0.7719273082915423, "learning_rate": 1.4909032726678001e-06, "loss": 0.0742, "step": 16274 }, { "epoch": 0.7562732342007435, "grad_norm": 0.6400615567446186, "learning_rate": 1.4903672557993675e-06, "loss": 0.0542, "step": 16275 }, { "epoch": 0.7563197026022305, "grad_norm": 0.5516273075609337, "learning_rate": 1.4898313184278107e-06, "loss": 0.059, "step": 16276 }, { "epoch": 0.7563661710037175, "grad_norm": 0.8649410732096998, "learning_rate": 1.4892954605652688e-06, "loss": 0.0441, "step": 16277 }, { "epoch": 0.7564126394052044, "grad_norm": 0.5728982315934317, "learning_rate": 1.4887596822238792e-06, "loss": 0.0561, "step": 16278 }, { "epoch": 0.7564591078066915, "grad_norm": 0.74612157198599, "learning_rate": 1.4882239834157791e-06, "loss": 0.0766, "step": 16279 }, { "epoch": 0.7565055762081785, "grad_norm": 0.5407094946581735, "learning_rate": 1.4876883641530998e-06, "loss": 0.041, "step": 16280 }, { "epoch": 0.7565520446096654, "grad_norm": 0.42341996263075726, "learning_rate": 1.4871528244479772e-06, "loss": 0.0615, "step": 16281 }, { "epoch": 0.7565985130111524, "grad_norm": 0.7937945284870395, "learning_rate": 1.4866173643125403e-06, "loss": 0.0657, "step": 16282 }, { "epoch": 0.7566449814126394, "grad_norm": 0.4508690140907932, "learning_rate": 1.486081983758918e-06, "loss": 0.0528, "step": 16283 }, { "epoch": 0.7566914498141264, "grad_norm": 0.4848942326484842, "learning_rate": 1.4855466827992371e-06, "loss": 0.0366, "step": 16284 }, { "epoch": 0.7567379182156134, "grad_norm": 0.34600594891083525, "learning_rate": 1.485011461445623e-06, "loss": 0.0378, "step": 16285 }, { "epoch": 0.7567843866171003, "grad_norm": 0.710709614867803, "learning_rate": 1.4844763197101985e-06, "loss": 0.0536, "step": 16286 }, { "epoch": 0.7568308550185874, "grad_norm": 0.5450066608740994, "learning_rate": 1.4839412576050839e-06, "loss": 0.062, "step": 16287 }, { "epoch": 0.7568773234200743, "grad_norm": 0.5510656127138543, "learning_rate": 1.4834062751424018e-06, "loss": 0.0507, "step": 16288 }, { "epoch": 0.7569237918215613, "grad_norm": 0.7916249272588068, "learning_rate": 1.4828713723342703e-06, "loss": 0.0681, "step": 16289 }, { "epoch": 0.7569702602230484, "grad_norm": 0.6065384727931006, "learning_rate": 1.4823365491928011e-06, "loss": 0.0546, "step": 16290 }, { "epoch": 0.7570167286245353, "grad_norm": 0.5419772946688813, "learning_rate": 1.4818018057301126e-06, "loss": 0.0547, "step": 16291 }, { "epoch": 0.7570631970260223, "grad_norm": 0.4496046714323128, "learning_rate": 1.4812671419583157e-06, "loss": 0.0576, "step": 16292 }, { "epoch": 0.7571096654275093, "grad_norm": 0.5921562733551352, "learning_rate": 1.4807325578895221e-06, "loss": 0.0488, "step": 16293 }, { "epoch": 0.7571561338289963, "grad_norm": 1.1714470476568808, "learning_rate": 1.4801980535358384e-06, "loss": 0.0609, "step": 16294 }, { "epoch": 0.7572026022304833, "grad_norm": 0.6907461050621005, "learning_rate": 1.4796636289093758e-06, "loss": 0.0551, "step": 16295 }, { "epoch": 0.7572490706319702, "grad_norm": 0.6573079656908896, "learning_rate": 1.479129284022236e-06, "loss": 0.0614, "step": 16296 }, { "epoch": 0.7572955390334573, "grad_norm": 0.6537174638054963, "learning_rate": 1.4785950188865223e-06, "loss": 0.0734, "step": 16297 }, { "epoch": 0.7573420074349443, "grad_norm": 0.4647262159745089, "learning_rate": 1.4780608335143387e-06, "loss": 0.0499, "step": 16298 }, { "epoch": 0.7573884758364312, "grad_norm": 0.6634391472507901, "learning_rate": 1.477526727917784e-06, "loss": 0.0542, "step": 16299 }, { "epoch": 0.7574349442379182, "grad_norm": 0.38968265302819066, "learning_rate": 1.4769927021089564e-06, "loss": 0.0399, "step": 16300 }, { "epoch": 0.7574814126394052, "grad_norm": 0.6448704446574655, "learning_rate": 1.476458756099952e-06, "loss": 0.0626, "step": 16301 }, { "epoch": 0.7575278810408922, "grad_norm": 0.48230200631014675, "learning_rate": 1.4759248899028654e-06, "loss": 0.0504, "step": 16302 }, { "epoch": 0.7575743494423792, "grad_norm": 0.6346294076698358, "learning_rate": 1.4753911035297892e-06, "loss": 0.0542, "step": 16303 }, { "epoch": 0.7576208178438661, "grad_norm": 0.7454211157634092, "learning_rate": 1.474857396992812e-06, "loss": 0.0578, "step": 16304 }, { "epoch": 0.7576672862453532, "grad_norm": 0.5566733999166406, "learning_rate": 1.4743237703040264e-06, "loss": 0.0652, "step": 16305 }, { "epoch": 0.7577137546468401, "grad_norm": 0.6155666815273838, "learning_rate": 1.4737902234755203e-06, "loss": 0.0573, "step": 16306 }, { "epoch": 0.7577602230483271, "grad_norm": 0.735568616072671, "learning_rate": 1.4732567565193735e-06, "loss": 0.0583, "step": 16307 }, { "epoch": 0.7578066914498142, "grad_norm": 0.9218509568205779, "learning_rate": 1.4727233694476746e-06, "loss": 0.0633, "step": 16308 }, { "epoch": 0.7578531598513011, "grad_norm": 0.46711239513292196, "learning_rate": 1.4721900622725032e-06, "loss": 0.0521, "step": 16309 }, { "epoch": 0.7578996282527881, "grad_norm": 0.4857143539465759, "learning_rate": 1.471656835005939e-06, "loss": 0.0426, "step": 16310 }, { "epoch": 0.7579460966542751, "grad_norm": 0.6174784876139761, "learning_rate": 1.4711236876600621e-06, "loss": 0.0555, "step": 16311 }, { "epoch": 0.7579925650557621, "grad_norm": 0.5706487492216039, "learning_rate": 1.47059062024695e-06, "loss": 0.0501, "step": 16312 }, { "epoch": 0.7580390334572491, "grad_norm": 0.5924321435056944, "learning_rate": 1.4700576327786725e-06, "loss": 0.063, "step": 16313 }, { "epoch": 0.758085501858736, "grad_norm": 0.558993598338958, "learning_rate": 1.469524725267304e-06, "loss": 0.0626, "step": 16314 }, { "epoch": 0.758131970260223, "grad_norm": 0.3587572730844645, "learning_rate": 1.4689918977249175e-06, "loss": 0.0441, "step": 16315 }, { "epoch": 0.7581784386617101, "grad_norm": 0.760690632932352, "learning_rate": 1.468459150163581e-06, "loss": 0.0607, "step": 16316 }, { "epoch": 0.758224907063197, "grad_norm": 0.6860271105467616, "learning_rate": 1.4679264825953599e-06, "loss": 0.0427, "step": 16317 }, { "epoch": 0.758271375464684, "grad_norm": 0.4883603201560077, "learning_rate": 1.4673938950323242e-06, "loss": 0.0643, "step": 16318 }, { "epoch": 0.758317843866171, "grad_norm": 0.32649927048468375, "learning_rate": 1.4668613874865335e-06, "loss": 0.0367, "step": 16319 }, { "epoch": 0.758364312267658, "grad_norm": 0.42565451692153244, "learning_rate": 1.4663289599700509e-06, "loss": 0.0761, "step": 16320 }, { "epoch": 0.758410780669145, "grad_norm": 0.48982645270982705, "learning_rate": 1.4657966124949351e-06, "loss": 0.059, "step": 16321 }, { "epoch": 0.758457249070632, "grad_norm": 0.45617565055179726, "learning_rate": 1.4652643450732473e-06, "loss": 0.0603, "step": 16322 }, { "epoch": 0.758503717472119, "grad_norm": 0.5763961655578413, "learning_rate": 1.464732157717042e-06, "loss": 0.0571, "step": 16323 }, { "epoch": 0.7585501858736059, "grad_norm": 0.7264678215445664, "learning_rate": 1.4642000504383746e-06, "loss": 0.0614, "step": 16324 }, { "epoch": 0.7585966542750929, "grad_norm": 0.318703552148421, "learning_rate": 1.4636680232492973e-06, "loss": 0.0524, "step": 16325 }, { "epoch": 0.75864312267658, "grad_norm": 0.5742895456175124, "learning_rate": 1.4631360761618606e-06, "loss": 0.0608, "step": 16326 }, { "epoch": 0.7586895910780669, "grad_norm": 0.3470492660243618, "learning_rate": 1.4626042091881132e-06, "loss": 0.0373, "step": 16327 }, { "epoch": 0.7587360594795539, "grad_norm": 0.5645289232381652, "learning_rate": 1.462072422340105e-06, "loss": 0.0615, "step": 16328 }, { "epoch": 0.758782527881041, "grad_norm": 0.7522286813497904, "learning_rate": 1.46154071562988e-06, "loss": 0.0467, "step": 16329 }, { "epoch": 0.7588289962825279, "grad_norm": 0.46441845334912935, "learning_rate": 1.461009089069484e-06, "loss": 0.0402, "step": 16330 }, { "epoch": 0.7588754646840149, "grad_norm": 0.6320979002262739, "learning_rate": 1.4604775426709527e-06, "loss": 0.063, "step": 16331 }, { "epoch": 0.7589219330855018, "grad_norm": 0.8704596083597044, "learning_rate": 1.459946076446333e-06, "loss": 0.0714, "step": 16332 }, { "epoch": 0.7589684014869889, "grad_norm": 0.4980848098270974, "learning_rate": 1.45941469040766e-06, "loss": 0.0741, "step": 16333 }, { "epoch": 0.7590148698884759, "grad_norm": 0.38251898611327584, "learning_rate": 1.4588833845669687e-06, "loss": 0.0229, "step": 16334 }, { "epoch": 0.7590613382899628, "grad_norm": 0.31395867399776795, "learning_rate": 1.4583521589362975e-06, "loss": 0.0441, "step": 16335 }, { "epoch": 0.7591078066914498, "grad_norm": 0.33993228662210495, "learning_rate": 1.4578210135276794e-06, "loss": 0.034, "step": 16336 }, { "epoch": 0.7591542750929368, "grad_norm": 0.507157322788995, "learning_rate": 1.4572899483531416e-06, "loss": 0.037, "step": 16337 }, { "epoch": 0.7592007434944238, "grad_norm": 0.6013962720612961, "learning_rate": 1.4567589634247136e-06, "loss": 0.0417, "step": 16338 }, { "epoch": 0.7592472118959108, "grad_norm": 0.4723453910516765, "learning_rate": 1.4562280587544258e-06, "loss": 0.0601, "step": 16339 }, { "epoch": 0.7592936802973977, "grad_norm": 0.4299302765907133, "learning_rate": 1.4556972343543018e-06, "loss": 0.0476, "step": 16340 }, { "epoch": 0.7593401486988848, "grad_norm": 0.43806805365952345, "learning_rate": 1.455166490236365e-06, "loss": 0.0427, "step": 16341 }, { "epoch": 0.7593866171003717, "grad_norm": 0.5092225010340196, "learning_rate": 1.454635826412641e-06, "loss": 0.0503, "step": 16342 }, { "epoch": 0.7594330855018587, "grad_norm": 0.5987375384724806, "learning_rate": 1.4541052428951446e-06, "loss": 0.0508, "step": 16343 }, { "epoch": 0.7594795539033458, "grad_norm": 0.5074213943557746, "learning_rate": 1.4535747396958955e-06, "loss": 0.0535, "step": 16344 }, { "epoch": 0.7595260223048327, "grad_norm": 0.38149810804241346, "learning_rate": 1.4530443168269127e-06, "loss": 0.0652, "step": 16345 }, { "epoch": 0.7595724907063197, "grad_norm": 0.8751392588642541, "learning_rate": 1.4525139743002092e-06, "loss": 0.0671, "step": 16346 }, { "epoch": 0.7596189591078067, "grad_norm": 0.6215493521663789, "learning_rate": 1.4519837121277974e-06, "loss": 0.0485, "step": 16347 }, { "epoch": 0.7596654275092937, "grad_norm": 0.3233493977420448, "learning_rate": 1.4514535303216893e-06, "loss": 0.0353, "step": 16348 }, { "epoch": 0.7597118959107807, "grad_norm": 1.3920252336449779, "learning_rate": 1.4509234288938934e-06, "loss": 0.0621, "step": 16349 }, { "epoch": 0.7597583643122676, "grad_norm": 0.4265998612882598, "learning_rate": 1.4503934078564168e-06, "loss": 0.0367, "step": 16350 }, { "epoch": 0.7598048327137547, "grad_norm": 0.32337593994585834, "learning_rate": 1.449863467221264e-06, "loss": 0.0481, "step": 16351 }, { "epoch": 0.7598513011152417, "grad_norm": 0.32054988707440746, "learning_rate": 1.449333607000442e-06, "loss": 0.0386, "step": 16352 }, { "epoch": 0.7598977695167286, "grad_norm": 0.885449945140759, "learning_rate": 1.4488038272059523e-06, "loss": 0.0358, "step": 16353 }, { "epoch": 0.7599442379182156, "grad_norm": 0.4800687244523047, "learning_rate": 1.4482741278497903e-06, "loss": 0.0524, "step": 16354 }, { "epoch": 0.7599907063197026, "grad_norm": 0.5557508348932099, "learning_rate": 1.447744508943959e-06, "loss": 0.0529, "step": 16355 }, { "epoch": 0.7600371747211896, "grad_norm": 0.6389446574949877, "learning_rate": 1.4472149705004529e-06, "loss": 0.0483, "step": 16356 }, { "epoch": 0.7600836431226766, "grad_norm": 0.5659226966798744, "learning_rate": 1.4466855125312673e-06, "loss": 0.0621, "step": 16357 }, { "epoch": 0.7601301115241635, "grad_norm": 0.7349328226274439, "learning_rate": 1.4461561350483931e-06, "loss": 0.053, "step": 16358 }, { "epoch": 0.7601765799256506, "grad_norm": 0.42604255360025284, "learning_rate": 1.4456268380638262e-06, "loss": 0.0498, "step": 16359 }, { "epoch": 0.7602230483271375, "grad_norm": 0.41012698906180145, "learning_rate": 1.44509762158955e-06, "loss": 0.052, "step": 16360 }, { "epoch": 0.7602695167286245, "grad_norm": 0.28525816482906385, "learning_rate": 1.444568485637553e-06, "loss": 0.0346, "step": 16361 }, { "epoch": 0.7603159851301116, "grad_norm": 0.5644390213857892, "learning_rate": 1.4440394302198234e-06, "loss": 0.0527, "step": 16362 }, { "epoch": 0.7603624535315985, "grad_norm": 0.4792418793943877, "learning_rate": 1.443510455348343e-06, "loss": 0.0525, "step": 16363 }, { "epoch": 0.7604089219330855, "grad_norm": 0.47272420931518466, "learning_rate": 1.4429815610350944e-06, "loss": 0.051, "step": 16364 }, { "epoch": 0.7604553903345725, "grad_norm": 0.6691240207473714, "learning_rate": 1.442452747292057e-06, "loss": 0.0633, "step": 16365 }, { "epoch": 0.7605018587360595, "grad_norm": 0.900035319011052, "learning_rate": 1.441924014131209e-06, "loss": 0.0591, "step": 16366 }, { "epoch": 0.7605483271375465, "grad_norm": 0.5533308247345851, "learning_rate": 1.441395361564527e-06, "loss": 0.0439, "step": 16367 }, { "epoch": 0.7605947955390334, "grad_norm": 0.37915371015140725, "learning_rate": 1.4408667896039836e-06, "loss": 0.0415, "step": 16368 }, { "epoch": 0.7606412639405205, "grad_norm": 0.4387917203002004, "learning_rate": 1.4403382982615549e-06, "loss": 0.0479, "step": 16369 }, { "epoch": 0.7606877323420075, "grad_norm": 0.6542053290834321, "learning_rate": 1.4398098875492106e-06, "loss": 0.0537, "step": 16370 }, { "epoch": 0.7607342007434944, "grad_norm": 1.4192223578272973, "learning_rate": 1.4392815574789192e-06, "loss": 0.0501, "step": 16371 }, { "epoch": 0.7607806691449814, "grad_norm": 0.6774113359948783, "learning_rate": 1.4387533080626482e-06, "loss": 0.0501, "step": 16372 }, { "epoch": 0.7608271375464684, "grad_norm": 0.38358811983997776, "learning_rate": 1.4382251393123625e-06, "loss": 0.0565, "step": 16373 }, { "epoch": 0.7608736059479554, "grad_norm": 0.8740957363065782, "learning_rate": 1.4376970512400263e-06, "loss": 0.0485, "step": 16374 }, { "epoch": 0.7609200743494424, "grad_norm": 0.48449551599661383, "learning_rate": 1.4371690438575997e-06, "loss": 0.0556, "step": 16375 }, { "epoch": 0.7609665427509293, "grad_norm": 0.510561517773049, "learning_rate": 1.4366411171770457e-06, "loss": 0.049, "step": 16376 }, { "epoch": 0.7610130111524164, "grad_norm": 0.6906840045976812, "learning_rate": 1.4361132712103222e-06, "loss": 0.0524, "step": 16377 }, { "epoch": 0.7610594795539033, "grad_norm": 0.7695459023625196, "learning_rate": 1.4355855059693813e-06, "loss": 0.0686, "step": 16378 }, { "epoch": 0.7611059479553903, "grad_norm": 0.5472835976462042, "learning_rate": 1.435057821466181e-06, "loss": 0.0552, "step": 16379 }, { "epoch": 0.7611524163568774, "grad_norm": 0.5569671459687837, "learning_rate": 1.4345302177126736e-06, "loss": 0.0611, "step": 16380 }, { "epoch": 0.7611988847583643, "grad_norm": 0.5990810199625998, "learning_rate": 1.4340026947208091e-06, "loss": 0.057, "step": 16381 }, { "epoch": 0.7612453531598513, "grad_norm": 0.4829170808307717, "learning_rate": 1.4334752525025353e-06, "loss": 0.0605, "step": 16382 }, { "epoch": 0.7612918215613383, "grad_norm": 0.37131025275921203, "learning_rate": 1.4329478910698036e-06, "loss": 0.0426, "step": 16383 }, { "epoch": 0.7613382899628253, "grad_norm": 0.5814682052468159, "learning_rate": 1.432420610434555e-06, "loss": 0.0668, "step": 16384 }, { "epoch": 0.7613847583643123, "grad_norm": 0.6466974514246439, "learning_rate": 1.4318934106087334e-06, "loss": 0.0538, "step": 16385 }, { "epoch": 0.7614312267657992, "grad_norm": 0.5798532388461309, "learning_rate": 1.431366291604282e-06, "loss": 0.0453, "step": 16386 }, { "epoch": 0.7614776951672863, "grad_norm": 0.4798894970566218, "learning_rate": 1.4308392534331406e-06, "loss": 0.0438, "step": 16387 }, { "epoch": 0.7615241635687733, "grad_norm": 0.6101776776304559, "learning_rate": 1.4303122961072462e-06, "loss": 0.0648, "step": 16388 }, { "epoch": 0.7615706319702602, "grad_norm": 0.433762503635393, "learning_rate": 1.4297854196385358e-06, "loss": 0.0407, "step": 16389 }, { "epoch": 0.7616171003717472, "grad_norm": 0.7256441402630693, "learning_rate": 1.4292586240389433e-06, "loss": 0.055, "step": 16390 }, { "epoch": 0.7616635687732342, "grad_norm": 0.9512691626783463, "learning_rate": 1.4287319093204006e-06, "loss": 0.0741, "step": 16391 }, { "epoch": 0.7617100371747212, "grad_norm": 0.49791750859424944, "learning_rate": 1.4282052754948378e-06, "loss": 0.0576, "step": 16392 }, { "epoch": 0.7617565055762082, "grad_norm": 0.6343404675366549, "learning_rate": 1.4276787225741856e-06, "loss": 0.0618, "step": 16393 }, { "epoch": 0.7618029739776951, "grad_norm": 0.6245413888516782, "learning_rate": 1.4271522505703705e-06, "loss": 0.0439, "step": 16394 }, { "epoch": 0.7618494423791822, "grad_norm": 0.3928989315018443, "learning_rate": 1.4266258594953175e-06, "loss": 0.024, "step": 16395 }, { "epoch": 0.7618959107806691, "grad_norm": 0.4975293577569245, "learning_rate": 1.4260995493609497e-06, "loss": 0.0472, "step": 16396 }, { "epoch": 0.7619423791821561, "grad_norm": 0.3127924243799268, "learning_rate": 1.4255733201791883e-06, "loss": 0.0267, "step": 16397 }, { "epoch": 0.7619888475836432, "grad_norm": 0.6411559991661225, "learning_rate": 1.425047171961953e-06, "loss": 0.0502, "step": 16398 }, { "epoch": 0.7620353159851301, "grad_norm": 0.834797079709841, "learning_rate": 1.424521104721161e-06, "loss": 0.0552, "step": 16399 }, { "epoch": 0.7620817843866171, "grad_norm": 0.5227732284232882, "learning_rate": 1.4239951184687302e-06, "loss": 0.0511, "step": 16400 }, { "epoch": 0.7621282527881041, "grad_norm": 0.6868557413962685, "learning_rate": 1.4234692132165751e-06, "loss": 0.0406, "step": 16401 }, { "epoch": 0.7621747211895911, "grad_norm": 0.45805457027735197, "learning_rate": 1.4229433889766036e-06, "loss": 0.0578, "step": 16402 }, { "epoch": 0.7622211895910781, "grad_norm": 0.38544563732661635, "learning_rate": 1.4224176457607303e-06, "loss": 0.0419, "step": 16403 }, { "epoch": 0.762267657992565, "grad_norm": 0.5977325156937487, "learning_rate": 1.4218919835808631e-06, "loss": 0.0696, "step": 16404 }, { "epoch": 0.762314126394052, "grad_norm": 0.48802916965448506, "learning_rate": 1.4213664024489081e-06, "loss": 0.057, "step": 16405 }, { "epoch": 0.7623605947955391, "grad_norm": 0.44893401257579707, "learning_rate": 1.4208409023767695e-06, "loss": 0.0508, "step": 16406 }, { "epoch": 0.762407063197026, "grad_norm": 0.3100475448421014, "learning_rate": 1.4203154833763538e-06, "loss": 0.032, "step": 16407 }, { "epoch": 0.762453531598513, "grad_norm": 0.5970507137392858, "learning_rate": 1.419790145459559e-06, "loss": 0.0473, "step": 16408 }, { "epoch": 0.7625, "grad_norm": 0.4279991417949599, "learning_rate": 1.4192648886382838e-06, "loss": 0.0431, "step": 16409 }, { "epoch": 0.762546468401487, "grad_norm": 0.41641508558998414, "learning_rate": 1.418739712924429e-06, "loss": 0.0503, "step": 16410 }, { "epoch": 0.762592936802974, "grad_norm": 0.6182457281841947, "learning_rate": 1.418214618329889e-06, "loss": 0.0489, "step": 16411 }, { "epoch": 0.7626394052044609, "grad_norm": 0.41258618129911684, "learning_rate": 1.4176896048665579e-06, "loss": 0.0421, "step": 16412 }, { "epoch": 0.762685873605948, "grad_norm": 0.41945631540313777, "learning_rate": 1.4171646725463266e-06, "loss": 0.0457, "step": 16413 }, { "epoch": 0.7627323420074349, "grad_norm": 0.5814081041909871, "learning_rate": 1.4166398213810872e-06, "loss": 0.0738, "step": 16414 }, { "epoch": 0.7627788104089219, "grad_norm": 0.5145778855225923, "learning_rate": 1.4161150513827271e-06, "loss": 0.0562, "step": 16415 }, { "epoch": 0.762825278810409, "grad_norm": 0.6197715301671538, "learning_rate": 1.4155903625631317e-06, "loss": 0.0507, "step": 16416 }, { "epoch": 0.7628717472118959, "grad_norm": 0.500107175716413, "learning_rate": 1.4150657549341885e-06, "loss": 0.0539, "step": 16417 }, { "epoch": 0.7629182156133829, "grad_norm": 0.5400493122948881, "learning_rate": 1.4145412285077804e-06, "loss": 0.048, "step": 16418 }, { "epoch": 0.7629646840148699, "grad_norm": 0.40529884914856046, "learning_rate": 1.414016783295784e-06, "loss": 0.0359, "step": 16419 }, { "epoch": 0.7630111524163569, "grad_norm": 0.5369482307093076, "learning_rate": 1.4134924193100835e-06, "loss": 0.0599, "step": 16420 }, { "epoch": 0.7630576208178439, "grad_norm": 0.5637826143714632, "learning_rate": 1.412968136562554e-06, "loss": 0.0318, "step": 16421 }, { "epoch": 0.7631040892193308, "grad_norm": 0.49857911338929783, "learning_rate": 1.4124439350650714e-06, "loss": 0.0462, "step": 16422 }, { "epoch": 0.7631505576208178, "grad_norm": 0.4259232121524556, "learning_rate": 1.4119198148295088e-06, "loss": 0.0486, "step": 16423 }, { "epoch": 0.7631970260223049, "grad_norm": 0.3825784395590442, "learning_rate": 1.411395775867741e-06, "loss": 0.0373, "step": 16424 }, { "epoch": 0.7632434944237918, "grad_norm": 0.6324372978735091, "learning_rate": 1.4108718181916348e-06, "loss": 0.0518, "step": 16425 }, { "epoch": 0.7632899628252788, "grad_norm": 0.366084434576406, "learning_rate": 1.4103479418130578e-06, "loss": 0.0402, "step": 16426 }, { "epoch": 0.7633364312267658, "grad_norm": 0.48126170466803664, "learning_rate": 1.4098241467438795e-06, "loss": 0.0634, "step": 16427 }, { "epoch": 0.7633828996282528, "grad_norm": 0.6088258876566068, "learning_rate": 1.4093004329959626e-06, "loss": 0.0516, "step": 16428 }, { "epoch": 0.7634293680297398, "grad_norm": 0.43662735912324413, "learning_rate": 1.4087768005811704e-06, "loss": 0.0478, "step": 16429 }, { "epoch": 0.7634758364312267, "grad_norm": 0.46276804889919737, "learning_rate": 1.4082532495113627e-06, "loss": 0.0458, "step": 16430 }, { "epoch": 0.7635223048327138, "grad_norm": 0.47359996425231204, "learning_rate": 1.4077297797983997e-06, "loss": 0.037, "step": 16431 }, { "epoch": 0.7635687732342007, "grad_norm": 0.32293930184421893, "learning_rate": 1.4072063914541378e-06, "loss": 0.0409, "step": 16432 }, { "epoch": 0.7636152416356877, "grad_norm": 0.5485885738983702, "learning_rate": 1.406683084490431e-06, "loss": 0.0535, "step": 16433 }, { "epoch": 0.7636617100371748, "grad_norm": 0.6142144628943459, "learning_rate": 1.4061598589191356e-06, "loss": 0.0604, "step": 16434 }, { "epoch": 0.7637081784386617, "grad_norm": 0.4325062563402752, "learning_rate": 1.405636714752102e-06, "loss": 0.0596, "step": 16435 }, { "epoch": 0.7637546468401487, "grad_norm": 0.41620319585886917, "learning_rate": 1.4051136520011798e-06, "loss": 0.0581, "step": 16436 }, { "epoch": 0.7638011152416357, "grad_norm": 0.6450081737901608, "learning_rate": 1.404590670678217e-06, "loss": 0.0501, "step": 16437 }, { "epoch": 0.7638475836431227, "grad_norm": 0.4169655049982293, "learning_rate": 1.4040677707950594e-06, "loss": 0.0426, "step": 16438 }, { "epoch": 0.7638940520446097, "grad_norm": 0.8133175990609854, "learning_rate": 1.4035449523635518e-06, "loss": 0.0573, "step": 16439 }, { "epoch": 0.7639405204460966, "grad_norm": 0.567335121012666, "learning_rate": 1.4030222153955347e-06, "loss": 0.0609, "step": 16440 }, { "epoch": 0.7639869888475836, "grad_norm": 0.5904531967544735, "learning_rate": 1.4024995599028513e-06, "loss": 0.0605, "step": 16441 }, { "epoch": 0.7640334572490707, "grad_norm": 0.5298837217600453, "learning_rate": 1.401976985897341e-06, "loss": 0.0623, "step": 16442 }, { "epoch": 0.7640799256505576, "grad_norm": 0.35004449009973104, "learning_rate": 1.4014544933908359e-06, "loss": 0.0584, "step": 16443 }, { "epoch": 0.7641263940520446, "grad_norm": 0.35644946384272086, "learning_rate": 1.4009320823951749e-06, "loss": 0.0509, "step": 16444 }, { "epoch": 0.7641728624535316, "grad_norm": 0.459255318513147, "learning_rate": 1.4004097529221906e-06, "loss": 0.0609, "step": 16445 }, { "epoch": 0.7642193308550186, "grad_norm": 0.685364143163228, "learning_rate": 1.3998875049837141e-06, "loss": 0.0374, "step": 16446 }, { "epoch": 0.7642657992565056, "grad_norm": 0.4551702070313161, "learning_rate": 1.3993653385915728e-06, "loss": 0.0493, "step": 16447 }, { "epoch": 0.7643122676579925, "grad_norm": 0.8657066589481288, "learning_rate": 1.3988432537575991e-06, "loss": 0.0547, "step": 16448 }, { "epoch": 0.7643587360594796, "grad_norm": 1.1776311281312941, "learning_rate": 1.3983212504936144e-06, "loss": 0.0796, "step": 16449 }, { "epoch": 0.7644052044609665, "grad_norm": 0.5879300449771748, "learning_rate": 1.3977993288114428e-06, "loss": 0.0549, "step": 16450 }, { "epoch": 0.7644516728624535, "grad_norm": 0.657002296456761, "learning_rate": 1.397277488722909e-06, "loss": 0.0492, "step": 16451 }, { "epoch": 0.7644981412639406, "grad_norm": 0.48925794648845083, "learning_rate": 1.3967557302398321e-06, "loss": 0.0658, "step": 16452 }, { "epoch": 0.7645446096654275, "grad_norm": 0.7230990601140636, "learning_rate": 1.3962340533740299e-06, "loss": 0.0535, "step": 16453 }, { "epoch": 0.7645910780669145, "grad_norm": 0.7085850772013934, "learning_rate": 1.3957124581373194e-06, "loss": 0.0531, "step": 16454 }, { "epoch": 0.7646375464684015, "grad_norm": 0.39669953433229, "learning_rate": 1.3951909445415151e-06, "loss": 0.0493, "step": 16455 }, { "epoch": 0.7646840148698885, "grad_norm": 0.5213095268307126, "learning_rate": 1.39466951259843e-06, "loss": 0.0499, "step": 16456 }, { "epoch": 0.7647304832713755, "grad_norm": 0.6211980925815754, "learning_rate": 1.3941481623198739e-06, "loss": 0.0346, "step": 16457 }, { "epoch": 0.7647769516728624, "grad_norm": 0.42107797756460635, "learning_rate": 1.393626893717658e-06, "loss": 0.0579, "step": 16458 }, { "epoch": 0.7648234200743494, "grad_norm": 0.6177542678709497, "learning_rate": 1.3931057068035886e-06, "loss": 0.064, "step": 16459 }, { "epoch": 0.7648698884758365, "grad_norm": 0.5096839457611321, "learning_rate": 1.3925846015894711e-06, "loss": 0.0476, "step": 16460 }, { "epoch": 0.7649163568773234, "grad_norm": 0.6548092590453094, "learning_rate": 1.3920635780871095e-06, "loss": 0.064, "step": 16461 }, { "epoch": 0.7649628252788104, "grad_norm": 0.48090629318063516, "learning_rate": 1.3915426363083045e-06, "loss": 0.0601, "step": 16462 }, { "epoch": 0.7650092936802974, "grad_norm": 0.34101392575748085, "learning_rate": 1.391021776264857e-06, "loss": 0.0303, "step": 16463 }, { "epoch": 0.7650557620817844, "grad_norm": 0.3615991145227942, "learning_rate": 1.3905009979685635e-06, "loss": 0.0341, "step": 16464 }, { "epoch": 0.7651022304832714, "grad_norm": 0.4705283243543163, "learning_rate": 1.3899803014312224e-06, "loss": 0.0461, "step": 16465 }, { "epoch": 0.7651486988847583, "grad_norm": 0.4849200916121399, "learning_rate": 1.389459686664628e-06, "loss": 0.0388, "step": 16466 }, { "epoch": 0.7651951672862454, "grad_norm": 0.4906059171347718, "learning_rate": 1.3889391536805696e-06, "loss": 0.0542, "step": 16467 }, { "epoch": 0.7652416356877323, "grad_norm": 0.47473664783942704, "learning_rate": 1.3884187024908414e-06, "loss": 0.0583, "step": 16468 }, { "epoch": 0.7652881040892193, "grad_norm": 0.7808062233151756, "learning_rate": 1.3878983331072298e-06, "loss": 0.0625, "step": 16469 }, { "epoch": 0.7653345724907064, "grad_norm": 0.7522418853397087, "learning_rate": 1.3873780455415238e-06, "loss": 0.0642, "step": 16470 }, { "epoch": 0.7653810408921933, "grad_norm": 0.36507360478345563, "learning_rate": 1.3868578398055066e-06, "loss": 0.0528, "step": 16471 }, { "epoch": 0.7654275092936803, "grad_norm": 0.6380072369527706, "learning_rate": 1.386337715910962e-06, "loss": 0.0715, "step": 16472 }, { "epoch": 0.7654739776951673, "grad_norm": 0.45951826360498516, "learning_rate": 1.3858176738696716e-06, "loss": 0.028, "step": 16473 }, { "epoch": 0.7655204460966543, "grad_norm": 0.6367145138890988, "learning_rate": 1.385297713693413e-06, "loss": 0.053, "step": 16474 }, { "epoch": 0.7655669144981413, "grad_norm": 0.43482021749104505, "learning_rate": 1.3847778353939678e-06, "loss": 0.0434, "step": 16475 }, { "epoch": 0.7656133828996282, "grad_norm": 0.4784707982673674, "learning_rate": 1.3842580389831089e-06, "loss": 0.0565, "step": 16476 }, { "epoch": 0.7656598513011152, "grad_norm": 0.5625210931753175, "learning_rate": 1.3837383244726117e-06, "loss": 0.0494, "step": 16477 }, { "epoch": 0.7657063197026023, "grad_norm": 0.6374417394762455, "learning_rate": 1.383218691874247e-06, "loss": 0.0625, "step": 16478 }, { "epoch": 0.7657527881040892, "grad_norm": 0.45725194763294885, "learning_rate": 1.3826991411997853e-06, "loss": 0.0537, "step": 16479 }, { "epoch": 0.7657992565055762, "grad_norm": 0.5117264680512824, "learning_rate": 1.3821796724609954e-06, "loss": 0.0516, "step": 16480 }, { "epoch": 0.7658457249070632, "grad_norm": 0.6805849255823735, "learning_rate": 1.381660285669642e-06, "loss": 0.0646, "step": 16481 }, { "epoch": 0.7658921933085502, "grad_norm": 0.6043146741126532, "learning_rate": 1.3811409808374926e-06, "loss": 0.0472, "step": 16482 }, { "epoch": 0.7659386617100372, "grad_norm": 0.7011854129542777, "learning_rate": 1.380621757976311e-06, "loss": 0.0306, "step": 16483 }, { "epoch": 0.7659851301115241, "grad_norm": 0.39454434501931546, "learning_rate": 1.380102617097852e-06, "loss": 0.0403, "step": 16484 }, { "epoch": 0.7660315985130112, "grad_norm": 0.6085570426278984, "learning_rate": 1.379583558213881e-06, "loss": 0.0546, "step": 16485 }, { "epoch": 0.7660780669144981, "grad_norm": 0.5354965185316213, "learning_rate": 1.3790645813361524e-06, "loss": 0.0518, "step": 16486 }, { "epoch": 0.7661245353159851, "grad_norm": 0.3628937319045692, "learning_rate": 1.3785456864764219e-06, "loss": 0.0502, "step": 16487 }, { "epoch": 0.7661710037174722, "grad_norm": 0.6200722131251749, "learning_rate": 1.378026873646442e-06, "loss": 0.0688, "step": 16488 }, { "epoch": 0.7662174721189591, "grad_norm": 0.6295617546501319, "learning_rate": 1.3775081428579678e-06, "loss": 0.0529, "step": 16489 }, { "epoch": 0.7662639405204461, "grad_norm": 0.43941821315415047, "learning_rate": 1.3769894941227458e-06, "loss": 0.0666, "step": 16490 }, { "epoch": 0.7663104089219331, "grad_norm": 0.5147172894889573, "learning_rate": 1.3764709274525235e-06, "loss": 0.0402, "step": 16491 }, { "epoch": 0.7663568773234201, "grad_norm": 0.4305092702332964, "learning_rate": 1.37595244285905e-06, "loss": 0.0473, "step": 16492 }, { "epoch": 0.7664033457249071, "grad_norm": 0.7181366497603309, "learning_rate": 1.3754340403540672e-06, "loss": 0.0676, "step": 16493 }, { "epoch": 0.766449814126394, "grad_norm": 0.49015959095879635, "learning_rate": 1.3749157199493184e-06, "loss": 0.0462, "step": 16494 }, { "epoch": 0.766496282527881, "grad_norm": 0.3523029673925569, "learning_rate": 1.3743974816565441e-06, "loss": 0.0326, "step": 16495 }, { "epoch": 0.7665427509293681, "grad_norm": 0.40923968941114275, "learning_rate": 1.3738793254874827e-06, "loss": 0.0508, "step": 16496 }, { "epoch": 0.766589219330855, "grad_norm": 0.8111365798656468, "learning_rate": 1.3733612514538714e-06, "loss": 0.0671, "step": 16497 }, { "epoch": 0.766635687732342, "grad_norm": 0.3899266609082955, "learning_rate": 1.3728432595674423e-06, "loss": 0.0541, "step": 16498 }, { "epoch": 0.766682156133829, "grad_norm": 0.7167759989628345, "learning_rate": 1.372325349839933e-06, "loss": 0.0604, "step": 16499 }, { "epoch": 0.766728624535316, "grad_norm": 0.3357523457973897, "learning_rate": 1.3718075222830724e-06, "loss": 0.0423, "step": 16500 }, { "epoch": 0.766775092936803, "grad_norm": 0.4969622302441401, "learning_rate": 1.3712897769085903e-06, "loss": 0.0484, "step": 16501 }, { "epoch": 0.7668215613382899, "grad_norm": 0.3418159745196788, "learning_rate": 1.3707721137282137e-06, "loss": 0.0339, "step": 16502 }, { "epoch": 0.766868029739777, "grad_norm": 0.5343592832209482, "learning_rate": 1.3702545327536682e-06, "loss": 0.0659, "step": 16503 }, { "epoch": 0.7669144981412639, "grad_norm": 0.5182166382966572, "learning_rate": 1.3697370339966782e-06, "loss": 0.0546, "step": 16504 }, { "epoch": 0.7669609665427509, "grad_norm": 0.37155292606937934, "learning_rate": 1.3692196174689632e-06, "loss": 0.0416, "step": 16505 }, { "epoch": 0.767007434944238, "grad_norm": 0.6887349233547987, "learning_rate": 1.3687022831822472e-06, "loss": 0.059, "step": 16506 }, { "epoch": 0.7670539033457249, "grad_norm": 0.5444240148530587, "learning_rate": 1.3681850311482476e-06, "loss": 0.0587, "step": 16507 }, { "epoch": 0.7671003717472119, "grad_norm": 0.25924487103006594, "learning_rate": 1.3676678613786759e-06, "loss": 0.0584, "step": 16508 }, { "epoch": 0.7671468401486989, "grad_norm": 0.4732321256807935, "learning_rate": 1.3671507738852524e-06, "loss": 0.0502, "step": 16509 }, { "epoch": 0.7671933085501859, "grad_norm": 1.5707766123378128, "learning_rate": 1.366633768679687e-06, "loss": 0.0537, "step": 16510 }, { "epoch": 0.7672397769516729, "grad_norm": 0.7981072203192859, "learning_rate": 1.3661168457736913e-06, "loss": 0.0653, "step": 16511 }, { "epoch": 0.7672862453531598, "grad_norm": 0.3866023665618129, "learning_rate": 1.3656000051789713e-06, "loss": 0.0419, "step": 16512 }, { "epoch": 0.7673327137546468, "grad_norm": 0.6108589758503322, "learning_rate": 1.3650832469072401e-06, "loss": 0.0595, "step": 16513 }, { "epoch": 0.7673791821561339, "grad_norm": 0.5352290431515438, "learning_rate": 1.3645665709701971e-06, "loss": 0.0424, "step": 16514 }, { "epoch": 0.7674256505576208, "grad_norm": 0.47671437148083096, "learning_rate": 1.3640499773795456e-06, "loss": 0.0617, "step": 16515 }, { "epoch": 0.7674721189591078, "grad_norm": 0.5377185328426747, "learning_rate": 1.3635334661469911e-06, "loss": 0.0628, "step": 16516 }, { "epoch": 0.7675185873605948, "grad_norm": 0.5192032958153153, "learning_rate": 1.3630170372842306e-06, "loss": 0.0482, "step": 16517 }, { "epoch": 0.7675650557620818, "grad_norm": 0.7456041001221664, "learning_rate": 1.3625006908029625e-06, "loss": 0.0571, "step": 16518 }, { "epoch": 0.7676115241635688, "grad_norm": 1.2077727597200414, "learning_rate": 1.361984426714882e-06, "loss": 0.0822, "step": 16519 }, { "epoch": 0.7676579925650557, "grad_norm": 0.8381849385235352, "learning_rate": 1.3614682450316823e-06, "loss": 0.062, "step": 16520 }, { "epoch": 0.7677044609665428, "grad_norm": 0.5638095626088173, "learning_rate": 1.360952145765057e-06, "loss": 0.0664, "step": 16521 }, { "epoch": 0.7677509293680297, "grad_norm": 0.27896425384810974, "learning_rate": 1.3604361289266944e-06, "loss": 0.0384, "step": 16522 }, { "epoch": 0.7677973977695167, "grad_norm": 0.8480754158873377, "learning_rate": 1.3599201945282852e-06, "loss": 0.0598, "step": 16523 }, { "epoch": 0.7678438661710038, "grad_norm": 0.6469055006931331, "learning_rate": 1.3594043425815145e-06, "loss": 0.0532, "step": 16524 }, { "epoch": 0.7678903345724907, "grad_norm": 0.43412049898248334, "learning_rate": 1.3588885730980678e-06, "loss": 0.0366, "step": 16525 }, { "epoch": 0.7679368029739777, "grad_norm": 0.40596161851916374, "learning_rate": 1.3583728860896268e-06, "loss": 0.0324, "step": 16526 }, { "epoch": 0.7679832713754647, "grad_norm": 0.6574302555865411, "learning_rate": 1.3578572815678726e-06, "loss": 0.0412, "step": 16527 }, { "epoch": 0.7680297397769517, "grad_norm": 0.34046972057071007, "learning_rate": 1.3573417595444839e-06, "loss": 0.0413, "step": 16528 }, { "epoch": 0.7680762081784387, "grad_norm": 0.5896744433308674, "learning_rate": 1.3568263200311372e-06, "loss": 0.0604, "step": 16529 }, { "epoch": 0.7681226765799256, "grad_norm": 0.4841943048796172, "learning_rate": 1.3563109630395117e-06, "loss": 0.0429, "step": 16530 }, { "epoch": 0.7681691449814126, "grad_norm": 0.5927463846779774, "learning_rate": 1.3557956885812755e-06, "loss": 0.0372, "step": 16531 }, { "epoch": 0.7682156133828997, "grad_norm": 0.4140185117477341, "learning_rate": 1.355280496668102e-06, "loss": 0.0319, "step": 16532 }, { "epoch": 0.7682620817843866, "grad_norm": 0.48953855748340214, "learning_rate": 1.3547653873116622e-06, "loss": 0.0483, "step": 16533 }, { "epoch": 0.7683085501858736, "grad_norm": 0.42316475762818995, "learning_rate": 1.3542503605236229e-06, "loss": 0.051, "step": 16534 }, { "epoch": 0.7683550185873605, "grad_norm": 0.5325155416724537, "learning_rate": 1.3537354163156502e-06, "loss": 0.0518, "step": 16535 }, { "epoch": 0.7684014869888476, "grad_norm": 0.37305552641760453, "learning_rate": 1.3532205546994076e-06, "loss": 0.055, "step": 16536 }, { "epoch": 0.7684479553903346, "grad_norm": 0.4618400384552417, "learning_rate": 1.352705775686558e-06, "loss": 0.0454, "step": 16537 }, { "epoch": 0.7684944237918215, "grad_norm": 0.8198843644065938, "learning_rate": 1.352191079288761e-06, "loss": 0.0651, "step": 16538 }, { "epoch": 0.7685408921933086, "grad_norm": 0.6354915598224479, "learning_rate": 1.351676465517674e-06, "loss": 0.0633, "step": 16539 }, { "epoch": 0.7685873605947955, "grad_norm": 0.48763637152356953, "learning_rate": 1.3511619343849564e-06, "loss": 0.0367, "step": 16540 }, { "epoch": 0.7686338289962825, "grad_norm": 0.34748763033319074, "learning_rate": 1.3506474859022612e-06, "loss": 0.0366, "step": 16541 }, { "epoch": 0.7686802973977696, "grad_norm": 0.5380776967529934, "learning_rate": 1.3501331200812411e-06, "loss": 0.0558, "step": 16542 }, { "epoch": 0.7687267657992565, "grad_norm": 0.6212751596174877, "learning_rate": 1.3496188369335477e-06, "loss": 0.0623, "step": 16543 }, { "epoch": 0.7687732342007435, "grad_norm": 0.7327478790842652, "learning_rate": 1.3491046364708294e-06, "loss": 0.0395, "step": 16544 }, { "epoch": 0.7688197026022305, "grad_norm": 1.0167769842777636, "learning_rate": 1.3485905187047337e-06, "loss": 0.0416, "step": 16545 }, { "epoch": 0.7688661710037175, "grad_norm": 1.1139034486865105, "learning_rate": 1.348076483646904e-06, "loss": 0.0636, "step": 16546 }, { "epoch": 0.7689126394052045, "grad_norm": 0.5951983673508496, "learning_rate": 1.3475625313089874e-06, "loss": 0.0738, "step": 16547 }, { "epoch": 0.7689591078066914, "grad_norm": 0.5388695009202528, "learning_rate": 1.3470486617026246e-06, "loss": 0.0474, "step": 16548 }, { "epoch": 0.7690055762081784, "grad_norm": 0.6899037598624582, "learning_rate": 1.3465348748394518e-06, "loss": 0.0478, "step": 16549 }, { "epoch": 0.7690520446096655, "grad_norm": 0.5806897608614885, "learning_rate": 1.3460211707311104e-06, "loss": 0.0563, "step": 16550 }, { "epoch": 0.7690985130111524, "grad_norm": 0.7680030483976633, "learning_rate": 1.3455075493892356e-06, "loss": 0.0522, "step": 16551 }, { "epoch": 0.7691449814126394, "grad_norm": 0.5209167095405375, "learning_rate": 1.3449940108254606e-06, "loss": 0.0466, "step": 16552 }, { "epoch": 0.7691914498141263, "grad_norm": 0.3964345979132377, "learning_rate": 1.3444805550514167e-06, "loss": 0.0393, "step": 16553 }, { "epoch": 0.7692379182156134, "grad_norm": 0.7057139743899392, "learning_rate": 1.3439671820787386e-06, "loss": 0.0484, "step": 16554 }, { "epoch": 0.7692843866171004, "grad_norm": 0.6784465511513453, "learning_rate": 1.3434538919190505e-06, "loss": 0.0549, "step": 16555 }, { "epoch": 0.7693308550185873, "grad_norm": 0.7057121309267379, "learning_rate": 1.3429406845839787e-06, "loss": 0.0585, "step": 16556 }, { "epoch": 0.7693773234200744, "grad_norm": 0.5493363921638119, "learning_rate": 1.3424275600851506e-06, "loss": 0.061, "step": 16557 }, { "epoch": 0.7694237918215613, "grad_norm": 0.38508529700008637, "learning_rate": 1.341914518434188e-06, "loss": 0.0469, "step": 16558 }, { "epoch": 0.7694702602230483, "grad_norm": 0.3814140151756856, "learning_rate": 1.3414015596427115e-06, "loss": 0.048, "step": 16559 }, { "epoch": 0.7695167286245354, "grad_norm": 0.43438648409186426, "learning_rate": 1.34088868372234e-06, "loss": 0.0378, "step": 16560 }, { "epoch": 0.7695631970260223, "grad_norm": 0.7529755590443584, "learning_rate": 1.3403758906846914e-06, "loss": 0.0522, "step": 16561 }, { "epoch": 0.7696096654275093, "grad_norm": 0.533617627505972, "learning_rate": 1.3398631805413808e-06, "loss": 0.0372, "step": 16562 }, { "epoch": 0.7696561338289963, "grad_norm": 0.5236896392269672, "learning_rate": 1.3393505533040191e-06, "loss": 0.0551, "step": 16563 }, { "epoch": 0.7697026022304833, "grad_norm": 0.3940931755135657, "learning_rate": 1.3388380089842222e-06, "loss": 0.0474, "step": 16564 }, { "epoch": 0.7697490706319703, "grad_norm": 0.6642652603028179, "learning_rate": 1.3383255475935975e-06, "loss": 0.0702, "step": 16565 }, { "epoch": 0.7697955390334572, "grad_norm": 0.859342011128743, "learning_rate": 1.3378131691437529e-06, "loss": 0.0823, "step": 16566 }, { "epoch": 0.7698420074349442, "grad_norm": 0.5395740937260262, "learning_rate": 1.337300873646295e-06, "loss": 0.0415, "step": 16567 }, { "epoch": 0.7698884758364313, "grad_norm": 0.7112900879945918, "learning_rate": 1.3367886611128266e-06, "loss": 0.0589, "step": 16568 }, { "epoch": 0.7699349442379182, "grad_norm": 0.285624006861433, "learning_rate": 1.336276531554951e-06, "loss": 0.0332, "step": 16569 }, { "epoch": 0.7699814126394052, "grad_norm": 0.4013377034198942, "learning_rate": 1.3357644849842655e-06, "loss": 0.0497, "step": 16570 }, { "epoch": 0.7700278810408921, "grad_norm": 0.551233159579677, "learning_rate": 1.3352525214123735e-06, "loss": 0.0805, "step": 16571 }, { "epoch": 0.7700743494423792, "grad_norm": 0.6078059317755238, "learning_rate": 1.3347406408508695e-06, "loss": 0.0541, "step": 16572 }, { "epoch": 0.7701208178438662, "grad_norm": 0.5176579353998888, "learning_rate": 1.3342288433113454e-06, "loss": 0.0377, "step": 16573 }, { "epoch": 0.7701672862453531, "grad_norm": 0.5579742892995984, "learning_rate": 1.3337171288053974e-06, "loss": 0.0626, "step": 16574 }, { "epoch": 0.7702137546468402, "grad_norm": 0.4428897647372504, "learning_rate": 1.333205497344615e-06, "loss": 0.0547, "step": 16575 }, { "epoch": 0.7702602230483271, "grad_norm": 0.4002258936060133, "learning_rate": 1.3326939489405878e-06, "loss": 0.0519, "step": 16576 }, { "epoch": 0.7703066914498141, "grad_norm": 0.7321737098920725, "learning_rate": 1.3321824836049001e-06, "loss": 0.0489, "step": 16577 }, { "epoch": 0.7703531598513012, "grad_norm": 0.8647070736437593, "learning_rate": 1.331671101349143e-06, "loss": 0.0582, "step": 16578 }, { "epoch": 0.7703996282527881, "grad_norm": 0.5347975391451338, "learning_rate": 1.3311598021848947e-06, "loss": 0.0472, "step": 16579 }, { "epoch": 0.7704460966542751, "grad_norm": 1.753974553363707, "learning_rate": 1.3306485861237368e-06, "loss": 0.0521, "step": 16580 }, { "epoch": 0.7704925650557621, "grad_norm": 0.675473061724782, "learning_rate": 1.330137453177252e-06, "loss": 0.0447, "step": 16581 }, { "epoch": 0.7705390334572491, "grad_norm": 0.6488654717603349, "learning_rate": 1.329626403357016e-06, "loss": 0.0592, "step": 16582 }, { "epoch": 0.7705855018587361, "grad_norm": 0.43432488265823604, "learning_rate": 1.329115436674605e-06, "loss": 0.0399, "step": 16583 }, { "epoch": 0.770631970260223, "grad_norm": 0.782282328261091, "learning_rate": 1.3286045531415932e-06, "loss": 0.0604, "step": 16584 }, { "epoch": 0.77067843866171, "grad_norm": 0.46592257073818966, "learning_rate": 1.328093752769552e-06, "loss": 0.0545, "step": 16585 }, { "epoch": 0.7707249070631971, "grad_norm": 0.7861133897007004, "learning_rate": 1.3275830355700519e-06, "loss": 0.062, "step": 16586 }, { "epoch": 0.770771375464684, "grad_norm": 0.5049921979494177, "learning_rate": 1.3270724015546604e-06, "loss": 0.0584, "step": 16587 }, { "epoch": 0.770817843866171, "grad_norm": 0.4658381536264191, "learning_rate": 1.3265618507349453e-06, "loss": 0.0414, "step": 16588 }, { "epoch": 0.770864312267658, "grad_norm": 0.4082154323099193, "learning_rate": 1.3260513831224725e-06, "loss": 0.0464, "step": 16589 }, { "epoch": 0.770910780669145, "grad_norm": 0.49338446969648275, "learning_rate": 1.3255409987287993e-06, "loss": 0.0491, "step": 16590 }, { "epoch": 0.770957249070632, "grad_norm": 0.4814808565165263, "learning_rate": 1.325030697565492e-06, "loss": 0.0558, "step": 16591 }, { "epoch": 0.7710037174721189, "grad_norm": 0.4614351679963105, "learning_rate": 1.3245204796441063e-06, "loss": 0.0613, "step": 16592 }, { "epoch": 0.771050185873606, "grad_norm": 0.5608856207279798, "learning_rate": 1.3240103449762004e-06, "loss": 0.0439, "step": 16593 }, { "epoch": 0.7710966542750929, "grad_norm": 0.39223560851644007, "learning_rate": 1.3235002935733282e-06, "loss": 0.0338, "step": 16594 }, { "epoch": 0.7711431226765799, "grad_norm": 0.38511197296215155, "learning_rate": 1.322990325447046e-06, "loss": 0.0439, "step": 16595 }, { "epoch": 0.771189591078067, "grad_norm": 0.6580986573778149, "learning_rate": 1.322480440608902e-06, "loss": 0.0541, "step": 16596 }, { "epoch": 0.7712360594795539, "grad_norm": 0.6710811890824943, "learning_rate": 1.3219706390704445e-06, "loss": 0.0507, "step": 16597 }, { "epoch": 0.7712825278810409, "grad_norm": 0.7597189420454892, "learning_rate": 1.321460920843225e-06, "loss": 0.0588, "step": 16598 }, { "epoch": 0.7713289962825279, "grad_norm": 0.6054140641993309, "learning_rate": 1.3209512859387869e-06, "loss": 0.0543, "step": 16599 }, { "epoch": 0.7713754646840149, "grad_norm": 0.3436670508334621, "learning_rate": 1.3204417343686743e-06, "loss": 0.0359, "step": 16600 }, { "epoch": 0.7714219330855019, "grad_norm": 1.0422678424353886, "learning_rate": 1.3199322661444297e-06, "loss": 0.0639, "step": 16601 }, { "epoch": 0.7714684014869888, "grad_norm": 1.1042138656437315, "learning_rate": 1.3194228812775917e-06, "loss": 0.0733, "step": 16602 }, { "epoch": 0.7715148698884758, "grad_norm": 0.7376787097927847, "learning_rate": 1.3189135797796998e-06, "loss": 0.0714, "step": 16603 }, { "epoch": 0.7715613382899629, "grad_norm": 0.47855516560783573, "learning_rate": 1.3184043616622876e-06, "loss": 0.066, "step": 16604 }, { "epoch": 0.7716078066914498, "grad_norm": 0.5573738331279258, "learning_rate": 1.3178952269368934e-06, "loss": 0.0558, "step": 16605 }, { "epoch": 0.7716542750929368, "grad_norm": 1.5155254645502159, "learning_rate": 1.317386175615047e-06, "loss": 0.0502, "step": 16606 }, { "epoch": 0.7717007434944237, "grad_norm": 0.752904605588082, "learning_rate": 1.3168772077082803e-06, "loss": 0.0667, "step": 16607 }, { "epoch": 0.7717472118959108, "grad_norm": 0.6036140118390897, "learning_rate": 1.3163683232281206e-06, "loss": 0.0551, "step": 16608 }, { "epoch": 0.7717936802973978, "grad_norm": 0.5152054010692676, "learning_rate": 1.315859522186096e-06, "loss": 0.0442, "step": 16609 }, { "epoch": 0.7718401486988847, "grad_norm": 0.3877225324908731, "learning_rate": 1.3153508045937303e-06, "loss": 0.0405, "step": 16610 }, { "epoch": 0.7718866171003718, "grad_norm": 0.7476174496043414, "learning_rate": 1.3148421704625452e-06, "loss": 0.0566, "step": 16611 }, { "epoch": 0.7719330855018587, "grad_norm": 0.6920745954732341, "learning_rate": 1.3143336198040652e-06, "loss": 0.0596, "step": 16612 }, { "epoch": 0.7719795539033457, "grad_norm": 0.3717797192410478, "learning_rate": 1.3138251526298096e-06, "loss": 0.0249, "step": 16613 }, { "epoch": 0.7720260223048327, "grad_norm": 0.5898128142564402, "learning_rate": 1.3133167689512905e-06, "loss": 0.0435, "step": 16614 }, { "epoch": 0.7720724907063197, "grad_norm": 0.761810831838436, "learning_rate": 1.3128084687800286e-06, "loss": 0.0394, "step": 16615 }, { "epoch": 0.7721189591078067, "grad_norm": 0.385901330170263, "learning_rate": 1.3123002521275352e-06, "loss": 0.0422, "step": 16616 }, { "epoch": 0.7721654275092937, "grad_norm": 0.5530915041342824, "learning_rate": 1.311792119005323e-06, "loss": 0.0473, "step": 16617 }, { "epoch": 0.7722118959107807, "grad_norm": 0.5138439002455933, "learning_rate": 1.311284069424899e-06, "loss": 0.0386, "step": 16618 }, { "epoch": 0.7722583643122677, "grad_norm": 0.5208894597288362, "learning_rate": 1.3107761033977762e-06, "loss": 0.0336, "step": 16619 }, { "epoch": 0.7723048327137546, "grad_norm": 0.555658374776494, "learning_rate": 1.310268220935456e-06, "loss": 0.0464, "step": 16620 }, { "epoch": 0.7723513011152416, "grad_norm": 0.49980360992709083, "learning_rate": 1.3097604220494425e-06, "loss": 0.0408, "step": 16621 }, { "epoch": 0.7723977695167287, "grad_norm": 0.7318267247686503, "learning_rate": 1.3092527067512412e-06, "loss": 0.0603, "step": 16622 }, { "epoch": 0.7724442379182156, "grad_norm": 0.6644999957318307, "learning_rate": 1.30874507505235e-06, "loss": 0.0428, "step": 16623 }, { "epoch": 0.7724907063197026, "grad_norm": 0.6304476110595963, "learning_rate": 1.3082375269642678e-06, "loss": 0.0586, "step": 16624 }, { "epoch": 0.7725371747211895, "grad_norm": 0.28539722684506663, "learning_rate": 1.3077300624984912e-06, "loss": 0.0382, "step": 16625 }, { "epoch": 0.7725836431226766, "grad_norm": 0.46660876659836104, "learning_rate": 1.3072226816665152e-06, "loss": 0.0421, "step": 16626 }, { "epoch": 0.7726301115241636, "grad_norm": 0.6252369081172808, "learning_rate": 1.306715384479832e-06, "loss": 0.0493, "step": 16627 }, { "epoch": 0.7726765799256505, "grad_norm": 0.6119742430771778, "learning_rate": 1.3062081709499302e-06, "loss": 0.0442, "step": 16628 }, { "epoch": 0.7727230483271376, "grad_norm": 0.6199316704641649, "learning_rate": 1.305701041088303e-06, "loss": 0.0608, "step": 16629 }, { "epoch": 0.7727695167286245, "grad_norm": 0.8887407840808927, "learning_rate": 1.3051939949064352e-06, "loss": 0.0666, "step": 16630 }, { "epoch": 0.7728159851301115, "grad_norm": 0.5095115232229611, "learning_rate": 1.3046870324158122e-06, "loss": 0.0575, "step": 16631 }, { "epoch": 0.7728624535315985, "grad_norm": 0.5735871469312064, "learning_rate": 1.3041801536279174e-06, "loss": 0.0491, "step": 16632 }, { "epoch": 0.7729089219330855, "grad_norm": 0.5418929250863694, "learning_rate": 1.3036733585542317e-06, "loss": 0.0562, "step": 16633 }, { "epoch": 0.7729553903345725, "grad_norm": 0.36867385589873214, "learning_rate": 1.3031666472062348e-06, "loss": 0.0413, "step": 16634 }, { "epoch": 0.7730018587360595, "grad_norm": 0.5330506849315582, "learning_rate": 1.3026600195954025e-06, "loss": 0.0729, "step": 16635 }, { "epoch": 0.7730483271375465, "grad_norm": 0.5774051456426101, "learning_rate": 1.3021534757332139e-06, "loss": 0.0537, "step": 16636 }, { "epoch": 0.7730947955390335, "grad_norm": 0.6324229456464461, "learning_rate": 1.3016470156311422e-06, "loss": 0.0538, "step": 16637 }, { "epoch": 0.7731412639405204, "grad_norm": 0.6651013817755502, "learning_rate": 1.3011406393006555e-06, "loss": 0.0478, "step": 16638 }, { "epoch": 0.7731877323420074, "grad_norm": 0.4190056686070385, "learning_rate": 1.3006343467532278e-06, "loss": 0.0552, "step": 16639 }, { "epoch": 0.7732342007434945, "grad_norm": 0.5947669169392221, "learning_rate": 1.3001281380003251e-06, "loss": 0.0547, "step": 16640 }, { "epoch": 0.7732806691449814, "grad_norm": 0.33172440291634975, "learning_rate": 1.2996220130534143e-06, "loss": 0.0515, "step": 16641 }, { "epoch": 0.7733271375464684, "grad_norm": 0.5444177391648419, "learning_rate": 1.2991159719239581e-06, "loss": 0.0472, "step": 16642 }, { "epoch": 0.7733736059479553, "grad_norm": 0.5683666243516949, "learning_rate": 1.298610014623423e-06, "loss": 0.0395, "step": 16643 }, { "epoch": 0.7734200743494424, "grad_norm": 0.4475125375115891, "learning_rate": 1.2981041411632656e-06, "loss": 0.0542, "step": 16644 }, { "epoch": 0.7734665427509294, "grad_norm": 0.7633679850387342, "learning_rate": 1.297598351554944e-06, "loss": 0.0655, "step": 16645 }, { "epoch": 0.7735130111524163, "grad_norm": 0.32559581466920057, "learning_rate": 1.297092645809918e-06, "loss": 0.0674, "step": 16646 }, { "epoch": 0.7735594795539034, "grad_norm": 0.5272728970672022, "learning_rate": 1.2965870239396406e-06, "loss": 0.0608, "step": 16647 }, { "epoch": 0.7736059479553903, "grad_norm": 0.6938571646443278, "learning_rate": 1.2960814859555648e-06, "loss": 0.0398, "step": 16648 }, { "epoch": 0.7736524163568773, "grad_norm": 0.46659448149974203, "learning_rate": 1.295576031869142e-06, "loss": 0.0513, "step": 16649 }, { "epoch": 0.7736988847583643, "grad_norm": 0.3813193090401813, "learning_rate": 1.2950706616918208e-06, "loss": 0.0508, "step": 16650 }, { "epoch": 0.7737453531598513, "grad_norm": 0.4904915205158511, "learning_rate": 1.2945653754350485e-06, "loss": 0.0541, "step": 16651 }, { "epoch": 0.7737918215613383, "grad_norm": 0.3874787991688008, "learning_rate": 1.2940601731102687e-06, "loss": 0.0342, "step": 16652 }, { "epoch": 0.7738382899628253, "grad_norm": 0.6001279817200202, "learning_rate": 1.2935550547289278e-06, "loss": 0.0704, "step": 16653 }, { "epoch": 0.7738847583643123, "grad_norm": 0.3155201073122889, "learning_rate": 1.2930500203024675e-06, "loss": 0.0337, "step": 16654 }, { "epoch": 0.7739312267657993, "grad_norm": 0.396682710145175, "learning_rate": 1.2925450698423226e-06, "loss": 0.0439, "step": 16655 }, { "epoch": 0.7739776951672862, "grad_norm": 0.44942097123045355, "learning_rate": 1.2920402033599355e-06, "loss": 0.0404, "step": 16656 }, { "epoch": 0.7740241635687732, "grad_norm": 0.46233768076500104, "learning_rate": 1.2915354208667409e-06, "loss": 0.0414, "step": 16657 }, { "epoch": 0.7740706319702603, "grad_norm": 0.5216548861677163, "learning_rate": 1.2910307223741714e-06, "loss": 0.0673, "step": 16658 }, { "epoch": 0.7741171003717472, "grad_norm": 0.4821827627060236, "learning_rate": 1.2905261078936582e-06, "loss": 0.0459, "step": 16659 }, { "epoch": 0.7741635687732342, "grad_norm": 0.4552828171953811, "learning_rate": 1.290021577436636e-06, "loss": 0.0457, "step": 16660 }, { "epoch": 0.7742100371747211, "grad_norm": 0.317871482506005, "learning_rate": 1.2895171310145282e-06, "loss": 0.0343, "step": 16661 }, { "epoch": 0.7742565055762082, "grad_norm": 0.46515007981622825, "learning_rate": 1.2890127686387611e-06, "loss": 0.0413, "step": 16662 }, { "epoch": 0.7743029739776952, "grad_norm": 0.4339892893723632, "learning_rate": 1.2885084903207624e-06, "loss": 0.0349, "step": 16663 }, { "epoch": 0.7743494423791821, "grad_norm": 0.4923319603478217, "learning_rate": 1.288004296071952e-06, "loss": 0.0535, "step": 16664 }, { "epoch": 0.7743959107806692, "grad_norm": 0.7389019272405815, "learning_rate": 1.2875001859037512e-06, "loss": 0.0566, "step": 16665 }, { "epoch": 0.7744423791821561, "grad_norm": 0.5146759352247766, "learning_rate": 1.2869961598275788e-06, "loss": 0.0532, "step": 16666 }, { "epoch": 0.7744888475836431, "grad_norm": 0.6429966797867646, "learning_rate": 1.2864922178548512e-06, "loss": 0.056, "step": 16667 }, { "epoch": 0.7745353159851301, "grad_norm": 0.5275407024057229, "learning_rate": 1.2859883599969835e-06, "loss": 0.0407, "step": 16668 }, { "epoch": 0.7745817843866171, "grad_norm": 0.5051032073863178, "learning_rate": 1.2854845862653864e-06, "loss": 0.0458, "step": 16669 }, { "epoch": 0.7746282527881041, "grad_norm": 0.6592304367978136, "learning_rate": 1.284980896671475e-06, "loss": 0.0696, "step": 16670 }, { "epoch": 0.7746747211895911, "grad_norm": 0.34363542577488415, "learning_rate": 1.2844772912266557e-06, "loss": 0.0271, "step": 16671 }, { "epoch": 0.774721189591078, "grad_norm": 0.37249037315700334, "learning_rate": 1.2839737699423366e-06, "loss": 0.0267, "step": 16672 }, { "epoch": 0.7747676579925651, "grad_norm": 0.5129411331254363, "learning_rate": 1.2834703328299219e-06, "loss": 0.0584, "step": 16673 }, { "epoch": 0.774814126394052, "grad_norm": 0.6945860441748957, "learning_rate": 1.2829669799008166e-06, "loss": 0.063, "step": 16674 }, { "epoch": 0.774860594795539, "grad_norm": 0.2802445101888413, "learning_rate": 1.2824637111664212e-06, "loss": 0.0402, "step": 16675 }, { "epoch": 0.7749070631970261, "grad_norm": 0.3719922914124536, "learning_rate": 1.2819605266381336e-06, "loss": 0.0425, "step": 16676 }, { "epoch": 0.774953531598513, "grad_norm": 0.3780350122229221, "learning_rate": 1.281457426327355e-06, "loss": 0.0516, "step": 16677 }, { "epoch": 0.775, "grad_norm": 0.6554836463269416, "learning_rate": 1.2809544102454806e-06, "loss": 0.0519, "step": 16678 }, { "epoch": 0.7750464684014869, "grad_norm": 0.7403409837161029, "learning_rate": 1.2804514784039e-06, "loss": 0.0526, "step": 16679 }, { "epoch": 0.775092936802974, "grad_norm": 0.616457199971061, "learning_rate": 1.2799486308140101e-06, "loss": 0.054, "step": 16680 }, { "epoch": 0.775139405204461, "grad_norm": 0.6974493991870419, "learning_rate": 1.279445867487199e-06, "loss": 0.067, "step": 16681 }, { "epoch": 0.7751858736059479, "grad_norm": 0.49304876140178394, "learning_rate": 1.2789431884348541e-06, "loss": 0.0421, "step": 16682 }, { "epoch": 0.775232342007435, "grad_norm": 0.6097612376897308, "learning_rate": 1.2784405936683614e-06, "loss": 0.0487, "step": 16683 }, { "epoch": 0.7752788104089219, "grad_norm": 0.299173134968794, "learning_rate": 1.277938083199109e-06, "loss": 0.0333, "step": 16684 }, { "epoch": 0.7753252788104089, "grad_norm": 0.3762598230608939, "learning_rate": 1.2774356570384749e-06, "loss": 0.0491, "step": 16685 }, { "epoch": 0.775371747211896, "grad_norm": 0.36402700559633827, "learning_rate": 1.2769333151978396e-06, "loss": 0.0272, "step": 16686 }, { "epoch": 0.7754182156133829, "grad_norm": 0.369418825139662, "learning_rate": 1.2764310576885847e-06, "loss": 0.0413, "step": 16687 }, { "epoch": 0.7754646840148699, "grad_norm": 0.3806159027257536, "learning_rate": 1.2759288845220852e-06, "loss": 0.056, "step": 16688 }, { "epoch": 0.7755111524163569, "grad_norm": 0.5815808477962735, "learning_rate": 1.2754267957097166e-06, "loss": 0.0506, "step": 16689 }, { "epoch": 0.7755576208178439, "grad_norm": 0.6248961842393403, "learning_rate": 1.2749247912628503e-06, "loss": 0.0571, "step": 16690 }, { "epoch": 0.7756040892193309, "grad_norm": 0.4947626378275562, "learning_rate": 1.2744228711928585e-06, "loss": 0.0624, "step": 16691 }, { "epoch": 0.7756505576208178, "grad_norm": 0.6946433216055404, "learning_rate": 1.2739210355111098e-06, "loss": 0.0782, "step": 16692 }, { "epoch": 0.7756970260223048, "grad_norm": 0.5177049783956965, "learning_rate": 1.2734192842289693e-06, "loss": 0.0479, "step": 16693 }, { "epoch": 0.7757434944237919, "grad_norm": 0.6459567869950507, "learning_rate": 1.2729176173578062e-06, "loss": 0.0649, "step": 16694 }, { "epoch": 0.7757899628252788, "grad_norm": 0.5377818817546539, "learning_rate": 1.2724160349089816e-06, "loss": 0.0509, "step": 16695 }, { "epoch": 0.7758364312267658, "grad_norm": 0.39511139943436585, "learning_rate": 1.2719145368938568e-06, "loss": 0.0489, "step": 16696 }, { "epoch": 0.7758828996282527, "grad_norm": 0.4680037808992507, "learning_rate": 1.2714131233237913e-06, "loss": 0.0598, "step": 16697 }, { "epoch": 0.7759293680297398, "grad_norm": 0.5551133871055305, "learning_rate": 1.2709117942101434e-06, "loss": 0.0638, "step": 16698 }, { "epoch": 0.7759758364312268, "grad_norm": 0.3721253254524406, "learning_rate": 1.2704105495642677e-06, "loss": 0.0441, "step": 16699 }, { "epoch": 0.7760223048327137, "grad_norm": 0.2985919858147372, "learning_rate": 1.2699093893975172e-06, "loss": 0.0534, "step": 16700 }, { "epoch": 0.7760687732342008, "grad_norm": 0.4514854630107356, "learning_rate": 1.2694083137212465e-06, "loss": 0.0639, "step": 16701 }, { "epoch": 0.7761152416356877, "grad_norm": 0.7021888796242827, "learning_rate": 1.2689073225468046e-06, "loss": 0.0537, "step": 16702 }, { "epoch": 0.7761617100371747, "grad_norm": 0.6684147233336258, "learning_rate": 1.2684064158855364e-06, "loss": 0.0596, "step": 16703 }, { "epoch": 0.7762081784386617, "grad_norm": 0.43601723943653, "learning_rate": 1.2679055937487916e-06, "loss": 0.0489, "step": 16704 }, { "epoch": 0.7762546468401487, "grad_norm": 1.0685985561756635, "learning_rate": 1.2674048561479135e-06, "loss": 0.0554, "step": 16705 }, { "epoch": 0.7763011152416357, "grad_norm": 0.4234453010831481, "learning_rate": 1.2669042030942436e-06, "loss": 0.0451, "step": 16706 }, { "epoch": 0.7763475836431227, "grad_norm": 0.6559003940252291, "learning_rate": 1.266403634599122e-06, "loss": 0.0503, "step": 16707 }, { "epoch": 0.7763940520446097, "grad_norm": 0.4588338688735731, "learning_rate": 1.2659031506738884e-06, "loss": 0.0403, "step": 16708 }, { "epoch": 0.7764405204460967, "grad_norm": 0.7760515897460135, "learning_rate": 1.2654027513298777e-06, "loss": 0.0453, "step": 16709 }, { "epoch": 0.7764869888475836, "grad_norm": 0.8609198457602237, "learning_rate": 1.2649024365784245e-06, "loss": 0.0649, "step": 16710 }, { "epoch": 0.7765334572490706, "grad_norm": 0.48265579815793, "learning_rate": 1.2644022064308637e-06, "loss": 0.043, "step": 16711 }, { "epoch": 0.7765799256505577, "grad_norm": 0.705895482833546, "learning_rate": 1.2639020608985242e-06, "loss": 0.055, "step": 16712 }, { "epoch": 0.7766263940520446, "grad_norm": 0.5809080566968069, "learning_rate": 1.2634019999927355e-06, "loss": 0.039, "step": 16713 }, { "epoch": 0.7766728624535316, "grad_norm": 0.30283369212193495, "learning_rate": 1.2629020237248241e-06, "loss": 0.0545, "step": 16714 }, { "epoch": 0.7767193308550185, "grad_norm": 0.5676042730227314, "learning_rate": 1.2624021321061147e-06, "loss": 0.0516, "step": 16715 }, { "epoch": 0.7767657992565056, "grad_norm": 0.36431116228141497, "learning_rate": 1.2619023251479311e-06, "loss": 0.0561, "step": 16716 }, { "epoch": 0.7768122676579926, "grad_norm": 0.5202715112626413, "learning_rate": 1.2614026028615928e-06, "loss": 0.0399, "step": 16717 }, { "epoch": 0.7768587360594795, "grad_norm": 0.38123919535015005, "learning_rate": 1.2609029652584214e-06, "loss": 0.0547, "step": 16718 }, { "epoch": 0.7769052044609666, "grad_norm": 0.5797443998460746, "learning_rate": 1.260403412349735e-06, "loss": 0.0485, "step": 16719 }, { "epoch": 0.7769516728624535, "grad_norm": 0.9986815084595669, "learning_rate": 1.259903944146844e-06, "loss": 0.0688, "step": 16720 }, { "epoch": 0.7769981412639405, "grad_norm": 0.5093445162087653, "learning_rate": 1.2594045606610666e-06, "loss": 0.0375, "step": 16721 }, { "epoch": 0.7770446096654275, "grad_norm": 0.5502446628189829, "learning_rate": 1.2589052619037128e-06, "loss": 0.0453, "step": 16722 }, { "epoch": 0.7770910780669145, "grad_norm": 0.5352411408113324, "learning_rate": 1.258406047886092e-06, "loss": 0.0418, "step": 16723 }, { "epoch": 0.7771375464684015, "grad_norm": 0.46511880346721696, "learning_rate": 1.2579069186195114e-06, "loss": 0.0344, "step": 16724 }, { "epoch": 0.7771840148698885, "grad_norm": 0.8456578477473368, "learning_rate": 1.2574078741152801e-06, "loss": 0.062, "step": 16725 }, { "epoch": 0.7772304832713754, "grad_norm": 0.553473421265052, "learning_rate": 1.256908914384698e-06, "loss": 0.0606, "step": 16726 }, { "epoch": 0.7772769516728625, "grad_norm": 0.4542125510005663, "learning_rate": 1.2564100394390666e-06, "loss": 0.042, "step": 16727 }, { "epoch": 0.7773234200743494, "grad_norm": 0.4946857207659207, "learning_rate": 1.2559112492896903e-06, "loss": 0.0551, "step": 16728 }, { "epoch": 0.7773698884758364, "grad_norm": 0.5154970971587116, "learning_rate": 1.2554125439478637e-06, "loss": 0.0433, "step": 16729 }, { "epoch": 0.7774163568773235, "grad_norm": 0.49299141890265996, "learning_rate": 1.2549139234248847e-06, "loss": 0.037, "step": 16730 }, { "epoch": 0.7774628252788104, "grad_norm": 0.5027030853072814, "learning_rate": 1.254415387732047e-06, "loss": 0.0433, "step": 16731 }, { "epoch": 0.7775092936802974, "grad_norm": 0.3867139939854025, "learning_rate": 1.2539169368806426e-06, "loss": 0.0318, "step": 16732 }, { "epoch": 0.7775557620817843, "grad_norm": 0.5948970938285427, "learning_rate": 1.2534185708819624e-06, "loss": 0.0587, "step": 16733 }, { "epoch": 0.7776022304832714, "grad_norm": 0.3381960452878064, "learning_rate": 1.2529202897472937e-06, "loss": 0.0353, "step": 16734 }, { "epoch": 0.7776486988847584, "grad_norm": 0.38394364310677304, "learning_rate": 1.252422093487925e-06, "loss": 0.0438, "step": 16735 }, { "epoch": 0.7776951672862453, "grad_norm": 0.6182260940926857, "learning_rate": 1.2519239821151408e-06, "loss": 0.0513, "step": 16736 }, { "epoch": 0.7777416356877324, "grad_norm": 0.41937914961587275, "learning_rate": 1.2514259556402225e-06, "loss": 0.0543, "step": 16737 }, { "epoch": 0.7777881040892193, "grad_norm": 0.24266485067292617, "learning_rate": 1.2509280140744522e-06, "loss": 0.0327, "step": 16738 }, { "epoch": 0.7778345724907063, "grad_norm": 0.6341526386609614, "learning_rate": 1.2504301574291077e-06, "loss": 0.07, "step": 16739 }, { "epoch": 0.7778810408921933, "grad_norm": 0.6407447730900988, "learning_rate": 1.249932385715467e-06, "loss": 0.0589, "step": 16740 }, { "epoch": 0.7779275092936803, "grad_norm": 0.4911102001350754, "learning_rate": 1.2494346989448031e-06, "loss": 0.0605, "step": 16741 }, { "epoch": 0.7779739776951673, "grad_norm": 0.6399210843126918, "learning_rate": 1.2489370971283916e-06, "loss": 0.0526, "step": 16742 }, { "epoch": 0.7780204460966543, "grad_norm": 0.5142026000752287, "learning_rate": 1.2484395802775045e-06, "loss": 0.0498, "step": 16743 }, { "epoch": 0.7780669144981412, "grad_norm": 0.6048056089041824, "learning_rate": 1.2479421484034066e-06, "loss": 0.0582, "step": 16744 }, { "epoch": 0.7781133828996283, "grad_norm": 0.5677689225051088, "learning_rate": 1.2474448015173695e-06, "loss": 0.0543, "step": 16745 }, { "epoch": 0.7781598513011152, "grad_norm": 0.5134825886550876, "learning_rate": 1.246947539630657e-06, "loss": 0.05, "step": 16746 }, { "epoch": 0.7782063197026022, "grad_norm": 0.28226090936022247, "learning_rate": 1.2464503627545327e-06, "loss": 0.0452, "step": 16747 }, { "epoch": 0.7782527881040893, "grad_norm": 1.154655417330777, "learning_rate": 1.245953270900257e-06, "loss": 0.0633, "step": 16748 }, { "epoch": 0.7782992565055762, "grad_norm": 0.9166423444988204, "learning_rate": 1.2454562640790935e-06, "loss": 0.0698, "step": 16749 }, { "epoch": 0.7783457249070632, "grad_norm": 0.8552037977810075, "learning_rate": 1.2449593423022955e-06, "loss": 0.0466, "step": 16750 }, { "epoch": 0.7783921933085501, "grad_norm": 0.216150577986314, "learning_rate": 1.2444625055811194e-06, "loss": 0.0214, "step": 16751 }, { "epoch": 0.7784386617100372, "grad_norm": 0.7274027263148509, "learning_rate": 1.2439657539268214e-06, "loss": 0.0812, "step": 16752 }, { "epoch": 0.7784851301115242, "grad_norm": 0.500254331673225, "learning_rate": 1.2434690873506528e-06, "loss": 0.0339, "step": 16753 }, { "epoch": 0.7785315985130111, "grad_norm": 0.42922512230607124, "learning_rate": 1.2429725058638625e-06, "loss": 0.039, "step": 16754 }, { "epoch": 0.7785780669144982, "grad_norm": 0.48716467461531365, "learning_rate": 1.242476009477699e-06, "loss": 0.0618, "step": 16755 }, { "epoch": 0.7786245353159851, "grad_norm": 0.500693121431686, "learning_rate": 1.2419795982034089e-06, "loss": 0.0404, "step": 16756 }, { "epoch": 0.7786710037174721, "grad_norm": 0.7111713576520278, "learning_rate": 1.2414832720522363e-06, "loss": 0.0463, "step": 16757 }, { "epoch": 0.7787174721189591, "grad_norm": 0.3894037407566429, "learning_rate": 1.2409870310354216e-06, "loss": 0.0307, "step": 16758 }, { "epoch": 0.7787639405204461, "grad_norm": 0.5192050289251146, "learning_rate": 1.2404908751642086e-06, "loss": 0.076, "step": 16759 }, { "epoch": 0.7788104089219331, "grad_norm": 0.46484652977136154, "learning_rate": 1.239994804449834e-06, "loss": 0.0594, "step": 16760 }, { "epoch": 0.7788568773234201, "grad_norm": 0.5388297241892238, "learning_rate": 1.239498818903534e-06, "loss": 0.0524, "step": 16761 }, { "epoch": 0.778903345724907, "grad_norm": 0.6095789721189209, "learning_rate": 1.2390029185365442e-06, "loss": 0.0451, "step": 16762 }, { "epoch": 0.7789498141263941, "grad_norm": 0.3081597477591256, "learning_rate": 1.2385071033600965e-06, "loss": 0.0328, "step": 16763 }, { "epoch": 0.778996282527881, "grad_norm": 0.4629234622178752, "learning_rate": 1.2380113733854215e-06, "loss": 0.0463, "step": 16764 }, { "epoch": 0.779042750929368, "grad_norm": 0.3919667653298015, "learning_rate": 1.2375157286237478e-06, "loss": 0.0518, "step": 16765 }, { "epoch": 0.7790892193308551, "grad_norm": 0.43608468677077605, "learning_rate": 1.2370201690863048e-06, "loss": 0.045, "step": 16766 }, { "epoch": 0.779135687732342, "grad_norm": 0.9965814949866666, "learning_rate": 1.2365246947843141e-06, "loss": 0.0726, "step": 16767 }, { "epoch": 0.779182156133829, "grad_norm": 0.4875657459161552, "learning_rate": 1.2360293057289991e-06, "loss": 0.0322, "step": 16768 }, { "epoch": 0.7792286245353159, "grad_norm": 0.5118086665313433, "learning_rate": 1.2355340019315826e-06, "loss": 0.0406, "step": 16769 }, { "epoch": 0.779275092936803, "grad_norm": 0.7955592939757272, "learning_rate": 1.2350387834032834e-06, "loss": 0.0609, "step": 16770 }, { "epoch": 0.77932156133829, "grad_norm": 0.5185326199833948, "learning_rate": 1.2345436501553183e-06, "loss": 0.0369, "step": 16771 }, { "epoch": 0.7793680297397769, "grad_norm": 0.421191324544129, "learning_rate": 1.2340486021989023e-06, "loss": 0.0409, "step": 16772 }, { "epoch": 0.779414498141264, "grad_norm": 0.5226858096708101, "learning_rate": 1.233553639545249e-06, "loss": 0.0544, "step": 16773 }, { "epoch": 0.7794609665427509, "grad_norm": 0.565662546524097, "learning_rate": 1.2330587622055707e-06, "loss": 0.0532, "step": 16774 }, { "epoch": 0.7795074349442379, "grad_norm": 0.682039919136759, "learning_rate": 1.2325639701910735e-06, "loss": 0.0514, "step": 16775 }, { "epoch": 0.7795539033457249, "grad_norm": 0.7792698064165097, "learning_rate": 1.23206926351297e-06, "loss": 0.0608, "step": 16776 }, { "epoch": 0.7796003717472119, "grad_norm": 0.5044810906543261, "learning_rate": 1.2315746421824627e-06, "loss": 0.0416, "step": 16777 }, { "epoch": 0.7796468401486989, "grad_norm": 0.4093168803878677, "learning_rate": 1.2310801062107558e-06, "loss": 0.0336, "step": 16778 }, { "epoch": 0.7796933085501859, "grad_norm": 0.5181596836385746, "learning_rate": 1.2305856556090517e-06, "loss": 0.0627, "step": 16779 }, { "epoch": 0.7797397769516728, "grad_norm": 0.5025858831905051, "learning_rate": 1.2300912903885497e-06, "loss": 0.0654, "step": 16780 }, { "epoch": 0.7797862453531599, "grad_norm": 0.29229287474982873, "learning_rate": 1.2295970105604476e-06, "loss": 0.0296, "step": 16781 }, { "epoch": 0.7798327137546468, "grad_norm": 0.5140117507392641, "learning_rate": 1.2291028161359404e-06, "loss": 0.0489, "step": 16782 }, { "epoch": 0.7798791821561338, "grad_norm": 0.45024843307712376, "learning_rate": 1.2286087071262237e-06, "loss": 0.0357, "step": 16783 }, { "epoch": 0.7799256505576209, "grad_norm": 0.47035077972252887, "learning_rate": 1.228114683542491e-06, "loss": 0.0488, "step": 16784 }, { "epoch": 0.7799721189591078, "grad_norm": 0.5372245055146152, "learning_rate": 1.2276207453959283e-06, "loss": 0.0528, "step": 16785 }, { "epoch": 0.7800185873605948, "grad_norm": 0.4124889156328049, "learning_rate": 1.227126892697727e-06, "loss": 0.0416, "step": 16786 }, { "epoch": 0.7800650557620817, "grad_norm": 0.36045581451131664, "learning_rate": 1.2266331254590724e-06, "loss": 0.0627, "step": 16787 }, { "epoch": 0.7801115241635688, "grad_norm": 0.5464395934936542, "learning_rate": 1.2261394436911478e-06, "loss": 0.0532, "step": 16788 }, { "epoch": 0.7801579925650558, "grad_norm": 0.47939061702571933, "learning_rate": 1.2256458474051375e-06, "loss": 0.0515, "step": 16789 }, { "epoch": 0.7802044609665427, "grad_norm": 0.7512427214708324, "learning_rate": 1.2251523366122232e-06, "loss": 0.0468, "step": 16790 }, { "epoch": 0.7802509293680298, "grad_norm": 0.48420511547772516, "learning_rate": 1.2246589113235802e-06, "loss": 0.028, "step": 16791 }, { "epoch": 0.7802973977695167, "grad_norm": 0.5133473520473923, "learning_rate": 1.2241655715503842e-06, "loss": 0.0683, "step": 16792 }, { "epoch": 0.7803438661710037, "grad_norm": 0.6907213582661678, "learning_rate": 1.2236723173038139e-06, "loss": 0.0849, "step": 16793 }, { "epoch": 0.7803903345724907, "grad_norm": 0.5933807877668428, "learning_rate": 1.22317914859504e-06, "loss": 0.0542, "step": 16794 }, { "epoch": 0.7804368029739777, "grad_norm": 0.27270402654783427, "learning_rate": 1.222686065435232e-06, "loss": 0.0307, "step": 16795 }, { "epoch": 0.7804832713754647, "grad_norm": 0.4900620792156846, "learning_rate": 1.2221930678355631e-06, "loss": 0.057, "step": 16796 }, { "epoch": 0.7805297397769517, "grad_norm": 0.417796739653353, "learning_rate": 1.2217001558071956e-06, "loss": 0.0449, "step": 16797 }, { "epoch": 0.7805762081784386, "grad_norm": 0.5454893470832975, "learning_rate": 1.2212073293612965e-06, "loss": 0.0574, "step": 16798 }, { "epoch": 0.7806226765799257, "grad_norm": 0.4600112632560122, "learning_rate": 1.2207145885090266e-06, "loss": 0.0431, "step": 16799 }, { "epoch": 0.7806691449814126, "grad_norm": 0.4258394449162746, "learning_rate": 1.2202219332615507e-06, "loss": 0.0404, "step": 16800 }, { "epoch": 0.7807156133828996, "grad_norm": 0.8071180607689988, "learning_rate": 1.2197293636300261e-06, "loss": 0.0483, "step": 16801 }, { "epoch": 0.7807620817843867, "grad_norm": 0.37877076865467757, "learning_rate": 1.2192368796256098e-06, "loss": 0.0532, "step": 16802 }, { "epoch": 0.7808085501858736, "grad_norm": 0.33407875374214857, "learning_rate": 1.2187444812594578e-06, "loss": 0.0376, "step": 16803 }, { "epoch": 0.7808550185873606, "grad_norm": 0.5237654814851297, "learning_rate": 1.2182521685427228e-06, "loss": 0.0566, "step": 16804 }, { "epoch": 0.7809014869888475, "grad_norm": 0.4525993351738582, "learning_rate": 1.2177599414865549e-06, "loss": 0.0387, "step": 16805 }, { "epoch": 0.7809479553903346, "grad_norm": 0.2807539134471545, "learning_rate": 1.2172678001021065e-06, "loss": 0.0357, "step": 16806 }, { "epoch": 0.7809944237918216, "grad_norm": 0.5342078429387831, "learning_rate": 1.216775744400524e-06, "loss": 0.0487, "step": 16807 }, { "epoch": 0.7810408921933085, "grad_norm": 0.43637976093655784, "learning_rate": 1.216283774392954e-06, "loss": 0.037, "step": 16808 }, { "epoch": 0.7810873605947956, "grad_norm": 0.5737264176459896, "learning_rate": 1.2157918900905358e-06, "loss": 0.0587, "step": 16809 }, { "epoch": 0.7811338289962825, "grad_norm": 0.5929263722455175, "learning_rate": 1.215300091504415e-06, "loss": 0.0696, "step": 16810 }, { "epoch": 0.7811802973977695, "grad_norm": 0.2779720183326675, "learning_rate": 1.214808378645731e-06, "loss": 0.0366, "step": 16811 }, { "epoch": 0.7812267657992565, "grad_norm": 0.5450689535089911, "learning_rate": 1.21431675152562e-06, "loss": 0.0605, "step": 16812 }, { "epoch": 0.7812732342007435, "grad_norm": 0.542887416610318, "learning_rate": 1.2138252101552195e-06, "loss": 0.0687, "step": 16813 }, { "epoch": 0.7813197026022305, "grad_norm": 0.49309891156788727, "learning_rate": 1.2133337545456653e-06, "loss": 0.055, "step": 16814 }, { "epoch": 0.7813661710037175, "grad_norm": 0.4525406967299516, "learning_rate": 1.2128423847080834e-06, "loss": 0.0381, "step": 16815 }, { "epoch": 0.7814126394052044, "grad_norm": 0.4109831387399182, "learning_rate": 1.2123511006536086e-06, "loss": 0.0237, "step": 16816 }, { "epoch": 0.7814591078066915, "grad_norm": 0.7656837614619989, "learning_rate": 1.2118599023933687e-06, "loss": 0.039, "step": 16817 }, { "epoch": 0.7815055762081784, "grad_norm": 0.771431152925328, "learning_rate": 1.2113687899384879e-06, "loss": 0.0343, "step": 16818 }, { "epoch": 0.7815520446096654, "grad_norm": 0.7401559174690902, "learning_rate": 1.2108777633000907e-06, "loss": 0.0618, "step": 16819 }, { "epoch": 0.7815985130111525, "grad_norm": 0.621700415510666, "learning_rate": 1.210386822489303e-06, "loss": 0.0577, "step": 16820 }, { "epoch": 0.7816449814126394, "grad_norm": 0.5890810500001858, "learning_rate": 1.2098959675172406e-06, "loss": 0.0381, "step": 16821 }, { "epoch": 0.7816914498141264, "grad_norm": 0.24683370799780122, "learning_rate": 1.2094051983950223e-06, "loss": 0.0327, "step": 16822 }, { "epoch": 0.7817379182156133, "grad_norm": 0.5154936385626963, "learning_rate": 1.2089145151337673e-06, "loss": 0.0453, "step": 16823 }, { "epoch": 0.7817843866171004, "grad_norm": 0.3993558773183421, "learning_rate": 1.2084239177445883e-06, "loss": 0.0324, "step": 16824 }, { "epoch": 0.7818308550185874, "grad_norm": 0.5741516321098029, "learning_rate": 1.2079334062385994e-06, "loss": 0.0541, "step": 16825 }, { "epoch": 0.7818773234200743, "grad_norm": 0.5321897159172355, "learning_rate": 1.2074429806269079e-06, "loss": 0.0681, "step": 16826 }, { "epoch": 0.7819237918215614, "grad_norm": 0.39597081971544046, "learning_rate": 1.206952640920625e-06, "loss": 0.0377, "step": 16827 }, { "epoch": 0.7819702602230483, "grad_norm": 0.42967848622490673, "learning_rate": 1.2064623871308573e-06, "loss": 0.0417, "step": 16828 }, { "epoch": 0.7820167286245353, "grad_norm": 0.6034752881069058, "learning_rate": 1.2059722192687073e-06, "loss": 0.0504, "step": 16829 }, { "epoch": 0.7820631970260223, "grad_norm": 0.9120536012135856, "learning_rate": 1.2054821373452818e-06, "loss": 0.0507, "step": 16830 }, { "epoch": 0.7821096654275093, "grad_norm": 0.5594648802353215, "learning_rate": 1.2049921413716809e-06, "loss": 0.0581, "step": 16831 }, { "epoch": 0.7821561338289963, "grad_norm": 0.32743750965624363, "learning_rate": 1.2045022313589988e-06, "loss": 0.0486, "step": 16832 }, { "epoch": 0.7822026022304833, "grad_norm": 0.5182156038449875, "learning_rate": 1.2040124073183373e-06, "loss": 0.0552, "step": 16833 }, { "epoch": 0.7822490706319702, "grad_norm": 0.7011690792434804, "learning_rate": 1.2035226692607904e-06, "loss": 0.0607, "step": 16834 }, { "epoch": 0.7822955390334573, "grad_norm": 0.5323825207588491, "learning_rate": 1.2030330171974508e-06, "loss": 0.0621, "step": 16835 }, { "epoch": 0.7823420074349442, "grad_norm": 0.5018811385474404, "learning_rate": 1.202543451139408e-06, "loss": 0.0536, "step": 16836 }, { "epoch": 0.7823884758364312, "grad_norm": 0.4691639763678464, "learning_rate": 1.2020539710977553e-06, "loss": 0.0509, "step": 16837 }, { "epoch": 0.7824349442379183, "grad_norm": 0.4035045657897456, "learning_rate": 1.2015645770835765e-06, "loss": 0.0549, "step": 16838 }, { "epoch": 0.7824814126394052, "grad_norm": 0.4987836731843938, "learning_rate": 1.2010752691079558e-06, "loss": 0.0417, "step": 16839 }, { "epoch": 0.7825278810408922, "grad_norm": 0.8566931444865501, "learning_rate": 1.2005860471819803e-06, "loss": 0.0749, "step": 16840 }, { "epoch": 0.7825743494423791, "grad_norm": 0.8050297407249979, "learning_rate": 1.2000969113167298e-06, "loss": 0.0547, "step": 16841 }, { "epoch": 0.7826208178438662, "grad_norm": 0.5267407838259313, "learning_rate": 1.1996078615232838e-06, "loss": 0.0623, "step": 16842 }, { "epoch": 0.7826672862453532, "grad_norm": 0.5006721459409089, "learning_rate": 1.199118897812719e-06, "loss": 0.0553, "step": 16843 }, { "epoch": 0.7827137546468401, "grad_norm": 0.49807064934736955, "learning_rate": 1.1986300201961115e-06, "loss": 0.05, "step": 16844 }, { "epoch": 0.7827602230483272, "grad_norm": 0.9961183210017927, "learning_rate": 1.1981412286845357e-06, "loss": 0.0778, "step": 16845 }, { "epoch": 0.7828066914498141, "grad_norm": 0.3677055860069177, "learning_rate": 1.1976525232890602e-06, "loss": 0.0451, "step": 16846 }, { "epoch": 0.7828531598513011, "grad_norm": 0.5729261768289217, "learning_rate": 1.197163904020759e-06, "loss": 0.0617, "step": 16847 }, { "epoch": 0.7828996282527881, "grad_norm": 0.40530028258192385, "learning_rate": 1.196675370890697e-06, "loss": 0.056, "step": 16848 }, { "epoch": 0.7829460966542751, "grad_norm": 0.5387067968654953, "learning_rate": 1.1961869239099416e-06, "loss": 0.057, "step": 16849 }, { "epoch": 0.7829925650557621, "grad_norm": 0.5243388467731112, "learning_rate": 1.1956985630895551e-06, "loss": 0.0473, "step": 16850 }, { "epoch": 0.7830390334572491, "grad_norm": 0.4514182652134433, "learning_rate": 1.1952102884406008e-06, "loss": 0.0502, "step": 16851 }, { "epoch": 0.783085501858736, "grad_norm": 0.36634771061785243, "learning_rate": 1.1947220999741372e-06, "loss": 0.0353, "step": 16852 }, { "epoch": 0.7831319702602231, "grad_norm": 0.4623031173093132, "learning_rate": 1.1942339977012225e-06, "loss": 0.0479, "step": 16853 }, { "epoch": 0.78317843866171, "grad_norm": 0.31700411003730916, "learning_rate": 1.193745981632914e-06, "loss": 0.0536, "step": 16854 }, { "epoch": 0.783224907063197, "grad_norm": 0.27406096991825374, "learning_rate": 1.1932580517802667e-06, "loss": 0.0353, "step": 16855 }, { "epoch": 0.7832713754646841, "grad_norm": 0.45628499637079756, "learning_rate": 1.1927702081543279e-06, "loss": 0.0414, "step": 16856 }, { "epoch": 0.783317843866171, "grad_norm": 0.565322366032473, "learning_rate": 1.1922824507661524e-06, "loss": 0.0527, "step": 16857 }, { "epoch": 0.783364312267658, "grad_norm": 0.559279817718259, "learning_rate": 1.191794779626787e-06, "loss": 0.0766, "step": 16858 }, { "epoch": 0.7834107806691449, "grad_norm": 0.48305205572440413, "learning_rate": 1.1913071947472783e-06, "loss": 0.0442, "step": 16859 }, { "epoch": 0.783457249070632, "grad_norm": 0.36637597319194554, "learning_rate": 1.1908196961386682e-06, "loss": 0.04, "step": 16860 }, { "epoch": 0.783503717472119, "grad_norm": 0.26724110500399695, "learning_rate": 1.1903322838120047e-06, "loss": 0.0272, "step": 16861 }, { "epoch": 0.7835501858736059, "grad_norm": 0.8222563449469477, "learning_rate": 1.1898449577783222e-06, "loss": 0.0485, "step": 16862 }, { "epoch": 0.783596654275093, "grad_norm": 0.36931185491601126, "learning_rate": 1.1893577180486605e-06, "loss": 0.0335, "step": 16863 }, { "epoch": 0.7836431226765799, "grad_norm": 0.4709427227910504, "learning_rate": 1.1888705646340587e-06, "loss": 0.0337, "step": 16864 }, { "epoch": 0.7836895910780669, "grad_norm": 0.5828377157211442, "learning_rate": 1.1883834975455499e-06, "loss": 0.0607, "step": 16865 }, { "epoch": 0.7837360594795539, "grad_norm": 0.8067104494448764, "learning_rate": 1.1878965167941658e-06, "loss": 0.0608, "step": 16866 }, { "epoch": 0.7837825278810409, "grad_norm": 0.45936912248549855, "learning_rate": 1.1874096223909388e-06, "loss": 0.0514, "step": 16867 }, { "epoch": 0.7838289962825279, "grad_norm": 0.4557389596117851, "learning_rate": 1.1869228143468958e-06, "loss": 0.0444, "step": 16868 }, { "epoch": 0.7838754646840149, "grad_norm": 0.6539683031142606, "learning_rate": 1.1864360926730645e-06, "loss": 0.0557, "step": 16869 }, { "epoch": 0.7839219330855018, "grad_norm": 0.49587648773126725, "learning_rate": 1.1859494573804676e-06, "loss": 0.051, "step": 16870 }, { "epoch": 0.7839684014869889, "grad_norm": 0.5674037917721211, "learning_rate": 1.1854629084801318e-06, "loss": 0.0461, "step": 16871 }, { "epoch": 0.7840148698884758, "grad_norm": 0.3942787009986941, "learning_rate": 1.1849764459830754e-06, "loss": 0.0564, "step": 16872 }, { "epoch": 0.7840613382899628, "grad_norm": 0.6536432826708565, "learning_rate": 1.1844900699003176e-06, "loss": 0.0486, "step": 16873 }, { "epoch": 0.7841078066914499, "grad_norm": 0.41625979615097886, "learning_rate": 1.1840037802428756e-06, "loss": 0.0419, "step": 16874 }, { "epoch": 0.7841542750929368, "grad_norm": 0.6689343237314315, "learning_rate": 1.1835175770217644e-06, "loss": 0.0729, "step": 16875 }, { "epoch": 0.7842007434944238, "grad_norm": 0.5734640561305503, "learning_rate": 1.183031460247997e-06, "loss": 0.055, "step": 16876 }, { "epoch": 0.7842472118959107, "grad_norm": 0.6618353128475257, "learning_rate": 1.1825454299325823e-06, "loss": 0.0493, "step": 16877 }, { "epoch": 0.7842936802973978, "grad_norm": 0.6297623824664078, "learning_rate": 1.1820594860865332e-06, "loss": 0.0443, "step": 16878 }, { "epoch": 0.7843401486988848, "grad_norm": 0.47715247306923986, "learning_rate": 1.1815736287208567e-06, "loss": 0.0676, "step": 16879 }, { "epoch": 0.7843866171003717, "grad_norm": 0.6395721283277055, "learning_rate": 1.1810878578465528e-06, "loss": 0.0484, "step": 16880 }, { "epoch": 0.7844330855018588, "grad_norm": 0.356919406188264, "learning_rate": 1.1806021734746304e-06, "loss": 0.061, "step": 16881 }, { "epoch": 0.7844795539033457, "grad_norm": 0.8050556685342143, "learning_rate": 1.1801165756160882e-06, "loss": 0.0493, "step": 16882 }, { "epoch": 0.7845260223048327, "grad_norm": 0.4596305939635498, "learning_rate": 1.1796310642819264e-06, "loss": 0.0716, "step": 16883 }, { "epoch": 0.7845724907063197, "grad_norm": 0.5703838747367612, "learning_rate": 1.1791456394831414e-06, "loss": 0.031, "step": 16884 }, { "epoch": 0.7846189591078067, "grad_norm": 0.5038313184493171, "learning_rate": 1.1786603012307292e-06, "loss": 0.0417, "step": 16885 }, { "epoch": 0.7846654275092937, "grad_norm": 0.5864156083228405, "learning_rate": 1.1781750495356831e-06, "loss": 0.0332, "step": 16886 }, { "epoch": 0.7847118959107807, "grad_norm": 0.639075258114908, "learning_rate": 1.177689884408993e-06, "loss": 0.0681, "step": 16887 }, { "epoch": 0.7847583643122676, "grad_norm": 0.7696570630499677, "learning_rate": 1.1772048058616514e-06, "loss": 0.0665, "step": 16888 }, { "epoch": 0.7848048327137547, "grad_norm": 0.5550679468160348, "learning_rate": 1.1767198139046444e-06, "loss": 0.0498, "step": 16889 }, { "epoch": 0.7848513011152416, "grad_norm": 0.6679669527543528, "learning_rate": 1.1762349085489572e-06, "loss": 0.0575, "step": 16890 }, { "epoch": 0.7848977695167286, "grad_norm": 0.4171476173426253, "learning_rate": 1.175750089805574e-06, "loss": 0.0617, "step": 16891 }, { "epoch": 0.7849442379182157, "grad_norm": 0.30913701055615345, "learning_rate": 1.1752653576854767e-06, "loss": 0.0463, "step": 16892 }, { "epoch": 0.7849907063197026, "grad_norm": 0.41313358221468005, "learning_rate": 1.174780712199644e-06, "loss": 0.0441, "step": 16893 }, { "epoch": 0.7850371747211896, "grad_norm": 1.2404842814120571, "learning_rate": 1.1742961533590524e-06, "loss": 0.0714, "step": 16894 }, { "epoch": 0.7850836431226765, "grad_norm": 0.6256407696979409, "learning_rate": 1.1738116811746813e-06, "loss": 0.0625, "step": 16895 }, { "epoch": 0.7851301115241636, "grad_norm": 0.5631932691488475, "learning_rate": 1.1733272956575043e-06, "loss": 0.052, "step": 16896 }, { "epoch": 0.7851765799256506, "grad_norm": 0.6302355871784802, "learning_rate": 1.1728429968184885e-06, "loss": 0.065, "step": 16897 }, { "epoch": 0.7852230483271375, "grad_norm": 0.44476948256445964, "learning_rate": 1.172358784668608e-06, "loss": 0.0381, "step": 16898 }, { "epoch": 0.7852695167286246, "grad_norm": 0.5733336501730724, "learning_rate": 1.1718746592188295e-06, "loss": 0.0611, "step": 16899 }, { "epoch": 0.7853159851301115, "grad_norm": 0.5993041880585661, "learning_rate": 1.171390620480119e-06, "loss": 0.0516, "step": 16900 }, { "epoch": 0.7853624535315985, "grad_norm": 0.22123972840359354, "learning_rate": 1.1709066684634396e-06, "loss": 0.0318, "step": 16901 }, { "epoch": 0.7854089219330855, "grad_norm": 0.5764703664833694, "learning_rate": 1.1704228031797566e-06, "loss": 0.0596, "step": 16902 }, { "epoch": 0.7854553903345725, "grad_norm": 0.32221875881106726, "learning_rate": 1.169939024640026e-06, "loss": 0.0432, "step": 16903 }, { "epoch": 0.7855018587360595, "grad_norm": 0.8518748540345256, "learning_rate": 1.1694553328552065e-06, "loss": 0.0584, "step": 16904 }, { "epoch": 0.7855483271375465, "grad_norm": 0.36635955646831575, "learning_rate": 1.1689717278362568e-06, "loss": 0.0465, "step": 16905 }, { "epoch": 0.7855947955390334, "grad_norm": 0.3296519394563366, "learning_rate": 1.168488209594129e-06, "loss": 0.0349, "step": 16906 }, { "epoch": 0.7856412639405205, "grad_norm": 0.4817132880537157, "learning_rate": 1.1680047781397768e-06, "loss": 0.0417, "step": 16907 }, { "epoch": 0.7856877323420074, "grad_norm": 0.6139172035939118, "learning_rate": 1.167521433484149e-06, "loss": 0.0696, "step": 16908 }, { "epoch": 0.7857342007434944, "grad_norm": 0.6107510217373926, "learning_rate": 1.1670381756381943e-06, "loss": 0.0731, "step": 16909 }, { "epoch": 0.7857806691449815, "grad_norm": 0.43604901207642704, "learning_rate": 1.1665550046128589e-06, "loss": 0.0439, "step": 16910 }, { "epoch": 0.7858271375464684, "grad_norm": 0.3235937062834256, "learning_rate": 1.1660719204190856e-06, "loss": 0.0402, "step": 16911 }, { "epoch": 0.7858736059479554, "grad_norm": 0.7912819673788598, "learning_rate": 1.1655889230678203e-06, "loss": 0.0312, "step": 16912 }, { "epoch": 0.7859200743494423, "grad_norm": 0.6885714399946856, "learning_rate": 1.1651060125700014e-06, "loss": 0.054, "step": 16913 }, { "epoch": 0.7859665427509294, "grad_norm": 0.4403033421508728, "learning_rate": 1.1646231889365678e-06, "loss": 0.0484, "step": 16914 }, { "epoch": 0.7860130111524164, "grad_norm": 0.4994802329364719, "learning_rate": 1.1641404521784554e-06, "loss": 0.0547, "step": 16915 }, { "epoch": 0.7860594795539033, "grad_norm": 0.6830821322183572, "learning_rate": 1.1636578023065986e-06, "loss": 0.0611, "step": 16916 }, { "epoch": 0.7861059479553903, "grad_norm": 0.6100633834562363, "learning_rate": 1.1631752393319306e-06, "loss": 0.0426, "step": 16917 }, { "epoch": 0.7861524163568773, "grad_norm": 0.6209734982316942, "learning_rate": 1.16269276326538e-06, "loss": 0.0434, "step": 16918 }, { "epoch": 0.7861988847583643, "grad_norm": 0.20004053474943692, "learning_rate": 1.1622103741178786e-06, "loss": 0.0311, "step": 16919 }, { "epoch": 0.7862453531598513, "grad_norm": 0.47547719067325894, "learning_rate": 1.1617280719003531e-06, "loss": 0.0578, "step": 16920 }, { "epoch": 0.7862918215613383, "grad_norm": 0.5729418711108416, "learning_rate": 1.1612458566237234e-06, "loss": 0.051, "step": 16921 }, { "epoch": 0.7863382899628253, "grad_norm": 0.48856978882876506, "learning_rate": 1.1607637282989165e-06, "loss": 0.0512, "step": 16922 }, { "epoch": 0.7863847583643123, "grad_norm": 0.45018387036519847, "learning_rate": 1.160281686936852e-06, "loss": 0.0472, "step": 16923 }, { "epoch": 0.7864312267657992, "grad_norm": 0.4252557515705642, "learning_rate": 1.1597997325484483e-06, "loss": 0.0498, "step": 16924 }, { "epoch": 0.7864776951672863, "grad_norm": 0.620041823333607, "learning_rate": 1.1593178651446213e-06, "loss": 0.0466, "step": 16925 }, { "epoch": 0.7865241635687732, "grad_norm": 0.587995797944944, "learning_rate": 1.15883608473629e-06, "loss": 0.0538, "step": 16926 }, { "epoch": 0.7865706319702602, "grad_norm": 0.29679076827346734, "learning_rate": 1.158354391334362e-06, "loss": 0.0282, "step": 16927 }, { "epoch": 0.7866171003717473, "grad_norm": 0.38532230451259897, "learning_rate": 1.1578727849497496e-06, "loss": 0.0497, "step": 16928 }, { "epoch": 0.7866635687732342, "grad_norm": 0.5276349777311222, "learning_rate": 1.1573912655933634e-06, "loss": 0.0707, "step": 16929 }, { "epoch": 0.7867100371747212, "grad_norm": 0.5330723690904648, "learning_rate": 1.1569098332761091e-06, "loss": 0.0492, "step": 16930 }, { "epoch": 0.7867565055762081, "grad_norm": 0.3813247413971602, "learning_rate": 1.1564284880088922e-06, "loss": 0.0481, "step": 16931 }, { "epoch": 0.7868029739776952, "grad_norm": 0.6458819182342235, "learning_rate": 1.1559472298026153e-06, "loss": 0.0562, "step": 16932 }, { "epoch": 0.7868494423791822, "grad_norm": 0.7035878920660915, "learning_rate": 1.1554660586681793e-06, "loss": 0.0695, "step": 16933 }, { "epoch": 0.7868959107806691, "grad_norm": 0.6905775396874146, "learning_rate": 1.1549849746164838e-06, "loss": 0.0538, "step": 16934 }, { "epoch": 0.7869423791821561, "grad_norm": 0.5214665378835877, "learning_rate": 1.1545039776584233e-06, "loss": 0.0595, "step": 16935 }, { "epoch": 0.7869888475836431, "grad_norm": 0.6282187843335209, "learning_rate": 1.1540230678048969e-06, "loss": 0.0665, "step": 16936 }, { "epoch": 0.7870353159851301, "grad_norm": 0.5195020796743633, "learning_rate": 1.1535422450667954e-06, "loss": 0.0408, "step": 16937 }, { "epoch": 0.7870817843866171, "grad_norm": 0.6272245526571203, "learning_rate": 1.1530615094550107e-06, "loss": 0.0675, "step": 16938 }, { "epoch": 0.787128252788104, "grad_norm": 0.5127701689675985, "learning_rate": 1.1525808609804318e-06, "loss": 0.0579, "step": 16939 }, { "epoch": 0.7871747211895911, "grad_norm": 0.3903731016570958, "learning_rate": 1.1521002996539455e-06, "loss": 0.0443, "step": 16940 }, { "epoch": 0.7872211895910781, "grad_norm": 0.4104659284073329, "learning_rate": 1.1516198254864374e-06, "loss": 0.0567, "step": 16941 }, { "epoch": 0.787267657992565, "grad_norm": 0.5320299861679885, "learning_rate": 1.1511394384887893e-06, "loss": 0.0405, "step": 16942 }, { "epoch": 0.7873141263940521, "grad_norm": 0.32592240488380575, "learning_rate": 1.1506591386718863e-06, "loss": 0.035, "step": 16943 }, { "epoch": 0.787360594795539, "grad_norm": 0.45035751131716056, "learning_rate": 1.150178926046604e-06, "loss": 0.0449, "step": 16944 }, { "epoch": 0.787407063197026, "grad_norm": 0.591067358933578, "learning_rate": 1.149698800623819e-06, "loss": 0.0571, "step": 16945 }, { "epoch": 0.7874535315985131, "grad_norm": 0.45138486732394434, "learning_rate": 1.1492187624144103e-06, "loss": 0.0561, "step": 16946 }, { "epoch": 0.7875, "grad_norm": 0.4308029174256281, "learning_rate": 1.1487388114292498e-06, "loss": 0.0521, "step": 16947 }, { "epoch": 0.787546468401487, "grad_norm": 0.5056770503640763, "learning_rate": 1.1482589476792083e-06, "loss": 0.0506, "step": 16948 }, { "epoch": 0.7875929368029739, "grad_norm": 0.4715449973193994, "learning_rate": 1.1477791711751552e-06, "loss": 0.0511, "step": 16949 }, { "epoch": 0.787639405204461, "grad_norm": 0.5559467126449182, "learning_rate": 1.1472994819279586e-06, "loss": 0.0434, "step": 16950 }, { "epoch": 0.787685873605948, "grad_norm": 0.4786254262335093, "learning_rate": 1.1468198799484835e-06, "loss": 0.0339, "step": 16951 }, { "epoch": 0.7877323420074349, "grad_norm": 0.38929969881462567, "learning_rate": 1.1463403652475923e-06, "loss": 0.0482, "step": 16952 }, { "epoch": 0.787778810408922, "grad_norm": 0.37566610185034427, "learning_rate": 1.145860937836149e-06, "loss": 0.0373, "step": 16953 }, { "epoch": 0.7878252788104089, "grad_norm": 0.4967529516174673, "learning_rate": 1.145381597725012e-06, "loss": 0.0542, "step": 16954 }, { "epoch": 0.7878717472118959, "grad_norm": 0.4906436186615222, "learning_rate": 1.1449023449250384e-06, "loss": 0.0454, "step": 16955 }, { "epoch": 0.7879182156133829, "grad_norm": 0.5752085820660195, "learning_rate": 1.1444231794470844e-06, "loss": 0.0544, "step": 16956 }, { "epoch": 0.7879646840148699, "grad_norm": 0.4152111773913505, "learning_rate": 1.1439441013020036e-06, "loss": 0.0352, "step": 16957 }, { "epoch": 0.7880111524163569, "grad_norm": 0.2660790160582906, "learning_rate": 1.1434651105006467e-06, "loss": 0.0538, "step": 16958 }, { "epoch": 0.7880576208178439, "grad_norm": 0.3946619394978254, "learning_rate": 1.1429862070538628e-06, "loss": 0.065, "step": 16959 }, { "epoch": 0.7881040892193308, "grad_norm": 0.5683708018146658, "learning_rate": 1.142507390972502e-06, "loss": 0.0698, "step": 16960 }, { "epoch": 0.7881505576208179, "grad_norm": 0.559703817575029, "learning_rate": 1.1420286622674105e-06, "loss": 0.0534, "step": 16961 }, { "epoch": 0.7881970260223048, "grad_norm": 0.7292297819514602, "learning_rate": 1.1415500209494273e-06, "loss": 0.0611, "step": 16962 }, { "epoch": 0.7882434944237918, "grad_norm": 0.46903933429240413, "learning_rate": 1.1410714670293987e-06, "loss": 0.048, "step": 16963 }, { "epoch": 0.7882899628252789, "grad_norm": 0.3781899401812507, "learning_rate": 1.1405930005181626e-06, "loss": 0.0375, "step": 16964 }, { "epoch": 0.7883364312267658, "grad_norm": 0.3941689562314028, "learning_rate": 1.1401146214265568e-06, "loss": 0.0566, "step": 16965 }, { "epoch": 0.7883828996282528, "grad_norm": 0.4465378785040717, "learning_rate": 1.139636329765416e-06, "loss": 0.047, "step": 16966 }, { "epoch": 0.7884293680297397, "grad_norm": 0.3903593960557237, "learning_rate": 1.139158125545578e-06, "loss": 0.0373, "step": 16967 }, { "epoch": 0.7884758364312268, "grad_norm": 1.0840313453979895, "learning_rate": 1.1386800087778705e-06, "loss": 0.0701, "step": 16968 }, { "epoch": 0.7885223048327138, "grad_norm": 0.5496318198702091, "learning_rate": 1.1382019794731236e-06, "loss": 0.0375, "step": 16969 }, { "epoch": 0.7885687732342007, "grad_norm": 0.2806153067259539, "learning_rate": 1.1377240376421678e-06, "loss": 0.0324, "step": 16970 }, { "epoch": 0.7886152416356877, "grad_norm": 0.6444331885730268, "learning_rate": 1.1372461832958271e-06, "loss": 0.0495, "step": 16971 }, { "epoch": 0.7886617100371747, "grad_norm": 0.6853330617777506, "learning_rate": 1.1367684164449261e-06, "loss": 0.0637, "step": 16972 }, { "epoch": 0.7887081784386617, "grad_norm": 0.6470573954689325, "learning_rate": 1.1362907371002868e-06, "loss": 0.0515, "step": 16973 }, { "epoch": 0.7887546468401487, "grad_norm": 0.9244352391168607, "learning_rate": 1.1358131452727284e-06, "loss": 0.0677, "step": 16974 }, { "epoch": 0.7888011152416357, "grad_norm": 0.6132405679415381, "learning_rate": 1.1353356409730692e-06, "loss": 0.0549, "step": 16975 }, { "epoch": 0.7888475836431227, "grad_norm": 0.3871986461597346, "learning_rate": 1.1348582242121237e-06, "loss": 0.0486, "step": 16976 }, { "epoch": 0.7888940520446097, "grad_norm": 0.6392424397660822, "learning_rate": 1.1343808950007091e-06, "loss": 0.0562, "step": 16977 }, { "epoch": 0.7889405204460966, "grad_norm": 0.6178910054324767, "learning_rate": 1.1339036533496356e-06, "loss": 0.055, "step": 16978 }, { "epoch": 0.7889869888475837, "grad_norm": 0.5804097758615999, "learning_rate": 1.133426499269713e-06, "loss": 0.0736, "step": 16979 }, { "epoch": 0.7890334572490706, "grad_norm": 0.5207513640822539, "learning_rate": 1.1329494327717499e-06, "loss": 0.0546, "step": 16980 }, { "epoch": 0.7890799256505576, "grad_norm": 0.5766323705934256, "learning_rate": 1.1324724538665527e-06, "loss": 0.0456, "step": 16981 }, { "epoch": 0.7891263940520447, "grad_norm": 0.3749701251501953, "learning_rate": 1.1319955625649243e-06, "loss": 0.0416, "step": 16982 }, { "epoch": 0.7891728624535316, "grad_norm": 0.643057548823072, "learning_rate": 1.131518758877666e-06, "loss": 0.0558, "step": 16983 }, { "epoch": 0.7892193308550186, "grad_norm": 0.20023901441568165, "learning_rate": 1.131042042815581e-06, "loss": 0.025, "step": 16984 }, { "epoch": 0.7892657992565055, "grad_norm": 0.4899904441882602, "learning_rate": 1.1305654143894674e-06, "loss": 0.0427, "step": 16985 }, { "epoch": 0.7893122676579926, "grad_norm": 0.3787267463573658, "learning_rate": 1.1300888736101163e-06, "loss": 0.0544, "step": 16986 }, { "epoch": 0.7893587360594796, "grad_norm": 0.9860384398485001, "learning_rate": 1.1296124204883269e-06, "loss": 0.0737, "step": 16987 }, { "epoch": 0.7894052044609665, "grad_norm": 0.3855703868857891, "learning_rate": 1.1291360550348894e-06, "loss": 0.059, "step": 16988 }, { "epoch": 0.7894516728624535, "grad_norm": 0.44629476137362256, "learning_rate": 1.1286597772605945e-06, "loss": 0.0309, "step": 16989 }, { "epoch": 0.7894981412639405, "grad_norm": 0.4755066869492727, "learning_rate": 1.1281835871762287e-06, "loss": 0.0451, "step": 16990 }, { "epoch": 0.7895446096654275, "grad_norm": 2.842704440140078, "learning_rate": 1.1277074847925828e-06, "loss": 0.072, "step": 16991 }, { "epoch": 0.7895910780669145, "grad_norm": 0.5190442627204226, "learning_rate": 1.1272314701204362e-06, "loss": 0.0231, "step": 16992 }, { "epoch": 0.7896375464684015, "grad_norm": 0.44401707082473946, "learning_rate": 1.1267555431705712e-06, "loss": 0.0478, "step": 16993 }, { "epoch": 0.7896840148698885, "grad_norm": 0.5463995096400889, "learning_rate": 1.1262797039537715e-06, "loss": 0.0449, "step": 16994 }, { "epoch": 0.7897304832713755, "grad_norm": 0.4079597669904825, "learning_rate": 1.1258039524808128e-06, "loss": 0.0315, "step": 16995 }, { "epoch": 0.7897769516728624, "grad_norm": 0.39217215338519795, "learning_rate": 1.1253282887624728e-06, "loss": 0.0383, "step": 16996 }, { "epoch": 0.7898234200743495, "grad_norm": 0.454612136770145, "learning_rate": 1.1248527128095249e-06, "loss": 0.0585, "step": 16997 }, { "epoch": 0.7898698884758364, "grad_norm": 0.3552245823877165, "learning_rate": 1.1243772246327416e-06, "loss": 0.0426, "step": 16998 }, { "epoch": 0.7899163568773234, "grad_norm": 0.4201353245762706, "learning_rate": 1.1239018242428928e-06, "loss": 0.0431, "step": 16999 }, { "epoch": 0.7899628252788105, "grad_norm": 0.4206260134950193, "learning_rate": 1.1234265116507464e-06, "loss": 0.0444, "step": 17000 }, { "epoch": 0.7900092936802974, "grad_norm": 0.3655844488350739, "learning_rate": 1.1229512868670706e-06, "loss": 0.0512, "step": 17001 }, { "epoch": 0.7900557620817844, "grad_norm": 0.3201030607673429, "learning_rate": 1.1224761499026305e-06, "loss": 0.0458, "step": 17002 }, { "epoch": 0.7901022304832713, "grad_norm": 0.4126440646628742, "learning_rate": 1.122001100768183e-06, "loss": 0.0429, "step": 17003 }, { "epoch": 0.7901486988847584, "grad_norm": 0.5894429211100927, "learning_rate": 1.1215261394744941e-06, "loss": 0.0552, "step": 17004 }, { "epoch": 0.7901951672862454, "grad_norm": 0.6120996053480434, "learning_rate": 1.1210512660323198e-06, "loss": 0.0573, "step": 17005 }, { "epoch": 0.7902416356877323, "grad_norm": 0.46475303086839675, "learning_rate": 1.1205764804524172e-06, "loss": 0.0371, "step": 17006 }, { "epoch": 0.7902881040892193, "grad_norm": 0.7124366507222846, "learning_rate": 1.1201017827455385e-06, "loss": 0.0748, "step": 17007 }, { "epoch": 0.7903345724907063, "grad_norm": 0.46004286615059803, "learning_rate": 1.1196271729224413e-06, "loss": 0.0541, "step": 17008 }, { "epoch": 0.7903810408921933, "grad_norm": 0.5629292970863534, "learning_rate": 1.119152650993871e-06, "loss": 0.0519, "step": 17009 }, { "epoch": 0.7904275092936803, "grad_norm": 0.3818267852064896, "learning_rate": 1.1186782169705762e-06, "loss": 0.0293, "step": 17010 }, { "epoch": 0.7904739776951673, "grad_norm": 0.2989066459767591, "learning_rate": 1.1182038708633064e-06, "loss": 0.033, "step": 17011 }, { "epoch": 0.7905204460966543, "grad_norm": 0.5603848574594988, "learning_rate": 1.117729612682804e-06, "loss": 0.0536, "step": 17012 }, { "epoch": 0.7905669144981413, "grad_norm": 0.46572921584966026, "learning_rate": 1.1172554424398124e-06, "loss": 0.044, "step": 17013 }, { "epoch": 0.7906133828996282, "grad_norm": 0.5167505212153125, "learning_rate": 1.116781360145071e-06, "loss": 0.0467, "step": 17014 }, { "epoch": 0.7906598513011153, "grad_norm": 0.6737911404746779, "learning_rate": 1.1163073658093194e-06, "loss": 0.0541, "step": 17015 }, { "epoch": 0.7907063197026022, "grad_norm": 0.43584974607804866, "learning_rate": 1.115833459443293e-06, "loss": 0.0469, "step": 17016 }, { "epoch": 0.7907527881040892, "grad_norm": 0.45844462773419653, "learning_rate": 1.1153596410577255e-06, "loss": 0.0312, "step": 17017 }, { "epoch": 0.7907992565055763, "grad_norm": 0.303932036977708, "learning_rate": 1.114885910663352e-06, "loss": 0.0481, "step": 17018 }, { "epoch": 0.7908457249070632, "grad_norm": 0.46180730927277097, "learning_rate": 1.1144122682709018e-06, "loss": 0.04, "step": 17019 }, { "epoch": 0.7908921933085502, "grad_norm": 0.44697482688213525, "learning_rate": 1.1139387138911034e-06, "loss": 0.0431, "step": 17020 }, { "epoch": 0.7909386617100371, "grad_norm": 0.6212281101312433, "learning_rate": 1.113465247534683e-06, "loss": 0.0608, "step": 17021 }, { "epoch": 0.7909851301115242, "grad_norm": 0.5266729873521141, "learning_rate": 1.1129918692123647e-06, "loss": 0.0468, "step": 17022 }, { "epoch": 0.7910315985130112, "grad_norm": 0.640595696727512, "learning_rate": 1.112518578934872e-06, "loss": 0.0677, "step": 17023 }, { "epoch": 0.7910780669144981, "grad_norm": 0.6293361956032534, "learning_rate": 1.1120453767129237e-06, "loss": 0.0596, "step": 17024 }, { "epoch": 0.7911245353159851, "grad_norm": 0.4194398581599287, "learning_rate": 1.1115722625572412e-06, "loss": 0.0379, "step": 17025 }, { "epoch": 0.7911710037174721, "grad_norm": 0.5577362088602011, "learning_rate": 1.1110992364785406e-06, "loss": 0.0639, "step": 17026 }, { "epoch": 0.7912174721189591, "grad_norm": 0.4937514647107466, "learning_rate": 1.1106262984875322e-06, "loss": 0.0573, "step": 17027 }, { "epoch": 0.7912639405204461, "grad_norm": 1.0832736430432839, "learning_rate": 1.1101534485949334e-06, "loss": 0.0558, "step": 17028 }, { "epoch": 0.791310408921933, "grad_norm": 1.935129330318913, "learning_rate": 1.1096806868114529e-06, "loss": 0.0727, "step": 17029 }, { "epoch": 0.7913568773234201, "grad_norm": 0.33680713640163845, "learning_rate": 1.1092080131477988e-06, "loss": 0.0392, "step": 17030 }, { "epoch": 0.7914033457249071, "grad_norm": 0.8937161108561882, "learning_rate": 1.1087354276146771e-06, "loss": 0.0575, "step": 17031 }, { "epoch": 0.791449814126394, "grad_norm": 0.4558091665215756, "learning_rate": 1.108262930222797e-06, "loss": 0.0612, "step": 17032 }, { "epoch": 0.7914962825278811, "grad_norm": 0.5604791565491436, "learning_rate": 1.1077905209828555e-06, "loss": 0.0551, "step": 17033 }, { "epoch": 0.791542750929368, "grad_norm": 0.4750245887022065, "learning_rate": 1.1073181999055538e-06, "loss": 0.0684, "step": 17034 }, { "epoch": 0.791589219330855, "grad_norm": 0.7267062586723632, "learning_rate": 1.106845967001593e-06, "loss": 0.0594, "step": 17035 }, { "epoch": 0.791635687732342, "grad_norm": 0.6111293590419262, "learning_rate": 1.106373822281669e-06, "loss": 0.0402, "step": 17036 }, { "epoch": 0.791682156133829, "grad_norm": 0.538057507154445, "learning_rate": 1.1059017657564752e-06, "loss": 0.0357, "step": 17037 }, { "epoch": 0.791728624535316, "grad_norm": 0.47234668740773983, "learning_rate": 1.1054297974367056e-06, "loss": 0.0565, "step": 17038 }, { "epoch": 0.7917750929368029, "grad_norm": 0.36462387816022246, "learning_rate": 1.1049579173330499e-06, "loss": 0.0427, "step": 17039 }, { "epoch": 0.79182156133829, "grad_norm": 0.5760409176502451, "learning_rate": 1.1044861254561966e-06, "loss": 0.0413, "step": 17040 }, { "epoch": 0.791868029739777, "grad_norm": 0.5786917451226357, "learning_rate": 1.1040144218168314e-06, "loss": 0.055, "step": 17041 }, { "epoch": 0.7919144981412639, "grad_norm": 0.489864501400403, "learning_rate": 1.103542806425641e-06, "loss": 0.0622, "step": 17042 }, { "epoch": 0.7919609665427509, "grad_norm": 0.5993746740296981, "learning_rate": 1.1030712792933074e-06, "loss": 0.064, "step": 17043 }, { "epoch": 0.7920074349442379, "grad_norm": 0.46971322659298803, "learning_rate": 1.1025998404305105e-06, "loss": 0.0553, "step": 17044 }, { "epoch": 0.7920539033457249, "grad_norm": 0.4199864673245579, "learning_rate": 1.102128489847929e-06, "loss": 0.0462, "step": 17045 }, { "epoch": 0.7921003717472119, "grad_norm": 0.42454882003977157, "learning_rate": 1.10165722755624e-06, "loss": 0.0558, "step": 17046 }, { "epoch": 0.7921468401486988, "grad_norm": 0.5337776121263831, "learning_rate": 1.1011860535661173e-06, "loss": 0.0388, "step": 17047 }, { "epoch": 0.7921933085501859, "grad_norm": 0.3905246614530623, "learning_rate": 1.100714967888233e-06, "loss": 0.0507, "step": 17048 }, { "epoch": 0.7922397769516729, "grad_norm": 0.5573501801066436, "learning_rate": 1.1002439705332597e-06, "loss": 0.0546, "step": 17049 }, { "epoch": 0.7922862453531598, "grad_norm": 0.5610863799753162, "learning_rate": 1.0997730615118663e-06, "loss": 0.0486, "step": 17050 }, { "epoch": 0.7923327137546469, "grad_norm": 0.4002684132089797, "learning_rate": 1.0993022408347155e-06, "loss": 0.0453, "step": 17051 }, { "epoch": 0.7923791821561338, "grad_norm": 0.7377726397915763, "learning_rate": 1.0988315085124752e-06, "loss": 0.0568, "step": 17052 }, { "epoch": 0.7924256505576208, "grad_norm": 0.44670761858303554, "learning_rate": 1.0983608645558074e-06, "loss": 0.0638, "step": 17053 }, { "epoch": 0.7924721189591079, "grad_norm": 0.4884442484031818, "learning_rate": 1.097890308975373e-06, "loss": 0.0455, "step": 17054 }, { "epoch": 0.7925185873605948, "grad_norm": 0.6969288769312735, "learning_rate": 1.0974198417818277e-06, "loss": 0.0493, "step": 17055 }, { "epoch": 0.7925650557620818, "grad_norm": 0.4129829601065565, "learning_rate": 1.0969494629858336e-06, "loss": 0.0529, "step": 17056 }, { "epoch": 0.7926115241635687, "grad_norm": 0.4795205349892816, "learning_rate": 1.0964791725980407e-06, "loss": 0.0449, "step": 17057 }, { "epoch": 0.7926579925650558, "grad_norm": 0.5264855160624593, "learning_rate": 1.096008970629101e-06, "loss": 0.0666, "step": 17058 }, { "epoch": 0.7927044609665428, "grad_norm": 0.521467790475729, "learning_rate": 1.0955388570896692e-06, "loss": 0.059, "step": 17059 }, { "epoch": 0.7927509293680297, "grad_norm": 0.528040382995255, "learning_rate": 1.0950688319903912e-06, "loss": 0.0622, "step": 17060 }, { "epoch": 0.7927973977695167, "grad_norm": 0.410600114322796, "learning_rate": 1.094598895341914e-06, "loss": 0.0686, "step": 17061 }, { "epoch": 0.7928438661710037, "grad_norm": 0.5844464568459372, "learning_rate": 1.0941290471548821e-06, "loss": 0.0655, "step": 17062 }, { "epoch": 0.7928903345724907, "grad_norm": 0.6215825997164827, "learning_rate": 1.0936592874399381e-06, "loss": 0.0633, "step": 17063 }, { "epoch": 0.7929368029739777, "grad_norm": 0.6031572127949584, "learning_rate": 1.0931896162077232e-06, "loss": 0.0406, "step": 17064 }, { "epoch": 0.7929832713754646, "grad_norm": 1.3544660283448, "learning_rate": 1.092720033468873e-06, "loss": 0.0552, "step": 17065 }, { "epoch": 0.7930297397769517, "grad_norm": 0.4122561908747025, "learning_rate": 1.0922505392340288e-06, "loss": 0.0437, "step": 17066 }, { "epoch": 0.7930762081784387, "grad_norm": 0.41530412673626776, "learning_rate": 1.0917811335138235e-06, "loss": 0.0282, "step": 17067 }, { "epoch": 0.7931226765799256, "grad_norm": 0.5614430688074594, "learning_rate": 1.0913118163188862e-06, "loss": 0.0393, "step": 17068 }, { "epoch": 0.7931691449814127, "grad_norm": 0.8483210505899479, "learning_rate": 1.0908425876598512e-06, "loss": 0.0471, "step": 17069 }, { "epoch": 0.7932156133828996, "grad_norm": 0.6291096429353831, "learning_rate": 1.0903734475473464e-06, "loss": 0.0515, "step": 17070 }, { "epoch": 0.7932620817843866, "grad_norm": 0.5393059833758338, "learning_rate": 1.0899043959919974e-06, "loss": 0.0506, "step": 17071 }, { "epoch": 0.7933085501858737, "grad_norm": 0.5402528227005655, "learning_rate": 1.0894354330044283e-06, "loss": 0.0603, "step": 17072 }, { "epoch": 0.7933550185873606, "grad_norm": 0.5191483316395343, "learning_rate": 1.0889665585952646e-06, "loss": 0.0447, "step": 17073 }, { "epoch": 0.7934014869888476, "grad_norm": 0.4501603622660188, "learning_rate": 1.0884977727751235e-06, "loss": 0.0504, "step": 17074 }, { "epoch": 0.7934479553903345, "grad_norm": 0.3733584440628065, "learning_rate": 1.0880290755546231e-06, "loss": 0.044, "step": 17075 }, { "epoch": 0.7934944237918216, "grad_norm": 0.39836249763953824, "learning_rate": 1.0875604669443828e-06, "loss": 0.0254, "step": 17076 }, { "epoch": 0.7935408921933086, "grad_norm": 0.8299757242747665, "learning_rate": 1.0870919469550156e-06, "loss": 0.0421, "step": 17077 }, { "epoch": 0.7935873605947955, "grad_norm": 0.6392564564554186, "learning_rate": 1.0866235155971345e-06, "loss": 0.0413, "step": 17078 }, { "epoch": 0.7936338289962825, "grad_norm": 0.5954705704056422, "learning_rate": 1.086155172881349e-06, "loss": 0.0627, "step": 17079 }, { "epoch": 0.7936802973977695, "grad_norm": 0.5052951938534905, "learning_rate": 1.085686918818269e-06, "loss": 0.0596, "step": 17080 }, { "epoch": 0.7937267657992565, "grad_norm": 0.4963073274967, "learning_rate": 1.0852187534184994e-06, "loss": 0.0508, "step": 17081 }, { "epoch": 0.7937732342007435, "grad_norm": 0.4030261105616868, "learning_rate": 1.0847506766926436e-06, "loss": 0.0523, "step": 17082 }, { "epoch": 0.7938197026022304, "grad_norm": 0.5378645834123164, "learning_rate": 1.0842826886513074e-06, "loss": 0.0397, "step": 17083 }, { "epoch": 0.7938661710037175, "grad_norm": 0.3120411614979716, "learning_rate": 1.0838147893050893e-06, "loss": 0.0318, "step": 17084 }, { "epoch": 0.7939126394052045, "grad_norm": 0.5605255454700938, "learning_rate": 1.0833469786645884e-06, "loss": 0.0522, "step": 17085 }, { "epoch": 0.7939591078066914, "grad_norm": 0.5437885315021797, "learning_rate": 1.0828792567404001e-06, "loss": 0.0567, "step": 17086 }, { "epoch": 0.7940055762081785, "grad_norm": 0.6067949204575949, "learning_rate": 1.0824116235431198e-06, "loss": 0.0578, "step": 17087 }, { "epoch": 0.7940520446096654, "grad_norm": 0.5148479050712722, "learning_rate": 1.081944079083339e-06, "loss": 0.0609, "step": 17088 }, { "epoch": 0.7940985130111524, "grad_norm": 0.5307934012926221, "learning_rate": 1.0814766233716473e-06, "loss": 0.048, "step": 17089 }, { "epoch": 0.7941449814126395, "grad_norm": 0.3345960126539922, "learning_rate": 1.0810092564186359e-06, "loss": 0.047, "step": 17090 }, { "epoch": 0.7941914498141264, "grad_norm": 0.5140145447756717, "learning_rate": 1.0805419782348902e-06, "loss": 0.0506, "step": 17091 }, { "epoch": 0.7942379182156134, "grad_norm": 0.5163327444414683, "learning_rate": 1.0800747888309915e-06, "loss": 0.0448, "step": 17092 }, { "epoch": 0.7942843866171003, "grad_norm": 0.5962635508446676, "learning_rate": 1.079607688217526e-06, "loss": 0.0514, "step": 17093 }, { "epoch": 0.7943308550185874, "grad_norm": 0.5894801066292154, "learning_rate": 1.0791406764050727e-06, "loss": 0.0622, "step": 17094 }, { "epoch": 0.7943773234200744, "grad_norm": 0.563329516322362, "learning_rate": 1.0786737534042091e-06, "loss": 0.0671, "step": 17095 }, { "epoch": 0.7944237918215613, "grad_norm": 0.47881430734850683, "learning_rate": 1.0782069192255112e-06, "loss": 0.0545, "step": 17096 }, { "epoch": 0.7944702602230483, "grad_norm": 0.5955110884361904, "learning_rate": 1.0777401738795567e-06, "loss": 0.0494, "step": 17097 }, { "epoch": 0.7945167286245353, "grad_norm": 0.6576539614979244, "learning_rate": 1.0772735173769138e-06, "loss": 0.0669, "step": 17098 }, { "epoch": 0.7945631970260223, "grad_norm": 0.42049100028943, "learning_rate": 1.0768069497281537e-06, "loss": 0.0434, "step": 17099 }, { "epoch": 0.7946096654275093, "grad_norm": 0.3943960224728292, "learning_rate": 1.0763404709438457e-06, "loss": 0.032, "step": 17100 }, { "epoch": 0.7946561338289962, "grad_norm": 0.6570814246199594, "learning_rate": 1.0758740810345564e-06, "loss": 0.0438, "step": 17101 }, { "epoch": 0.7947026022304833, "grad_norm": 0.5506340223130741, "learning_rate": 1.0754077800108488e-06, "loss": 0.0399, "step": 17102 }, { "epoch": 0.7947490706319703, "grad_norm": 0.4525838626614073, "learning_rate": 1.0749415678832853e-06, "loss": 0.065, "step": 17103 }, { "epoch": 0.7947955390334572, "grad_norm": 0.49184071294438225, "learning_rate": 1.0744754446624267e-06, "loss": 0.0492, "step": 17104 }, { "epoch": 0.7948420074349443, "grad_norm": 0.4322699160643819, "learning_rate": 1.0740094103588312e-06, "loss": 0.0589, "step": 17105 }, { "epoch": 0.7948884758364312, "grad_norm": 0.5436007803295979, "learning_rate": 1.0735434649830523e-06, "loss": 0.0692, "step": 17106 }, { "epoch": 0.7949349442379182, "grad_norm": 0.31506697304230363, "learning_rate": 1.0730776085456485e-06, "loss": 0.043, "step": 17107 }, { "epoch": 0.7949814126394052, "grad_norm": 0.41428542121024065, "learning_rate": 1.0726118410571696e-06, "loss": 0.0354, "step": 17108 }, { "epoch": 0.7950278810408922, "grad_norm": 0.47070325747693814, "learning_rate": 1.072146162528166e-06, "loss": 0.0482, "step": 17109 }, { "epoch": 0.7950743494423792, "grad_norm": 0.3295936822681078, "learning_rate": 1.071680572969186e-06, "loss": 0.0365, "step": 17110 }, { "epoch": 0.7951208178438661, "grad_norm": 0.3186787707228778, "learning_rate": 1.0712150723907755e-06, "loss": 0.0405, "step": 17111 }, { "epoch": 0.7951672862453532, "grad_norm": 0.7399737102047695, "learning_rate": 1.0707496608034784e-06, "loss": 0.0567, "step": 17112 }, { "epoch": 0.7952137546468402, "grad_norm": 0.5334636620583322, "learning_rate": 1.0702843382178358e-06, "loss": 0.0459, "step": 17113 }, { "epoch": 0.7952602230483271, "grad_norm": 0.5157499345100844, "learning_rate": 1.06981910464439e-06, "loss": 0.0566, "step": 17114 }, { "epoch": 0.7953066914498141, "grad_norm": 0.7438066690820075, "learning_rate": 1.0693539600936792e-06, "loss": 0.0656, "step": 17115 }, { "epoch": 0.7953531598513011, "grad_norm": 0.41444178364772033, "learning_rate": 1.0688889045762357e-06, "loss": 0.0541, "step": 17116 }, { "epoch": 0.7953996282527881, "grad_norm": 0.38019951979872735, "learning_rate": 1.0684239381025973e-06, "loss": 0.054, "step": 17117 }, { "epoch": 0.7954460966542751, "grad_norm": 0.39101045939999396, "learning_rate": 1.0679590606832946e-06, "loss": 0.0365, "step": 17118 }, { "epoch": 0.795492565055762, "grad_norm": 0.6854587228342632, "learning_rate": 1.0674942723288573e-06, "loss": 0.0696, "step": 17119 }, { "epoch": 0.7955390334572491, "grad_norm": 0.442180291391586, "learning_rate": 1.0670295730498137e-06, "loss": 0.0308, "step": 17120 }, { "epoch": 0.7955855018587361, "grad_norm": 0.4483352692455668, "learning_rate": 1.0665649628566898e-06, "loss": 0.0489, "step": 17121 }, { "epoch": 0.795631970260223, "grad_norm": 0.45941542210645503, "learning_rate": 1.0661004417600097e-06, "loss": 0.0543, "step": 17122 }, { "epoch": 0.7956784386617101, "grad_norm": 0.5946558685923011, "learning_rate": 1.0656360097702927e-06, "loss": 0.0406, "step": 17123 }, { "epoch": 0.795724907063197, "grad_norm": 2.0510364089256736, "learning_rate": 1.0651716668980628e-06, "loss": 0.0857, "step": 17124 }, { "epoch": 0.795771375464684, "grad_norm": 0.5749143416058337, "learning_rate": 1.064707413153836e-06, "loss": 0.0639, "step": 17125 }, { "epoch": 0.795817843866171, "grad_norm": 0.4863879957923312, "learning_rate": 1.064243248548128e-06, "loss": 0.0539, "step": 17126 }, { "epoch": 0.795864312267658, "grad_norm": 0.5454347357641335, "learning_rate": 1.0637791730914527e-06, "loss": 0.0477, "step": 17127 }, { "epoch": 0.795910780669145, "grad_norm": 0.49913875431166477, "learning_rate": 1.063315186794322e-06, "loss": 0.0484, "step": 17128 }, { "epoch": 0.7959572490706319, "grad_norm": 0.4753888473631388, "learning_rate": 1.062851289667246e-06, "loss": 0.0376, "step": 17129 }, { "epoch": 0.796003717472119, "grad_norm": 0.8196643506021332, "learning_rate": 1.0623874817207302e-06, "loss": 0.056, "step": 17130 }, { "epoch": 0.796050185873606, "grad_norm": 0.3748415595221181, "learning_rate": 1.0619237629652834e-06, "loss": 0.0462, "step": 17131 }, { "epoch": 0.7960966542750929, "grad_norm": 0.617857678611741, "learning_rate": 1.0614601334114099e-06, "loss": 0.0336, "step": 17132 }, { "epoch": 0.7961431226765799, "grad_norm": 0.4108318203168924, "learning_rate": 1.060996593069607e-06, "loss": 0.0322, "step": 17133 }, { "epoch": 0.7961895910780669, "grad_norm": 0.7399361988431669, "learning_rate": 1.0605331419503778e-06, "loss": 0.0508, "step": 17134 }, { "epoch": 0.7962360594795539, "grad_norm": 0.6802648392083158, "learning_rate": 1.0600697800642196e-06, "loss": 0.0523, "step": 17135 }, { "epoch": 0.7962825278810409, "grad_norm": 0.3645518413396003, "learning_rate": 1.0596065074216271e-06, "loss": 0.0292, "step": 17136 }, { "epoch": 0.7963289962825278, "grad_norm": 0.4394441358590219, "learning_rate": 1.059143324033094e-06, "loss": 0.056, "step": 17137 }, { "epoch": 0.7963754646840149, "grad_norm": 0.5590800290985598, "learning_rate": 1.0586802299091142e-06, "loss": 0.0493, "step": 17138 }, { "epoch": 0.7964219330855019, "grad_norm": 0.5239990604860804, "learning_rate": 1.058217225060174e-06, "loss": 0.0514, "step": 17139 }, { "epoch": 0.7964684014869888, "grad_norm": 0.4235237351759389, "learning_rate": 1.0577543094967613e-06, "loss": 0.0487, "step": 17140 }, { "epoch": 0.7965148698884759, "grad_norm": 0.40030257623192966, "learning_rate": 1.0572914832293635e-06, "loss": 0.0527, "step": 17141 }, { "epoch": 0.7965613382899628, "grad_norm": 5.699718202842388, "learning_rate": 1.056828746268464e-06, "loss": 0.0595, "step": 17142 }, { "epoch": 0.7966078066914498, "grad_norm": 0.2854403488955431, "learning_rate": 1.0563660986245428e-06, "loss": 0.0442, "step": 17143 }, { "epoch": 0.7966542750929368, "grad_norm": 0.433649710457334, "learning_rate": 1.0559035403080803e-06, "loss": 0.0509, "step": 17144 }, { "epoch": 0.7967007434944238, "grad_norm": 0.633414810694478, "learning_rate": 1.0554410713295537e-06, "loss": 0.0515, "step": 17145 }, { "epoch": 0.7967472118959108, "grad_norm": 0.2869554107976648, "learning_rate": 1.0549786916994387e-06, "loss": 0.0393, "step": 17146 }, { "epoch": 0.7967936802973977, "grad_norm": 0.4078276773721264, "learning_rate": 1.0545164014282066e-06, "loss": 0.0336, "step": 17147 }, { "epoch": 0.7968401486988848, "grad_norm": 0.5067406767839568, "learning_rate": 1.0540542005263322e-06, "loss": 0.0334, "step": 17148 }, { "epoch": 0.7968866171003718, "grad_norm": 0.4240686284895275, "learning_rate": 1.0535920890042828e-06, "loss": 0.0518, "step": 17149 }, { "epoch": 0.7969330855018587, "grad_norm": 0.4706368277545722, "learning_rate": 1.0531300668725265e-06, "loss": 0.0428, "step": 17150 }, { "epoch": 0.7969795539033457, "grad_norm": 0.3786131160291249, "learning_rate": 1.052668134141528e-06, "loss": 0.0502, "step": 17151 }, { "epoch": 0.7970260223048327, "grad_norm": 0.7681957836634102, "learning_rate": 1.0522062908217507e-06, "loss": 0.0497, "step": 17152 }, { "epoch": 0.7970724907063197, "grad_norm": 0.7022336064182753, "learning_rate": 1.0517445369236562e-06, "loss": 0.0513, "step": 17153 }, { "epoch": 0.7971189591078067, "grad_norm": 0.3969911931500145, "learning_rate": 1.0512828724577018e-06, "loss": 0.0588, "step": 17154 }, { "epoch": 0.7971654275092936, "grad_norm": 0.6324781366308025, "learning_rate": 1.0508212974343473e-06, "loss": 0.0698, "step": 17155 }, { "epoch": 0.7972118959107807, "grad_norm": 0.5420684102290436, "learning_rate": 1.050359811864049e-06, "loss": 0.0623, "step": 17156 }, { "epoch": 0.7972583643122677, "grad_norm": 0.44197445064433594, "learning_rate": 1.0498984157572544e-06, "loss": 0.0496, "step": 17157 }, { "epoch": 0.7973048327137546, "grad_norm": 0.6433430489487839, "learning_rate": 1.04943710912442e-06, "loss": 0.0512, "step": 17158 }, { "epoch": 0.7973513011152417, "grad_norm": 0.3902810189976694, "learning_rate": 1.048975891975993e-06, "loss": 0.0538, "step": 17159 }, { "epoch": 0.7973977695167286, "grad_norm": 0.5016020650960321, "learning_rate": 1.0485147643224203e-06, "loss": 0.0513, "step": 17160 }, { "epoch": 0.7974442379182156, "grad_norm": 0.462406413557812, "learning_rate": 1.0480537261741453e-06, "loss": 0.0541, "step": 17161 }, { "epoch": 0.7974907063197026, "grad_norm": 0.33532905866287316, "learning_rate": 1.0475927775416156e-06, "loss": 0.0451, "step": 17162 }, { "epoch": 0.7975371747211896, "grad_norm": 0.4832957224187967, "learning_rate": 1.0471319184352685e-06, "loss": 0.0554, "step": 17163 }, { "epoch": 0.7975836431226766, "grad_norm": 0.5040343077723752, "learning_rate": 1.0466711488655417e-06, "loss": 0.0471, "step": 17164 }, { "epoch": 0.7976301115241635, "grad_norm": 0.2792931299588869, "learning_rate": 1.0462104688428758e-06, "loss": 0.0383, "step": 17165 }, { "epoch": 0.7976765799256506, "grad_norm": 0.6143699185732453, "learning_rate": 1.0457498783777042e-06, "loss": 0.0539, "step": 17166 }, { "epoch": 0.7977230483271376, "grad_norm": 0.3851262797905392, "learning_rate": 1.0452893774804596e-06, "loss": 0.0494, "step": 17167 }, { "epoch": 0.7977695167286245, "grad_norm": 0.5401868324312704, "learning_rate": 1.044828966161573e-06, "loss": 0.0607, "step": 17168 }, { "epoch": 0.7978159851301115, "grad_norm": 0.5033358259176003, "learning_rate": 1.0443686444314727e-06, "loss": 0.0391, "step": 17169 }, { "epoch": 0.7978624535315985, "grad_norm": 0.4510685780279842, "learning_rate": 1.043908412300586e-06, "loss": 0.0452, "step": 17170 }, { "epoch": 0.7979089219330855, "grad_norm": 0.6303862479161295, "learning_rate": 1.0434482697793363e-06, "loss": 0.0626, "step": 17171 }, { "epoch": 0.7979553903345725, "grad_norm": 0.4382184747285179, "learning_rate": 1.0429882168781485e-06, "loss": 0.0277, "step": 17172 }, { "epoch": 0.7980018587360594, "grad_norm": 0.5271498887025726, "learning_rate": 1.0425282536074428e-06, "loss": 0.0529, "step": 17173 }, { "epoch": 0.7980483271375465, "grad_norm": 0.3118039685948736, "learning_rate": 1.0420683799776366e-06, "loss": 0.0322, "step": 17174 }, { "epoch": 0.7980947955390335, "grad_norm": 0.531846671419633, "learning_rate": 1.0416085959991478e-06, "loss": 0.0414, "step": 17175 }, { "epoch": 0.7981412639405204, "grad_norm": 0.8613432959077753, "learning_rate": 1.04114890168239e-06, "loss": 0.0388, "step": 17176 }, { "epoch": 0.7981877323420075, "grad_norm": 0.5674712528031244, "learning_rate": 1.0406892970377763e-06, "loss": 0.0447, "step": 17177 }, { "epoch": 0.7982342007434944, "grad_norm": 0.4967030603766091, "learning_rate": 1.0402297820757163e-06, "loss": 0.0661, "step": 17178 }, { "epoch": 0.7982806691449814, "grad_norm": 0.6392216850557642, "learning_rate": 1.0397703568066214e-06, "loss": 0.0816, "step": 17179 }, { "epoch": 0.7983271375464684, "grad_norm": 0.48199322620825547, "learning_rate": 1.0393110212408947e-06, "loss": 0.0413, "step": 17180 }, { "epoch": 0.7983736059479554, "grad_norm": 0.47013020906703373, "learning_rate": 1.038851775388941e-06, "loss": 0.0505, "step": 17181 }, { "epoch": 0.7984200743494424, "grad_norm": 0.48192949432815885, "learning_rate": 1.038392619261165e-06, "loss": 0.0423, "step": 17182 }, { "epoch": 0.7984665427509293, "grad_norm": 0.6848039501593688, "learning_rate": 1.0379335528679652e-06, "loss": 0.0701, "step": 17183 }, { "epoch": 0.7985130111524164, "grad_norm": 0.4213206512890819, "learning_rate": 1.0374745762197402e-06, "loss": 0.0582, "step": 17184 }, { "epoch": 0.7985594795539034, "grad_norm": 0.41177465323262924, "learning_rate": 1.0370156893268874e-06, "loss": 0.0547, "step": 17185 }, { "epoch": 0.7986059479553903, "grad_norm": 0.7327251327351382, "learning_rate": 1.0365568921997993e-06, "loss": 0.0736, "step": 17186 }, { "epoch": 0.7986524163568773, "grad_norm": 0.5850325797655048, "learning_rate": 1.0360981848488694e-06, "loss": 0.0596, "step": 17187 }, { "epoch": 0.7986988847583643, "grad_norm": 0.3757694317469318, "learning_rate": 1.0356395672844865e-06, "loss": 0.0387, "step": 17188 }, { "epoch": 0.7987453531598513, "grad_norm": 0.7850088290607313, "learning_rate": 1.0351810395170408e-06, "loss": 0.0723, "step": 17189 }, { "epoch": 0.7987918215613383, "grad_norm": 0.5960250006523286, "learning_rate": 1.0347226015569178e-06, "loss": 0.0516, "step": 17190 }, { "epoch": 0.7988382899628252, "grad_norm": 0.4400801106728091, "learning_rate": 1.0342642534145009e-06, "loss": 0.0407, "step": 17191 }, { "epoch": 0.7988847583643123, "grad_norm": 0.5193854874755202, "learning_rate": 1.0338059951001729e-06, "loss": 0.0687, "step": 17192 }, { "epoch": 0.7989312267657993, "grad_norm": 0.6461702676863161, "learning_rate": 1.033347826624313e-06, "loss": 0.0571, "step": 17193 }, { "epoch": 0.7989776951672862, "grad_norm": 0.637168034851779, "learning_rate": 1.0328897479972998e-06, "loss": 0.049, "step": 17194 }, { "epoch": 0.7990241635687733, "grad_norm": 0.5047690271064356, "learning_rate": 1.032431759229508e-06, "loss": 0.0682, "step": 17195 }, { "epoch": 0.7990706319702602, "grad_norm": 0.6799809376184647, "learning_rate": 1.031973860331314e-06, "loss": 0.0518, "step": 17196 }, { "epoch": 0.7991171003717472, "grad_norm": 0.4660429842011808, "learning_rate": 1.0315160513130889e-06, "loss": 0.0381, "step": 17197 }, { "epoch": 0.7991635687732342, "grad_norm": 0.29164723210042154, "learning_rate": 1.0310583321852002e-06, "loss": 0.032, "step": 17198 }, { "epoch": 0.7992100371747212, "grad_norm": 0.5220939851269641, "learning_rate": 1.0306007029580183e-06, "loss": 0.0515, "step": 17199 }, { "epoch": 0.7992565055762082, "grad_norm": 0.46004973787982784, "learning_rate": 1.0301431636419085e-06, "loss": 0.0524, "step": 17200 }, { "epoch": 0.7993029739776951, "grad_norm": 0.4725594567105585, "learning_rate": 1.0296857142472344e-06, "loss": 0.0432, "step": 17201 }, { "epoch": 0.7993494423791822, "grad_norm": 0.460276453701274, "learning_rate": 1.0292283547843563e-06, "loss": 0.0539, "step": 17202 }, { "epoch": 0.7993959107806692, "grad_norm": 0.4492198546677331, "learning_rate": 1.0287710852636373e-06, "loss": 0.0566, "step": 17203 }, { "epoch": 0.7994423791821561, "grad_norm": 0.3662287710252251, "learning_rate": 1.0283139056954317e-06, "loss": 0.0383, "step": 17204 }, { "epoch": 0.7994888475836431, "grad_norm": 0.3232767081054963, "learning_rate": 1.0278568160900949e-06, "loss": 0.0377, "step": 17205 }, { "epoch": 0.7995353159851301, "grad_norm": 0.5300618445244253, "learning_rate": 1.0273998164579834e-06, "loss": 0.0453, "step": 17206 }, { "epoch": 0.7995817843866171, "grad_norm": 0.576159128070359, "learning_rate": 1.0269429068094466e-06, "loss": 0.0513, "step": 17207 }, { "epoch": 0.7996282527881041, "grad_norm": 0.702458458268541, "learning_rate": 1.026486087154835e-06, "loss": 0.0679, "step": 17208 }, { "epoch": 0.799674721189591, "grad_norm": 0.357046886610302, "learning_rate": 1.0260293575044955e-06, "loss": 0.052, "step": 17209 }, { "epoch": 0.7997211895910781, "grad_norm": 0.5011774396773372, "learning_rate": 1.0255727178687735e-06, "loss": 0.0637, "step": 17210 }, { "epoch": 0.7997676579925651, "grad_norm": 0.40089970252230844, "learning_rate": 1.0251161682580125e-06, "loss": 0.0492, "step": 17211 }, { "epoch": 0.799814126394052, "grad_norm": 0.3612363049365153, "learning_rate": 1.0246597086825528e-06, "loss": 0.0474, "step": 17212 }, { "epoch": 0.7998605947955391, "grad_norm": 0.4537273192123256, "learning_rate": 1.0242033391527357e-06, "loss": 0.0411, "step": 17213 }, { "epoch": 0.799907063197026, "grad_norm": 0.4091726486322954, "learning_rate": 1.0237470596788967e-06, "loss": 0.0405, "step": 17214 }, { "epoch": 0.799953531598513, "grad_norm": 0.39956254177067246, "learning_rate": 1.0232908702713728e-06, "loss": 0.0326, "step": 17215 }, { "epoch": 0.8, "grad_norm": 0.7113303382506745, "learning_rate": 1.0228347709404957e-06, "loss": 0.0515, "step": 17216 }, { "epoch": 0.800046468401487, "grad_norm": 0.3900612412472651, "learning_rate": 1.0223787616965968e-06, "loss": 0.0526, "step": 17217 }, { "epoch": 0.800092936802974, "grad_norm": 0.3630009230700556, "learning_rate": 1.0219228425500056e-06, "loss": 0.0529, "step": 17218 }, { "epoch": 0.8001394052044609, "grad_norm": 0.5166511421643315, "learning_rate": 1.0214670135110471e-06, "loss": 0.047, "step": 17219 }, { "epoch": 0.800185873605948, "grad_norm": 0.6314310248688887, "learning_rate": 1.0210112745900496e-06, "loss": 0.0512, "step": 17220 }, { "epoch": 0.800232342007435, "grad_norm": 0.4639828171823272, "learning_rate": 1.020555625797336e-06, "loss": 0.0451, "step": 17221 }, { "epoch": 0.8002788104089219, "grad_norm": 0.36522274568368407, "learning_rate": 1.0201000671432232e-06, "loss": 0.0459, "step": 17222 }, { "epoch": 0.8003252788104089, "grad_norm": 0.45423877892836007, "learning_rate": 1.0196445986380339e-06, "loss": 0.0515, "step": 17223 }, { "epoch": 0.8003717472118959, "grad_norm": 0.6333453950134708, "learning_rate": 1.0191892202920834e-06, "loss": 0.0518, "step": 17224 }, { "epoch": 0.8004182156133829, "grad_norm": 0.27033217680460103, "learning_rate": 1.0187339321156868e-06, "loss": 0.0444, "step": 17225 }, { "epoch": 0.8004646840148699, "grad_norm": 0.4757758220867701, "learning_rate": 1.0182787341191557e-06, "loss": 0.0436, "step": 17226 }, { "epoch": 0.8005111524163568, "grad_norm": 0.4874103105310611, "learning_rate": 1.017823626312805e-06, "loss": 0.0449, "step": 17227 }, { "epoch": 0.8005576208178439, "grad_norm": 0.5912162789819205, "learning_rate": 1.0173686087069384e-06, "loss": 0.0449, "step": 17228 }, { "epoch": 0.8006040892193309, "grad_norm": 0.6276807375497252, "learning_rate": 1.0169136813118635e-06, "loss": 0.0694, "step": 17229 }, { "epoch": 0.8006505576208178, "grad_norm": 0.6226759669594047, "learning_rate": 1.016458844137887e-06, "loss": 0.049, "step": 17230 }, { "epoch": 0.8006970260223049, "grad_norm": 0.5896156745216342, "learning_rate": 1.0160040971953106e-06, "loss": 0.0582, "step": 17231 }, { "epoch": 0.8007434944237918, "grad_norm": 0.6328352280025835, "learning_rate": 1.015549440494434e-06, "loss": 0.0684, "step": 17232 }, { "epoch": 0.8007899628252788, "grad_norm": 0.3310434029696532, "learning_rate": 1.0150948740455563e-06, "loss": 0.0327, "step": 17233 }, { "epoch": 0.8008364312267658, "grad_norm": 0.4554527203589746, "learning_rate": 1.0146403978589737e-06, "loss": 0.0437, "step": 17234 }, { "epoch": 0.8008828996282528, "grad_norm": 0.4469030537086313, "learning_rate": 1.0141860119449808e-06, "loss": 0.0354, "step": 17235 }, { "epoch": 0.8009293680297398, "grad_norm": 0.5307984996095974, "learning_rate": 1.0137317163138683e-06, "loss": 0.0428, "step": 17236 }, { "epoch": 0.8009758364312267, "grad_norm": 0.46363792821555655, "learning_rate": 1.0132775109759292e-06, "loss": 0.0528, "step": 17237 }, { "epoch": 0.8010223048327137, "grad_norm": 0.5914129340421576, "learning_rate": 1.0128233959414512e-06, "loss": 0.0547, "step": 17238 }, { "epoch": 0.8010687732342008, "grad_norm": 0.4508793354276931, "learning_rate": 1.0123693712207177e-06, "loss": 0.043, "step": 17239 }, { "epoch": 0.8011152416356877, "grad_norm": 0.5752885611963939, "learning_rate": 1.0119154368240164e-06, "loss": 0.0455, "step": 17240 }, { "epoch": 0.8011617100371747, "grad_norm": 0.5710650105712116, "learning_rate": 1.0114615927616272e-06, "loss": 0.0703, "step": 17241 }, { "epoch": 0.8012081784386617, "grad_norm": 0.5884301587249364, "learning_rate": 1.011007839043831e-06, "loss": 0.0531, "step": 17242 }, { "epoch": 0.8012546468401487, "grad_norm": 0.36669979146226606, "learning_rate": 1.0105541756809046e-06, "loss": 0.0494, "step": 17243 }, { "epoch": 0.8013011152416357, "grad_norm": 0.6919818853912544, "learning_rate": 1.0101006026831273e-06, "loss": 0.0515, "step": 17244 }, { "epoch": 0.8013475836431226, "grad_norm": 1.4010454133212267, "learning_rate": 1.0096471200607694e-06, "loss": 0.0845, "step": 17245 }, { "epoch": 0.8013940520446097, "grad_norm": 0.893133995036031, "learning_rate": 1.0091937278241027e-06, "loss": 0.058, "step": 17246 }, { "epoch": 0.8014405204460967, "grad_norm": 0.5853137624601378, "learning_rate": 1.0087404259833993e-06, "loss": 0.0573, "step": 17247 }, { "epoch": 0.8014869888475836, "grad_norm": 0.3220024908275001, "learning_rate": 1.0082872145489264e-06, "loss": 0.0676, "step": 17248 }, { "epoch": 0.8015334572490707, "grad_norm": 0.3426221109629648, "learning_rate": 1.0078340935309478e-06, "loss": 0.0384, "step": 17249 }, { "epoch": 0.8015799256505576, "grad_norm": 0.43382006667741546, "learning_rate": 1.0073810629397312e-06, "loss": 0.0345, "step": 17250 }, { "epoch": 0.8016263940520446, "grad_norm": 0.5512618547310928, "learning_rate": 1.0069281227855343e-06, "loss": 0.0471, "step": 17251 }, { "epoch": 0.8016728624535316, "grad_norm": 0.3662178785640137, "learning_rate": 1.0064752730786186e-06, "loss": 0.0353, "step": 17252 }, { "epoch": 0.8017193308550186, "grad_norm": 0.5556332358051391, "learning_rate": 1.006022513829239e-06, "loss": 0.059, "step": 17253 }, { "epoch": 0.8017657992565056, "grad_norm": 0.4059988145699869, "learning_rate": 1.0055698450476549e-06, "loss": 0.0372, "step": 17254 }, { "epoch": 0.8018122676579925, "grad_norm": 0.33936257553822435, "learning_rate": 1.0051172667441179e-06, "loss": 0.0477, "step": 17255 }, { "epoch": 0.8018587360594795, "grad_norm": 0.5643050755848948, "learning_rate": 1.0046647789288788e-06, "loss": 0.0674, "step": 17256 }, { "epoch": 0.8019052044609666, "grad_norm": 0.6321052417543128, "learning_rate": 1.004212381612188e-06, "loss": 0.0639, "step": 17257 }, { "epoch": 0.8019516728624535, "grad_norm": 0.5747332843147822, "learning_rate": 1.003760074804292e-06, "loss": 0.0618, "step": 17258 }, { "epoch": 0.8019981412639405, "grad_norm": 0.545000195608012, "learning_rate": 1.0033078585154366e-06, "loss": 0.0384, "step": 17259 }, { "epoch": 0.8020446096654275, "grad_norm": 0.49515541560290427, "learning_rate": 1.002855732755863e-06, "loss": 0.0473, "step": 17260 }, { "epoch": 0.8020910780669145, "grad_norm": 0.3203731909948345, "learning_rate": 1.0024036975358154e-06, "loss": 0.0348, "step": 17261 }, { "epoch": 0.8021375464684015, "grad_norm": 0.5907091835571762, "learning_rate": 1.0019517528655326e-06, "loss": 0.0485, "step": 17262 }, { "epoch": 0.8021840148698884, "grad_norm": 0.472679379497605, "learning_rate": 1.0014998987552482e-06, "loss": 0.0667, "step": 17263 }, { "epoch": 0.8022304832713755, "grad_norm": 0.2858653142853019, "learning_rate": 1.0010481352152002e-06, "loss": 0.0358, "step": 17264 }, { "epoch": 0.8022769516728625, "grad_norm": 0.4593550626393206, "learning_rate": 1.0005964622556213e-06, "loss": 0.0463, "step": 17265 }, { "epoch": 0.8023234200743494, "grad_norm": 0.5301652251586606, "learning_rate": 1.0001448798867402e-06, "loss": 0.062, "step": 17266 }, { "epoch": 0.8023698884758365, "grad_norm": 1.1470100346813763, "learning_rate": 9.996933881187886e-07, "loss": 0.0676, "step": 17267 }, { "epoch": 0.8024163568773234, "grad_norm": 0.4796062441604723, "learning_rate": 9.992419869619935e-07, "loss": 0.0375, "step": 17268 }, { "epoch": 0.8024628252788104, "grad_norm": 0.25734310150070466, "learning_rate": 9.98790676426577e-07, "loss": 0.0331, "step": 17269 }, { "epoch": 0.8025092936802974, "grad_norm": 0.42608809325882335, "learning_rate": 9.98339456522761e-07, "loss": 0.0451, "step": 17270 }, { "epoch": 0.8025557620817844, "grad_norm": 0.48328734657213174, "learning_rate": 9.978883272607698e-07, "loss": 0.0465, "step": 17271 }, { "epoch": 0.8026022304832714, "grad_norm": 0.3261014185043865, "learning_rate": 9.974372886508198e-07, "loss": 0.036, "step": 17272 }, { "epoch": 0.8026486988847583, "grad_norm": 0.46889272504882484, "learning_rate": 9.969863407031266e-07, "loss": 0.0415, "step": 17273 }, { "epoch": 0.8026951672862453, "grad_norm": 0.5881384517011847, "learning_rate": 9.965354834279083e-07, "loss": 0.0488, "step": 17274 }, { "epoch": 0.8027416356877324, "grad_norm": 0.7819909549532437, "learning_rate": 9.960847168353743e-07, "loss": 0.0585, "step": 17275 }, { "epoch": 0.8027881040892193, "grad_norm": 0.43158268192414223, "learning_rate": 9.956340409357352e-07, "loss": 0.0365, "step": 17276 }, { "epoch": 0.8028345724907063, "grad_norm": 0.6982820724903483, "learning_rate": 9.951834557391976e-07, "loss": 0.0489, "step": 17277 }, { "epoch": 0.8028810408921933, "grad_norm": 0.5696117851279429, "learning_rate": 9.947329612559719e-07, "loss": 0.0642, "step": 17278 }, { "epoch": 0.8029275092936803, "grad_norm": 0.5108203465476314, "learning_rate": 9.942825574962595e-07, "loss": 0.0423, "step": 17279 }, { "epoch": 0.8029739776951673, "grad_norm": 0.44982355503258353, "learning_rate": 9.938322444702636e-07, "loss": 0.0341, "step": 17280 }, { "epoch": 0.8030204460966542, "grad_norm": 0.5915472464449124, "learning_rate": 9.933820221881835e-07, "loss": 0.0688, "step": 17281 }, { "epoch": 0.8030669144981413, "grad_norm": 0.4477862357255565, "learning_rate": 9.929318906602176e-07, "loss": 0.0561, "step": 17282 }, { "epoch": 0.8031133828996283, "grad_norm": 0.7483141581018743, "learning_rate": 9.9248184989656e-07, "loss": 0.0625, "step": 17283 }, { "epoch": 0.8031598513011152, "grad_norm": 0.6585437809425683, "learning_rate": 9.92031899907408e-07, "loss": 0.0481, "step": 17284 }, { "epoch": 0.8032063197026023, "grad_norm": 0.4319631335667718, "learning_rate": 9.91582040702952e-07, "loss": 0.0542, "step": 17285 }, { "epoch": 0.8032527881040892, "grad_norm": 0.4642480621478205, "learning_rate": 9.911322722933825e-07, "loss": 0.0555, "step": 17286 }, { "epoch": 0.8032992565055762, "grad_norm": 0.3860489464490782, "learning_rate": 9.906825946888837e-07, "loss": 0.0305, "step": 17287 }, { "epoch": 0.8033457249070632, "grad_norm": 0.5258079608763508, "learning_rate": 9.902330078996452e-07, "loss": 0.0603, "step": 17288 }, { "epoch": 0.8033921933085502, "grad_norm": 0.3557382870416863, "learning_rate": 9.897835119358495e-07, "loss": 0.0437, "step": 17289 }, { "epoch": 0.8034386617100372, "grad_norm": 0.41398570054609707, "learning_rate": 9.893341068076767e-07, "loss": 0.0441, "step": 17290 }, { "epoch": 0.8034851301115241, "grad_norm": 0.5166162158020453, "learning_rate": 9.888847925253087e-07, "loss": 0.0438, "step": 17291 }, { "epoch": 0.8035315985130111, "grad_norm": 0.7160588681902181, "learning_rate": 9.884355690989228e-07, "loss": 0.0667, "step": 17292 }, { "epoch": 0.8035780669144982, "grad_norm": 0.3744918517233296, "learning_rate": 9.87986436538691e-07, "loss": 0.0418, "step": 17293 }, { "epoch": 0.8036245353159851, "grad_norm": 0.5772087582027425, "learning_rate": 9.8753739485479e-07, "loss": 0.0676, "step": 17294 }, { "epoch": 0.8036710037174721, "grad_norm": 0.6839295058036549, "learning_rate": 9.870884440573901e-07, "loss": 0.0462, "step": 17295 }, { "epoch": 0.803717472118959, "grad_norm": 0.43004215272861734, "learning_rate": 9.86639584156661e-07, "loss": 0.0356, "step": 17296 }, { "epoch": 0.8037639405204461, "grad_norm": 0.6846699999669027, "learning_rate": 9.861908151627692e-07, "loss": 0.0605, "step": 17297 }, { "epoch": 0.8038104089219331, "grad_norm": 0.5796064361399982, "learning_rate": 9.857421370858794e-07, "loss": 0.0491, "step": 17298 }, { "epoch": 0.80385687732342, "grad_norm": 0.37694794330062936, "learning_rate": 9.852935499361555e-07, "loss": 0.0428, "step": 17299 }, { "epoch": 0.8039033457249071, "grad_norm": 0.6302289801474326, "learning_rate": 9.848450537237563e-07, "loss": 0.0595, "step": 17300 }, { "epoch": 0.8039498141263941, "grad_norm": 0.39935050442643855, "learning_rate": 9.84396648458844e-07, "loss": 0.0558, "step": 17301 }, { "epoch": 0.803996282527881, "grad_norm": 0.34834738827780565, "learning_rate": 9.839483341515742e-07, "loss": 0.0379, "step": 17302 }, { "epoch": 0.804042750929368, "grad_norm": 0.48599592675038394, "learning_rate": 9.835001108121028e-07, "loss": 0.0417, "step": 17303 }, { "epoch": 0.804089219330855, "grad_norm": 0.6166546879112174, "learning_rate": 9.830519784505788e-07, "loss": 0.0522, "step": 17304 }, { "epoch": 0.804135687732342, "grad_norm": 0.29970203636056075, "learning_rate": 9.826039370771561e-07, "loss": 0.0373, "step": 17305 }, { "epoch": 0.804182156133829, "grad_norm": 0.6218451887692276, "learning_rate": 9.821559867019825e-07, "loss": 0.0653, "step": 17306 }, { "epoch": 0.804228624535316, "grad_norm": 0.40557857825973403, "learning_rate": 9.817081273352031e-07, "loss": 0.0467, "step": 17307 }, { "epoch": 0.804275092936803, "grad_norm": 0.39473064559134485, "learning_rate": 9.812603589869651e-07, "loss": 0.0508, "step": 17308 }, { "epoch": 0.8043215613382899, "grad_norm": 0.4797653318662781, "learning_rate": 9.808126816674113e-07, "loss": 0.036, "step": 17309 }, { "epoch": 0.804368029739777, "grad_norm": 0.5226569051716529, "learning_rate": 9.803650953866772e-07, "loss": 0.0343, "step": 17310 }, { "epoch": 0.804414498141264, "grad_norm": 0.3631817953936473, "learning_rate": 9.79917600154906e-07, "loss": 0.0408, "step": 17311 }, { "epoch": 0.8044609665427509, "grad_norm": 0.8106269790007945, "learning_rate": 9.794701959822322e-07, "loss": 0.0526, "step": 17312 }, { "epoch": 0.8045074349442379, "grad_norm": 0.555808351388141, "learning_rate": 9.790228828787902e-07, "loss": 0.0622, "step": 17313 }, { "epoch": 0.8045539033457249, "grad_norm": 0.29575331692619167, "learning_rate": 9.785756608547103e-07, "loss": 0.039, "step": 17314 }, { "epoch": 0.8046003717472119, "grad_norm": 0.648871950891134, "learning_rate": 9.781285299201264e-07, "loss": 0.0642, "step": 17315 }, { "epoch": 0.8046468401486989, "grad_norm": 0.7862326968862672, "learning_rate": 9.776814900851634e-07, "loss": 0.0622, "step": 17316 }, { "epoch": 0.8046933085501858, "grad_norm": 0.4675264767450193, "learning_rate": 9.77234541359947e-07, "loss": 0.0536, "step": 17317 }, { "epoch": 0.8047397769516729, "grad_norm": 0.42500655342621, "learning_rate": 9.76787683754603e-07, "loss": 0.0408, "step": 17318 }, { "epoch": 0.8047862453531599, "grad_norm": 0.8314693657470005, "learning_rate": 9.763409172792525e-07, "loss": 0.0645, "step": 17319 }, { "epoch": 0.8048327137546468, "grad_norm": 0.5443056721440446, "learning_rate": 9.758942419440148e-07, "loss": 0.0435, "step": 17320 }, { "epoch": 0.8048791821561339, "grad_norm": 0.43625543655438864, "learning_rate": 9.754476577590082e-07, "loss": 0.0464, "step": 17321 }, { "epoch": 0.8049256505576208, "grad_norm": 0.3825025106444582, "learning_rate": 9.750011647343477e-07, "loss": 0.0382, "step": 17322 }, { "epoch": 0.8049721189591078, "grad_norm": 0.735455786334743, "learning_rate": 9.745547628801472e-07, "loss": 0.0674, "step": 17323 }, { "epoch": 0.8050185873605948, "grad_norm": 0.4116796426021077, "learning_rate": 9.741084522065165e-07, "loss": 0.0347, "step": 17324 }, { "epoch": 0.8050650557620818, "grad_norm": 0.4523766415167883, "learning_rate": 9.736622327235678e-07, "loss": 0.0534, "step": 17325 }, { "epoch": 0.8051115241635688, "grad_norm": 0.7936505227237203, "learning_rate": 9.732161044414073e-07, "loss": 0.0528, "step": 17326 }, { "epoch": 0.8051579925650557, "grad_norm": 0.6787713532234257, "learning_rate": 9.7277006737014e-07, "loss": 0.0627, "step": 17327 }, { "epoch": 0.8052044609665427, "grad_norm": 0.40222161849005433, "learning_rate": 9.723241215198692e-07, "loss": 0.0587, "step": 17328 }, { "epoch": 0.8052509293680298, "grad_norm": 0.43077843951459216, "learning_rate": 9.718782669006966e-07, "loss": 0.0499, "step": 17329 }, { "epoch": 0.8052973977695167, "grad_norm": 0.5641962946344757, "learning_rate": 9.714325035227201e-07, "loss": 0.0504, "step": 17330 }, { "epoch": 0.8053438661710037, "grad_norm": 0.651084436703278, "learning_rate": 9.70986831396037e-07, "loss": 0.0604, "step": 17331 }, { "epoch": 0.8053903345724907, "grad_norm": 0.5256874685359225, "learning_rate": 9.705412505307432e-07, "loss": 0.0672, "step": 17332 }, { "epoch": 0.8054368029739777, "grad_norm": 0.349716458543945, "learning_rate": 9.70095760936932e-07, "loss": 0.0494, "step": 17333 }, { "epoch": 0.8054832713754647, "grad_norm": 0.3683860196382135, "learning_rate": 9.696503626246916e-07, "loss": 0.0522, "step": 17334 }, { "epoch": 0.8055297397769516, "grad_norm": 0.6817460233446415, "learning_rate": 9.692050556041132e-07, "loss": 0.0537, "step": 17335 }, { "epoch": 0.8055762081784387, "grad_norm": 0.7911364225217622, "learning_rate": 9.68759839885282e-07, "loss": 0.0617, "step": 17336 }, { "epoch": 0.8056226765799257, "grad_norm": 0.5566771956626272, "learning_rate": 9.68314715478284e-07, "loss": 0.0381, "step": 17337 }, { "epoch": 0.8056691449814126, "grad_norm": 0.3537624487909311, "learning_rate": 9.678696823932e-07, "loss": 0.0317, "step": 17338 }, { "epoch": 0.8057156133828997, "grad_norm": 0.5970158692300253, "learning_rate": 9.674247406401132e-07, "loss": 0.0511, "step": 17339 }, { "epoch": 0.8057620817843866, "grad_norm": 0.7270120286172211, "learning_rate": 9.669798902290995e-07, "loss": 0.0659, "step": 17340 }, { "epoch": 0.8058085501858736, "grad_norm": 0.5791280713661664, "learning_rate": 9.665351311702342e-07, "loss": 0.0494, "step": 17341 }, { "epoch": 0.8058550185873606, "grad_norm": 0.32793660415354414, "learning_rate": 9.660904634735945e-07, "loss": 0.0347, "step": 17342 }, { "epoch": 0.8059014869888476, "grad_norm": 0.5543008468569379, "learning_rate": 9.656458871492519e-07, "loss": 0.0536, "step": 17343 }, { "epoch": 0.8059479553903346, "grad_norm": 0.43380776983896835, "learning_rate": 9.652014022072758e-07, "loss": 0.0336, "step": 17344 }, { "epoch": 0.8059944237918215, "grad_norm": 0.37713135123795266, "learning_rate": 9.647570086577346e-07, "loss": 0.0489, "step": 17345 }, { "epoch": 0.8060408921933085, "grad_norm": 0.5805271473495641, "learning_rate": 9.643127065106943e-07, "loss": 0.0597, "step": 17346 }, { "epoch": 0.8060873605947956, "grad_norm": 0.44870931903952765, "learning_rate": 9.638684957762184e-07, "loss": 0.055, "step": 17347 }, { "epoch": 0.8061338289962825, "grad_norm": 0.7466386620799094, "learning_rate": 9.63424376464368e-07, "loss": 0.0521, "step": 17348 }, { "epoch": 0.8061802973977695, "grad_norm": 0.3794546892878787, "learning_rate": 9.629803485852047e-07, "loss": 0.0432, "step": 17349 }, { "epoch": 0.8062267657992565, "grad_norm": 0.5162295759275173, "learning_rate": 9.625364121487856e-07, "loss": 0.0601, "step": 17350 }, { "epoch": 0.8062732342007435, "grad_norm": 0.5430955474225002, "learning_rate": 9.620925671651665e-07, "loss": 0.0617, "step": 17351 }, { "epoch": 0.8063197026022305, "grad_norm": 0.6204925024302469, "learning_rate": 9.616488136444002e-07, "loss": 0.0439, "step": 17352 }, { "epoch": 0.8063661710037174, "grad_norm": 0.45452674255539827, "learning_rate": 9.612051515965388e-07, "loss": 0.0479, "step": 17353 }, { "epoch": 0.8064126394052045, "grad_norm": 0.6576795289799773, "learning_rate": 9.607615810316312e-07, "loss": 0.0447, "step": 17354 }, { "epoch": 0.8064591078066915, "grad_norm": 0.6326303407408209, "learning_rate": 9.603181019597242e-07, "loss": 0.0637, "step": 17355 }, { "epoch": 0.8065055762081784, "grad_norm": 0.6020854356169962, "learning_rate": 9.59874714390866e-07, "loss": 0.0421, "step": 17356 }, { "epoch": 0.8065520446096655, "grad_norm": 0.6190619974403028, "learning_rate": 9.594314183350967e-07, "loss": 0.0549, "step": 17357 }, { "epoch": 0.8065985130111524, "grad_norm": 0.6158449286040444, "learning_rate": 9.589882138024565e-07, "loss": 0.0618, "step": 17358 }, { "epoch": 0.8066449814126394, "grad_norm": 0.4558300170962514, "learning_rate": 9.585451008029884e-07, "loss": 0.0449, "step": 17359 }, { "epoch": 0.8066914498141264, "grad_norm": 0.6677543499789076, "learning_rate": 9.581020793467266e-07, "loss": 0.0587, "step": 17360 }, { "epoch": 0.8067379182156134, "grad_norm": 0.5724305989438722, "learning_rate": 9.57659149443707e-07, "loss": 0.0461, "step": 17361 }, { "epoch": 0.8067843866171004, "grad_norm": 0.508907347035182, "learning_rate": 9.572163111039624e-07, "loss": 0.0367, "step": 17362 }, { "epoch": 0.8068308550185873, "grad_norm": 0.37989959776994286, "learning_rate": 9.56773564337523e-07, "loss": 0.0472, "step": 17363 }, { "epoch": 0.8068773234200743, "grad_norm": 0.6576618493356373, "learning_rate": 9.563309091544182e-07, "loss": 0.0499, "step": 17364 }, { "epoch": 0.8069237918215614, "grad_norm": 0.3840025916585451, "learning_rate": 9.558883455646728e-07, "loss": 0.0449, "step": 17365 }, { "epoch": 0.8069702602230483, "grad_norm": 0.4471187781158691, "learning_rate": 9.554458735783135e-07, "loss": 0.0559, "step": 17366 }, { "epoch": 0.8070167286245353, "grad_norm": 0.5463213438006747, "learning_rate": 9.550034932053625e-07, "loss": 0.0562, "step": 17367 }, { "epoch": 0.8070631970260222, "grad_norm": 0.549403882789839, "learning_rate": 9.545612044558395e-07, "loss": 0.045, "step": 17368 }, { "epoch": 0.8071096654275093, "grad_norm": 0.5950478931907068, "learning_rate": 9.541190073397626e-07, "loss": 0.0478, "step": 17369 }, { "epoch": 0.8071561338289963, "grad_norm": 0.475926685994462, "learning_rate": 9.536769018671487e-07, "loss": 0.0625, "step": 17370 }, { "epoch": 0.8072026022304832, "grad_norm": 0.4265233671815316, "learning_rate": 9.532348880480119e-07, "loss": 0.0631, "step": 17371 }, { "epoch": 0.8072490706319703, "grad_norm": 0.7876088803618019, "learning_rate": 9.527929658923629e-07, "loss": 0.0426, "step": 17372 }, { "epoch": 0.8072955390334573, "grad_norm": 0.8236443599748492, "learning_rate": 9.523511354102133e-07, "loss": 0.0508, "step": 17373 }, { "epoch": 0.8073420074349442, "grad_norm": 0.4352417033424375, "learning_rate": 9.519093966115728e-07, "loss": 0.0698, "step": 17374 }, { "epoch": 0.8073884758364313, "grad_norm": 0.5219734085372831, "learning_rate": 9.514677495064418e-07, "loss": 0.0708, "step": 17375 }, { "epoch": 0.8074349442379182, "grad_norm": 0.38625845983288215, "learning_rate": 9.510261941048288e-07, "loss": 0.0319, "step": 17376 }, { "epoch": 0.8074814126394052, "grad_norm": 0.6761533876613088, "learning_rate": 9.505847304167337e-07, "loss": 0.0515, "step": 17377 }, { "epoch": 0.8075278810408922, "grad_norm": 0.7445473613716184, "learning_rate": 9.501433584521563e-07, "loss": 0.0529, "step": 17378 }, { "epoch": 0.8075743494423792, "grad_norm": 0.6276971850684304, "learning_rate": 9.497020782210936e-07, "loss": 0.0736, "step": 17379 }, { "epoch": 0.8076208178438662, "grad_norm": 0.4657098074544949, "learning_rate": 9.492608897335437e-07, "loss": 0.0463, "step": 17380 }, { "epoch": 0.8076672862453531, "grad_norm": 0.41323228103950865, "learning_rate": 9.488197929994969e-07, "loss": 0.0397, "step": 17381 }, { "epoch": 0.8077137546468401, "grad_norm": 0.41732328565246307, "learning_rate": 9.483787880289441e-07, "loss": 0.0455, "step": 17382 }, { "epoch": 0.8077602230483272, "grad_norm": 0.46535796600280777, "learning_rate": 9.479378748318774e-07, "loss": 0.0414, "step": 17383 }, { "epoch": 0.8078066914498141, "grad_norm": 0.501106714487689, "learning_rate": 9.474970534182826e-07, "loss": 0.0598, "step": 17384 }, { "epoch": 0.8078531598513011, "grad_norm": 0.4114611565374025, "learning_rate": 9.470563237981445e-07, "loss": 0.0316, "step": 17385 }, { "epoch": 0.807899628252788, "grad_norm": 0.5318690382488248, "learning_rate": 9.466156859814468e-07, "loss": 0.0489, "step": 17386 }, { "epoch": 0.8079460966542751, "grad_norm": 0.6258255804772581, "learning_rate": 9.461751399781694e-07, "loss": 0.0668, "step": 17387 }, { "epoch": 0.8079925650557621, "grad_norm": 0.48624928350722485, "learning_rate": 9.457346857982919e-07, "loss": 0.043, "step": 17388 }, { "epoch": 0.808039033457249, "grad_norm": 0.7267101746935296, "learning_rate": 9.452943234517897e-07, "loss": 0.0571, "step": 17389 }, { "epoch": 0.8080855018587361, "grad_norm": 0.42793522273718737, "learning_rate": 9.448540529486394e-07, "loss": 0.0515, "step": 17390 }, { "epoch": 0.8081319702602231, "grad_norm": 0.35189034772270417, "learning_rate": 9.444138742988135e-07, "loss": 0.0494, "step": 17391 }, { "epoch": 0.80817843866171, "grad_norm": 0.8377091278054347, "learning_rate": 9.439737875122812e-07, "loss": 0.0754, "step": 17392 }, { "epoch": 0.808224907063197, "grad_norm": 0.5083676640636954, "learning_rate": 9.435337925990118e-07, "loss": 0.0541, "step": 17393 }, { "epoch": 0.808271375464684, "grad_norm": 0.5196720877871525, "learning_rate": 9.430938895689706e-07, "loss": 0.031, "step": 17394 }, { "epoch": 0.808317843866171, "grad_norm": 0.6819725309952848, "learning_rate": 9.426540784321237e-07, "loss": 0.0701, "step": 17395 }, { "epoch": 0.808364312267658, "grad_norm": 0.3623437805031542, "learning_rate": 9.422143591984301e-07, "loss": 0.0431, "step": 17396 }, { "epoch": 0.808410780669145, "grad_norm": 0.38711146772606425, "learning_rate": 9.417747318778536e-07, "loss": 0.0316, "step": 17397 }, { "epoch": 0.808457249070632, "grad_norm": 0.7668636782798386, "learning_rate": 9.413351964803518e-07, "loss": 0.0561, "step": 17398 }, { "epoch": 0.8085037174721189, "grad_norm": 0.6395408255964333, "learning_rate": 9.408957530158769e-07, "loss": 0.0526, "step": 17399 }, { "epoch": 0.8085501858736059, "grad_norm": 0.46478895143641796, "learning_rate": 9.404564014943867e-07, "loss": 0.0539, "step": 17400 }, { "epoch": 0.808596654275093, "grad_norm": 0.6877668390462071, "learning_rate": 9.400171419258319e-07, "loss": 0.0568, "step": 17401 }, { "epoch": 0.8086431226765799, "grad_norm": 0.4277381323203475, "learning_rate": 9.395779743201616e-07, "loss": 0.045, "step": 17402 }, { "epoch": 0.8086895910780669, "grad_norm": 1.0193521182296632, "learning_rate": 9.391388986873223e-07, "loss": 0.0674, "step": 17403 }, { "epoch": 0.8087360594795538, "grad_norm": 0.4703525942204317, "learning_rate": 9.386999150372639e-07, "loss": 0.044, "step": 17404 }, { "epoch": 0.8087825278810409, "grad_norm": 0.47528437365110837, "learning_rate": 9.38261023379925e-07, "loss": 0.0473, "step": 17405 }, { "epoch": 0.8088289962825279, "grad_norm": 0.5710407588942393, "learning_rate": 9.378222237252477e-07, "loss": 0.035, "step": 17406 }, { "epoch": 0.8088754646840148, "grad_norm": 0.4104845854671458, "learning_rate": 9.37383516083174e-07, "loss": 0.0204, "step": 17407 }, { "epoch": 0.8089219330855019, "grad_norm": 0.40243160834770825, "learning_rate": 9.369449004636394e-07, "loss": 0.0563, "step": 17408 }, { "epoch": 0.8089684014869889, "grad_norm": 0.5552567551925431, "learning_rate": 9.365063768765786e-07, "loss": 0.0767, "step": 17409 }, { "epoch": 0.8090148698884758, "grad_norm": 0.3737685811096653, "learning_rate": 9.360679453319254e-07, "loss": 0.0513, "step": 17410 }, { "epoch": 0.8090613382899629, "grad_norm": 0.4414593519886069, "learning_rate": 9.356296058396108e-07, "loss": 0.047, "step": 17411 }, { "epoch": 0.8091078066914498, "grad_norm": 0.4637967974328105, "learning_rate": 9.351913584095628e-07, "loss": 0.0605, "step": 17412 }, { "epoch": 0.8091542750929368, "grad_norm": 0.479175009026822, "learning_rate": 9.347532030517082e-07, "loss": 0.0531, "step": 17413 }, { "epoch": 0.8092007434944238, "grad_norm": 0.5455193502248887, "learning_rate": 9.343151397759726e-07, "loss": 0.0449, "step": 17414 }, { "epoch": 0.8092472118959108, "grad_norm": 0.6000691256688928, "learning_rate": 9.338771685922798e-07, "loss": 0.0361, "step": 17415 }, { "epoch": 0.8092936802973978, "grad_norm": 0.661550566444724, "learning_rate": 9.33439289510546e-07, "loss": 0.0471, "step": 17416 }, { "epoch": 0.8093401486988847, "grad_norm": 0.3631110584692983, "learning_rate": 9.330015025406936e-07, "loss": 0.0572, "step": 17417 }, { "epoch": 0.8093866171003717, "grad_norm": 0.7777382511423947, "learning_rate": 9.325638076926374e-07, "loss": 0.0498, "step": 17418 }, { "epoch": 0.8094330855018588, "grad_norm": 0.42684860724592105, "learning_rate": 9.321262049762919e-07, "loss": 0.0467, "step": 17419 }, { "epoch": 0.8094795539033457, "grad_norm": 0.5862323292144516, "learning_rate": 9.316886944015679e-07, "loss": 0.0539, "step": 17420 }, { "epoch": 0.8095260223048327, "grad_norm": 0.5354496864864462, "learning_rate": 9.312512759783793e-07, "loss": 0.0491, "step": 17421 }, { "epoch": 0.8095724907063196, "grad_norm": 0.5069718029902213, "learning_rate": 9.308139497166308e-07, "loss": 0.064, "step": 17422 }, { "epoch": 0.8096189591078067, "grad_norm": 0.6101735609329918, "learning_rate": 9.303767156262272e-07, "loss": 0.0582, "step": 17423 }, { "epoch": 0.8096654275092937, "grad_norm": 0.5327380433859414, "learning_rate": 9.299395737170758e-07, "loss": 0.0378, "step": 17424 }, { "epoch": 0.8097118959107806, "grad_norm": 0.8241498739752176, "learning_rate": 9.295025239990768e-07, "loss": 0.0857, "step": 17425 }, { "epoch": 0.8097583643122677, "grad_norm": 0.5670456790567275, "learning_rate": 9.290655664821296e-07, "loss": 0.0596, "step": 17426 }, { "epoch": 0.8098048327137547, "grad_norm": 0.5107895770367353, "learning_rate": 9.28628701176132e-07, "loss": 0.0604, "step": 17427 }, { "epoch": 0.8098513011152416, "grad_norm": 0.3988198100209549, "learning_rate": 9.28191928090979e-07, "loss": 0.0495, "step": 17428 }, { "epoch": 0.8098977695167286, "grad_norm": 0.670441780348067, "learning_rate": 9.277552472365648e-07, "loss": 0.0613, "step": 17429 }, { "epoch": 0.8099442379182156, "grad_norm": 0.5618619165227742, "learning_rate": 9.273186586227784e-07, "loss": 0.0439, "step": 17430 }, { "epoch": 0.8099907063197026, "grad_norm": 0.4751732714418549, "learning_rate": 9.268821622595125e-07, "loss": 0.0486, "step": 17431 }, { "epoch": 0.8100371747211896, "grad_norm": 0.3612329358173956, "learning_rate": 9.264457581566522e-07, "loss": 0.0293, "step": 17432 }, { "epoch": 0.8100836431226766, "grad_norm": 0.4598470277736964, "learning_rate": 9.260094463240831e-07, "loss": 0.0613, "step": 17433 }, { "epoch": 0.8101301115241636, "grad_norm": 0.5717153034249521, "learning_rate": 9.255732267716877e-07, "loss": 0.0392, "step": 17434 }, { "epoch": 0.8101765799256505, "grad_norm": 0.4394299769769655, "learning_rate": 9.251370995093472e-07, "loss": 0.0549, "step": 17435 }, { "epoch": 0.8102230483271375, "grad_norm": 0.5056157502527759, "learning_rate": 9.2470106454694e-07, "loss": 0.0623, "step": 17436 }, { "epoch": 0.8102695167286246, "grad_norm": 0.6874576561684198, "learning_rate": 9.242651218943416e-07, "loss": 0.0488, "step": 17437 }, { "epoch": 0.8103159851301115, "grad_norm": 0.49845086059301275, "learning_rate": 9.238292715614288e-07, "loss": 0.0319, "step": 17438 }, { "epoch": 0.8103624535315985, "grad_norm": 0.42584205435681843, "learning_rate": 9.233935135580746e-07, "loss": 0.0537, "step": 17439 }, { "epoch": 0.8104089219330854, "grad_norm": 0.5210616041977225, "learning_rate": 9.229578478941453e-07, "loss": 0.0367, "step": 17440 }, { "epoch": 0.8104553903345725, "grad_norm": 0.5844233192936666, "learning_rate": 9.225222745795131e-07, "loss": 0.0579, "step": 17441 }, { "epoch": 0.8105018587360595, "grad_norm": 0.34325783607212385, "learning_rate": 9.220867936240429e-07, "loss": 0.0435, "step": 17442 }, { "epoch": 0.8105483271375464, "grad_norm": 0.8286472940055125, "learning_rate": 9.216514050375985e-07, "loss": 0.0605, "step": 17443 }, { "epoch": 0.8105947955390335, "grad_norm": 0.6138707980807909, "learning_rate": 9.212161088300409e-07, "loss": 0.0518, "step": 17444 }, { "epoch": 0.8106412639405205, "grad_norm": 0.42547011911286786, "learning_rate": 9.207809050112343e-07, "loss": 0.024, "step": 17445 }, { "epoch": 0.8106877323420074, "grad_norm": 0.27619331805573155, "learning_rate": 9.203457935910315e-07, "loss": 0.0446, "step": 17446 }, { "epoch": 0.8107342007434944, "grad_norm": 0.3842108304284196, "learning_rate": 9.199107745792895e-07, "loss": 0.0321, "step": 17447 }, { "epoch": 0.8107806691449814, "grad_norm": 0.36024745501940225, "learning_rate": 9.19475847985864e-07, "loss": 0.036, "step": 17448 }, { "epoch": 0.8108271375464684, "grad_norm": 0.9321042150669772, "learning_rate": 9.190410138206046e-07, "loss": 0.0623, "step": 17449 }, { "epoch": 0.8108736059479554, "grad_norm": 1.6172055429771315, "learning_rate": 9.18606272093362e-07, "loss": 0.0698, "step": 17450 }, { "epoch": 0.8109200743494424, "grad_norm": 0.44948815083036786, "learning_rate": 9.181716228139825e-07, "loss": 0.042, "step": 17451 }, { "epoch": 0.8109665427509294, "grad_norm": 0.32909334241392646, "learning_rate": 9.177370659923123e-07, "loss": 0.0381, "step": 17452 }, { "epoch": 0.8110130111524163, "grad_norm": 0.7035360334266345, "learning_rate": 9.173026016381936e-07, "loss": 0.053, "step": 17453 }, { "epoch": 0.8110594795539033, "grad_norm": 0.5890592242643605, "learning_rate": 9.168682297614667e-07, "loss": 0.056, "step": 17454 }, { "epoch": 0.8111059479553904, "grad_norm": 0.5793410223654483, "learning_rate": 9.164339503719733e-07, "loss": 0.053, "step": 17455 }, { "epoch": 0.8111524163568773, "grad_norm": 0.5525911739654191, "learning_rate": 9.159997634795486e-07, "loss": 0.0629, "step": 17456 }, { "epoch": 0.8111988847583643, "grad_norm": 0.39417292601413884, "learning_rate": 9.155656690940279e-07, "loss": 0.0447, "step": 17457 }, { "epoch": 0.8112453531598512, "grad_norm": 0.41456752620786674, "learning_rate": 9.15131667225243e-07, "loss": 0.0424, "step": 17458 }, { "epoch": 0.8112918215613383, "grad_norm": 0.44852457202601753, "learning_rate": 9.146977578830257e-07, "loss": 0.0438, "step": 17459 }, { "epoch": 0.8113382899628253, "grad_norm": 0.6934447163595964, "learning_rate": 9.142639410772036e-07, "loss": 0.0459, "step": 17460 }, { "epoch": 0.8113847583643122, "grad_norm": 0.6311217949775064, "learning_rate": 9.138302168176022e-07, "loss": 0.0648, "step": 17461 }, { "epoch": 0.8114312267657993, "grad_norm": 0.7113141372433314, "learning_rate": 9.13396585114048e-07, "loss": 0.0484, "step": 17462 }, { "epoch": 0.8114776951672863, "grad_norm": 0.4495946876917021, "learning_rate": 9.129630459763638e-07, "loss": 0.0552, "step": 17463 }, { "epoch": 0.8115241635687732, "grad_norm": 0.6352816487545423, "learning_rate": 9.125295994143657e-07, "loss": 0.0441, "step": 17464 }, { "epoch": 0.8115706319702602, "grad_norm": 0.4831473383837284, "learning_rate": 9.120962454378751e-07, "loss": 0.0599, "step": 17465 }, { "epoch": 0.8116171003717472, "grad_norm": 0.6006570114991333, "learning_rate": 9.11662984056707e-07, "loss": 0.0613, "step": 17466 }, { "epoch": 0.8116635687732342, "grad_norm": 0.4602415289543933, "learning_rate": 9.112298152806754e-07, "loss": 0.0534, "step": 17467 }, { "epoch": 0.8117100371747212, "grad_norm": 0.5859817905094276, "learning_rate": 9.107967391195904e-07, "loss": 0.08, "step": 17468 }, { "epoch": 0.8117565055762082, "grad_norm": 0.6591027066024999, "learning_rate": 9.103637555832656e-07, "loss": 0.046, "step": 17469 }, { "epoch": 0.8118029739776952, "grad_norm": 0.908395103308425, "learning_rate": 9.099308646815052e-07, "loss": 0.0646, "step": 17470 }, { "epoch": 0.8118494423791821, "grad_norm": 0.7800706815897728, "learning_rate": 9.094980664241137e-07, "loss": 0.0604, "step": 17471 }, { "epoch": 0.8118959107806691, "grad_norm": 0.7319628927823673, "learning_rate": 9.090653608208977e-07, "loss": 0.0431, "step": 17472 }, { "epoch": 0.8119423791821562, "grad_norm": 0.49062789969156545, "learning_rate": 9.086327478816564e-07, "loss": 0.0434, "step": 17473 }, { "epoch": 0.8119888475836431, "grad_norm": 0.3713103146840326, "learning_rate": 9.0820022761619e-07, "loss": 0.046, "step": 17474 }, { "epoch": 0.8120353159851301, "grad_norm": 0.53304683078429, "learning_rate": 9.077678000342949e-07, "loss": 0.0468, "step": 17475 }, { "epoch": 0.812081784386617, "grad_norm": 0.37135939263370943, "learning_rate": 9.073354651457661e-07, "loss": 0.0676, "step": 17476 }, { "epoch": 0.8121282527881041, "grad_norm": 0.30577207422680214, "learning_rate": 9.069032229603963e-07, "loss": 0.0292, "step": 17477 }, { "epoch": 0.8121747211895911, "grad_norm": 0.39605166396908803, "learning_rate": 9.064710734879751e-07, "loss": 0.073, "step": 17478 }, { "epoch": 0.812221189591078, "grad_norm": 0.4037359447271793, "learning_rate": 9.060390167382938e-07, "loss": 0.0645, "step": 17479 }, { "epoch": 0.8122676579925651, "grad_norm": 0.5877157606724595, "learning_rate": 9.056070527211386e-07, "loss": 0.0624, "step": 17480 }, { "epoch": 0.8123141263940521, "grad_norm": 0.645003036951655, "learning_rate": 9.051751814462912e-07, "loss": 0.0548, "step": 17481 }, { "epoch": 0.812360594795539, "grad_norm": 0.4811176307575209, "learning_rate": 9.047434029235358e-07, "loss": 0.0473, "step": 17482 }, { "epoch": 0.812407063197026, "grad_norm": 0.5156205810391992, "learning_rate": 9.043117171626536e-07, "loss": 0.0533, "step": 17483 }, { "epoch": 0.812453531598513, "grad_norm": 0.24889053326410365, "learning_rate": 9.038801241734207e-07, "loss": 0.0308, "step": 17484 }, { "epoch": 0.8125, "grad_norm": 0.599463340941689, "learning_rate": 9.034486239656137e-07, "loss": 0.0445, "step": 17485 }, { "epoch": 0.812546468401487, "grad_norm": 0.7127275892711729, "learning_rate": 9.030172165490086e-07, "loss": 0.0471, "step": 17486 }, { "epoch": 0.812592936802974, "grad_norm": 0.7660234877268484, "learning_rate": 9.025859019333749e-07, "loss": 0.0516, "step": 17487 }, { "epoch": 0.812639405204461, "grad_norm": 0.6089890435344004, "learning_rate": 9.021546801284814e-07, "loss": 0.043, "step": 17488 }, { "epoch": 0.8126858736059479, "grad_norm": 0.5802138324647594, "learning_rate": 9.017235511440986e-07, "loss": 0.053, "step": 17489 }, { "epoch": 0.8127323420074349, "grad_norm": 0.4301103625744383, "learning_rate": 9.012925149899909e-07, "loss": 0.0465, "step": 17490 }, { "epoch": 0.812778810408922, "grad_norm": 0.4031932629320762, "learning_rate": 9.008615716759211e-07, "loss": 0.027, "step": 17491 }, { "epoch": 0.8128252788104089, "grad_norm": 0.30417617468150693, "learning_rate": 9.004307212116514e-07, "loss": 0.0328, "step": 17492 }, { "epoch": 0.8128717472118959, "grad_norm": 0.41322222408069353, "learning_rate": 8.999999636069406e-07, "loss": 0.0479, "step": 17493 }, { "epoch": 0.812918215613383, "grad_norm": 0.7271333503941271, "learning_rate": 8.995692988715454e-07, "loss": 0.0569, "step": 17494 }, { "epoch": 0.8129646840148699, "grad_norm": 0.5489541798252779, "learning_rate": 8.991387270152202e-07, "loss": 0.0419, "step": 17495 }, { "epoch": 0.8130111524163569, "grad_norm": 0.649063885516695, "learning_rate": 8.987082480477205e-07, "loss": 0.0582, "step": 17496 }, { "epoch": 0.8130576208178438, "grad_norm": 0.796849352904268, "learning_rate": 8.982778619787952e-07, "loss": 0.0653, "step": 17497 }, { "epoch": 0.8131040892193309, "grad_norm": 0.7331941954037695, "learning_rate": 8.978475688181931e-07, "loss": 0.0609, "step": 17498 }, { "epoch": 0.8131505576208179, "grad_norm": 0.37430648422451424, "learning_rate": 8.974173685756616e-07, "loss": 0.0346, "step": 17499 }, { "epoch": 0.8131970260223048, "grad_norm": 0.6653881469123, "learning_rate": 8.969872612609438e-07, "loss": 0.0528, "step": 17500 }, { "epoch": 0.8132434944237918, "grad_norm": 0.6142709435409303, "learning_rate": 8.965572468837835e-07, "loss": 0.0608, "step": 17501 }, { "epoch": 0.8132899628252788, "grad_norm": 0.38739550076656565, "learning_rate": 8.961273254539188e-07, "loss": 0.0468, "step": 17502 }, { "epoch": 0.8133364312267658, "grad_norm": 0.44290731039924663, "learning_rate": 8.956974969810905e-07, "loss": 0.0552, "step": 17503 }, { "epoch": 0.8133828996282528, "grad_norm": 0.5829191520457551, "learning_rate": 8.952677614750354e-07, "loss": 0.0656, "step": 17504 }, { "epoch": 0.8134293680297398, "grad_norm": 0.34045017453651516, "learning_rate": 8.948381189454824e-07, "loss": 0.0351, "step": 17505 }, { "epoch": 0.8134758364312268, "grad_norm": 0.7261321711897215, "learning_rate": 8.944085694021687e-07, "loss": 0.0512, "step": 17506 }, { "epoch": 0.8135223048327137, "grad_norm": 0.5087131728000716, "learning_rate": 8.939791128548214e-07, "loss": 0.0396, "step": 17507 }, { "epoch": 0.8135687732342007, "grad_norm": 0.49406650840299043, "learning_rate": 8.93549749313169e-07, "loss": 0.0474, "step": 17508 }, { "epoch": 0.8136152416356878, "grad_norm": 0.5380870169543173, "learning_rate": 8.931204787869351e-07, "loss": 0.0491, "step": 17509 }, { "epoch": 0.8136617100371747, "grad_norm": 0.6180695900016243, "learning_rate": 8.926913012858479e-07, "loss": 0.0666, "step": 17510 }, { "epoch": 0.8137081784386617, "grad_norm": 0.6710232101824917, "learning_rate": 8.922622168196238e-07, "loss": 0.0539, "step": 17511 }, { "epoch": 0.8137546468401488, "grad_norm": 0.653120930978799, "learning_rate": 8.918332253979828e-07, "loss": 0.0515, "step": 17512 }, { "epoch": 0.8138011152416357, "grad_norm": 0.61230486804974, "learning_rate": 8.914043270306438e-07, "loss": 0.061, "step": 17513 }, { "epoch": 0.8138475836431227, "grad_norm": 0.878833575504614, "learning_rate": 8.909755217273214e-07, "loss": 0.0472, "step": 17514 }, { "epoch": 0.8138940520446096, "grad_norm": 0.77992647638639, "learning_rate": 8.905468094977282e-07, "loss": 0.0527, "step": 17515 }, { "epoch": 0.8139405204460967, "grad_norm": 0.48480891739327675, "learning_rate": 8.901181903515749e-07, "loss": 0.0554, "step": 17516 }, { "epoch": 0.8139869888475837, "grad_norm": 0.6492901903088157, "learning_rate": 8.896896642985702e-07, "loss": 0.0577, "step": 17517 }, { "epoch": 0.8140334572490706, "grad_norm": 0.9126955612817051, "learning_rate": 8.8926123134842e-07, "loss": 0.0694, "step": 17518 }, { "epoch": 0.8140799256505576, "grad_norm": 0.7021519775897692, "learning_rate": 8.888328915108291e-07, "loss": 0.0677, "step": 17519 }, { "epoch": 0.8141263940520446, "grad_norm": 0.3343817433115112, "learning_rate": 8.884046447955008e-07, "loss": 0.0387, "step": 17520 }, { "epoch": 0.8141728624535316, "grad_norm": 0.4177448160722351, "learning_rate": 8.879764912121347e-07, "loss": 0.0543, "step": 17521 }, { "epoch": 0.8142193308550186, "grad_norm": 0.5058048296500142, "learning_rate": 8.875484307704291e-07, "loss": 0.0492, "step": 17522 }, { "epoch": 0.8142657992565056, "grad_norm": 0.42132679944649887, "learning_rate": 8.871204634800795e-07, "loss": 0.0507, "step": 17523 }, { "epoch": 0.8143122676579926, "grad_norm": 0.45131540605018133, "learning_rate": 8.866925893507805e-07, "loss": 0.0684, "step": 17524 }, { "epoch": 0.8143587360594795, "grad_norm": 0.6978786453656763, "learning_rate": 8.862648083922232e-07, "loss": 0.0438, "step": 17525 }, { "epoch": 0.8144052044609665, "grad_norm": 0.3982247541097953, "learning_rate": 8.858371206140965e-07, "loss": 0.0467, "step": 17526 }, { "epoch": 0.8144516728624536, "grad_norm": 0.3814361311421661, "learning_rate": 8.854095260260903e-07, "loss": 0.033, "step": 17527 }, { "epoch": 0.8144981412639405, "grad_norm": 0.5546033567426115, "learning_rate": 8.849820246378898e-07, "loss": 0.0469, "step": 17528 }, { "epoch": 0.8145446096654275, "grad_norm": 0.46707496398747844, "learning_rate": 8.845546164591756e-07, "loss": 0.0413, "step": 17529 }, { "epoch": 0.8145910780669146, "grad_norm": 0.39902120472103164, "learning_rate": 8.841273014996316e-07, "loss": 0.0428, "step": 17530 }, { "epoch": 0.8146375464684015, "grad_norm": 0.48794960988725095, "learning_rate": 8.837000797689361e-07, "loss": 0.058, "step": 17531 }, { "epoch": 0.8146840148698885, "grad_norm": 0.5472043598775455, "learning_rate": 8.832729512767662e-07, "loss": 0.0338, "step": 17532 }, { "epoch": 0.8147304832713754, "grad_norm": 0.4584092420803232, "learning_rate": 8.828459160327967e-07, "loss": 0.0486, "step": 17533 }, { "epoch": 0.8147769516728625, "grad_norm": 0.4611137065682736, "learning_rate": 8.824189740467004e-07, "loss": 0.0494, "step": 17534 }, { "epoch": 0.8148234200743495, "grad_norm": 0.4934459556678593, "learning_rate": 8.819921253281483e-07, "loss": 0.0527, "step": 17535 }, { "epoch": 0.8148698884758364, "grad_norm": 0.6340730302567339, "learning_rate": 8.815653698868071e-07, "loss": 0.0748, "step": 17536 }, { "epoch": 0.8149163568773234, "grad_norm": 0.305181747004395, "learning_rate": 8.811387077323463e-07, "loss": 0.0325, "step": 17537 }, { "epoch": 0.8149628252788104, "grad_norm": 0.37689745498155797, "learning_rate": 8.807121388744288e-07, "loss": 0.0455, "step": 17538 }, { "epoch": 0.8150092936802974, "grad_norm": 0.7428186055096964, "learning_rate": 8.80285663322717e-07, "loss": 0.0512, "step": 17539 }, { "epoch": 0.8150557620817844, "grad_norm": 0.34266462183048635, "learning_rate": 8.7985928108687e-07, "loss": 0.0543, "step": 17540 }, { "epoch": 0.8151022304832714, "grad_norm": 0.48788268785078487, "learning_rate": 8.794329921765476e-07, "loss": 0.0456, "step": 17541 }, { "epoch": 0.8151486988847584, "grad_norm": 0.4615242923518723, "learning_rate": 8.790067966014044e-07, "loss": 0.058, "step": 17542 }, { "epoch": 0.8151951672862453, "grad_norm": 0.4536675907800141, "learning_rate": 8.785806943710934e-07, "loss": 0.0476, "step": 17543 }, { "epoch": 0.8152416356877323, "grad_norm": 0.42459635190438805, "learning_rate": 8.781546854952683e-07, "loss": 0.0363, "step": 17544 }, { "epoch": 0.8152881040892194, "grad_norm": 0.4994826009546738, "learning_rate": 8.77728769983579e-07, "loss": 0.0596, "step": 17545 }, { "epoch": 0.8153345724907063, "grad_norm": 1.2905373614767695, "learning_rate": 8.773029478456696e-07, "loss": 0.0447, "step": 17546 }, { "epoch": 0.8153810408921933, "grad_norm": 0.3210739638619285, "learning_rate": 8.768772190911884e-07, "loss": 0.0379, "step": 17547 }, { "epoch": 0.8154275092936804, "grad_norm": 0.7351954919055715, "learning_rate": 8.764515837297777e-07, "loss": 0.0534, "step": 17548 }, { "epoch": 0.8154739776951673, "grad_norm": 0.6442867696009328, "learning_rate": 8.760260417710786e-07, "loss": 0.0466, "step": 17549 }, { "epoch": 0.8155204460966543, "grad_norm": 0.5987326841141657, "learning_rate": 8.756005932247291e-07, "loss": 0.0544, "step": 17550 }, { "epoch": 0.8155669144981412, "grad_norm": 0.4651279044893108, "learning_rate": 8.751752381003692e-07, "loss": 0.0376, "step": 17551 }, { "epoch": 0.8156133828996283, "grad_norm": 0.3614823612220983, "learning_rate": 8.7474997640763e-07, "loss": 0.0319, "step": 17552 }, { "epoch": 0.8156598513011153, "grad_norm": 0.285836013749377, "learning_rate": 8.74324808156145e-07, "loss": 0.0384, "step": 17553 }, { "epoch": 0.8157063197026022, "grad_norm": 0.6425602469543894, "learning_rate": 8.73899733355546e-07, "loss": 0.0613, "step": 17554 }, { "epoch": 0.8157527881040892, "grad_norm": 0.4286308436521079, "learning_rate": 8.734747520154608e-07, "loss": 0.0456, "step": 17555 }, { "epoch": 0.8157992565055762, "grad_norm": 0.3893135492774164, "learning_rate": 8.73049864145516e-07, "loss": 0.0392, "step": 17556 }, { "epoch": 0.8158457249070632, "grad_norm": 0.4151522474522741, "learning_rate": 8.726250697553345e-07, "loss": 0.0651, "step": 17557 }, { "epoch": 0.8158921933085502, "grad_norm": 1.3052175510810573, "learning_rate": 8.722003688545394e-07, "loss": 0.0791, "step": 17558 }, { "epoch": 0.8159386617100371, "grad_norm": 0.2858514554622739, "learning_rate": 8.717757614527506e-07, "loss": 0.0428, "step": 17559 }, { "epoch": 0.8159851301115242, "grad_norm": 0.4787857964442074, "learning_rate": 8.713512475595848e-07, "loss": 0.0553, "step": 17560 }, { "epoch": 0.8160315985130111, "grad_norm": 0.4358805434543752, "learning_rate": 8.709268271846593e-07, "loss": 0.0435, "step": 17561 }, { "epoch": 0.8160780669144981, "grad_norm": 0.4937412903400991, "learning_rate": 8.705025003375866e-07, "loss": 0.0443, "step": 17562 }, { "epoch": 0.8161245353159852, "grad_norm": 0.45229815146165436, "learning_rate": 8.700782670279795e-07, "loss": 0.0375, "step": 17563 }, { "epoch": 0.8161710037174721, "grad_norm": 0.5895730917061869, "learning_rate": 8.696541272654458e-07, "loss": 0.0521, "step": 17564 }, { "epoch": 0.8162174721189591, "grad_norm": 0.34367386342632955, "learning_rate": 8.692300810595927e-07, "loss": 0.0261, "step": 17565 }, { "epoch": 0.8162639405204462, "grad_norm": 0.5751021968585005, "learning_rate": 8.688061284200266e-07, "loss": 0.05, "step": 17566 }, { "epoch": 0.8163104089219331, "grad_norm": 0.44032907800521737, "learning_rate": 8.683822693563482e-07, "loss": 0.0656, "step": 17567 }, { "epoch": 0.8163568773234201, "grad_norm": 0.6109504376177379, "learning_rate": 8.679585038781607e-07, "loss": 0.0425, "step": 17568 }, { "epoch": 0.816403345724907, "grad_norm": 0.6238342471157178, "learning_rate": 8.675348319950638e-07, "loss": 0.0552, "step": 17569 }, { "epoch": 0.8164498141263941, "grad_norm": 0.22368061880636617, "learning_rate": 8.671112537166493e-07, "loss": 0.0318, "step": 17570 }, { "epoch": 0.8164962825278811, "grad_norm": 0.4356721225665293, "learning_rate": 8.66687769052516e-07, "loss": 0.048, "step": 17571 }, { "epoch": 0.816542750929368, "grad_norm": 0.5952566355488355, "learning_rate": 8.662643780122554e-07, "loss": 0.0535, "step": 17572 }, { "epoch": 0.816589219330855, "grad_norm": 0.6238822523550929, "learning_rate": 8.658410806054568e-07, "loss": 0.0612, "step": 17573 }, { "epoch": 0.816635687732342, "grad_norm": 0.5794027012151495, "learning_rate": 8.654178768417082e-07, "loss": 0.0466, "step": 17574 }, { "epoch": 0.816682156133829, "grad_norm": 0.4383459024493584, "learning_rate": 8.649947667305986e-07, "loss": 0.0611, "step": 17575 }, { "epoch": 0.816728624535316, "grad_norm": 0.5292330835737962, "learning_rate": 8.645717502817081e-07, "loss": 0.0584, "step": 17576 }, { "epoch": 0.816775092936803, "grad_norm": 0.5182178953166914, "learning_rate": 8.641488275046184e-07, "loss": 0.0474, "step": 17577 }, { "epoch": 0.81682156133829, "grad_norm": 0.5913150049476018, "learning_rate": 8.637259984089125e-07, "loss": 0.0408, "step": 17578 }, { "epoch": 0.8168680297397769, "grad_norm": 0.3840374453379663, "learning_rate": 8.633032630041654e-07, "loss": 0.0574, "step": 17579 }, { "epoch": 0.8169144981412639, "grad_norm": 0.4508078453316077, "learning_rate": 8.628806212999536e-07, "loss": 0.061, "step": 17580 }, { "epoch": 0.816960966542751, "grad_norm": 0.439754806734011, "learning_rate": 8.624580733058502e-07, "loss": 0.0438, "step": 17581 }, { "epoch": 0.8170074349442379, "grad_norm": 0.5683352352084025, "learning_rate": 8.620356190314256e-07, "loss": 0.0434, "step": 17582 }, { "epoch": 0.8170539033457249, "grad_norm": 0.63585093981512, "learning_rate": 8.616132584862497e-07, "loss": 0.0713, "step": 17583 }, { "epoch": 0.817100371747212, "grad_norm": 0.6624965646438149, "learning_rate": 8.611909916798878e-07, "loss": 0.054, "step": 17584 }, { "epoch": 0.8171468401486989, "grad_norm": 0.7294059803638421, "learning_rate": 8.607688186219071e-07, "loss": 0.0552, "step": 17585 }, { "epoch": 0.8171933085501859, "grad_norm": 0.41010504973896117, "learning_rate": 8.60346739321869e-07, "loss": 0.0646, "step": 17586 }, { "epoch": 0.8172397769516728, "grad_norm": 0.4960150018854577, "learning_rate": 8.599247537893346e-07, "loss": 0.0501, "step": 17587 }, { "epoch": 0.8172862453531599, "grad_norm": 0.33269795089592, "learning_rate": 8.595028620338613e-07, "loss": 0.0462, "step": 17588 }, { "epoch": 0.8173327137546469, "grad_norm": 0.7137710325025814, "learning_rate": 8.590810640650066e-07, "loss": 0.0563, "step": 17589 }, { "epoch": 0.8173791821561338, "grad_norm": 0.44313533361344465, "learning_rate": 8.586593598923238e-07, "loss": 0.0725, "step": 17590 }, { "epoch": 0.8174256505576208, "grad_norm": 0.688144193330654, "learning_rate": 8.582377495253635e-07, "loss": 0.0596, "step": 17591 }, { "epoch": 0.8174721189591078, "grad_norm": 0.3834873310531174, "learning_rate": 8.5781623297368e-07, "loss": 0.0372, "step": 17592 }, { "epoch": 0.8175185873605948, "grad_norm": 0.39793467896643814, "learning_rate": 8.573948102468177e-07, "loss": 0.0647, "step": 17593 }, { "epoch": 0.8175650557620818, "grad_norm": 0.5154175793932838, "learning_rate": 8.569734813543212e-07, "loss": 0.0469, "step": 17594 }, { "epoch": 0.8176115241635687, "grad_norm": 0.5372203046860036, "learning_rate": 8.565522463057368e-07, "loss": 0.0453, "step": 17595 }, { "epoch": 0.8176579925650558, "grad_norm": 0.46340550174838707, "learning_rate": 8.561311051106053e-07, "loss": 0.0504, "step": 17596 }, { "epoch": 0.8177044609665427, "grad_norm": 0.6433278125984746, "learning_rate": 8.557100577784655e-07, "loss": 0.0645, "step": 17597 }, { "epoch": 0.8177509293680297, "grad_norm": 0.4416746261804209, "learning_rate": 8.552891043188549e-07, "loss": 0.0525, "step": 17598 }, { "epoch": 0.8177973977695168, "grad_norm": 0.47172930241574496, "learning_rate": 8.548682447413082e-07, "loss": 0.044, "step": 17599 }, { "epoch": 0.8178438661710037, "grad_norm": 0.5034426198544948, "learning_rate": 8.544474790553587e-07, "loss": 0.0348, "step": 17600 }, { "epoch": 0.8178903345724907, "grad_norm": 0.6360676260450818, "learning_rate": 8.540268072705348e-07, "loss": 0.0504, "step": 17601 }, { "epoch": 0.8179368029739778, "grad_norm": 0.6541274431755006, "learning_rate": 8.536062293963693e-07, "loss": 0.0439, "step": 17602 }, { "epoch": 0.8179832713754647, "grad_norm": 0.4792028173564189, "learning_rate": 8.531857454423865e-07, "loss": 0.0403, "step": 17603 }, { "epoch": 0.8180297397769517, "grad_norm": 0.5969629474250353, "learning_rate": 8.527653554181109e-07, "loss": 0.0497, "step": 17604 }, { "epoch": 0.8180762081784386, "grad_norm": 0.4245566555360342, "learning_rate": 8.523450593330651e-07, "loss": 0.0532, "step": 17605 }, { "epoch": 0.8181226765799257, "grad_norm": 0.6758555574538342, "learning_rate": 8.519248571967693e-07, "loss": 0.0659, "step": 17606 }, { "epoch": 0.8181691449814127, "grad_norm": 0.5777026857470785, "learning_rate": 8.515047490187406e-07, "loss": 0.0553, "step": 17607 }, { "epoch": 0.8182156133828996, "grad_norm": 0.487744273719521, "learning_rate": 8.510847348084944e-07, "loss": 0.068, "step": 17608 }, { "epoch": 0.8182620817843866, "grad_norm": 0.44886349997987823, "learning_rate": 8.506648145755469e-07, "loss": 0.0517, "step": 17609 }, { "epoch": 0.8183085501858736, "grad_norm": 0.4554662410096959, "learning_rate": 8.5024498832941e-07, "loss": 0.0631, "step": 17610 }, { "epoch": 0.8183550185873606, "grad_norm": 0.6571531230060069, "learning_rate": 8.498252560795888e-07, "loss": 0.0621, "step": 17611 }, { "epoch": 0.8184014869888476, "grad_norm": 0.5289378981601511, "learning_rate": 8.494056178355953e-07, "loss": 0.0471, "step": 17612 }, { "epoch": 0.8184479553903345, "grad_norm": 0.4129632545728372, "learning_rate": 8.489860736069327e-07, "loss": 0.0381, "step": 17613 }, { "epoch": 0.8184944237918216, "grad_norm": 0.45945644975284955, "learning_rate": 8.485666234031042e-07, "loss": 0.0355, "step": 17614 }, { "epoch": 0.8185408921933085, "grad_norm": 0.34714107246880227, "learning_rate": 8.4814726723361e-07, "loss": 0.0348, "step": 17615 }, { "epoch": 0.8185873605947955, "grad_norm": 0.5880985764608779, "learning_rate": 8.477280051079522e-07, "loss": 0.0592, "step": 17616 }, { "epoch": 0.8186338289962826, "grad_norm": 0.3678400456880697, "learning_rate": 8.473088370356242e-07, "loss": 0.0588, "step": 17617 }, { "epoch": 0.8186802973977695, "grad_norm": 0.6292794491604693, "learning_rate": 8.468897630261203e-07, "loss": 0.0625, "step": 17618 }, { "epoch": 0.8187267657992565, "grad_norm": 0.5080870273552934, "learning_rate": 8.464707830889357e-07, "loss": 0.0511, "step": 17619 }, { "epoch": 0.8187732342007435, "grad_norm": 0.3572160845246143, "learning_rate": 8.460518972335591e-07, "loss": 0.0262, "step": 17620 }, { "epoch": 0.8188197026022305, "grad_norm": 0.5039626260981389, "learning_rate": 8.456331054694794e-07, "loss": 0.0504, "step": 17621 }, { "epoch": 0.8188661710037175, "grad_norm": 0.6573508173162653, "learning_rate": 8.45214407806182e-07, "loss": 0.0651, "step": 17622 }, { "epoch": 0.8189126394052044, "grad_norm": 0.46330879447389806, "learning_rate": 8.447958042531512e-07, "loss": 0.0583, "step": 17623 }, { "epoch": 0.8189591078066915, "grad_norm": 0.6090810352989298, "learning_rate": 8.443772948198686e-07, "loss": 0.0409, "step": 17624 }, { "epoch": 0.8190055762081785, "grad_norm": 0.4269547693689669, "learning_rate": 8.439588795158127e-07, "loss": 0.0413, "step": 17625 }, { "epoch": 0.8190520446096654, "grad_norm": 0.7123929685687165, "learning_rate": 8.435405583504636e-07, "loss": 0.0693, "step": 17626 }, { "epoch": 0.8190985130111524, "grad_norm": 0.5317633374427954, "learning_rate": 8.431223313332953e-07, "loss": 0.0514, "step": 17627 }, { "epoch": 0.8191449814126394, "grad_norm": 0.5364209292121404, "learning_rate": 8.427041984737816e-07, "loss": 0.0609, "step": 17628 }, { "epoch": 0.8191914498141264, "grad_norm": 0.6227745653929548, "learning_rate": 8.42286159781393e-07, "loss": 0.068, "step": 17629 }, { "epoch": 0.8192379182156134, "grad_norm": 0.3321490560393465, "learning_rate": 8.418682152655988e-07, "loss": 0.053, "step": 17630 }, { "epoch": 0.8192843866171003, "grad_norm": 0.43386006358000195, "learning_rate": 8.414503649358663e-07, "loss": 0.056, "step": 17631 }, { "epoch": 0.8193308550185874, "grad_norm": 0.3912330402389509, "learning_rate": 8.410326088016579e-07, "loss": 0.0421, "step": 17632 }, { "epoch": 0.8193773234200743, "grad_norm": 0.35813013471111593, "learning_rate": 8.406149468724401e-07, "loss": 0.0491, "step": 17633 }, { "epoch": 0.8194237918215613, "grad_norm": 0.5644295687425169, "learning_rate": 8.401973791576723e-07, "loss": 0.0788, "step": 17634 }, { "epoch": 0.8194702602230484, "grad_norm": 0.350528041068897, "learning_rate": 8.397799056668094e-07, "loss": 0.039, "step": 17635 }, { "epoch": 0.8195167286245353, "grad_norm": 0.4901776159621978, "learning_rate": 8.393625264093114e-07, "loss": 0.0623, "step": 17636 }, { "epoch": 0.8195631970260223, "grad_norm": 0.4897445836018016, "learning_rate": 8.389452413946314e-07, "loss": 0.0443, "step": 17637 }, { "epoch": 0.8196096654275093, "grad_norm": 0.5858845618103784, "learning_rate": 8.385280506322213e-07, "loss": 0.0557, "step": 17638 }, { "epoch": 0.8196561338289963, "grad_norm": 0.5080290898675163, "learning_rate": 8.381109541315291e-07, "loss": 0.0536, "step": 17639 }, { "epoch": 0.8197026022304833, "grad_norm": 0.5493843232175096, "learning_rate": 8.376939519020067e-07, "loss": 0.0299, "step": 17640 }, { "epoch": 0.8197490706319702, "grad_norm": 0.6356677994969355, "learning_rate": 8.372770439530958e-07, "loss": 0.0572, "step": 17641 }, { "epoch": 0.8197955390334573, "grad_norm": 0.4423882273503729, "learning_rate": 8.368602302942397e-07, "loss": 0.0455, "step": 17642 }, { "epoch": 0.8198420074349443, "grad_norm": 0.4032894901738115, "learning_rate": 8.364435109348822e-07, "loss": 0.0524, "step": 17643 }, { "epoch": 0.8198884758364312, "grad_norm": 0.522813957819945, "learning_rate": 8.360268858844617e-07, "loss": 0.0449, "step": 17644 }, { "epoch": 0.8199349442379182, "grad_norm": 0.5267843370077263, "learning_rate": 8.356103551524142e-07, "loss": 0.0473, "step": 17645 }, { "epoch": 0.8199814126394052, "grad_norm": 0.42725069086357886, "learning_rate": 8.351939187481755e-07, "loss": 0.0395, "step": 17646 }, { "epoch": 0.8200278810408922, "grad_norm": 0.6581752850463145, "learning_rate": 8.347775766811778e-07, "loss": 0.0687, "step": 17647 }, { "epoch": 0.8200743494423792, "grad_norm": 0.6722735171400968, "learning_rate": 8.343613289608515e-07, "loss": 0.0499, "step": 17648 }, { "epoch": 0.8201208178438661, "grad_norm": 0.7967066215485996, "learning_rate": 8.339451755966243e-07, "loss": 0.0704, "step": 17649 }, { "epoch": 0.8201672862453532, "grad_norm": 0.5074354745366639, "learning_rate": 8.335291165979248e-07, "loss": 0.0561, "step": 17650 }, { "epoch": 0.8202137546468401, "grad_norm": 0.45596083321700004, "learning_rate": 8.331131519741775e-07, "loss": 0.0467, "step": 17651 }, { "epoch": 0.8202602230483271, "grad_norm": 0.5793803282078503, "learning_rate": 8.326972817348e-07, "loss": 0.0486, "step": 17652 }, { "epoch": 0.8203066914498142, "grad_norm": 0.5244978833771627, "learning_rate": 8.322815058892165e-07, "loss": 0.0615, "step": 17653 }, { "epoch": 0.8203531598513011, "grad_norm": 0.5712000909712742, "learning_rate": 8.318658244468431e-07, "loss": 0.063, "step": 17654 }, { "epoch": 0.8203996282527881, "grad_norm": 0.7127411619098015, "learning_rate": 8.314502374170963e-07, "loss": 0.0608, "step": 17655 }, { "epoch": 0.8204460966542751, "grad_norm": 0.5650857151265772, "learning_rate": 8.31034744809387e-07, "loss": 0.0541, "step": 17656 }, { "epoch": 0.8204925650557621, "grad_norm": 0.48150898043540025, "learning_rate": 8.306193466331314e-07, "loss": 0.0428, "step": 17657 }, { "epoch": 0.8205390334572491, "grad_norm": 0.3127750786895445, "learning_rate": 8.302040428977343e-07, "loss": 0.0295, "step": 17658 }, { "epoch": 0.820585501858736, "grad_norm": 0.5176605029658317, "learning_rate": 8.29788833612603e-07, "loss": 0.0644, "step": 17659 }, { "epoch": 0.820631970260223, "grad_norm": 0.5341669644194668, "learning_rate": 8.293737187871448e-07, "loss": 0.0472, "step": 17660 }, { "epoch": 0.8206784386617101, "grad_norm": 0.5266566462547887, "learning_rate": 8.289586984307618e-07, "loss": 0.0532, "step": 17661 }, { "epoch": 0.820724907063197, "grad_norm": 0.5496225767097981, "learning_rate": 8.285437725528539e-07, "loss": 0.0581, "step": 17662 }, { "epoch": 0.820771375464684, "grad_norm": 0.6969493039143723, "learning_rate": 8.281289411628202e-07, "loss": 0.0456, "step": 17663 }, { "epoch": 0.820817843866171, "grad_norm": 0.31161243351026424, "learning_rate": 8.277142042700564e-07, "loss": 0.0309, "step": 17664 }, { "epoch": 0.820864312267658, "grad_norm": 0.46927301343748756, "learning_rate": 8.272995618839574e-07, "loss": 0.0447, "step": 17665 }, { "epoch": 0.820910780669145, "grad_norm": 0.38917064474763885, "learning_rate": 8.268850140139139e-07, "loss": 0.0453, "step": 17666 }, { "epoch": 0.820957249070632, "grad_norm": 0.5363557384529846, "learning_rate": 8.26470560669318e-07, "loss": 0.0652, "step": 17667 }, { "epoch": 0.821003717472119, "grad_norm": 0.5294800307412769, "learning_rate": 8.260562018595564e-07, "loss": 0.049, "step": 17668 }, { "epoch": 0.8210501858736059, "grad_norm": 0.7511249740173903, "learning_rate": 8.256419375940151e-07, "loss": 0.0522, "step": 17669 }, { "epoch": 0.8210966542750929, "grad_norm": 0.5633344476016253, "learning_rate": 8.252277678820769e-07, "loss": 0.0552, "step": 17670 }, { "epoch": 0.82114312267658, "grad_norm": 0.5001991107766355, "learning_rate": 8.248136927331241e-07, "loss": 0.0604, "step": 17671 }, { "epoch": 0.8211895910780669, "grad_norm": 0.6945838775599785, "learning_rate": 8.24399712156535e-07, "loss": 0.0735, "step": 17672 }, { "epoch": 0.8212360594795539, "grad_norm": 0.5772361013438657, "learning_rate": 8.239858261616862e-07, "loss": 0.0324, "step": 17673 }, { "epoch": 0.821282527881041, "grad_norm": 0.5037772682284344, "learning_rate": 8.235720347579551e-07, "loss": 0.0613, "step": 17674 }, { "epoch": 0.8213289962825279, "grad_norm": 0.6068524912888364, "learning_rate": 8.231583379547137e-07, "loss": 0.0524, "step": 17675 }, { "epoch": 0.8213754646840149, "grad_norm": 0.7022789460930181, "learning_rate": 8.2274473576133e-07, "loss": 0.0603, "step": 17676 }, { "epoch": 0.8214219330855018, "grad_norm": 0.5031303776960004, "learning_rate": 8.223312281871754e-07, "loss": 0.049, "step": 17677 }, { "epoch": 0.8214684014869889, "grad_norm": 0.6917004336311023, "learning_rate": 8.219178152416158e-07, "loss": 0.0667, "step": 17678 }, { "epoch": 0.8215148698884759, "grad_norm": 0.6936273905489229, "learning_rate": 8.215044969340146e-07, "loss": 0.0454, "step": 17679 }, { "epoch": 0.8215613382899628, "grad_norm": 0.3304999554277844, "learning_rate": 8.210912732737331e-07, "loss": 0.0328, "step": 17680 }, { "epoch": 0.8216078066914498, "grad_norm": 0.4245350415348589, "learning_rate": 8.206781442701351e-07, "loss": 0.0516, "step": 17681 }, { "epoch": 0.8216542750929368, "grad_norm": 0.5724039716031036, "learning_rate": 8.202651099325748e-07, "loss": 0.0505, "step": 17682 }, { "epoch": 0.8217007434944238, "grad_norm": 0.5525620857404564, "learning_rate": 8.198521702704071e-07, "loss": 0.0526, "step": 17683 }, { "epoch": 0.8217472118959108, "grad_norm": 0.6089262334441168, "learning_rate": 8.194393252929889e-07, "loss": 0.0562, "step": 17684 }, { "epoch": 0.8217936802973977, "grad_norm": 0.7427304419604055, "learning_rate": 8.190265750096699e-07, "loss": 0.0489, "step": 17685 }, { "epoch": 0.8218401486988848, "grad_norm": 0.9551728503954183, "learning_rate": 8.186139194297993e-07, "loss": 0.0561, "step": 17686 }, { "epoch": 0.8218866171003717, "grad_norm": 0.6224680490377107, "learning_rate": 8.182013585627241e-07, "loss": 0.0496, "step": 17687 }, { "epoch": 0.8219330855018587, "grad_norm": 0.2784657884890208, "learning_rate": 8.177888924177895e-07, "loss": 0.0546, "step": 17688 }, { "epoch": 0.8219795539033458, "grad_norm": 0.4420556397465218, "learning_rate": 8.173765210043382e-07, "loss": 0.0517, "step": 17689 }, { "epoch": 0.8220260223048327, "grad_norm": 0.5680710528293648, "learning_rate": 8.169642443317099e-07, "loss": 0.0642, "step": 17690 }, { "epoch": 0.8220724907063197, "grad_norm": 0.8557789511956919, "learning_rate": 8.165520624092449e-07, "loss": 0.0555, "step": 17691 }, { "epoch": 0.8221189591078067, "grad_norm": 0.7238769458593357, "learning_rate": 8.161399752462789e-07, "loss": 0.0535, "step": 17692 }, { "epoch": 0.8221654275092937, "grad_norm": 0.6334611134660929, "learning_rate": 8.157279828521458e-07, "loss": 0.0504, "step": 17693 }, { "epoch": 0.8222118959107807, "grad_norm": 0.4054100744616316, "learning_rate": 8.15316085236178e-07, "loss": 0.0458, "step": 17694 }, { "epoch": 0.8222583643122676, "grad_norm": 0.511875741441803, "learning_rate": 8.149042824077047e-07, "loss": 0.0667, "step": 17695 }, { "epoch": 0.8223048327137547, "grad_norm": 0.732682198141487, "learning_rate": 8.144925743760545e-07, "loss": 0.0509, "step": 17696 }, { "epoch": 0.8223513011152417, "grad_norm": 0.28848107525929656, "learning_rate": 8.140809611505513e-07, "loss": 0.0377, "step": 17697 }, { "epoch": 0.8223977695167286, "grad_norm": 0.7896431613215439, "learning_rate": 8.136694427405211e-07, "loss": 0.0664, "step": 17698 }, { "epoch": 0.8224442379182156, "grad_norm": 0.3649861120649746, "learning_rate": 8.132580191552853e-07, "loss": 0.0403, "step": 17699 }, { "epoch": 0.8224907063197026, "grad_norm": 0.5294885378405663, "learning_rate": 8.128466904041593e-07, "loss": 0.0524, "step": 17700 }, { "epoch": 0.8225371747211896, "grad_norm": 0.3479363192857228, "learning_rate": 8.124354564964637e-07, "loss": 0.0459, "step": 17701 }, { "epoch": 0.8225836431226766, "grad_norm": 0.3810200269982051, "learning_rate": 8.120243174415127e-07, "loss": 0.054, "step": 17702 }, { "epoch": 0.8226301115241635, "grad_norm": 0.43374100434145907, "learning_rate": 8.116132732486187e-07, "loss": 0.0408, "step": 17703 }, { "epoch": 0.8226765799256506, "grad_norm": 0.5324413915003353, "learning_rate": 8.112023239270911e-07, "loss": 0.0551, "step": 17704 }, { "epoch": 0.8227230483271375, "grad_norm": 0.42521175587274285, "learning_rate": 8.107914694862418e-07, "loss": 0.0315, "step": 17705 }, { "epoch": 0.8227695167286245, "grad_norm": 0.7772322065168048, "learning_rate": 8.103807099353733e-07, "loss": 0.0616, "step": 17706 }, { "epoch": 0.8228159851301116, "grad_norm": 0.41625770684797564, "learning_rate": 8.099700452837905e-07, "loss": 0.0395, "step": 17707 }, { "epoch": 0.8228624535315985, "grad_norm": 0.6166274068919076, "learning_rate": 8.095594755407971e-07, "loss": 0.0707, "step": 17708 }, { "epoch": 0.8229089219330855, "grad_norm": 0.537605802337476, "learning_rate": 8.091490007156916e-07, "loss": 0.056, "step": 17709 }, { "epoch": 0.8229553903345725, "grad_norm": 0.6451752761725343, "learning_rate": 8.087386208177722e-07, "loss": 0.0534, "step": 17710 }, { "epoch": 0.8230018587360595, "grad_norm": 0.3724681462740733, "learning_rate": 8.083283358563343e-07, "loss": 0.051, "step": 17711 }, { "epoch": 0.8230483271375465, "grad_norm": 0.4485903924376202, "learning_rate": 8.079181458406715e-07, "loss": 0.0546, "step": 17712 }, { "epoch": 0.8230947955390334, "grad_norm": 0.7934206519708809, "learning_rate": 8.075080507800748e-07, "loss": 0.0586, "step": 17713 }, { "epoch": 0.8231412639405205, "grad_norm": 0.4641954835758938, "learning_rate": 8.070980506838316e-07, "loss": 0.0507, "step": 17714 }, { "epoch": 0.8231877323420075, "grad_norm": 0.46233969407897924, "learning_rate": 8.066881455612319e-07, "loss": 0.0432, "step": 17715 }, { "epoch": 0.8232342007434944, "grad_norm": 0.4599662034246647, "learning_rate": 8.062783354215603e-07, "loss": 0.0363, "step": 17716 }, { "epoch": 0.8232806691449814, "grad_norm": 0.5347207476971212, "learning_rate": 8.058686202740961e-07, "loss": 0.044, "step": 17717 }, { "epoch": 0.8233271375464684, "grad_norm": 0.6252588928010885, "learning_rate": 8.054590001281226e-07, "loss": 0.0528, "step": 17718 }, { "epoch": 0.8233736059479554, "grad_norm": 0.7345272059535894, "learning_rate": 8.050494749929178e-07, "loss": 0.0448, "step": 17719 }, { "epoch": 0.8234200743494424, "grad_norm": 0.5732580045752819, "learning_rate": 8.046400448777575e-07, "loss": 0.0709, "step": 17720 }, { "epoch": 0.8234665427509293, "grad_norm": 0.34485845641293983, "learning_rate": 8.042307097919144e-07, "loss": 0.0403, "step": 17721 }, { "epoch": 0.8235130111524164, "grad_norm": 0.5065406061994069, "learning_rate": 8.038214697446639e-07, "loss": 0.0661, "step": 17722 }, { "epoch": 0.8235594795539033, "grad_norm": 0.5573777926211365, "learning_rate": 8.034123247452725e-07, "loss": 0.0598, "step": 17723 }, { "epoch": 0.8236059479553903, "grad_norm": 0.6245394854213068, "learning_rate": 8.030032748030076e-07, "loss": 0.0516, "step": 17724 }, { "epoch": 0.8236524163568774, "grad_norm": 0.5267284238229168, "learning_rate": 8.025943199271368e-07, "loss": 0.0575, "step": 17725 }, { "epoch": 0.8236988847583643, "grad_norm": 0.3719288893883828, "learning_rate": 8.021854601269225e-07, "loss": 0.0458, "step": 17726 }, { "epoch": 0.8237453531598513, "grad_norm": 0.5692690714632119, "learning_rate": 8.017766954116246e-07, "loss": 0.0457, "step": 17727 }, { "epoch": 0.8237918215613383, "grad_norm": 0.34477340733282014, "learning_rate": 8.013680257905055e-07, "loss": 0.0424, "step": 17728 }, { "epoch": 0.8238382899628253, "grad_norm": 0.5953432076075047, "learning_rate": 8.00959451272818e-07, "loss": 0.0387, "step": 17729 }, { "epoch": 0.8238847583643123, "grad_norm": 0.4402432084641882, "learning_rate": 8.00550971867819e-07, "loss": 0.0489, "step": 17730 }, { "epoch": 0.8239312267657992, "grad_norm": 0.461564632166614, "learning_rate": 8.001425875847585e-07, "loss": 0.0592, "step": 17731 }, { "epoch": 0.8239776951672863, "grad_norm": 0.6344543486948232, "learning_rate": 7.997342984328899e-07, "loss": 0.0601, "step": 17732 }, { "epoch": 0.8240241635687733, "grad_norm": 0.6649936318399975, "learning_rate": 7.9932610442146e-07, "loss": 0.0558, "step": 17733 }, { "epoch": 0.8240706319702602, "grad_norm": 0.5840820463935666, "learning_rate": 7.989180055597156e-07, "loss": 0.067, "step": 17734 }, { "epoch": 0.8241171003717472, "grad_norm": 0.35195826603802294, "learning_rate": 7.985100018568997e-07, "loss": 0.0445, "step": 17735 }, { "epoch": 0.8241635687732342, "grad_norm": 0.46706406042828325, "learning_rate": 7.981020933222539e-07, "loss": 0.0459, "step": 17736 }, { "epoch": 0.8242100371747212, "grad_norm": 0.6329783089973926, "learning_rate": 7.976942799650184e-07, "loss": 0.0533, "step": 17737 }, { "epoch": 0.8242565055762082, "grad_norm": 0.5556697087917523, "learning_rate": 7.972865617944292e-07, "loss": 0.0442, "step": 17738 }, { "epoch": 0.8243029739776951, "grad_norm": 0.5368607551019721, "learning_rate": 7.968789388197234e-07, "loss": 0.0539, "step": 17739 }, { "epoch": 0.8243494423791822, "grad_norm": 0.8115765501294057, "learning_rate": 7.964714110501348e-07, "loss": 0.0467, "step": 17740 }, { "epoch": 0.8243959107806691, "grad_norm": 0.5394284966475361, "learning_rate": 7.960639784948909e-07, "loss": 0.0544, "step": 17741 }, { "epoch": 0.8244423791821561, "grad_norm": 0.4436866911376694, "learning_rate": 7.956566411632233e-07, "loss": 0.0474, "step": 17742 }, { "epoch": 0.8244888475836432, "grad_norm": 0.6977120926130499, "learning_rate": 7.952493990643578e-07, "loss": 0.071, "step": 17743 }, { "epoch": 0.8245353159851301, "grad_norm": 0.5916905936425904, "learning_rate": 7.948422522075172e-07, "loss": 0.0526, "step": 17744 }, { "epoch": 0.8245817843866171, "grad_norm": 0.4206507381571294, "learning_rate": 7.944352006019274e-07, "loss": 0.045, "step": 17745 }, { "epoch": 0.8246282527881041, "grad_norm": 0.7052399378619153, "learning_rate": 7.940282442568071e-07, "loss": 0.0544, "step": 17746 }, { "epoch": 0.8246747211895911, "grad_norm": 0.40321343067435206, "learning_rate": 7.936213831813727e-07, "loss": 0.0335, "step": 17747 }, { "epoch": 0.8247211895910781, "grad_norm": 0.46051278534958595, "learning_rate": 7.932146173848404e-07, "loss": 0.0316, "step": 17748 }, { "epoch": 0.824767657992565, "grad_norm": 0.559624735041186, "learning_rate": 7.928079468764255e-07, "loss": 0.0408, "step": 17749 }, { "epoch": 0.824814126394052, "grad_norm": 0.802263422580228, "learning_rate": 7.924013716653384e-07, "loss": 0.0583, "step": 17750 }, { "epoch": 0.8248605947955391, "grad_norm": 0.566775747579134, "learning_rate": 7.919948917607873e-07, "loss": 0.0594, "step": 17751 }, { "epoch": 0.824907063197026, "grad_norm": 0.7876704324751472, "learning_rate": 7.915885071719837e-07, "loss": 0.055, "step": 17752 }, { "epoch": 0.824953531598513, "grad_norm": 0.663818949448008, "learning_rate": 7.91182217908128e-07, "loss": 0.0592, "step": 17753 }, { "epoch": 0.825, "grad_norm": 0.3060569409538673, "learning_rate": 7.90776023978424e-07, "loss": 0.0324, "step": 17754 }, { "epoch": 0.825046468401487, "grad_norm": 0.5411915166521215, "learning_rate": 7.903699253920738e-07, "loss": 0.07, "step": 17755 }, { "epoch": 0.825092936802974, "grad_norm": 0.5095949801930251, "learning_rate": 7.899639221582761e-07, "loss": 0.0439, "step": 17756 }, { "epoch": 0.8251394052044609, "grad_norm": 0.45663970440340784, "learning_rate": 7.895580142862258e-07, "loss": 0.0578, "step": 17757 }, { "epoch": 0.825185873605948, "grad_norm": 0.3793847201784484, "learning_rate": 7.891522017851183e-07, "loss": 0.0284, "step": 17758 }, { "epoch": 0.8252323420074349, "grad_norm": 0.28475425999684945, "learning_rate": 7.887464846641451e-07, "loss": 0.0367, "step": 17759 }, { "epoch": 0.8252788104089219, "grad_norm": 0.2901114848026418, "learning_rate": 7.883408629324968e-07, "loss": 0.0491, "step": 17760 }, { "epoch": 0.825325278810409, "grad_norm": 0.33958721207518555, "learning_rate": 7.879353365993586e-07, "loss": 0.0451, "step": 17761 }, { "epoch": 0.8253717472118959, "grad_norm": 0.6183581173470729, "learning_rate": 7.8752990567392e-07, "loss": 0.0535, "step": 17762 }, { "epoch": 0.8254182156133829, "grad_norm": 0.6409393089286726, "learning_rate": 7.871245701653624e-07, "loss": 0.0622, "step": 17763 }, { "epoch": 0.8254646840148699, "grad_norm": 0.4805926824925087, "learning_rate": 7.867193300828685e-07, "loss": 0.0526, "step": 17764 }, { "epoch": 0.8255111524163569, "grad_norm": 0.6181425976341274, "learning_rate": 7.863141854356138e-07, "loss": 0.0569, "step": 17765 }, { "epoch": 0.8255576208178439, "grad_norm": 0.5594018160831377, "learning_rate": 7.859091362327786e-07, "loss": 0.0441, "step": 17766 }, { "epoch": 0.8256040892193308, "grad_norm": 0.3010763940331743, "learning_rate": 7.855041824835369e-07, "loss": 0.0351, "step": 17767 }, { "epoch": 0.8256505576208178, "grad_norm": 0.5746607325010562, "learning_rate": 7.850993241970595e-07, "loss": 0.0588, "step": 17768 }, { "epoch": 0.8256970260223049, "grad_norm": 0.42045850267271084, "learning_rate": 7.846945613825213e-07, "loss": 0.043, "step": 17769 }, { "epoch": 0.8257434944237918, "grad_norm": 0.6540264859913067, "learning_rate": 7.842898940490868e-07, "loss": 0.055, "step": 17770 }, { "epoch": 0.8257899628252788, "grad_norm": 0.7587336872725109, "learning_rate": 7.838853222059217e-07, "loss": 0.0531, "step": 17771 }, { "epoch": 0.8258364312267658, "grad_norm": 0.427929754763314, "learning_rate": 7.834808458621929e-07, "loss": 0.0487, "step": 17772 }, { "epoch": 0.8258828996282528, "grad_norm": 0.5112262472170254, "learning_rate": 7.830764650270606e-07, "loss": 0.0391, "step": 17773 }, { "epoch": 0.8259293680297398, "grad_norm": 0.4178762065913888, "learning_rate": 7.826721797096842e-07, "loss": 0.0377, "step": 17774 }, { "epoch": 0.8259758364312267, "grad_norm": 0.5138306940564974, "learning_rate": 7.82267989919222e-07, "loss": 0.0407, "step": 17775 }, { "epoch": 0.8260223048327138, "grad_norm": 0.5071851119829476, "learning_rate": 7.818638956648289e-07, "loss": 0.0577, "step": 17776 }, { "epoch": 0.8260687732342007, "grad_norm": 0.2466428770910591, "learning_rate": 7.814598969556581e-07, "loss": 0.0246, "step": 17777 }, { "epoch": 0.8261152416356877, "grad_norm": 0.6057555594270582, "learning_rate": 7.810559938008588e-07, "loss": 0.0537, "step": 17778 }, { "epoch": 0.8261617100371748, "grad_norm": 0.386895564697083, "learning_rate": 7.806521862095834e-07, "loss": 0.039, "step": 17779 }, { "epoch": 0.8262081784386617, "grad_norm": 0.304940925529253, "learning_rate": 7.802484741909766e-07, "loss": 0.0366, "step": 17780 }, { "epoch": 0.8262546468401487, "grad_norm": 0.5297882924789885, "learning_rate": 7.798448577541834e-07, "loss": 0.0535, "step": 17781 }, { "epoch": 0.8263011152416357, "grad_norm": 0.7646140831465266, "learning_rate": 7.794413369083453e-07, "loss": 0.0492, "step": 17782 }, { "epoch": 0.8263475836431227, "grad_norm": 0.43221247173803334, "learning_rate": 7.790379116626029e-07, "loss": 0.0578, "step": 17783 }, { "epoch": 0.8263940520446097, "grad_norm": 0.4758289575078014, "learning_rate": 7.786345820260949e-07, "loss": 0.038, "step": 17784 }, { "epoch": 0.8264405204460966, "grad_norm": 1.7505697042933017, "learning_rate": 7.78231348007955e-07, "loss": 0.0484, "step": 17785 }, { "epoch": 0.8264869888475836, "grad_norm": 0.3620494615359632, "learning_rate": 7.778282096173196e-07, "loss": 0.0495, "step": 17786 }, { "epoch": 0.8265334572490707, "grad_norm": 0.5109542302362327, "learning_rate": 7.774251668633198e-07, "loss": 0.0527, "step": 17787 }, { "epoch": 0.8265799256505576, "grad_norm": 0.38967801172761735, "learning_rate": 7.770222197550825e-07, "loss": 0.0415, "step": 17788 }, { "epoch": 0.8266263940520446, "grad_norm": 0.4442389087512534, "learning_rate": 7.76619368301737e-07, "loss": 0.0488, "step": 17789 }, { "epoch": 0.8266728624535316, "grad_norm": 0.3769414382713555, "learning_rate": 7.762166125124082e-07, "loss": 0.038, "step": 17790 }, { "epoch": 0.8267193308550186, "grad_norm": 0.5004541951663951, "learning_rate": 7.758139523962188e-07, "loss": 0.0532, "step": 17791 }, { "epoch": 0.8267657992565056, "grad_norm": 0.5088135425126269, "learning_rate": 7.754113879622877e-07, "loss": 0.0577, "step": 17792 }, { "epoch": 0.8268122676579925, "grad_norm": 0.5135948857549559, "learning_rate": 7.750089192197374e-07, "loss": 0.0418, "step": 17793 }, { "epoch": 0.8268587360594796, "grad_norm": 0.3969264276071888, "learning_rate": 7.746065461776808e-07, "loss": 0.0433, "step": 17794 }, { "epoch": 0.8269052044609665, "grad_norm": 0.5928744461039129, "learning_rate": 7.742042688452317e-07, "loss": 0.0402, "step": 17795 }, { "epoch": 0.8269516728624535, "grad_norm": 0.8463447467658832, "learning_rate": 7.73802087231505e-07, "loss": 0.0645, "step": 17796 }, { "epoch": 0.8269981412639406, "grad_norm": 0.39491483094046315, "learning_rate": 7.734000013456089e-07, "loss": 0.0247, "step": 17797 }, { "epoch": 0.8270446096654275, "grad_norm": 0.9174798123444293, "learning_rate": 7.729980111966506e-07, "loss": 0.0563, "step": 17798 }, { "epoch": 0.8270910780669145, "grad_norm": 0.5822323440335045, "learning_rate": 7.725961167937368e-07, "loss": 0.0416, "step": 17799 }, { "epoch": 0.8271375464684015, "grad_norm": 0.4485701169493745, "learning_rate": 7.721943181459702e-07, "loss": 0.0328, "step": 17800 }, { "epoch": 0.8271840148698885, "grad_norm": 0.4170736111576029, "learning_rate": 7.717926152624516e-07, "loss": 0.0499, "step": 17801 }, { "epoch": 0.8272304832713755, "grad_norm": 0.5975045192476975, "learning_rate": 7.713910081522796e-07, "loss": 0.044, "step": 17802 }, { "epoch": 0.8272769516728624, "grad_norm": 0.4855741362715623, "learning_rate": 7.709894968245529e-07, "loss": 0.0436, "step": 17803 }, { "epoch": 0.8273234200743494, "grad_norm": 0.7193109511626683, "learning_rate": 7.705880812883648e-07, "loss": 0.0598, "step": 17804 }, { "epoch": 0.8273698884758365, "grad_norm": 0.6098295534330606, "learning_rate": 7.70186761552808e-07, "loss": 0.0596, "step": 17805 }, { "epoch": 0.8274163568773234, "grad_norm": 0.40496167943267386, "learning_rate": 7.69785537626973e-07, "loss": 0.0395, "step": 17806 }, { "epoch": 0.8274628252788104, "grad_norm": 1.301438223710374, "learning_rate": 7.693844095199482e-07, "loss": 0.079, "step": 17807 }, { "epoch": 0.8275092936802974, "grad_norm": 0.5875812410724333, "learning_rate": 7.689833772408184e-07, "loss": 0.0553, "step": 17808 }, { "epoch": 0.8275557620817844, "grad_norm": 0.5553311846162837, "learning_rate": 7.685824407986676e-07, "loss": 0.0422, "step": 17809 }, { "epoch": 0.8276022304832714, "grad_norm": 0.4339044063537075, "learning_rate": 7.681816002025788e-07, "loss": 0.0579, "step": 17810 }, { "epoch": 0.8276486988847583, "grad_norm": 1.0276147836893317, "learning_rate": 7.677808554616323e-07, "loss": 0.056, "step": 17811 }, { "epoch": 0.8276951672862454, "grad_norm": 0.45315448610706566, "learning_rate": 7.673802065849006e-07, "loss": 0.0419, "step": 17812 }, { "epoch": 0.8277416356877323, "grad_norm": 0.4467133389438779, "learning_rate": 7.669796535814639e-07, "loss": 0.0504, "step": 17813 }, { "epoch": 0.8277881040892193, "grad_norm": 0.33259350868280263, "learning_rate": 7.66579196460393e-07, "loss": 0.043, "step": 17814 }, { "epoch": 0.8278345724907064, "grad_norm": 0.4328742572286502, "learning_rate": 7.661788352307586e-07, "loss": 0.0495, "step": 17815 }, { "epoch": 0.8278810408921933, "grad_norm": 0.43250984736946974, "learning_rate": 7.65778569901629e-07, "loss": 0.0573, "step": 17816 }, { "epoch": 0.8279275092936803, "grad_norm": 0.3840682872393718, "learning_rate": 7.65378400482073e-07, "loss": 0.0542, "step": 17817 }, { "epoch": 0.8279739776951673, "grad_norm": 0.41009144715902024, "learning_rate": 7.649783269811523e-07, "loss": 0.0443, "step": 17818 }, { "epoch": 0.8280204460966543, "grad_norm": 0.3925618948046528, "learning_rate": 7.645783494079284e-07, "loss": 0.0381, "step": 17819 }, { "epoch": 0.8280669144981413, "grad_norm": 0.6827519388346878, "learning_rate": 7.641784677714642e-07, "loss": 0.0444, "step": 17820 }, { "epoch": 0.8281133828996282, "grad_norm": 0.6566605635122797, "learning_rate": 7.637786820808152e-07, "loss": 0.0503, "step": 17821 }, { "epoch": 0.8281598513011152, "grad_norm": 0.7232609438074712, "learning_rate": 7.63378992345038e-07, "loss": 0.0491, "step": 17822 }, { "epoch": 0.8282063197026023, "grad_norm": 0.5583062079265857, "learning_rate": 7.629793985731853e-07, "loss": 0.051, "step": 17823 }, { "epoch": 0.8282527881040892, "grad_norm": 0.5371293824803899, "learning_rate": 7.62579900774309e-07, "loss": 0.0529, "step": 17824 }, { "epoch": 0.8282992565055762, "grad_norm": 0.41052699373642043, "learning_rate": 7.621804989574572e-07, "loss": 0.0454, "step": 17825 }, { "epoch": 0.8283457249070632, "grad_norm": 0.4542254892310984, "learning_rate": 7.617811931316765e-07, "loss": 0.0563, "step": 17826 }, { "epoch": 0.8283921933085502, "grad_norm": 0.49658801685076986, "learning_rate": 7.613819833060137e-07, "loss": 0.0411, "step": 17827 }, { "epoch": 0.8284386617100372, "grad_norm": 0.8952580384577058, "learning_rate": 7.60982869489511e-07, "loss": 0.053, "step": 17828 }, { "epoch": 0.8284851301115241, "grad_norm": 0.41363379495741864, "learning_rate": 7.605838516912056e-07, "loss": 0.046, "step": 17829 }, { "epoch": 0.8285315985130112, "grad_norm": 0.5132491042876463, "learning_rate": 7.601849299201386e-07, "loss": 0.042, "step": 17830 }, { "epoch": 0.8285780669144981, "grad_norm": 0.34444625370172044, "learning_rate": 7.597861041853455e-07, "loss": 0.039, "step": 17831 }, { "epoch": 0.8286245353159851, "grad_norm": 0.3107906353175809, "learning_rate": 7.593873744958602e-07, "loss": 0.0531, "step": 17832 }, { "epoch": 0.8286710037174722, "grad_norm": 0.603644283071207, "learning_rate": 7.589887408607127e-07, "loss": 0.0441, "step": 17833 }, { "epoch": 0.8287174721189591, "grad_norm": 0.5647201968866207, "learning_rate": 7.585902032889358e-07, "loss": 0.052, "step": 17834 }, { "epoch": 0.8287639405204461, "grad_norm": 2.0883287942041, "learning_rate": 7.58191761789554e-07, "loss": 0.0621, "step": 17835 }, { "epoch": 0.8288104089219331, "grad_norm": 0.5143559231471483, "learning_rate": 7.577934163715922e-07, "loss": 0.0493, "step": 17836 }, { "epoch": 0.8288568773234201, "grad_norm": 0.7769118907879833, "learning_rate": 7.573951670440754e-07, "loss": 0.0655, "step": 17837 }, { "epoch": 0.8289033457249071, "grad_norm": 0.4055121213447772, "learning_rate": 7.569970138160232e-07, "loss": 0.0573, "step": 17838 }, { "epoch": 0.828949814126394, "grad_norm": 0.764486642459961, "learning_rate": 7.565989566964543e-07, "loss": 0.0466, "step": 17839 }, { "epoch": 0.828996282527881, "grad_norm": 0.35386073874144586, "learning_rate": 7.56200995694385e-07, "loss": 0.0529, "step": 17840 }, { "epoch": 0.8290427509293681, "grad_norm": 0.5598295665370515, "learning_rate": 7.558031308188296e-07, "loss": 0.0569, "step": 17841 }, { "epoch": 0.829089219330855, "grad_norm": 0.407664582761565, "learning_rate": 7.554053620788004e-07, "loss": 0.04, "step": 17842 }, { "epoch": 0.829135687732342, "grad_norm": 0.7797656105031687, "learning_rate": 7.550076894833064e-07, "loss": 0.0627, "step": 17843 }, { "epoch": 0.829182156133829, "grad_norm": 0.44906086730537215, "learning_rate": 7.546101130413569e-07, "loss": 0.0419, "step": 17844 }, { "epoch": 0.829228624535316, "grad_norm": 0.5840912936320342, "learning_rate": 7.542126327619559e-07, "loss": 0.0539, "step": 17845 }, { "epoch": 0.829275092936803, "grad_norm": 0.5646805373428452, "learning_rate": 7.538152486541078e-07, "loss": 0.0361, "step": 17846 }, { "epoch": 0.8293215613382899, "grad_norm": 0.8766318238194692, "learning_rate": 7.534179607268139e-07, "loss": 0.0621, "step": 17847 }, { "epoch": 0.829368029739777, "grad_norm": 0.4253223781874198, "learning_rate": 7.530207689890718e-07, "loss": 0.0491, "step": 17848 }, { "epoch": 0.8294144981412639, "grad_norm": 0.43950478909028945, "learning_rate": 7.526236734498799e-07, "loss": 0.0354, "step": 17849 }, { "epoch": 0.8294609665427509, "grad_norm": 0.42619770792627515, "learning_rate": 7.522266741182305e-07, "loss": 0.0576, "step": 17850 }, { "epoch": 0.829507434944238, "grad_norm": 0.5234333664210602, "learning_rate": 7.518297710031186e-07, "loss": 0.0554, "step": 17851 }, { "epoch": 0.8295539033457249, "grad_norm": 4.8770717241012616, "learning_rate": 7.514329641135349e-07, "loss": 0.0426, "step": 17852 }, { "epoch": 0.8296003717472119, "grad_norm": 0.6172143250910493, "learning_rate": 7.510362534584637e-07, "loss": 0.057, "step": 17853 }, { "epoch": 0.8296468401486989, "grad_norm": 0.4562761882262507, "learning_rate": 7.506396390468946e-07, "loss": 0.0516, "step": 17854 }, { "epoch": 0.8296933085501859, "grad_norm": 0.5207716292356731, "learning_rate": 7.502431208878097e-07, "loss": 0.0421, "step": 17855 }, { "epoch": 0.8297397769516729, "grad_norm": 0.5007988292267499, "learning_rate": 7.498466989901915e-07, "loss": 0.0455, "step": 17856 }, { "epoch": 0.8297862453531598, "grad_norm": 0.31713239136546095, "learning_rate": 7.494503733630171e-07, "loss": 0.0372, "step": 17857 }, { "epoch": 0.8298327137546468, "grad_norm": 0.551147171351582, "learning_rate": 7.490541440152682e-07, "loss": 0.0584, "step": 17858 }, { "epoch": 0.8298791821561339, "grad_norm": 0.33292646487751215, "learning_rate": 7.48658010955915e-07, "loss": 0.0491, "step": 17859 }, { "epoch": 0.8299256505576208, "grad_norm": 0.6994418498103451, "learning_rate": 7.482619741939318e-07, "loss": 0.0608, "step": 17860 }, { "epoch": 0.8299721189591078, "grad_norm": 0.3999211682130539, "learning_rate": 7.47866033738291e-07, "loss": 0.0415, "step": 17861 }, { "epoch": 0.8300185873605948, "grad_norm": 0.5231061244951305, "learning_rate": 7.474701895979591e-07, "loss": 0.0522, "step": 17862 }, { "epoch": 0.8300650557620818, "grad_norm": 0.8164274343493488, "learning_rate": 7.470744417819037e-07, "loss": 0.0686, "step": 17863 }, { "epoch": 0.8301115241635688, "grad_norm": 0.38253986472986584, "learning_rate": 7.466787902990885e-07, "loss": 0.0507, "step": 17864 }, { "epoch": 0.8301579925650557, "grad_norm": 0.43534485913143744, "learning_rate": 7.462832351584748e-07, "loss": 0.0455, "step": 17865 }, { "epoch": 0.8302044609665428, "grad_norm": 0.5725115726754199, "learning_rate": 7.458877763690231e-07, "loss": 0.0412, "step": 17866 }, { "epoch": 0.8302509293680297, "grad_norm": 0.4280583950111443, "learning_rate": 7.45492413939689e-07, "loss": 0.0522, "step": 17867 }, { "epoch": 0.8302973977695167, "grad_norm": 0.48932980165716633, "learning_rate": 7.450971478794311e-07, "loss": 0.0541, "step": 17868 }, { "epoch": 0.8303438661710038, "grad_norm": 0.6315644957523442, "learning_rate": 7.447019781972009e-07, "loss": 0.0703, "step": 17869 }, { "epoch": 0.8303903345724907, "grad_norm": 0.3975297326949541, "learning_rate": 7.443069049019497e-07, "loss": 0.0537, "step": 17870 }, { "epoch": 0.8304368029739777, "grad_norm": 0.43108082267482895, "learning_rate": 7.439119280026258e-07, "loss": 0.0455, "step": 17871 }, { "epoch": 0.8304832713754647, "grad_norm": 0.4734019157596387, "learning_rate": 7.43517047508176e-07, "loss": 0.0557, "step": 17872 }, { "epoch": 0.8305297397769517, "grad_norm": 0.6027309933314108, "learning_rate": 7.431222634275454e-07, "loss": 0.0378, "step": 17873 }, { "epoch": 0.8305762081784387, "grad_norm": 0.7969363774988302, "learning_rate": 7.427275757696745e-07, "loss": 0.0779, "step": 17874 }, { "epoch": 0.8306226765799256, "grad_norm": 0.36290844350738666, "learning_rate": 7.423329845435062e-07, "loss": 0.036, "step": 17875 }, { "epoch": 0.8306691449814126, "grad_norm": 0.731752591057487, "learning_rate": 7.419384897579773e-07, "loss": 0.0449, "step": 17876 }, { "epoch": 0.8307156133828997, "grad_norm": 0.5749147030981204, "learning_rate": 7.415440914220218e-07, "loss": 0.0661, "step": 17877 }, { "epoch": 0.8307620817843866, "grad_norm": 0.4609887906262899, "learning_rate": 7.411497895445751e-07, "loss": 0.0474, "step": 17878 }, { "epoch": 0.8308085501858736, "grad_norm": 0.5619483424198366, "learning_rate": 7.407555841345681e-07, "loss": 0.0481, "step": 17879 }, { "epoch": 0.8308550185873605, "grad_norm": 0.5358854360484482, "learning_rate": 7.403614752009297e-07, "loss": 0.052, "step": 17880 }, { "epoch": 0.8309014869888476, "grad_norm": 0.6313520415828617, "learning_rate": 7.399674627525861e-07, "loss": 0.0535, "step": 17881 }, { "epoch": 0.8309479553903346, "grad_norm": 0.5879815449299615, "learning_rate": 7.395735467984655e-07, "loss": 0.0491, "step": 17882 }, { "epoch": 0.8309944237918215, "grad_norm": 0.46149910063980765, "learning_rate": 7.391797273474871e-07, "loss": 0.0489, "step": 17883 }, { "epoch": 0.8310408921933086, "grad_norm": 0.6755257038997177, "learning_rate": 7.387860044085704e-07, "loss": 0.0522, "step": 17884 }, { "epoch": 0.8310873605947955, "grad_norm": 0.5040322742836975, "learning_rate": 7.383923779906377e-07, "loss": 0.0565, "step": 17885 }, { "epoch": 0.8311338289962825, "grad_norm": 0.3135995049889422, "learning_rate": 7.379988481026024e-07, "loss": 0.0507, "step": 17886 }, { "epoch": 0.8311802973977696, "grad_norm": 0.5224961937038543, "learning_rate": 7.376054147533784e-07, "loss": 0.0469, "step": 17887 }, { "epoch": 0.8312267657992565, "grad_norm": 0.5618083999037917, "learning_rate": 7.372120779518788e-07, "loss": 0.0401, "step": 17888 }, { "epoch": 0.8312732342007435, "grad_norm": 0.6257995055295192, "learning_rate": 7.368188377070112e-07, "loss": 0.059, "step": 17889 }, { "epoch": 0.8313197026022305, "grad_norm": 0.506946428017383, "learning_rate": 7.364256940276843e-07, "loss": 0.0605, "step": 17890 }, { "epoch": 0.8313661710037175, "grad_norm": 0.47893527349484527, "learning_rate": 7.360326469228018e-07, "loss": 0.0602, "step": 17891 }, { "epoch": 0.8314126394052045, "grad_norm": 0.4459826545816401, "learning_rate": 7.356396964012685e-07, "loss": 0.0421, "step": 17892 }, { "epoch": 0.8314591078066914, "grad_norm": 0.5396020280459379, "learning_rate": 7.352468424719856e-07, "loss": 0.0549, "step": 17893 }, { "epoch": 0.8315055762081784, "grad_norm": 0.5764383739535047, "learning_rate": 7.348540851438479e-07, "loss": 0.0546, "step": 17894 }, { "epoch": 0.8315520446096655, "grad_norm": 0.5142274802847241, "learning_rate": 7.344614244257547e-07, "loss": 0.0362, "step": 17895 }, { "epoch": 0.8315985130111524, "grad_norm": 0.5051138492121191, "learning_rate": 7.340688603266006e-07, "loss": 0.0539, "step": 17896 }, { "epoch": 0.8316449814126394, "grad_norm": 0.43314754964236185, "learning_rate": 7.336763928552759e-07, "loss": 0.0323, "step": 17897 }, { "epoch": 0.8316914498141263, "grad_norm": 0.39679950778238016, "learning_rate": 7.332840220206705e-07, "loss": 0.0456, "step": 17898 }, { "epoch": 0.8317379182156134, "grad_norm": 0.6072902546080352, "learning_rate": 7.328917478316744e-07, "loss": 0.0377, "step": 17899 }, { "epoch": 0.8317843866171004, "grad_norm": 0.5346001893366144, "learning_rate": 7.324995702971704e-07, "loss": 0.0395, "step": 17900 }, { "epoch": 0.8318308550185873, "grad_norm": 0.853429249768102, "learning_rate": 7.321074894260416e-07, "loss": 0.0563, "step": 17901 }, { "epoch": 0.8318773234200744, "grad_norm": 0.8337920106597504, "learning_rate": 7.317155052271713e-07, "loss": 0.0578, "step": 17902 }, { "epoch": 0.8319237918215613, "grad_norm": 0.38921764436279166, "learning_rate": 7.313236177094368e-07, "loss": 0.0377, "step": 17903 }, { "epoch": 0.8319702602230483, "grad_norm": 0.5300127911992516, "learning_rate": 7.30931826881715e-07, "loss": 0.0449, "step": 17904 }, { "epoch": 0.8320167286245354, "grad_norm": 0.4250191600946332, "learning_rate": 7.305401327528805e-07, "loss": 0.0275, "step": 17905 }, { "epoch": 0.8320631970260223, "grad_norm": 0.5691191602537105, "learning_rate": 7.301485353318056e-07, "loss": 0.0452, "step": 17906 }, { "epoch": 0.8321096654275093, "grad_norm": 0.6004436157589504, "learning_rate": 7.297570346273602e-07, "loss": 0.0343, "step": 17907 }, { "epoch": 0.8321561338289963, "grad_norm": 0.43548663458666687, "learning_rate": 7.293656306484109e-07, "loss": 0.0413, "step": 17908 }, { "epoch": 0.8322026022304833, "grad_norm": 0.674909086656309, "learning_rate": 7.289743234038266e-07, "loss": 0.0506, "step": 17909 }, { "epoch": 0.8322490706319703, "grad_norm": 0.46762895476023253, "learning_rate": 7.28583112902469e-07, "loss": 0.0446, "step": 17910 }, { "epoch": 0.8322955390334572, "grad_norm": 0.41441815788693454, "learning_rate": 7.281919991531988e-07, "loss": 0.0355, "step": 17911 }, { "epoch": 0.8323420074349442, "grad_norm": 0.4100702848813366, "learning_rate": 7.278009821648762e-07, "loss": 0.0336, "step": 17912 }, { "epoch": 0.8323884758364313, "grad_norm": 0.45106363037081715, "learning_rate": 7.274100619463581e-07, "loss": 0.0563, "step": 17913 }, { "epoch": 0.8324349442379182, "grad_norm": 0.5904718412958773, "learning_rate": 7.270192385064984e-07, "loss": 0.0508, "step": 17914 }, { "epoch": 0.8324814126394052, "grad_norm": 0.3452041102512647, "learning_rate": 7.266285118541494e-07, "loss": 0.0534, "step": 17915 }, { "epoch": 0.8325278810408921, "grad_norm": 0.7136464739750294, "learning_rate": 7.26237881998163e-07, "loss": 0.0531, "step": 17916 }, { "epoch": 0.8325743494423792, "grad_norm": 0.5559004979852926, "learning_rate": 7.258473489473883e-07, "loss": 0.0617, "step": 17917 }, { "epoch": 0.8326208178438662, "grad_norm": 0.39549269894769645, "learning_rate": 7.25456912710667e-07, "loss": 0.028, "step": 17918 }, { "epoch": 0.8326672862453531, "grad_norm": 0.6187209444513757, "learning_rate": 7.250665732968465e-07, "loss": 0.0416, "step": 17919 }, { "epoch": 0.8327137546468402, "grad_norm": 0.5014323725900195, "learning_rate": 7.246763307147675e-07, "loss": 0.061, "step": 17920 }, { "epoch": 0.8327602230483271, "grad_norm": 0.4692182012098971, "learning_rate": 7.242861849732696e-07, "loss": 0.045, "step": 17921 }, { "epoch": 0.8328066914498141, "grad_norm": 0.5154940180056674, "learning_rate": 7.238961360811886e-07, "loss": 0.0368, "step": 17922 }, { "epoch": 0.8328531598513012, "grad_norm": 0.7692461907723622, "learning_rate": 7.235061840473623e-07, "loss": 0.0491, "step": 17923 }, { "epoch": 0.8328996282527881, "grad_norm": 0.42476587190111875, "learning_rate": 7.231163288806215e-07, "loss": 0.0516, "step": 17924 }, { "epoch": 0.8329460966542751, "grad_norm": 0.3743158491109178, "learning_rate": 7.227265705897957e-07, "loss": 0.0486, "step": 17925 }, { "epoch": 0.8329925650557621, "grad_norm": 0.5018727970902331, "learning_rate": 7.223369091837157e-07, "loss": 0.0452, "step": 17926 }, { "epoch": 0.8330390334572491, "grad_norm": 0.48100860206023016, "learning_rate": 7.219473446712071e-07, "loss": 0.0451, "step": 17927 }, { "epoch": 0.8330855018587361, "grad_norm": 0.3625663688752777, "learning_rate": 7.215578770610937e-07, "loss": 0.0357, "step": 17928 }, { "epoch": 0.833131970260223, "grad_norm": 0.7078783665072602, "learning_rate": 7.211685063621976e-07, "loss": 0.0567, "step": 17929 }, { "epoch": 0.83317843866171, "grad_norm": 0.6936303118686922, "learning_rate": 7.207792325833379e-07, "loss": 0.0602, "step": 17930 }, { "epoch": 0.8332249070631971, "grad_norm": 0.5132655912139634, "learning_rate": 7.203900557333321e-07, "loss": 0.0402, "step": 17931 }, { "epoch": 0.833271375464684, "grad_norm": 0.5495384255077035, "learning_rate": 7.200009758209953e-07, "loss": 0.0515, "step": 17932 }, { "epoch": 0.833317843866171, "grad_norm": 0.510419761959949, "learning_rate": 7.196119928551415e-07, "loss": 0.0389, "step": 17933 }, { "epoch": 0.833364312267658, "grad_norm": 0.6285470181527611, "learning_rate": 7.192231068445815e-07, "loss": 0.073, "step": 17934 }, { "epoch": 0.833410780669145, "grad_norm": 0.37585498311758586, "learning_rate": 7.188343177981233e-07, "loss": 0.0421, "step": 17935 }, { "epoch": 0.833457249070632, "grad_norm": 0.7259916424882855, "learning_rate": 7.184456257245742e-07, "loss": 0.0665, "step": 17936 }, { "epoch": 0.8335037174721189, "grad_norm": 0.5829664483956251, "learning_rate": 7.180570306327372e-07, "loss": 0.0558, "step": 17937 }, { "epoch": 0.833550185873606, "grad_norm": 0.6088035252977229, "learning_rate": 7.176685325314159e-07, "loss": 0.0729, "step": 17938 }, { "epoch": 0.8335966542750929, "grad_norm": 0.598350311760162, "learning_rate": 7.172801314294075e-07, "loss": 0.0607, "step": 17939 }, { "epoch": 0.8336431226765799, "grad_norm": 0.6384901396077879, "learning_rate": 7.16891827335513e-07, "loss": 0.0761, "step": 17940 }, { "epoch": 0.833689591078067, "grad_norm": 0.371731240533372, "learning_rate": 7.16503620258528e-07, "loss": 0.0403, "step": 17941 }, { "epoch": 0.8337360594795539, "grad_norm": 0.7663488363464593, "learning_rate": 7.161155102072415e-07, "loss": 0.0627, "step": 17942 }, { "epoch": 0.8337825278810409, "grad_norm": 0.34632297001066137, "learning_rate": 7.157274971904488e-07, "loss": 0.0552, "step": 17943 }, { "epoch": 0.8338289962825279, "grad_norm": 0.6543149496910743, "learning_rate": 7.153395812169373e-07, "loss": 0.0666, "step": 17944 }, { "epoch": 0.8338754646840149, "grad_norm": 0.6274376035740424, "learning_rate": 7.149517622954944e-07, "loss": 0.0636, "step": 17945 }, { "epoch": 0.8339219330855019, "grad_norm": 0.33129913334191885, "learning_rate": 7.14564040434903e-07, "loss": 0.0313, "step": 17946 }, { "epoch": 0.8339684014869888, "grad_norm": 0.4509831630617526, "learning_rate": 7.141764156439473e-07, "loss": 0.0502, "step": 17947 }, { "epoch": 0.8340148698884758, "grad_norm": 0.5051777762108678, "learning_rate": 7.137888879314059e-07, "loss": 0.0337, "step": 17948 }, { "epoch": 0.8340613382899629, "grad_norm": 0.5645944175579113, "learning_rate": 7.134014573060566e-07, "loss": 0.0614, "step": 17949 }, { "epoch": 0.8341078066914498, "grad_norm": 0.48147203206812605, "learning_rate": 7.130141237766769e-07, "loss": 0.0646, "step": 17950 }, { "epoch": 0.8341542750929368, "grad_norm": 0.4410938115181967, "learning_rate": 7.126268873520393e-07, "loss": 0.0474, "step": 17951 }, { "epoch": 0.8342007434944237, "grad_norm": 0.5266382348172916, "learning_rate": 7.122397480409149e-07, "loss": 0.0439, "step": 17952 }, { "epoch": 0.8342472118959108, "grad_norm": 0.5415114391050655, "learning_rate": 7.11852705852073e-07, "loss": 0.0607, "step": 17953 }, { "epoch": 0.8342936802973978, "grad_norm": 0.40639333819477447, "learning_rate": 7.11465760794281e-07, "loss": 0.0267, "step": 17954 }, { "epoch": 0.8343401486988847, "grad_norm": 0.4297921308464815, "learning_rate": 7.110789128763024e-07, "loss": 0.0631, "step": 17955 }, { "epoch": 0.8343866171003718, "grad_norm": 0.37462793527404137, "learning_rate": 7.106921621068997e-07, "loss": 0.0701, "step": 17956 }, { "epoch": 0.8344330855018587, "grad_norm": 0.4483318407124652, "learning_rate": 7.103055084948351e-07, "loss": 0.045, "step": 17957 }, { "epoch": 0.8344795539033457, "grad_norm": 0.3500723252164339, "learning_rate": 7.099189520488664e-07, "loss": 0.0443, "step": 17958 }, { "epoch": 0.8345260223048327, "grad_norm": 0.522392722003359, "learning_rate": 7.09532492777747e-07, "loss": 0.049, "step": 17959 }, { "epoch": 0.8345724907063197, "grad_norm": 0.6489014193185344, "learning_rate": 7.091461306902325e-07, "loss": 0.0552, "step": 17960 }, { "epoch": 0.8346189591078067, "grad_norm": 0.5007699594572849, "learning_rate": 7.087598657950751e-07, "loss": 0.0433, "step": 17961 }, { "epoch": 0.8346654275092937, "grad_norm": 0.4711684604138935, "learning_rate": 7.083736981010225e-07, "loss": 0.058, "step": 17962 }, { "epoch": 0.8347118959107807, "grad_norm": 0.6819043829134457, "learning_rate": 7.079876276168218e-07, "loss": 0.056, "step": 17963 }, { "epoch": 0.8347583643122677, "grad_norm": 0.5678129635004515, "learning_rate": 7.076016543512204e-07, "loss": 0.0474, "step": 17964 }, { "epoch": 0.8348048327137546, "grad_norm": 0.29106282159789887, "learning_rate": 7.072157783129585e-07, "loss": 0.0415, "step": 17965 }, { "epoch": 0.8348513011152416, "grad_norm": 0.4121827004967967, "learning_rate": 7.068299995107758e-07, "loss": 0.0656, "step": 17966 }, { "epoch": 0.8348977695167287, "grad_norm": 0.47260696578137595, "learning_rate": 7.064443179534136e-07, "loss": 0.0493, "step": 17967 }, { "epoch": 0.8349442379182156, "grad_norm": 0.4497606646163414, "learning_rate": 7.060587336496066e-07, "loss": 0.057, "step": 17968 }, { "epoch": 0.8349907063197026, "grad_norm": 0.41341268795735325, "learning_rate": 7.056732466080879e-07, "loss": 0.0607, "step": 17969 }, { "epoch": 0.8350371747211895, "grad_norm": 0.5815588720425767, "learning_rate": 7.052878568375904e-07, "loss": 0.0588, "step": 17970 }, { "epoch": 0.8350836431226766, "grad_norm": 0.6163348439165016, "learning_rate": 7.049025643468426e-07, "loss": 0.0572, "step": 17971 }, { "epoch": 0.8351301115241636, "grad_norm": 0.5090835467839575, "learning_rate": 7.045173691445729e-07, "loss": 0.0611, "step": 17972 }, { "epoch": 0.8351765799256505, "grad_norm": 0.6360590280805677, "learning_rate": 7.041322712395043e-07, "loss": 0.051, "step": 17973 }, { "epoch": 0.8352230483271376, "grad_norm": 0.5817356951019401, "learning_rate": 7.03747270640362e-07, "loss": 0.0424, "step": 17974 }, { "epoch": 0.8352695167286245, "grad_norm": 0.42365550122655765, "learning_rate": 7.033623673558659e-07, "loss": 0.0404, "step": 17975 }, { "epoch": 0.8353159851301115, "grad_norm": 0.6042107859568668, "learning_rate": 7.029775613947343e-07, "loss": 0.0673, "step": 17976 }, { "epoch": 0.8353624535315985, "grad_norm": 0.5018988688057794, "learning_rate": 7.025928527656833e-07, "loss": 0.0511, "step": 17977 }, { "epoch": 0.8354089219330855, "grad_norm": 0.4360431441038279, "learning_rate": 7.022082414774273e-07, "loss": 0.0295, "step": 17978 }, { "epoch": 0.8354553903345725, "grad_norm": 0.8004694696186511, "learning_rate": 7.01823727538678e-07, "loss": 0.0512, "step": 17979 }, { "epoch": 0.8355018587360595, "grad_norm": 0.45962202210277936, "learning_rate": 7.014393109581435e-07, "loss": 0.0552, "step": 17980 }, { "epoch": 0.8355483271375465, "grad_norm": 0.6710132033626641, "learning_rate": 7.010549917445342e-07, "loss": 0.0592, "step": 17981 }, { "epoch": 0.8355947955390335, "grad_norm": 0.5589067538816271, "learning_rate": 7.006707699065551e-07, "loss": 0.0525, "step": 17982 }, { "epoch": 0.8356412639405204, "grad_norm": 0.19953888668934536, "learning_rate": 7.002866454529056e-07, "loss": 0.0318, "step": 17983 }, { "epoch": 0.8356877323420074, "grad_norm": 0.4083254268051237, "learning_rate": 6.999026183922902e-07, "loss": 0.0417, "step": 17984 }, { "epoch": 0.8357342007434945, "grad_norm": 0.5330739469494767, "learning_rate": 6.995186887334065e-07, "loss": 0.0531, "step": 17985 }, { "epoch": 0.8357806691449814, "grad_norm": 0.5461697222206259, "learning_rate": 6.991348564849504e-07, "loss": 0.0653, "step": 17986 }, { "epoch": 0.8358271375464684, "grad_norm": 0.7760361453414658, "learning_rate": 6.987511216556153e-07, "loss": 0.0719, "step": 17987 }, { "epoch": 0.8358736059479553, "grad_norm": 0.5442627059590187, "learning_rate": 6.983674842540966e-07, "loss": 0.0557, "step": 17988 }, { "epoch": 0.8359200743494424, "grad_norm": 0.38065290569818006, "learning_rate": 6.979839442890801e-07, "loss": 0.0576, "step": 17989 }, { "epoch": 0.8359665427509294, "grad_norm": 0.6279891793292267, "learning_rate": 6.976005017692544e-07, "loss": 0.0607, "step": 17990 }, { "epoch": 0.8360130111524163, "grad_norm": 0.6228630522648878, "learning_rate": 6.972171567033064e-07, "loss": 0.0511, "step": 17991 }, { "epoch": 0.8360594795539034, "grad_norm": 0.5601052892767673, "learning_rate": 6.968339090999188e-07, "loss": 0.0356, "step": 17992 }, { "epoch": 0.8361059479553903, "grad_norm": 0.5690245478519576, "learning_rate": 6.964507589677721e-07, "loss": 0.0473, "step": 17993 }, { "epoch": 0.8361524163568773, "grad_norm": 0.410622373863008, "learning_rate": 6.960677063155452e-07, "loss": 0.0491, "step": 17994 }, { "epoch": 0.8361988847583643, "grad_norm": 0.5435646969986118, "learning_rate": 6.956847511519143e-07, "loss": 0.0503, "step": 17995 }, { "epoch": 0.8362453531598513, "grad_norm": 0.33884361705057775, "learning_rate": 6.953018934855537e-07, "loss": 0.0372, "step": 17996 }, { "epoch": 0.8362918215613383, "grad_norm": 0.5007570301637019, "learning_rate": 6.949191333251354e-07, "loss": 0.0352, "step": 17997 }, { "epoch": 0.8363382899628253, "grad_norm": 0.37129874971797255, "learning_rate": 6.945364706793301e-07, "loss": 0.0536, "step": 17998 }, { "epoch": 0.8363847583643123, "grad_norm": 0.42089976180425914, "learning_rate": 6.941539055568058e-07, "loss": 0.0341, "step": 17999 }, { "epoch": 0.8364312267657993, "grad_norm": 0.3577504458478587, "learning_rate": 6.93771437966227e-07, "loss": 0.0284, "step": 18000 }, { "epoch": 0.8364776951672862, "grad_norm": 0.3561247907481803, "learning_rate": 6.933890679162575e-07, "loss": 0.0361, "step": 18001 }, { "epoch": 0.8365241635687732, "grad_norm": 0.5472356681134772, "learning_rate": 6.930067954155583e-07, "loss": 0.0642, "step": 18002 }, { "epoch": 0.8365706319702603, "grad_norm": 0.3489610084836997, "learning_rate": 6.926246204727883e-07, "loss": 0.044, "step": 18003 }, { "epoch": 0.8366171003717472, "grad_norm": 0.4105286098416883, "learning_rate": 6.922425430966023e-07, "loss": 0.0495, "step": 18004 }, { "epoch": 0.8366635687732342, "grad_norm": 0.4958433122579824, "learning_rate": 6.918605632956593e-07, "loss": 0.0546, "step": 18005 }, { "epoch": 0.8367100371747211, "grad_norm": 0.5212579212666988, "learning_rate": 6.914786810786079e-07, "loss": 0.0774, "step": 18006 }, { "epoch": 0.8367565055762082, "grad_norm": 0.4704562339158889, "learning_rate": 6.910968964540976e-07, "loss": 0.0586, "step": 18007 }, { "epoch": 0.8368029739776952, "grad_norm": 0.4246585562881453, "learning_rate": 6.907152094307784e-07, "loss": 0.0651, "step": 18008 }, { "epoch": 0.8368494423791821, "grad_norm": 0.3576374745934941, "learning_rate": 6.903336200172955e-07, "loss": 0.0518, "step": 18009 }, { "epoch": 0.8368959107806692, "grad_norm": 0.6150953280817048, "learning_rate": 6.899521282222921e-07, "loss": 0.0527, "step": 18010 }, { "epoch": 0.8369423791821561, "grad_norm": 0.5618029090096123, "learning_rate": 6.895707340544094e-07, "loss": 0.0502, "step": 18011 }, { "epoch": 0.8369888475836431, "grad_norm": 0.40123323383070486, "learning_rate": 6.891894375222857e-07, "loss": 0.0449, "step": 18012 }, { "epoch": 0.8370353159851301, "grad_norm": 0.47659663635463856, "learning_rate": 6.888082386345584e-07, "loss": 0.0462, "step": 18013 }, { "epoch": 0.8370817843866171, "grad_norm": 0.4106223597314679, "learning_rate": 6.884271373998608e-07, "loss": 0.0478, "step": 18014 }, { "epoch": 0.8371282527881041, "grad_norm": 0.3896201957621576, "learning_rate": 6.880461338268279e-07, "loss": 0.0379, "step": 18015 }, { "epoch": 0.8371747211895911, "grad_norm": 0.8467076048688484, "learning_rate": 6.87665227924088e-07, "loss": 0.0385, "step": 18016 }, { "epoch": 0.837221189591078, "grad_norm": 0.41616081213011463, "learning_rate": 6.872844197002692e-07, "loss": 0.0535, "step": 18017 }, { "epoch": 0.8372676579925651, "grad_norm": 0.6642815944929289, "learning_rate": 6.869037091639974e-07, "loss": 0.0475, "step": 18018 }, { "epoch": 0.837314126394052, "grad_norm": 0.5119276656112692, "learning_rate": 6.865230963238961e-07, "loss": 0.0526, "step": 18019 }, { "epoch": 0.837360594795539, "grad_norm": 0.6909271573385907, "learning_rate": 6.861425811885869e-07, "loss": 0.0573, "step": 18020 }, { "epoch": 0.8374070631970261, "grad_norm": 0.490018068473922, "learning_rate": 6.857621637666867e-07, "loss": 0.0536, "step": 18021 }, { "epoch": 0.837453531598513, "grad_norm": 0.5883240163671444, "learning_rate": 6.853818440668159e-07, "loss": 0.0621, "step": 18022 }, { "epoch": 0.8375, "grad_norm": 0.6607119026336918, "learning_rate": 6.85001622097588e-07, "loss": 0.0407, "step": 18023 }, { "epoch": 0.8375464684014869, "grad_norm": 0.6119992511816672, "learning_rate": 6.84621497867613e-07, "loss": 0.0367, "step": 18024 }, { "epoch": 0.837592936802974, "grad_norm": 0.6831299293455144, "learning_rate": 6.842414713855039e-07, "loss": 0.0495, "step": 18025 }, { "epoch": 0.837639405204461, "grad_norm": 0.36766382102391415, "learning_rate": 6.838615426598677e-07, "loss": 0.041, "step": 18026 }, { "epoch": 0.8376858736059479, "grad_norm": 0.5463063255848375, "learning_rate": 6.834817116993098e-07, "loss": 0.0836, "step": 18027 }, { "epoch": 0.837732342007435, "grad_norm": 0.7007767042314558, "learning_rate": 6.831019785124338e-07, "loss": 0.0609, "step": 18028 }, { "epoch": 0.8377788104089219, "grad_norm": 0.5191247367464984, "learning_rate": 6.827223431078434e-07, "loss": 0.037, "step": 18029 }, { "epoch": 0.8378252788104089, "grad_norm": 0.6530903000314623, "learning_rate": 6.823428054941344e-07, "loss": 0.0444, "step": 18030 }, { "epoch": 0.837871747211896, "grad_norm": 0.3300510086750748, "learning_rate": 6.819633656799041e-07, "loss": 0.0421, "step": 18031 }, { "epoch": 0.8379182156133829, "grad_norm": 0.6270172212671409, "learning_rate": 6.815840236737492e-07, "loss": 0.0489, "step": 18032 }, { "epoch": 0.8379646840148699, "grad_norm": 0.5689790002081269, "learning_rate": 6.812047794842619e-07, "loss": 0.0433, "step": 18033 }, { "epoch": 0.8380111524163569, "grad_norm": 0.5527653799870426, "learning_rate": 6.808256331200308e-07, "loss": 0.0592, "step": 18034 }, { "epoch": 0.8380576208178439, "grad_norm": 0.47628789123276555, "learning_rate": 6.804465845896458e-07, "loss": 0.0551, "step": 18035 }, { "epoch": 0.8381040892193309, "grad_norm": 0.46812599580504893, "learning_rate": 6.800676339016915e-07, "loss": 0.0492, "step": 18036 }, { "epoch": 0.8381505576208178, "grad_norm": 0.4156652737897041, "learning_rate": 6.796887810647517e-07, "loss": 0.0456, "step": 18037 }, { "epoch": 0.8381970260223048, "grad_norm": 0.3617297305340891, "learning_rate": 6.793100260874075e-07, "loss": 0.0391, "step": 18038 }, { "epoch": 0.8382434944237919, "grad_norm": 0.383287919982744, "learning_rate": 6.789313689782396e-07, "loss": 0.0462, "step": 18039 }, { "epoch": 0.8382899628252788, "grad_norm": 0.6149721988026882, "learning_rate": 6.785528097458238e-07, "loss": 0.0545, "step": 18040 }, { "epoch": 0.8383364312267658, "grad_norm": 0.5469159950574288, "learning_rate": 6.781743483987352e-07, "loss": 0.0567, "step": 18041 }, { "epoch": 0.8383828996282527, "grad_norm": 0.44032843885700257, "learning_rate": 6.777959849455462e-07, "loss": 0.0541, "step": 18042 }, { "epoch": 0.8384293680297398, "grad_norm": 0.6195362336700351, "learning_rate": 6.774177193948272e-07, "loss": 0.0664, "step": 18043 }, { "epoch": 0.8384758364312268, "grad_norm": 0.39651259139268524, "learning_rate": 6.770395517551464e-07, "loss": 0.0512, "step": 18044 }, { "epoch": 0.8385223048327137, "grad_norm": 0.5192258926824399, "learning_rate": 6.766614820350687e-07, "loss": 0.0522, "step": 18045 }, { "epoch": 0.8385687732342008, "grad_norm": 0.47189242018129796, "learning_rate": 6.762835102431598e-07, "loss": 0.0486, "step": 18046 }, { "epoch": 0.8386152416356877, "grad_norm": 0.5570181531287808, "learning_rate": 6.759056363879807e-07, "loss": 0.0575, "step": 18047 }, { "epoch": 0.8386617100371747, "grad_norm": 0.6826396893249065, "learning_rate": 6.755278604780879e-07, "loss": 0.057, "step": 18048 }, { "epoch": 0.8387081784386617, "grad_norm": 0.5318529750486006, "learning_rate": 6.751501825220414e-07, "loss": 0.0577, "step": 18049 }, { "epoch": 0.8387546468401487, "grad_norm": 0.5578899024722275, "learning_rate": 6.747726025283952e-07, "loss": 0.0415, "step": 18050 }, { "epoch": 0.8388011152416357, "grad_norm": 0.3693830644848635, "learning_rate": 6.74395120505702e-07, "loss": 0.0323, "step": 18051 }, { "epoch": 0.8388475836431227, "grad_norm": 0.43248676504661904, "learning_rate": 6.740177364625106e-07, "loss": 0.0511, "step": 18052 }, { "epoch": 0.8388940520446097, "grad_norm": 0.5304131315491859, "learning_rate": 6.73640450407373e-07, "loss": 0.05, "step": 18053 }, { "epoch": 0.8389405204460967, "grad_norm": 0.48052996554154653, "learning_rate": 6.732632623488312e-07, "loss": 0.0454, "step": 18054 }, { "epoch": 0.8389869888475836, "grad_norm": 1.0000477804608081, "learning_rate": 6.728861722954294e-07, "loss": 0.0563, "step": 18055 }, { "epoch": 0.8390334572490706, "grad_norm": 0.5815175343788849, "learning_rate": 6.72509180255711e-07, "loss": 0.054, "step": 18056 }, { "epoch": 0.8390799256505577, "grad_norm": 0.33024728042363216, "learning_rate": 6.72132286238214e-07, "loss": 0.049, "step": 18057 }, { "epoch": 0.8391263940520446, "grad_norm": 0.7880478238066443, "learning_rate": 6.717554902514761e-07, "loss": 0.0524, "step": 18058 }, { "epoch": 0.8391728624535316, "grad_norm": 0.40415244535867245, "learning_rate": 6.713787923040316e-07, "loss": 0.0631, "step": 18059 }, { "epoch": 0.8392193308550185, "grad_norm": 0.42027338477536264, "learning_rate": 6.71002192404413e-07, "loss": 0.0549, "step": 18060 }, { "epoch": 0.8392657992565056, "grad_norm": 0.369979816485302, "learning_rate": 6.706256905611514e-07, "loss": 0.037, "step": 18061 }, { "epoch": 0.8393122676579926, "grad_norm": 0.39938316056730455, "learning_rate": 6.702492867827731e-07, "loss": 0.0557, "step": 18062 }, { "epoch": 0.8393587360594795, "grad_norm": 0.4176155087966131, "learning_rate": 6.698729810778065e-07, "loss": 0.0545, "step": 18063 }, { "epoch": 0.8394052044609666, "grad_norm": 0.4984505565531921, "learning_rate": 6.694967734547759e-07, "loss": 0.0429, "step": 18064 }, { "epoch": 0.8394516728624535, "grad_norm": 0.5346166452718487, "learning_rate": 6.691206639221981e-07, "loss": 0.0492, "step": 18065 }, { "epoch": 0.8394981412639405, "grad_norm": 0.31813269875398487, "learning_rate": 6.687446524885971e-07, "loss": 0.0366, "step": 18066 }, { "epoch": 0.8395446096654275, "grad_norm": 0.5337819447059453, "learning_rate": 6.683687391624883e-07, "loss": 0.0651, "step": 18067 }, { "epoch": 0.8395910780669145, "grad_norm": 0.4770042090981652, "learning_rate": 6.679929239523863e-07, "loss": 0.0509, "step": 18068 }, { "epoch": 0.8396375464684015, "grad_norm": 1.5153066401539017, "learning_rate": 6.676172068668029e-07, "loss": 0.043, "step": 18069 }, { "epoch": 0.8396840148698885, "grad_norm": 0.5295084160836685, "learning_rate": 6.672415879142518e-07, "loss": 0.0567, "step": 18070 }, { "epoch": 0.8397304832713754, "grad_norm": 0.5194274657143554, "learning_rate": 6.668660671032373e-07, "loss": 0.0451, "step": 18071 }, { "epoch": 0.8397769516728625, "grad_norm": 0.317730330316021, "learning_rate": 6.664906444422664e-07, "loss": 0.0394, "step": 18072 }, { "epoch": 0.8398234200743494, "grad_norm": 0.4668765680370901, "learning_rate": 6.66115319939844e-07, "loss": 0.0582, "step": 18073 }, { "epoch": 0.8398698884758364, "grad_norm": 0.7816979232758856, "learning_rate": 6.657400936044717e-07, "loss": 0.0683, "step": 18074 }, { "epoch": 0.8399163568773235, "grad_norm": 0.4237670922640725, "learning_rate": 6.65364965444647e-07, "loss": 0.0421, "step": 18075 }, { "epoch": 0.8399628252788104, "grad_norm": 0.7229159813679914, "learning_rate": 6.649899354688688e-07, "loss": 0.0532, "step": 18076 }, { "epoch": 0.8400092936802974, "grad_norm": 0.5556733991730523, "learning_rate": 6.646150036856308e-07, "loss": 0.0523, "step": 18077 }, { "epoch": 0.8400557620817843, "grad_norm": 0.4013776858446647, "learning_rate": 6.642401701034256e-07, "loss": 0.0416, "step": 18078 }, { "epoch": 0.8401022304832714, "grad_norm": 0.36379371716219067, "learning_rate": 6.638654347307427e-07, "loss": 0.0335, "step": 18079 }, { "epoch": 0.8401486988847584, "grad_norm": 0.6665753686531289, "learning_rate": 6.63490797576073e-07, "loss": 0.0579, "step": 18080 }, { "epoch": 0.8401951672862453, "grad_norm": 0.3916553494044635, "learning_rate": 6.631162586479006e-07, "loss": 0.0625, "step": 18081 }, { "epoch": 0.8402416356877324, "grad_norm": 0.5464905163397878, "learning_rate": 6.62741817954709e-07, "loss": 0.0713, "step": 18082 }, { "epoch": 0.8402881040892193, "grad_norm": 0.43925592918340195, "learning_rate": 6.623674755049808e-07, "loss": 0.0405, "step": 18083 }, { "epoch": 0.8403345724907063, "grad_norm": 0.4513014138573519, "learning_rate": 6.619932313071942e-07, "loss": 0.0619, "step": 18084 }, { "epoch": 0.8403810408921933, "grad_norm": 0.5915645511180927, "learning_rate": 6.616190853698268e-07, "loss": 0.0608, "step": 18085 }, { "epoch": 0.8404275092936803, "grad_norm": 0.4369446442791274, "learning_rate": 6.612450377013519e-07, "loss": 0.0487, "step": 18086 }, { "epoch": 0.8404739776951673, "grad_norm": 0.32931593827263284, "learning_rate": 6.608710883102443e-07, "loss": 0.0426, "step": 18087 }, { "epoch": 0.8405204460966543, "grad_norm": 0.42093548273596143, "learning_rate": 6.604972372049745e-07, "loss": 0.0415, "step": 18088 }, { "epoch": 0.8405669144981412, "grad_norm": 0.561051032290143, "learning_rate": 6.60123484394008e-07, "loss": 0.0553, "step": 18089 }, { "epoch": 0.8406133828996283, "grad_norm": 0.40531929100888503, "learning_rate": 6.597498298858124e-07, "loss": 0.0481, "step": 18090 }, { "epoch": 0.8406598513011152, "grad_norm": 0.5153238122013144, "learning_rate": 6.593762736888514e-07, "loss": 0.0633, "step": 18091 }, { "epoch": 0.8407063197026022, "grad_norm": 0.5710487041230007, "learning_rate": 6.590028158115868e-07, "loss": 0.0439, "step": 18092 }, { "epoch": 0.8407527881040893, "grad_norm": 0.24575974300918713, "learning_rate": 6.586294562624757e-07, "loss": 0.0374, "step": 18093 }, { "epoch": 0.8407992565055762, "grad_norm": 0.47730786830552624, "learning_rate": 6.582561950499788e-07, "loss": 0.0761, "step": 18094 }, { "epoch": 0.8408457249070632, "grad_norm": 0.7532172578574924, "learning_rate": 6.578830321825474e-07, "loss": 0.0683, "step": 18095 }, { "epoch": 0.8408921933085501, "grad_norm": 0.47678675743160515, "learning_rate": 6.575099676686347e-07, "loss": 0.0322, "step": 18096 }, { "epoch": 0.8409386617100372, "grad_norm": 0.6222574732786582, "learning_rate": 6.571370015166928e-07, "loss": 0.0361, "step": 18097 }, { "epoch": 0.8409851301115242, "grad_norm": 0.6714327286831303, "learning_rate": 6.567641337351682e-07, "loss": 0.0636, "step": 18098 }, { "epoch": 0.8410315985130111, "grad_norm": 0.6072006389105296, "learning_rate": 6.56391364332507e-07, "loss": 0.0553, "step": 18099 }, { "epoch": 0.8410780669144982, "grad_norm": 0.7897984025874377, "learning_rate": 6.560186933171531e-07, "loss": 0.056, "step": 18100 }, { "epoch": 0.8411245353159851, "grad_norm": 1.8605013209595695, "learning_rate": 6.556461206975479e-07, "loss": 0.0797, "step": 18101 }, { "epoch": 0.8411710037174721, "grad_norm": 0.4929449581248939, "learning_rate": 6.552736464821307e-07, "loss": 0.0471, "step": 18102 }, { "epoch": 0.8412174721189591, "grad_norm": 0.4891573722519699, "learning_rate": 6.549012706793368e-07, "loss": 0.0541, "step": 18103 }, { "epoch": 0.8412639405204461, "grad_norm": 0.4868727190750892, "learning_rate": 6.545289932976029e-07, "loss": 0.0346, "step": 18104 }, { "epoch": 0.8413104089219331, "grad_norm": 0.4958637136486549, "learning_rate": 6.541568143453613e-07, "loss": 0.0459, "step": 18105 }, { "epoch": 0.8413568773234201, "grad_norm": 0.5495777121855753, "learning_rate": 6.537847338310416e-07, "loss": 0.0444, "step": 18106 }, { "epoch": 0.841403345724907, "grad_norm": 0.7471443666366188, "learning_rate": 6.534127517630722e-07, "loss": 0.048, "step": 18107 }, { "epoch": 0.8414498141263941, "grad_norm": 0.19968081903128207, "learning_rate": 6.530408681498784e-07, "loss": 0.0532, "step": 18108 }, { "epoch": 0.841496282527881, "grad_norm": 0.4135912146537987, "learning_rate": 6.526690829998844e-07, "loss": 0.0288, "step": 18109 }, { "epoch": 0.841542750929368, "grad_norm": 0.550998258552678, "learning_rate": 6.522973963215095e-07, "loss": 0.0408, "step": 18110 }, { "epoch": 0.8415892193308551, "grad_norm": 0.413312838160306, "learning_rate": 6.51925808123176e-07, "loss": 0.0476, "step": 18111 }, { "epoch": 0.841635687732342, "grad_norm": 0.416926299735115, "learning_rate": 6.515543184133e-07, "loss": 0.0442, "step": 18112 }, { "epoch": 0.841682156133829, "grad_norm": 0.43501090096176687, "learning_rate": 6.511829272002934e-07, "loss": 0.0632, "step": 18113 }, { "epoch": 0.8417286245353159, "grad_norm": 0.7784714614013594, "learning_rate": 6.508116344925719e-07, "loss": 0.0379, "step": 18114 }, { "epoch": 0.841775092936803, "grad_norm": 0.5235144383141468, "learning_rate": 6.504404402985436e-07, "loss": 0.0671, "step": 18115 }, { "epoch": 0.84182156133829, "grad_norm": 0.47096080204495255, "learning_rate": 6.500693446266176e-07, "loss": 0.0343, "step": 18116 }, { "epoch": 0.8418680297397769, "grad_norm": 0.606582117759722, "learning_rate": 6.496983474851981e-07, "loss": 0.052, "step": 18117 }, { "epoch": 0.841914498141264, "grad_norm": 0.5970030159884543, "learning_rate": 6.493274488826918e-07, "loss": 0.0445, "step": 18118 }, { "epoch": 0.8419609665427509, "grad_norm": 0.39553623586286907, "learning_rate": 6.489566488274968e-07, "loss": 0.0575, "step": 18119 }, { "epoch": 0.8420074349442379, "grad_norm": 0.5670985502416629, "learning_rate": 6.485859473280115e-07, "loss": 0.0507, "step": 18120 }, { "epoch": 0.8420539033457249, "grad_norm": 0.5174318028820059, "learning_rate": 6.482153443926359e-07, "loss": 0.0556, "step": 18121 }, { "epoch": 0.8421003717472119, "grad_norm": 0.5321565082988398, "learning_rate": 6.478448400297627e-07, "loss": 0.0353, "step": 18122 }, { "epoch": 0.8421468401486989, "grad_norm": 0.6978889012192179, "learning_rate": 6.47474434247784e-07, "loss": 0.0724, "step": 18123 }, { "epoch": 0.8421933085501859, "grad_norm": 0.5398995508072437, "learning_rate": 6.47104127055091e-07, "loss": 0.0519, "step": 18124 }, { "epoch": 0.8422397769516728, "grad_norm": 0.3362509903020575, "learning_rate": 6.467339184600707e-07, "loss": 0.0548, "step": 18125 }, { "epoch": 0.8422862453531599, "grad_norm": 0.6497203925611427, "learning_rate": 6.463638084711088e-07, "loss": 0.0533, "step": 18126 }, { "epoch": 0.8423327137546468, "grad_norm": 0.34148507748315066, "learning_rate": 6.459937970965874e-07, "loss": 0.0321, "step": 18127 }, { "epoch": 0.8423791821561338, "grad_norm": 0.8258930472725752, "learning_rate": 6.456238843448904e-07, "loss": 0.0622, "step": 18128 }, { "epoch": 0.8424256505576209, "grad_norm": 0.6618781939683329, "learning_rate": 6.452540702243964e-07, "loss": 0.0567, "step": 18129 }, { "epoch": 0.8424721189591078, "grad_norm": 0.5058713551521388, "learning_rate": 6.448843547434786e-07, "loss": 0.0504, "step": 18130 }, { "epoch": 0.8425185873605948, "grad_norm": 0.5570640251410495, "learning_rate": 6.445147379105154e-07, "loss": 0.044, "step": 18131 }, { "epoch": 0.8425650557620817, "grad_norm": 0.49043595668903683, "learning_rate": 6.441452197338771e-07, "loss": 0.0478, "step": 18132 }, { "epoch": 0.8426115241635688, "grad_norm": 0.3887893375084975, "learning_rate": 6.43775800221934e-07, "loss": 0.0472, "step": 18133 }, { "epoch": 0.8426579925650558, "grad_norm": 0.5271860704417052, "learning_rate": 6.43406479383053e-07, "loss": 0.0488, "step": 18134 }, { "epoch": 0.8427044609665427, "grad_norm": 0.515161982510035, "learning_rate": 6.430372572256022e-07, "loss": 0.055, "step": 18135 }, { "epoch": 0.8427509293680298, "grad_norm": 0.6768978556490013, "learning_rate": 6.426681337579427e-07, "loss": 0.0531, "step": 18136 }, { "epoch": 0.8427973977695167, "grad_norm": 0.42296178551285213, "learning_rate": 6.422991089884345e-07, "loss": 0.0388, "step": 18137 }, { "epoch": 0.8428438661710037, "grad_norm": 0.4563414313213904, "learning_rate": 6.419301829254387e-07, "loss": 0.0512, "step": 18138 }, { "epoch": 0.8428903345724907, "grad_norm": 0.4157573812433711, "learning_rate": 6.415613555773114e-07, "loss": 0.0462, "step": 18139 }, { "epoch": 0.8429368029739777, "grad_norm": 0.5187262609773028, "learning_rate": 6.411926269524066e-07, "loss": 0.058, "step": 18140 }, { "epoch": 0.8429832713754647, "grad_norm": 0.4384071186593826, "learning_rate": 6.408239970590763e-07, "loss": 0.0763, "step": 18141 }, { "epoch": 0.8430297397769517, "grad_norm": 0.6182860975835327, "learning_rate": 6.404554659056706e-07, "loss": 0.0424, "step": 18142 }, { "epoch": 0.8430762081784386, "grad_norm": 0.4613981742630577, "learning_rate": 6.400870335005365e-07, "loss": 0.0666, "step": 18143 }, { "epoch": 0.8431226765799257, "grad_norm": 0.5736470984962434, "learning_rate": 6.397186998520194e-07, "loss": 0.0516, "step": 18144 }, { "epoch": 0.8431691449814126, "grad_norm": 0.5790346120685573, "learning_rate": 6.393504649684634e-07, "loss": 0.0482, "step": 18145 }, { "epoch": 0.8432156133828996, "grad_norm": 0.6305476889748778, "learning_rate": 6.389823288582092e-07, "loss": 0.0626, "step": 18146 }, { "epoch": 0.8432620817843867, "grad_norm": 0.36054500089070196, "learning_rate": 6.386142915295945e-07, "loss": 0.0433, "step": 18147 }, { "epoch": 0.8433085501858736, "grad_norm": 0.6495040211280694, "learning_rate": 6.382463529909572e-07, "loss": 0.0459, "step": 18148 }, { "epoch": 0.8433550185873606, "grad_norm": 0.5528207176392665, "learning_rate": 6.378785132506299e-07, "loss": 0.0694, "step": 18149 }, { "epoch": 0.8434014869888475, "grad_norm": 0.5171336541920306, "learning_rate": 6.37510772316946e-07, "loss": 0.0513, "step": 18150 }, { "epoch": 0.8434479553903346, "grad_norm": 0.38368843544070974, "learning_rate": 6.371431301982328e-07, "loss": 0.0293, "step": 18151 }, { "epoch": 0.8434944237918216, "grad_norm": 0.4837428318316767, "learning_rate": 6.367755869028213e-07, "loss": 0.047, "step": 18152 }, { "epoch": 0.8435408921933085, "grad_norm": 0.5724574836197822, "learning_rate": 6.364081424390355e-07, "loss": 0.0526, "step": 18153 }, { "epoch": 0.8435873605947956, "grad_norm": 0.3194319747762387, "learning_rate": 6.36040796815196e-07, "loss": 0.039, "step": 18154 }, { "epoch": 0.8436338289962825, "grad_norm": 0.7528625592516749, "learning_rate": 6.356735500396266e-07, "loss": 0.0398, "step": 18155 }, { "epoch": 0.8436802973977695, "grad_norm": 0.4915720228040844, "learning_rate": 6.353064021206445e-07, "loss": 0.0463, "step": 18156 }, { "epoch": 0.8437267657992565, "grad_norm": 0.4964487879471706, "learning_rate": 6.34939353066566e-07, "loss": 0.052, "step": 18157 }, { "epoch": 0.8437732342007435, "grad_norm": 0.5518993735631101, "learning_rate": 6.345724028857042e-07, "loss": 0.0479, "step": 18158 }, { "epoch": 0.8438197026022305, "grad_norm": 0.5480940279560915, "learning_rate": 6.342055515863744e-07, "loss": 0.063, "step": 18159 }, { "epoch": 0.8438661710037175, "grad_norm": 0.8349870351234719, "learning_rate": 6.338387991768829e-07, "loss": 0.0711, "step": 18160 }, { "epoch": 0.8439126394052044, "grad_norm": 0.48610964171290005, "learning_rate": 6.334721456655363e-07, "loss": 0.0381, "step": 18161 }, { "epoch": 0.8439591078066915, "grad_norm": 0.4841976917295525, "learning_rate": 6.331055910606427e-07, "loss": 0.0454, "step": 18162 }, { "epoch": 0.8440055762081784, "grad_norm": 0.3468415184650404, "learning_rate": 6.327391353705031e-07, "loss": 0.0578, "step": 18163 }, { "epoch": 0.8440520446096654, "grad_norm": 0.23764930350304472, "learning_rate": 6.32372778603419e-07, "loss": 0.031, "step": 18164 }, { "epoch": 0.8440985130111525, "grad_norm": 0.5980324381551064, "learning_rate": 6.320065207676884e-07, "loss": 0.0448, "step": 18165 }, { "epoch": 0.8441449814126394, "grad_norm": 0.6857296537437227, "learning_rate": 6.316403618716072e-07, "loss": 0.0595, "step": 18166 }, { "epoch": 0.8441914498141264, "grad_norm": 0.35781034685682567, "learning_rate": 6.31274301923469e-07, "loss": 0.0557, "step": 18167 }, { "epoch": 0.8442379182156133, "grad_norm": 0.31133356494564063, "learning_rate": 6.309083409315653e-07, "loss": 0.0398, "step": 18168 }, { "epoch": 0.8442843866171004, "grad_norm": 0.7042817248352746, "learning_rate": 6.30542478904187e-07, "loss": 0.0569, "step": 18169 }, { "epoch": 0.8443308550185874, "grad_norm": 0.5320284939775543, "learning_rate": 6.301767158496203e-07, "loss": 0.0503, "step": 18170 }, { "epoch": 0.8443773234200743, "grad_norm": 0.700417658629965, "learning_rate": 6.298110517761502e-07, "loss": 0.0633, "step": 18171 }, { "epoch": 0.8444237918215614, "grad_norm": 0.3615842349514479, "learning_rate": 6.294454866920591e-07, "loss": 0.0488, "step": 18172 }, { "epoch": 0.8444702602230483, "grad_norm": 0.42195523161654336, "learning_rate": 6.29080020605628e-07, "loss": 0.0618, "step": 18173 }, { "epoch": 0.8445167286245353, "grad_norm": 0.5271169450072372, "learning_rate": 6.287146535251343e-07, "loss": 0.0645, "step": 18174 }, { "epoch": 0.8445631970260223, "grad_norm": 0.4279889475822757, "learning_rate": 6.28349385458854e-07, "loss": 0.0518, "step": 18175 }, { "epoch": 0.8446096654275093, "grad_norm": 0.4607089494777329, "learning_rate": 6.279842164150618e-07, "loss": 0.0545, "step": 18176 }, { "epoch": 0.8446561338289963, "grad_norm": 0.3891097362874949, "learning_rate": 6.276191464020293e-07, "loss": 0.0452, "step": 18177 }, { "epoch": 0.8447026022304833, "grad_norm": 0.8902804079801829, "learning_rate": 6.272541754280231e-07, "loss": 0.0748, "step": 18178 }, { "epoch": 0.8447490706319702, "grad_norm": 0.4332480629758495, "learning_rate": 6.268893035013135e-07, "loss": 0.0497, "step": 18179 }, { "epoch": 0.8447955390334573, "grad_norm": 0.417290592951028, "learning_rate": 6.265245306301637e-07, "loss": 0.0512, "step": 18180 }, { "epoch": 0.8448420074349442, "grad_norm": 0.6877374148877239, "learning_rate": 6.261598568228361e-07, "loss": 0.0604, "step": 18181 }, { "epoch": 0.8448884758364312, "grad_norm": 0.4022899679736035, "learning_rate": 6.257952820875912e-07, "loss": 0.0545, "step": 18182 }, { "epoch": 0.8449349442379183, "grad_norm": 0.41729677701269136, "learning_rate": 6.254308064326869e-07, "loss": 0.0701, "step": 18183 }, { "epoch": 0.8449814126394052, "grad_norm": 0.41389414015781106, "learning_rate": 6.250664298663789e-07, "loss": 0.0626, "step": 18184 }, { "epoch": 0.8450278810408922, "grad_norm": 0.5221015145781023, "learning_rate": 6.247021523969199e-07, "loss": 0.0437, "step": 18185 }, { "epoch": 0.8450743494423791, "grad_norm": 0.3785255684391764, "learning_rate": 6.243379740325628e-07, "loss": 0.0425, "step": 18186 }, { "epoch": 0.8451208178438662, "grad_norm": 0.7918735945525396, "learning_rate": 6.23973894781556e-07, "loss": 0.0496, "step": 18187 }, { "epoch": 0.8451672862453532, "grad_norm": 0.4140212258314343, "learning_rate": 6.236099146521462e-07, "loss": 0.0392, "step": 18188 }, { "epoch": 0.8452137546468401, "grad_norm": 0.4623111539442541, "learning_rate": 6.23246033652578e-07, "loss": 0.0492, "step": 18189 }, { "epoch": 0.8452602230483272, "grad_norm": 0.5345424598865602, "learning_rate": 6.228822517910937e-07, "loss": 0.0469, "step": 18190 }, { "epoch": 0.8453066914498141, "grad_norm": 0.6431080949189643, "learning_rate": 6.225185690759328e-07, "loss": 0.0473, "step": 18191 }, { "epoch": 0.8453531598513011, "grad_norm": 0.363599706900152, "learning_rate": 6.221549855153331e-07, "loss": 0.0231, "step": 18192 }, { "epoch": 0.8453996282527881, "grad_norm": 0.41527416881559504, "learning_rate": 6.217915011175313e-07, "loss": 0.032, "step": 18193 }, { "epoch": 0.8454460966542751, "grad_norm": 0.8573846617153054, "learning_rate": 6.214281158907609e-07, "loss": 0.051, "step": 18194 }, { "epoch": 0.8454925650557621, "grad_norm": 0.6243224961410746, "learning_rate": 6.210648298432498e-07, "loss": 0.0484, "step": 18195 }, { "epoch": 0.8455390334572491, "grad_norm": 0.483052586723509, "learning_rate": 6.207016429832307e-07, "loss": 0.0502, "step": 18196 }, { "epoch": 0.845585501858736, "grad_norm": 0.3319879420227017, "learning_rate": 6.203385553189278e-07, "loss": 0.0419, "step": 18197 }, { "epoch": 0.8456319702602231, "grad_norm": 0.5984440614771198, "learning_rate": 6.199755668585667e-07, "loss": 0.0618, "step": 18198 }, { "epoch": 0.84567843866171, "grad_norm": 0.6315439681876458, "learning_rate": 6.196126776103678e-07, "loss": 0.0415, "step": 18199 }, { "epoch": 0.845724907063197, "grad_norm": 0.7330964613335403, "learning_rate": 6.192498875825536e-07, "loss": 0.0518, "step": 18200 }, { "epoch": 0.8457713754646841, "grad_norm": 0.49046878279272144, "learning_rate": 6.188871967833398e-07, "loss": 0.0584, "step": 18201 }, { "epoch": 0.845817843866171, "grad_norm": 0.4646173906873636, "learning_rate": 6.185246052209404e-07, "loss": 0.042, "step": 18202 }, { "epoch": 0.845864312267658, "grad_norm": 0.8214688518872467, "learning_rate": 6.181621129035714e-07, "loss": 0.0641, "step": 18203 }, { "epoch": 0.8459107806691449, "grad_norm": 0.575959477456987, "learning_rate": 6.177997198394426e-07, "loss": 0.0581, "step": 18204 }, { "epoch": 0.845957249070632, "grad_norm": 0.7264032526404169, "learning_rate": 6.174374260367611e-07, "loss": 0.0642, "step": 18205 }, { "epoch": 0.846003717472119, "grad_norm": 0.36376901116043875, "learning_rate": 6.170752315037365e-07, "loss": 0.0425, "step": 18206 }, { "epoch": 0.8460501858736059, "grad_norm": 0.47180424989999936, "learning_rate": 6.167131362485695e-07, "loss": 0.0331, "step": 18207 }, { "epoch": 0.846096654275093, "grad_norm": 0.5388736701473386, "learning_rate": 6.163511402794636e-07, "loss": 0.0489, "step": 18208 }, { "epoch": 0.8461431226765799, "grad_norm": 0.7462269097959635, "learning_rate": 6.159892436046166e-07, "loss": 0.0559, "step": 18209 }, { "epoch": 0.8461895910780669, "grad_norm": 0.33384951758842085, "learning_rate": 6.156274462322292e-07, "loss": 0.0404, "step": 18210 }, { "epoch": 0.8462360594795539, "grad_norm": 0.7571071536719282, "learning_rate": 6.152657481704938e-07, "loss": 0.048, "step": 18211 }, { "epoch": 0.8462825278810409, "grad_norm": 0.45921763865013815, "learning_rate": 6.149041494276048e-07, "loss": 0.0552, "step": 18212 }, { "epoch": 0.8463289962825279, "grad_norm": 0.3966881322412887, "learning_rate": 6.145426500117512e-07, "loss": 0.043, "step": 18213 }, { "epoch": 0.8463754646840149, "grad_norm": 0.6235101298002917, "learning_rate": 6.141812499311228e-07, "loss": 0.0524, "step": 18214 }, { "epoch": 0.8464219330855018, "grad_norm": 0.5989590384526698, "learning_rate": 6.13819949193904e-07, "loss": 0.0558, "step": 18215 }, { "epoch": 0.8464684014869889, "grad_norm": 0.6350413815328656, "learning_rate": 6.134587478082804e-07, "loss": 0.0463, "step": 18216 }, { "epoch": 0.8465148698884758, "grad_norm": 0.5283980583493415, "learning_rate": 6.130976457824334e-07, "loss": 0.0674, "step": 18217 }, { "epoch": 0.8465613382899628, "grad_norm": 0.3628136237981256, "learning_rate": 6.12736643124543e-07, "loss": 0.0427, "step": 18218 }, { "epoch": 0.8466078066914499, "grad_norm": 0.4406396544745307, "learning_rate": 6.123757398427827e-07, "loss": 0.0436, "step": 18219 }, { "epoch": 0.8466542750929368, "grad_norm": 0.5109044032412939, "learning_rate": 6.120149359453304e-07, "loss": 0.0456, "step": 18220 }, { "epoch": 0.8467007434944238, "grad_norm": 0.4140935587491194, "learning_rate": 6.116542314403584e-07, "loss": 0.039, "step": 18221 }, { "epoch": 0.8467472118959107, "grad_norm": 0.5144110827937827, "learning_rate": 6.112936263360358e-07, "loss": 0.0502, "step": 18222 }, { "epoch": 0.8467936802973978, "grad_norm": 0.39976177756626696, "learning_rate": 6.109331206405328e-07, "loss": 0.0417, "step": 18223 }, { "epoch": 0.8468401486988848, "grad_norm": 0.40093867179822096, "learning_rate": 6.105727143620149e-07, "loss": 0.0334, "step": 18224 }, { "epoch": 0.8468866171003717, "grad_norm": 0.6900945393767022, "learning_rate": 6.10212407508644e-07, "loss": 0.0734, "step": 18225 }, { "epoch": 0.8469330855018588, "grad_norm": 0.5443081209307606, "learning_rate": 6.098522000885809e-07, "loss": 0.0416, "step": 18226 }, { "epoch": 0.8469795539033457, "grad_norm": 0.5734561562738493, "learning_rate": 6.094920921099868e-07, "loss": 0.0422, "step": 18227 }, { "epoch": 0.8470260223048327, "grad_norm": 0.47691420306994176, "learning_rate": 6.091320835810183e-07, "loss": 0.0518, "step": 18228 }, { "epoch": 0.8470724907063197, "grad_norm": 0.43239932196296477, "learning_rate": 6.087721745098279e-07, "loss": 0.0367, "step": 18229 }, { "epoch": 0.8471189591078067, "grad_norm": 0.6835979956165155, "learning_rate": 6.08412364904572e-07, "loss": 0.0434, "step": 18230 }, { "epoch": 0.8471654275092937, "grad_norm": 0.7661847883624118, "learning_rate": 6.080526547733967e-07, "loss": 0.0668, "step": 18231 }, { "epoch": 0.8472118959107807, "grad_norm": 0.4377512455868262, "learning_rate": 6.0769304412445e-07, "loss": 0.047, "step": 18232 }, { "epoch": 0.8472583643122676, "grad_norm": 0.6174580075519545, "learning_rate": 6.073335329658802e-07, "loss": 0.0535, "step": 18233 }, { "epoch": 0.8473048327137547, "grad_norm": 0.6500106474878223, "learning_rate": 6.069741213058289e-07, "loss": 0.0645, "step": 18234 }, { "epoch": 0.8473513011152416, "grad_norm": 0.47125402232763147, "learning_rate": 6.066148091524371e-07, "loss": 0.0421, "step": 18235 }, { "epoch": 0.8473977695167286, "grad_norm": 0.575624921801062, "learning_rate": 6.06255596513844e-07, "loss": 0.0569, "step": 18236 }, { "epoch": 0.8474442379182157, "grad_norm": 0.7120177033173763, "learning_rate": 6.058964833981862e-07, "loss": 0.0652, "step": 18237 }, { "epoch": 0.8474907063197026, "grad_norm": 0.5937558469622104, "learning_rate": 6.055374698135974e-07, "loss": 0.0446, "step": 18238 }, { "epoch": 0.8475371747211896, "grad_norm": 0.5559826720974622, "learning_rate": 6.051785557682089e-07, "loss": 0.0547, "step": 18239 }, { "epoch": 0.8475836431226765, "grad_norm": 0.6048618515128884, "learning_rate": 6.048197412701528e-07, "loss": 0.07, "step": 18240 }, { "epoch": 0.8476301115241636, "grad_norm": 0.4878420197841483, "learning_rate": 6.044610263275563e-07, "loss": 0.0623, "step": 18241 }, { "epoch": 0.8476765799256506, "grad_norm": 0.49157414172985675, "learning_rate": 6.041024109485433e-07, "loss": 0.0374, "step": 18242 }, { "epoch": 0.8477230483271375, "grad_norm": 0.6185803879799586, "learning_rate": 6.037438951412356e-07, "loss": 0.05, "step": 18243 }, { "epoch": 0.8477695167286246, "grad_norm": 0.42771091536041855, "learning_rate": 6.033854789137567e-07, "loss": 0.0435, "step": 18244 }, { "epoch": 0.8478159851301115, "grad_norm": 0.46685517230923335, "learning_rate": 6.030271622742245e-07, "loss": 0.0516, "step": 18245 }, { "epoch": 0.8478624535315985, "grad_norm": 0.5752617708629776, "learning_rate": 6.026689452307532e-07, "loss": 0.0514, "step": 18246 }, { "epoch": 0.8479089219330855, "grad_norm": 0.38997156222896534, "learning_rate": 6.023108277914614e-07, "loss": 0.0534, "step": 18247 }, { "epoch": 0.8479553903345725, "grad_norm": 0.451764510767877, "learning_rate": 6.019528099644561e-07, "loss": 0.0371, "step": 18248 }, { "epoch": 0.8480018587360595, "grad_norm": 0.4969365474844008, "learning_rate": 6.015948917578473e-07, "loss": 0.0526, "step": 18249 }, { "epoch": 0.8480483271375465, "grad_norm": 0.3611216999612019, "learning_rate": 6.012370731797451e-07, "loss": 0.0375, "step": 18250 }, { "epoch": 0.8480947955390334, "grad_norm": 0.6967583471772534, "learning_rate": 6.008793542382524e-07, "loss": 0.0637, "step": 18251 }, { "epoch": 0.8481412639405205, "grad_norm": 0.5255581855619126, "learning_rate": 6.005217349414721e-07, "loss": 0.0459, "step": 18252 }, { "epoch": 0.8481877323420074, "grad_norm": 0.5114898101741523, "learning_rate": 6.001642152975052e-07, "loss": 0.0502, "step": 18253 }, { "epoch": 0.8482342007434944, "grad_norm": 0.6348551373468717, "learning_rate": 5.998067953144498e-07, "loss": 0.0637, "step": 18254 }, { "epoch": 0.8482806691449815, "grad_norm": 0.5768319431739402, "learning_rate": 5.99449475000401e-07, "loss": 0.0522, "step": 18255 }, { "epoch": 0.8483271375464684, "grad_norm": 0.5641252994968088, "learning_rate": 5.990922543634525e-07, "loss": 0.0558, "step": 18256 }, { "epoch": 0.8483736059479554, "grad_norm": 0.6106636555190459, "learning_rate": 5.98735133411697e-07, "loss": 0.0664, "step": 18257 }, { "epoch": 0.8484200743494423, "grad_norm": 0.3028142149058592, "learning_rate": 5.983781121532234e-07, "loss": 0.0423, "step": 18258 }, { "epoch": 0.8484665427509294, "grad_norm": 0.5493439868990142, "learning_rate": 5.980211905961181e-07, "loss": 0.0523, "step": 18259 }, { "epoch": 0.8485130111524164, "grad_norm": 0.797009574828762, "learning_rate": 5.976643687484657e-07, "loss": 0.0474, "step": 18260 }, { "epoch": 0.8485594795539033, "grad_norm": 0.6039284825397181, "learning_rate": 5.973076466183486e-07, "loss": 0.0632, "step": 18261 }, { "epoch": 0.8486059479553903, "grad_norm": 0.5423584939636027, "learning_rate": 5.969510242138471e-07, "loss": 0.0591, "step": 18262 }, { "epoch": 0.8486524163568773, "grad_norm": 1.0711445282561487, "learning_rate": 5.96594501543038e-07, "loss": 0.043, "step": 18263 }, { "epoch": 0.8486988847583643, "grad_norm": 0.5121763138688237, "learning_rate": 5.962380786139993e-07, "loss": 0.0481, "step": 18264 }, { "epoch": 0.8487453531598513, "grad_norm": 0.4757321822330724, "learning_rate": 5.95881755434804e-07, "loss": 0.0564, "step": 18265 }, { "epoch": 0.8487918215613383, "grad_norm": 0.5035236751545654, "learning_rate": 5.955255320135195e-07, "loss": 0.0511, "step": 18266 }, { "epoch": 0.8488382899628253, "grad_norm": 0.5125580470443417, "learning_rate": 5.951694083582188e-07, "loss": 0.0569, "step": 18267 }, { "epoch": 0.8488847583643123, "grad_norm": 0.6243341514595534, "learning_rate": 5.948133844769672e-07, "loss": 0.0577, "step": 18268 }, { "epoch": 0.8489312267657992, "grad_norm": 0.8171267565474176, "learning_rate": 5.944574603778286e-07, "loss": 0.0552, "step": 18269 }, { "epoch": 0.8489776951672863, "grad_norm": 0.3726752629079614, "learning_rate": 5.94101636068864e-07, "loss": 0.0322, "step": 18270 }, { "epoch": 0.8490241635687732, "grad_norm": 0.3414827257792305, "learning_rate": 5.937459115581368e-07, "loss": 0.0215, "step": 18271 }, { "epoch": 0.8490706319702602, "grad_norm": 0.6530933016414068, "learning_rate": 5.933902868537011e-07, "loss": 0.0476, "step": 18272 }, { "epoch": 0.8491171003717473, "grad_norm": 0.5819012056421844, "learning_rate": 5.930347619636123e-07, "loss": 0.0685, "step": 18273 }, { "epoch": 0.8491635687732342, "grad_norm": 0.3541577509827281, "learning_rate": 5.926793368959254e-07, "loss": 0.0387, "step": 18274 }, { "epoch": 0.8492100371747212, "grad_norm": 0.6205430338504261, "learning_rate": 5.923240116586904e-07, "loss": 0.0839, "step": 18275 }, { "epoch": 0.8492565055762081, "grad_norm": 0.4198464796012495, "learning_rate": 5.919687862599549e-07, "loss": 0.0483, "step": 18276 }, { "epoch": 0.8493029739776952, "grad_norm": 0.6948124394477837, "learning_rate": 5.916136607077661e-07, "loss": 0.0534, "step": 18277 }, { "epoch": 0.8493494423791822, "grad_norm": 0.3331823781429385, "learning_rate": 5.912586350101679e-07, "loss": 0.0614, "step": 18278 }, { "epoch": 0.8493959107806691, "grad_norm": 0.5241394263324491, "learning_rate": 5.909037091752013e-07, "loss": 0.0606, "step": 18279 }, { "epoch": 0.8494423791821561, "grad_norm": 0.6407371027562205, "learning_rate": 5.905488832109053e-07, "loss": 0.0549, "step": 18280 }, { "epoch": 0.8494888475836431, "grad_norm": 0.3514132387053662, "learning_rate": 5.901941571253189e-07, "loss": 0.0505, "step": 18281 }, { "epoch": 0.8495353159851301, "grad_norm": 0.36613575815134797, "learning_rate": 5.898395309264765e-07, "loss": 0.0305, "step": 18282 }, { "epoch": 0.8495817843866171, "grad_norm": 0.3892365087532127, "learning_rate": 5.894850046224099e-07, "loss": 0.0494, "step": 18283 }, { "epoch": 0.849628252788104, "grad_norm": 0.6038993356162334, "learning_rate": 5.891305782211498e-07, "loss": 0.0587, "step": 18284 }, { "epoch": 0.8496747211895911, "grad_norm": 0.6733403278438652, "learning_rate": 5.88776251730725e-07, "loss": 0.0605, "step": 18285 }, { "epoch": 0.8497211895910781, "grad_norm": 0.4505056251365476, "learning_rate": 5.884220251591605e-07, "loss": 0.0579, "step": 18286 }, { "epoch": 0.849767657992565, "grad_norm": 0.3811956662903923, "learning_rate": 5.880678985144794e-07, "loss": 0.0381, "step": 18287 }, { "epoch": 0.8498141263940521, "grad_norm": 0.5538803098100317, "learning_rate": 5.877138718047048e-07, "loss": 0.0622, "step": 18288 }, { "epoch": 0.849860594795539, "grad_norm": 0.5590941110871662, "learning_rate": 5.873599450378559e-07, "loss": 0.0437, "step": 18289 }, { "epoch": 0.849907063197026, "grad_norm": 0.4857276805793528, "learning_rate": 5.87006118221946e-07, "loss": 0.0295, "step": 18290 }, { "epoch": 0.8499535315985131, "grad_norm": 0.5800242509775893, "learning_rate": 5.866523913649935e-07, "loss": 0.0594, "step": 18291 }, { "epoch": 0.85, "grad_norm": 0.6437967219294581, "learning_rate": 5.862987644750096e-07, "loss": 0.0503, "step": 18292 }, { "epoch": 0.850046468401487, "grad_norm": 0.5121675964225428, "learning_rate": 5.859452375600033e-07, "loss": 0.0536, "step": 18293 }, { "epoch": 0.8500929368029739, "grad_norm": 0.6494630898315653, "learning_rate": 5.855918106279823e-07, "loss": 0.0468, "step": 18294 }, { "epoch": 0.850139405204461, "grad_norm": 0.7044261849578499, "learning_rate": 5.852384836869551e-07, "loss": 0.0643, "step": 18295 }, { "epoch": 0.850185873605948, "grad_norm": 0.5727039689423427, "learning_rate": 5.848852567449221e-07, "loss": 0.0667, "step": 18296 }, { "epoch": 0.8502323420074349, "grad_norm": 0.394328653940404, "learning_rate": 5.845321298098833e-07, "loss": 0.0619, "step": 18297 }, { "epoch": 0.850278810408922, "grad_norm": 0.6602579408770605, "learning_rate": 5.841791028898397e-07, "loss": 0.0638, "step": 18298 }, { "epoch": 0.8503252788104089, "grad_norm": 0.6109546225584611, "learning_rate": 5.838261759927877e-07, "loss": 0.0485, "step": 18299 }, { "epoch": 0.8503717472118959, "grad_norm": 0.6150694093035236, "learning_rate": 5.834733491267203e-07, "loss": 0.0619, "step": 18300 }, { "epoch": 0.8504182156133829, "grad_norm": 0.562758156344198, "learning_rate": 5.8312062229963e-07, "loss": 0.0577, "step": 18301 }, { "epoch": 0.8504646840148699, "grad_norm": 0.5593827925864755, "learning_rate": 5.82767995519506e-07, "loss": 0.068, "step": 18302 }, { "epoch": 0.8505111524163569, "grad_norm": 0.5333982705577953, "learning_rate": 5.824154687943362e-07, "loss": 0.0469, "step": 18303 }, { "epoch": 0.8505576208178439, "grad_norm": 0.35721134430120516, "learning_rate": 5.820630421321044e-07, "loss": 0.054, "step": 18304 }, { "epoch": 0.8506040892193308, "grad_norm": 0.36874731955121437, "learning_rate": 5.817107155407958e-07, "loss": 0.0586, "step": 18305 }, { "epoch": 0.8506505576208179, "grad_norm": 0.5300023121357587, "learning_rate": 5.813584890283903e-07, "loss": 0.0457, "step": 18306 }, { "epoch": 0.8506970260223048, "grad_norm": 0.506227889473566, "learning_rate": 5.810063626028633e-07, "loss": 0.0592, "step": 18307 }, { "epoch": 0.8507434944237918, "grad_norm": 0.5207806974068857, "learning_rate": 5.806543362721945e-07, "loss": 0.0583, "step": 18308 }, { "epoch": 0.8507899628252789, "grad_norm": 0.35740630665018497, "learning_rate": 5.803024100443566e-07, "loss": 0.0405, "step": 18309 }, { "epoch": 0.8508364312267658, "grad_norm": 0.5126520394793157, "learning_rate": 5.799505839273201e-07, "loss": 0.0546, "step": 18310 }, { "epoch": 0.8508828996282528, "grad_norm": 0.48486729429173453, "learning_rate": 5.79598857929054e-07, "loss": 0.0364, "step": 18311 }, { "epoch": 0.8509293680297397, "grad_norm": 0.5815042384375227, "learning_rate": 5.792472320575282e-07, "loss": 0.0626, "step": 18312 }, { "epoch": 0.8509758364312268, "grad_norm": 0.5018251023062165, "learning_rate": 5.788957063207045e-07, "loss": 0.0571, "step": 18313 }, { "epoch": 0.8510223048327138, "grad_norm": 0.6431645598249704, "learning_rate": 5.785442807265451e-07, "loss": 0.0569, "step": 18314 }, { "epoch": 0.8510687732342007, "grad_norm": 0.38462936914656426, "learning_rate": 5.781929552830123e-07, "loss": 0.0461, "step": 18315 }, { "epoch": 0.8511152416356877, "grad_norm": 0.4350181914702347, "learning_rate": 5.778417299980632e-07, "loss": 0.0592, "step": 18316 }, { "epoch": 0.8511617100371747, "grad_norm": 0.38479506395919744, "learning_rate": 5.77490604879653e-07, "loss": 0.0471, "step": 18317 }, { "epoch": 0.8512081784386617, "grad_norm": 0.2885810804175737, "learning_rate": 5.771395799357349e-07, "loss": 0.0276, "step": 18318 }, { "epoch": 0.8512546468401487, "grad_norm": 0.49574773973231057, "learning_rate": 5.767886551742602e-07, "loss": 0.0568, "step": 18319 }, { "epoch": 0.8513011152416357, "grad_norm": 0.5779398250938, "learning_rate": 5.764378306031781e-07, "loss": 0.0655, "step": 18320 }, { "epoch": 0.8513475836431227, "grad_norm": 0.47354631647579176, "learning_rate": 5.76087106230434e-07, "loss": 0.0607, "step": 18321 }, { "epoch": 0.8513940520446097, "grad_norm": 0.550970520340849, "learning_rate": 5.757364820639738e-07, "loss": 0.0385, "step": 18322 }, { "epoch": 0.8514405204460966, "grad_norm": 0.4566431001198866, "learning_rate": 5.753859581117388e-07, "loss": 0.0477, "step": 18323 }, { "epoch": 0.8514869888475837, "grad_norm": 0.5410056528711918, "learning_rate": 5.750355343816689e-07, "loss": 0.0584, "step": 18324 }, { "epoch": 0.8515334572490706, "grad_norm": 0.31405072207908724, "learning_rate": 5.746852108817008e-07, "loss": 0.0406, "step": 18325 }, { "epoch": 0.8515799256505576, "grad_norm": 0.505924199111845, "learning_rate": 5.743349876197706e-07, "loss": 0.0591, "step": 18326 }, { "epoch": 0.8516263940520447, "grad_norm": 0.6358836379316464, "learning_rate": 5.739848646038104e-07, "loss": 0.0539, "step": 18327 }, { "epoch": 0.8516728624535316, "grad_norm": 0.6774642948792203, "learning_rate": 5.736348418417509e-07, "loss": 0.0513, "step": 18328 }, { "epoch": 0.8517193308550186, "grad_norm": 0.5437580429448639, "learning_rate": 5.732849193415219e-07, "loss": 0.0364, "step": 18329 }, { "epoch": 0.8517657992565055, "grad_norm": 0.4652023429383277, "learning_rate": 5.729350971110492e-07, "loss": 0.0359, "step": 18330 }, { "epoch": 0.8518122676579926, "grad_norm": 0.6448795646457653, "learning_rate": 5.725853751582538e-07, "loss": 0.0642, "step": 18331 }, { "epoch": 0.8518587360594796, "grad_norm": 0.5057266348581493, "learning_rate": 5.722357534910611e-07, "loss": 0.0433, "step": 18332 }, { "epoch": 0.8519052044609665, "grad_norm": 0.5740468462320438, "learning_rate": 5.718862321173879e-07, "loss": 0.0662, "step": 18333 }, { "epoch": 0.8519516728624535, "grad_norm": 0.725370951629316, "learning_rate": 5.715368110451524e-07, "loss": 0.0656, "step": 18334 }, { "epoch": 0.8519981412639405, "grad_norm": 0.439201001623934, "learning_rate": 5.711874902822684e-07, "loss": 0.0513, "step": 18335 }, { "epoch": 0.8520446096654275, "grad_norm": 0.6772975843880363, "learning_rate": 5.708382698366505e-07, "loss": 0.0524, "step": 18336 }, { "epoch": 0.8520910780669145, "grad_norm": 0.5077330108033934, "learning_rate": 5.704891497162063e-07, "loss": 0.0596, "step": 18337 }, { "epoch": 0.8521375464684015, "grad_norm": 0.6209359305979649, "learning_rate": 5.70140129928844e-07, "loss": 0.0551, "step": 18338 }, { "epoch": 0.8521840148698885, "grad_norm": 0.5507941386497789, "learning_rate": 5.697912104824716e-07, "loss": 0.0534, "step": 18339 }, { "epoch": 0.8522304832713755, "grad_norm": 0.5087798394603719, "learning_rate": 5.694423913849901e-07, "loss": 0.0601, "step": 18340 }, { "epoch": 0.8522769516728624, "grad_norm": 0.46674909694458977, "learning_rate": 5.69093672644302e-07, "loss": 0.0543, "step": 18341 }, { "epoch": 0.8523234200743495, "grad_norm": 0.454834674538136, "learning_rate": 5.68745054268306e-07, "loss": 0.0539, "step": 18342 }, { "epoch": 0.8523698884758364, "grad_norm": 0.6108707581705616, "learning_rate": 5.683965362648974e-07, "loss": 0.0489, "step": 18343 }, { "epoch": 0.8524163568773234, "grad_norm": 0.5707038468133906, "learning_rate": 5.680481186419723e-07, "loss": 0.0602, "step": 18344 }, { "epoch": 0.8524628252788105, "grad_norm": 0.45498292261603296, "learning_rate": 5.676998014074203e-07, "loss": 0.0397, "step": 18345 }, { "epoch": 0.8525092936802974, "grad_norm": 0.4201307654900819, "learning_rate": 5.673515845691341e-07, "loss": 0.046, "step": 18346 }, { "epoch": 0.8525557620817844, "grad_norm": 0.7508636735603407, "learning_rate": 5.670034681349995e-07, "loss": 0.0542, "step": 18347 }, { "epoch": 0.8526022304832713, "grad_norm": 0.4473200667945755, "learning_rate": 5.666554521129025e-07, "loss": 0.0423, "step": 18348 }, { "epoch": 0.8526486988847584, "grad_norm": 0.668997104459614, "learning_rate": 5.663075365107251e-07, "loss": 0.0451, "step": 18349 }, { "epoch": 0.8526951672862454, "grad_norm": 0.562078537852199, "learning_rate": 5.659597213363483e-07, "loss": 0.0423, "step": 18350 }, { "epoch": 0.8527416356877323, "grad_norm": 0.6041383964863812, "learning_rate": 5.656120065976506e-07, "loss": 0.0456, "step": 18351 }, { "epoch": 0.8527881040892193, "grad_norm": 0.6164437999398845, "learning_rate": 5.652643923025069e-07, "loss": 0.0576, "step": 18352 }, { "epoch": 0.8528345724907063, "grad_norm": 0.46259173222330613, "learning_rate": 5.649168784587938e-07, "loss": 0.0357, "step": 18353 }, { "epoch": 0.8528810408921933, "grad_norm": 0.5559933081850231, "learning_rate": 5.645694650743816e-07, "loss": 0.046, "step": 18354 }, { "epoch": 0.8529275092936803, "grad_norm": 0.6594844626756539, "learning_rate": 5.642221521571379e-07, "loss": 0.0546, "step": 18355 }, { "epoch": 0.8529739776951673, "grad_norm": 0.47337323927761765, "learning_rate": 5.638749397149312e-07, "loss": 0.065, "step": 18356 }, { "epoch": 0.8530204460966543, "grad_norm": 0.47504927967466176, "learning_rate": 5.635278277556267e-07, "loss": 0.0591, "step": 18357 }, { "epoch": 0.8530669144981413, "grad_norm": 0.44614303941561634, "learning_rate": 5.631808162870861e-07, "loss": 0.0492, "step": 18358 }, { "epoch": 0.8531133828996282, "grad_norm": 0.29161747552200323, "learning_rate": 5.628339053171694e-07, "loss": 0.049, "step": 18359 }, { "epoch": 0.8531598513011153, "grad_norm": 0.9088247878343193, "learning_rate": 5.624870948537348e-07, "loss": 0.065, "step": 18360 }, { "epoch": 0.8532063197026022, "grad_norm": 0.5371611664849826, "learning_rate": 5.621403849046381e-07, "loss": 0.0488, "step": 18361 }, { "epoch": 0.8532527881040892, "grad_norm": 0.5186083356414649, "learning_rate": 5.617937754777314e-07, "loss": 0.0423, "step": 18362 }, { "epoch": 0.8532992565055763, "grad_norm": 0.3197079180120488, "learning_rate": 5.614472665808678e-07, "loss": 0.042, "step": 18363 }, { "epoch": 0.8533457249070632, "grad_norm": 0.7079134508622116, "learning_rate": 5.611008582218952e-07, "loss": 0.0537, "step": 18364 }, { "epoch": 0.8533921933085502, "grad_norm": 0.5656655546064967, "learning_rate": 5.607545504086598e-07, "loss": 0.0466, "step": 18365 }, { "epoch": 0.8534386617100371, "grad_norm": 0.5306473815824212, "learning_rate": 5.60408343149006e-07, "loss": 0.0683, "step": 18366 }, { "epoch": 0.8534851301115242, "grad_norm": 0.6606673746595423, "learning_rate": 5.600622364507763e-07, "loss": 0.0498, "step": 18367 }, { "epoch": 0.8535315985130112, "grad_norm": 0.6546541396286645, "learning_rate": 5.597162303218095e-07, "loss": 0.049, "step": 18368 }, { "epoch": 0.8535780669144981, "grad_norm": 0.30641447057600535, "learning_rate": 5.593703247699423e-07, "loss": 0.044, "step": 18369 }, { "epoch": 0.8536245353159851, "grad_norm": 0.32358562040195105, "learning_rate": 5.59024519803012e-07, "loss": 0.047, "step": 18370 }, { "epoch": 0.8536710037174721, "grad_norm": 0.47143711832840957, "learning_rate": 5.586788154288514e-07, "loss": 0.0609, "step": 18371 }, { "epoch": 0.8537174721189591, "grad_norm": 0.6082421788017884, "learning_rate": 5.583332116552881e-07, "loss": 0.057, "step": 18372 }, { "epoch": 0.8537639405204461, "grad_norm": 0.37936944385503313, "learning_rate": 5.579877084901536e-07, "loss": 0.0184, "step": 18373 }, { "epoch": 0.853810408921933, "grad_norm": 0.6384040053134221, "learning_rate": 5.576423059412723e-07, "loss": 0.0311, "step": 18374 }, { "epoch": 0.8538568773234201, "grad_norm": 0.3346578374149391, "learning_rate": 5.572970040164682e-07, "loss": 0.0192, "step": 18375 }, { "epoch": 0.8539033457249071, "grad_norm": 0.41977014106093025, "learning_rate": 5.569518027235621e-07, "loss": 0.039, "step": 18376 }, { "epoch": 0.853949814126394, "grad_norm": 0.3940548437498464, "learning_rate": 5.566067020703758e-07, "loss": 0.0583, "step": 18377 }, { "epoch": 0.8539962825278811, "grad_norm": 0.38150609773242516, "learning_rate": 5.562617020647232e-07, "loss": 0.0334, "step": 18378 }, { "epoch": 0.854042750929368, "grad_norm": 0.571373067342218, "learning_rate": 5.55916802714419e-07, "loss": 0.053, "step": 18379 }, { "epoch": 0.854089219330855, "grad_norm": 0.5721076131466664, "learning_rate": 5.555720040272772e-07, "loss": 0.0451, "step": 18380 }, { "epoch": 0.854135687732342, "grad_norm": 0.261576867997932, "learning_rate": 5.552273060111074e-07, "loss": 0.0316, "step": 18381 }, { "epoch": 0.854182156133829, "grad_norm": 0.45153731238042644, "learning_rate": 5.548827086737174e-07, "loss": 0.0666, "step": 18382 }, { "epoch": 0.854228624535316, "grad_norm": 0.6765166361292414, "learning_rate": 5.545382120229126e-07, "loss": 0.0608, "step": 18383 }, { "epoch": 0.8542750929368029, "grad_norm": 0.47595754027434095, "learning_rate": 5.541938160664956e-07, "loss": 0.0517, "step": 18384 }, { "epoch": 0.85432156133829, "grad_norm": 0.5490324592463273, "learning_rate": 5.538495208122679e-07, "loss": 0.0597, "step": 18385 }, { "epoch": 0.854368029739777, "grad_norm": 0.44667523871836495, "learning_rate": 5.535053262680273e-07, "loss": 0.0628, "step": 18386 }, { "epoch": 0.8544144981412639, "grad_norm": 0.6787633405936042, "learning_rate": 5.53161232441572e-07, "loss": 0.0659, "step": 18387 }, { "epoch": 0.8544609665427509, "grad_norm": 0.4246646565821442, "learning_rate": 5.528172393406944e-07, "loss": 0.0493, "step": 18388 }, { "epoch": 0.8545074349442379, "grad_norm": 0.3877093134945314, "learning_rate": 5.524733469731875e-07, "loss": 0.0591, "step": 18389 }, { "epoch": 0.8545539033457249, "grad_norm": 0.5045109126731243, "learning_rate": 5.521295553468398e-07, "loss": 0.0496, "step": 18390 }, { "epoch": 0.8546003717472119, "grad_norm": 0.3233641214497801, "learning_rate": 5.51785864469439e-07, "loss": 0.0376, "step": 18391 }, { "epoch": 0.8546468401486988, "grad_norm": 0.5598787443122942, "learning_rate": 5.514422743487702e-07, "loss": 0.0516, "step": 18392 }, { "epoch": 0.8546933085501859, "grad_norm": 0.34615256352918605, "learning_rate": 5.510987849926147e-07, "loss": 0.0476, "step": 18393 }, { "epoch": 0.8547397769516729, "grad_norm": 0.7427666771501762, "learning_rate": 5.507553964087553e-07, "loss": 0.0603, "step": 18394 }, { "epoch": 0.8547862453531598, "grad_norm": 0.4734140748916917, "learning_rate": 5.504121086049701e-07, "loss": 0.0477, "step": 18395 }, { "epoch": 0.8548327137546469, "grad_norm": 0.5730733803520844, "learning_rate": 5.500689215890314e-07, "loss": 0.0561, "step": 18396 }, { "epoch": 0.8548791821561338, "grad_norm": 0.40563722339546293, "learning_rate": 5.497258353687162e-07, "loss": 0.0512, "step": 18397 }, { "epoch": 0.8549256505576208, "grad_norm": 0.6180032180348342, "learning_rate": 5.493828499517951e-07, "loss": 0.0508, "step": 18398 }, { "epoch": 0.8549721189591079, "grad_norm": 0.4127465001929542, "learning_rate": 5.49039965346036e-07, "loss": 0.0391, "step": 18399 }, { "epoch": 0.8550185873605948, "grad_norm": 0.43284021873138157, "learning_rate": 5.486971815592057e-07, "loss": 0.0592, "step": 18400 }, { "epoch": 0.8550650557620818, "grad_norm": 0.39328111818950223, "learning_rate": 5.483544985990707e-07, "loss": 0.0483, "step": 18401 }, { "epoch": 0.8551115241635687, "grad_norm": 0.6026142519024638, "learning_rate": 5.480119164733911e-07, "loss": 0.049, "step": 18402 }, { "epoch": 0.8551579925650558, "grad_norm": 0.45779455054138196, "learning_rate": 5.476694351899259e-07, "loss": 0.0513, "step": 18403 }, { "epoch": 0.8552044609665428, "grad_norm": 0.35873460897515214, "learning_rate": 5.473270547564346e-07, "loss": 0.0475, "step": 18404 }, { "epoch": 0.8552509293680297, "grad_norm": 0.43538570036408725, "learning_rate": 5.469847751806723e-07, "loss": 0.0398, "step": 18405 }, { "epoch": 0.8552973977695167, "grad_norm": 0.6811563425004307, "learning_rate": 5.466425964703914e-07, "loss": 0.0384, "step": 18406 }, { "epoch": 0.8553438661710037, "grad_norm": 0.4926121722001393, "learning_rate": 5.463005186333431e-07, "loss": 0.0561, "step": 18407 }, { "epoch": 0.8553903345724907, "grad_norm": 0.4345080400953395, "learning_rate": 5.459585416772756e-07, "loss": 0.048, "step": 18408 }, { "epoch": 0.8554368029739777, "grad_norm": 0.38223912821581413, "learning_rate": 5.456166656099344e-07, "loss": 0.0511, "step": 18409 }, { "epoch": 0.8554832713754646, "grad_norm": 0.6114274071544074, "learning_rate": 5.452748904390631e-07, "loss": 0.0613, "step": 18410 }, { "epoch": 0.8555297397769517, "grad_norm": 0.39477509093260166, "learning_rate": 5.449332161724047e-07, "loss": 0.0486, "step": 18411 }, { "epoch": 0.8555762081784387, "grad_norm": 0.6593318877380027, "learning_rate": 5.445916428176984e-07, "loss": 0.0537, "step": 18412 }, { "epoch": 0.8556226765799256, "grad_norm": 0.4433251515072079, "learning_rate": 5.442501703826803e-07, "loss": 0.0548, "step": 18413 }, { "epoch": 0.8556691449814127, "grad_norm": 0.5693433958834492, "learning_rate": 5.439087988750858e-07, "loss": 0.0432, "step": 18414 }, { "epoch": 0.8557156133828996, "grad_norm": 0.7106174910338794, "learning_rate": 5.435675283026465e-07, "loss": 0.0612, "step": 18415 }, { "epoch": 0.8557620817843866, "grad_norm": 0.32260104353982844, "learning_rate": 5.432263586730929e-07, "loss": 0.0356, "step": 18416 }, { "epoch": 0.8558085501858737, "grad_norm": 0.4126453599195301, "learning_rate": 5.428852899941522e-07, "loss": 0.048, "step": 18417 }, { "epoch": 0.8558550185873606, "grad_norm": 0.5440911324595377, "learning_rate": 5.425443222735527e-07, "loss": 0.0446, "step": 18418 }, { "epoch": 0.8559014869888476, "grad_norm": 0.38365653678698963, "learning_rate": 5.422034555190148e-07, "loss": 0.0429, "step": 18419 }, { "epoch": 0.8559479553903345, "grad_norm": 0.3287094783893082, "learning_rate": 5.418626897382595e-07, "loss": 0.047, "step": 18420 }, { "epoch": 0.8559944237918216, "grad_norm": 0.5508775961679884, "learning_rate": 5.415220249390068e-07, "loss": 0.0406, "step": 18421 }, { "epoch": 0.8560408921933086, "grad_norm": 0.3869524896362962, "learning_rate": 5.411814611289735e-07, "loss": 0.0523, "step": 18422 }, { "epoch": 0.8560873605947955, "grad_norm": 0.4695407418650634, "learning_rate": 5.408409983158725e-07, "loss": 0.042, "step": 18423 }, { "epoch": 0.8561338289962825, "grad_norm": 0.2944439598790982, "learning_rate": 5.405006365074155e-07, "loss": 0.041, "step": 18424 }, { "epoch": 0.8561802973977695, "grad_norm": 0.6738534098382362, "learning_rate": 5.401603757113133e-07, "loss": 0.0575, "step": 18425 }, { "epoch": 0.8562267657992565, "grad_norm": 0.6182253042443536, "learning_rate": 5.398202159352722e-07, "loss": 0.0661, "step": 18426 }, { "epoch": 0.8562732342007435, "grad_norm": 0.4777634783601264, "learning_rate": 5.394801571869967e-07, "loss": 0.0567, "step": 18427 }, { "epoch": 0.8563197026022304, "grad_norm": 0.39773915248386754, "learning_rate": 5.391401994741913e-07, "loss": 0.0418, "step": 18428 }, { "epoch": 0.8563661710037175, "grad_norm": 0.638473660117358, "learning_rate": 5.388003428045552e-07, "loss": 0.0498, "step": 18429 }, { "epoch": 0.8564126394052045, "grad_norm": 0.31274001359954, "learning_rate": 5.384605871857867e-07, "loss": 0.046, "step": 18430 }, { "epoch": 0.8564591078066914, "grad_norm": 0.47155630121170605, "learning_rate": 5.381209326255815e-07, "loss": 0.0484, "step": 18431 }, { "epoch": 0.8565055762081785, "grad_norm": 0.6567709698070558, "learning_rate": 5.377813791316333e-07, "loss": 0.0634, "step": 18432 }, { "epoch": 0.8565520446096654, "grad_norm": 0.6127827335162236, "learning_rate": 5.374419267116332e-07, "loss": 0.0547, "step": 18433 }, { "epoch": 0.8565985130111524, "grad_norm": 0.5019040065971077, "learning_rate": 5.371025753732695e-07, "loss": 0.0383, "step": 18434 }, { "epoch": 0.8566449814126395, "grad_norm": 0.7132461465823348, "learning_rate": 5.367633251242304e-07, "loss": 0.0437, "step": 18435 }, { "epoch": 0.8566914498141264, "grad_norm": 0.4280671976498462, "learning_rate": 5.364241759722006e-07, "loss": 0.048, "step": 18436 }, { "epoch": 0.8567379182156134, "grad_norm": 0.3680017308034574, "learning_rate": 5.360851279248585e-07, "loss": 0.0521, "step": 18437 }, { "epoch": 0.8567843866171003, "grad_norm": 0.3519481054535556, "learning_rate": 5.357461809898878e-07, "loss": 0.0527, "step": 18438 }, { "epoch": 0.8568308550185874, "grad_norm": 0.8904881242294949, "learning_rate": 5.354073351749644e-07, "loss": 0.0681, "step": 18439 }, { "epoch": 0.8568773234200744, "grad_norm": 0.3408271862141975, "learning_rate": 5.350685904877645e-07, "loss": 0.0374, "step": 18440 }, { "epoch": 0.8569237918215613, "grad_norm": 0.4130078105950428, "learning_rate": 5.347299469359585e-07, "loss": 0.0394, "step": 18441 }, { "epoch": 0.8569702602230483, "grad_norm": 0.4115229951152397, "learning_rate": 5.343914045272208e-07, "loss": 0.0353, "step": 18442 }, { "epoch": 0.8570167286245353, "grad_norm": 0.501999695283609, "learning_rate": 5.340529632692171e-07, "loss": 0.0457, "step": 18443 }, { "epoch": 0.8570631970260223, "grad_norm": 0.5941793671873112, "learning_rate": 5.337146231696133e-07, "loss": 0.0647, "step": 18444 }, { "epoch": 0.8571096654275093, "grad_norm": 0.489574591016941, "learning_rate": 5.333763842360745e-07, "loss": 0.04, "step": 18445 }, { "epoch": 0.8571561338289962, "grad_norm": 0.6621861055100905, "learning_rate": 5.330382464762618e-07, "loss": 0.0661, "step": 18446 }, { "epoch": 0.8572026022304833, "grad_norm": 0.4061886339498913, "learning_rate": 5.327002098978345e-07, "loss": 0.0586, "step": 18447 }, { "epoch": 0.8572490706319703, "grad_norm": 0.5893047265527436, "learning_rate": 5.323622745084489e-07, "loss": 0.0567, "step": 18448 }, { "epoch": 0.8572955390334572, "grad_norm": 0.6085934928408037, "learning_rate": 5.320244403157599e-07, "loss": 0.0679, "step": 18449 }, { "epoch": 0.8573420074349443, "grad_norm": 0.42070042110246847, "learning_rate": 5.316867073274201e-07, "loss": 0.0393, "step": 18450 }, { "epoch": 0.8573884758364312, "grad_norm": 0.7129558256465316, "learning_rate": 5.313490755510776e-07, "loss": 0.0669, "step": 18451 }, { "epoch": 0.8574349442379182, "grad_norm": 0.29143091575513935, "learning_rate": 5.310115449943831e-07, "loss": 0.0371, "step": 18452 }, { "epoch": 0.8574814126394052, "grad_norm": 0.32178196339449056, "learning_rate": 5.306741156649802e-07, "loss": 0.0503, "step": 18453 }, { "epoch": 0.8575278810408922, "grad_norm": 0.2400864899763705, "learning_rate": 5.303367875705129e-07, "loss": 0.0319, "step": 18454 }, { "epoch": 0.8575743494423792, "grad_norm": 0.5182444147146076, "learning_rate": 5.29999560718622e-07, "loss": 0.0373, "step": 18455 }, { "epoch": 0.8576208178438661, "grad_norm": 0.49268906055058487, "learning_rate": 5.29662435116945e-07, "loss": 0.0557, "step": 18456 }, { "epoch": 0.8576672862453532, "grad_norm": 0.3689476702462766, "learning_rate": 5.293254107731189e-07, "loss": 0.0298, "step": 18457 }, { "epoch": 0.8577137546468402, "grad_norm": 0.2928193190726653, "learning_rate": 5.289884876947765e-07, "loss": 0.0315, "step": 18458 }, { "epoch": 0.8577602230483271, "grad_norm": 0.5260370553159303, "learning_rate": 5.286516658895519e-07, "loss": 0.0502, "step": 18459 }, { "epoch": 0.8578066914498141, "grad_norm": 0.5192470501563686, "learning_rate": 5.283149453650732e-07, "loss": 0.0465, "step": 18460 }, { "epoch": 0.8578531598513011, "grad_norm": 0.6163993994852496, "learning_rate": 5.279783261289662e-07, "loss": 0.065, "step": 18461 }, { "epoch": 0.8578996282527881, "grad_norm": 0.22101932087220436, "learning_rate": 5.276418081888574e-07, "loss": 0.0339, "step": 18462 }, { "epoch": 0.8579460966542751, "grad_norm": 0.5017707620140764, "learning_rate": 5.27305391552369e-07, "loss": 0.0532, "step": 18463 }, { "epoch": 0.857992565055762, "grad_norm": 0.3750771623295821, "learning_rate": 5.269690762271207e-07, "loss": 0.0424, "step": 18464 }, { "epoch": 0.8580390334572491, "grad_norm": 0.5173688983452278, "learning_rate": 5.266328622207295e-07, "loss": 0.0447, "step": 18465 }, { "epoch": 0.8580855018587361, "grad_norm": 0.44283798774870214, "learning_rate": 5.262967495408145e-07, "loss": 0.0474, "step": 18466 }, { "epoch": 0.858131970260223, "grad_norm": 0.5311313252686767, "learning_rate": 5.259607381949855e-07, "loss": 0.0565, "step": 18467 }, { "epoch": 0.8581784386617101, "grad_norm": 0.7135406513304338, "learning_rate": 5.256248281908532e-07, "loss": 0.0588, "step": 18468 }, { "epoch": 0.858224907063197, "grad_norm": 0.550311023202812, "learning_rate": 5.252890195360294e-07, "loss": 0.0548, "step": 18469 }, { "epoch": 0.858271375464684, "grad_norm": 0.47804308729373074, "learning_rate": 5.249533122381184e-07, "loss": 0.0556, "step": 18470 }, { "epoch": 0.858317843866171, "grad_norm": 0.5386467799051776, "learning_rate": 5.246177063047247e-07, "loss": 0.0508, "step": 18471 }, { "epoch": 0.858364312267658, "grad_norm": 0.43016809144400187, "learning_rate": 5.242822017434507e-07, "loss": 0.0495, "step": 18472 }, { "epoch": 0.858410780669145, "grad_norm": 0.36577686646106555, "learning_rate": 5.239467985618951e-07, "loss": 0.0307, "step": 18473 }, { "epoch": 0.8584572490706319, "grad_norm": 0.624463517539801, "learning_rate": 5.236114967676553e-07, "loss": 0.0786, "step": 18474 }, { "epoch": 0.858503717472119, "grad_norm": 0.6637005855154254, "learning_rate": 5.232762963683257e-07, "loss": 0.0517, "step": 18475 }, { "epoch": 0.858550185873606, "grad_norm": 0.57860053493605, "learning_rate": 5.229411973715004e-07, "loss": 0.0594, "step": 18476 }, { "epoch": 0.8585966542750929, "grad_norm": 0.559286011722769, "learning_rate": 5.226061997847703e-07, "loss": 0.0628, "step": 18477 }, { "epoch": 0.8586431226765799, "grad_norm": 0.3664017029345797, "learning_rate": 5.2227130361572e-07, "loss": 0.0569, "step": 18478 }, { "epoch": 0.8586895910780669, "grad_norm": 0.49814201234039185, "learning_rate": 5.219365088719386e-07, "loss": 0.0403, "step": 18479 }, { "epoch": 0.8587360594795539, "grad_norm": 0.5695695347041283, "learning_rate": 5.216018155610081e-07, "loss": 0.0453, "step": 18480 }, { "epoch": 0.8587825278810409, "grad_norm": 0.40362592995174335, "learning_rate": 5.212672236905097e-07, "loss": 0.0446, "step": 18481 }, { "epoch": 0.8588289962825278, "grad_norm": 0.7068997341506073, "learning_rate": 5.20932733268022e-07, "loss": 0.043, "step": 18482 }, { "epoch": 0.8588754646840149, "grad_norm": 0.5706060627998858, "learning_rate": 5.205983443011237e-07, "loss": 0.0445, "step": 18483 }, { "epoch": 0.8589219330855019, "grad_norm": 0.9140108837530998, "learning_rate": 5.20264056797386e-07, "loss": 0.0614, "step": 18484 }, { "epoch": 0.8589684014869888, "grad_norm": 0.4323725005151894, "learning_rate": 5.199298707643818e-07, "loss": 0.0426, "step": 18485 }, { "epoch": 0.8590148698884759, "grad_norm": 0.37074883850611245, "learning_rate": 5.195957862096823e-07, "loss": 0.0335, "step": 18486 }, { "epoch": 0.8590613382899628, "grad_norm": 0.7468227608934492, "learning_rate": 5.192618031408536e-07, "loss": 0.0586, "step": 18487 }, { "epoch": 0.8591078066914498, "grad_norm": 0.7636231168126615, "learning_rate": 5.189279215654608e-07, "loss": 0.0574, "step": 18488 }, { "epoch": 0.8591542750929368, "grad_norm": 0.6621015788784774, "learning_rate": 5.185941414910673e-07, "loss": 0.062, "step": 18489 }, { "epoch": 0.8592007434944238, "grad_norm": 0.5669982701020582, "learning_rate": 5.182604629252324e-07, "loss": 0.0691, "step": 18490 }, { "epoch": 0.8592472118959108, "grad_norm": 0.4598976707536961, "learning_rate": 5.179268858755154e-07, "loss": 0.0366, "step": 18491 }, { "epoch": 0.8592936802973977, "grad_norm": 0.38675902941869333, "learning_rate": 5.175934103494706e-07, "loss": 0.0586, "step": 18492 }, { "epoch": 0.8593401486988848, "grad_norm": 0.4851638827355692, "learning_rate": 5.172600363546537e-07, "loss": 0.0449, "step": 18493 }, { "epoch": 0.8593866171003718, "grad_norm": 0.6239496941100371, "learning_rate": 5.169267638986148e-07, "loss": 0.0624, "step": 18494 }, { "epoch": 0.8594330855018587, "grad_norm": 0.5019353223631285, "learning_rate": 5.165935929889032e-07, "loss": 0.0526, "step": 18495 }, { "epoch": 0.8594795539033457, "grad_norm": 0.39144547010431197, "learning_rate": 5.162605236330658e-07, "loss": 0.0422, "step": 18496 }, { "epoch": 0.8595260223048327, "grad_norm": 0.5427573440559433, "learning_rate": 5.159275558386462e-07, "loss": 0.0467, "step": 18497 }, { "epoch": 0.8595724907063197, "grad_norm": 0.4754013791292516, "learning_rate": 5.155946896131869e-07, "loss": 0.045, "step": 18498 }, { "epoch": 0.8596189591078067, "grad_norm": 0.40126236156589873, "learning_rate": 5.152619249642265e-07, "loss": 0.0457, "step": 18499 }, { "epoch": 0.8596654275092936, "grad_norm": 0.45467261332201975, "learning_rate": 5.149292618993051e-07, "loss": 0.0468, "step": 18500 }, { "epoch": 0.8597118959107807, "grad_norm": 0.37186521315534277, "learning_rate": 5.145967004259567e-07, "loss": 0.0567, "step": 18501 }, { "epoch": 0.8597583643122677, "grad_norm": 0.3988016824497661, "learning_rate": 5.142642405517123e-07, "loss": 0.0508, "step": 18502 }, { "epoch": 0.8598048327137546, "grad_norm": 0.7548240623481566, "learning_rate": 5.139318822841049e-07, "loss": 0.053, "step": 18503 }, { "epoch": 0.8598513011152417, "grad_norm": 0.8736485112241714, "learning_rate": 5.135996256306619e-07, "loss": 0.0657, "step": 18504 }, { "epoch": 0.8598977695167286, "grad_norm": 0.5471030587123206, "learning_rate": 5.132674705989088e-07, "loss": 0.0517, "step": 18505 }, { "epoch": 0.8599442379182156, "grad_norm": 0.5163229059619097, "learning_rate": 5.129354171963691e-07, "loss": 0.0566, "step": 18506 }, { "epoch": 0.8599907063197026, "grad_norm": 0.4762664700975491, "learning_rate": 5.126034654305667e-07, "loss": 0.0431, "step": 18507 }, { "epoch": 0.8600371747211896, "grad_norm": 0.6416135707436328, "learning_rate": 5.12271615309018e-07, "loss": 0.0465, "step": 18508 }, { "epoch": 0.8600836431226766, "grad_norm": 0.6415089093751073, "learning_rate": 5.119398668392389e-07, "loss": 0.0616, "step": 18509 }, { "epoch": 0.8601301115241635, "grad_norm": 0.5061306948790891, "learning_rate": 5.11608220028747e-07, "loss": 0.0379, "step": 18510 }, { "epoch": 0.8601765799256506, "grad_norm": 0.6275113305330039, "learning_rate": 5.112766748850523e-07, "loss": 0.0344, "step": 18511 }, { "epoch": 0.8602230483271376, "grad_norm": 0.4633776950980294, "learning_rate": 5.109452314156654e-07, "loss": 0.0528, "step": 18512 }, { "epoch": 0.8602695167286245, "grad_norm": 0.49589851952789055, "learning_rate": 5.106138896280938e-07, "loss": 0.0569, "step": 18513 }, { "epoch": 0.8603159851301115, "grad_norm": 0.6912839609188567, "learning_rate": 5.102826495298424e-07, "loss": 0.062, "step": 18514 }, { "epoch": 0.8603624535315985, "grad_norm": 0.6048695442296534, "learning_rate": 5.099515111284148e-07, "loss": 0.0491, "step": 18515 }, { "epoch": 0.8604089219330855, "grad_norm": 0.3455253535373504, "learning_rate": 5.096204744313099e-07, "loss": 0.0472, "step": 18516 }, { "epoch": 0.8604553903345725, "grad_norm": 0.5574854998333981, "learning_rate": 5.092895394460279e-07, "loss": 0.0447, "step": 18517 }, { "epoch": 0.8605018587360594, "grad_norm": 0.5081577428019254, "learning_rate": 5.089587061800644e-07, "loss": 0.0428, "step": 18518 }, { "epoch": 0.8605483271375465, "grad_norm": 0.6090998976800902, "learning_rate": 5.086279746409134e-07, "loss": 0.0578, "step": 18519 }, { "epoch": 0.8605947955390335, "grad_norm": 0.47077669153805496, "learning_rate": 5.082973448360651e-07, "loss": 0.049, "step": 18520 }, { "epoch": 0.8606412639405204, "grad_norm": 0.47267119478796527, "learning_rate": 5.079668167730101e-07, "loss": 0.0553, "step": 18521 }, { "epoch": 0.8606877323420075, "grad_norm": 0.43916607122718004, "learning_rate": 5.07636390459234e-07, "loss": 0.043, "step": 18522 }, { "epoch": 0.8607342007434944, "grad_norm": 0.5455531453209658, "learning_rate": 5.073060659022211e-07, "loss": 0.0358, "step": 18523 }, { "epoch": 0.8607806691449814, "grad_norm": 0.2990346533191326, "learning_rate": 5.069758431094551e-07, "loss": 0.0348, "step": 18524 }, { "epoch": 0.8608271375464684, "grad_norm": 0.605178929282392, "learning_rate": 5.066457220884158e-07, "loss": 0.048, "step": 18525 }, { "epoch": 0.8608736059479554, "grad_norm": 0.2593302662975262, "learning_rate": 5.063157028465787e-07, "loss": 0.0283, "step": 18526 }, { "epoch": 0.8609200743494424, "grad_norm": 0.530698641771854, "learning_rate": 5.059857853914207e-07, "loss": 0.0672, "step": 18527 }, { "epoch": 0.8609665427509293, "grad_norm": 0.5357209367139236, "learning_rate": 5.056559697304153e-07, "loss": 0.0451, "step": 18528 }, { "epoch": 0.8610130111524164, "grad_norm": 0.4660724131310166, "learning_rate": 5.05326255871032e-07, "loss": 0.0547, "step": 18529 }, { "epoch": 0.8610594795539034, "grad_norm": 0.4710187227943708, "learning_rate": 5.049966438207387e-07, "loss": 0.0653, "step": 18530 }, { "epoch": 0.8611059479553903, "grad_norm": 0.7333830455010983, "learning_rate": 5.04667133587004e-07, "loss": 0.0616, "step": 18531 }, { "epoch": 0.8611524163568773, "grad_norm": 0.2362348680570736, "learning_rate": 5.043377251772891e-07, "loss": 0.0228, "step": 18532 }, { "epoch": 0.8611988847583643, "grad_norm": 0.7316058208085353, "learning_rate": 5.040084185990557e-07, "loss": 0.0684, "step": 18533 }, { "epoch": 0.8612453531598513, "grad_norm": 0.4372759263010631, "learning_rate": 5.036792138597641e-07, "loss": 0.038, "step": 18534 }, { "epoch": 0.8612918215613383, "grad_norm": 0.530378349169805, "learning_rate": 5.03350110966871e-07, "loss": 0.0406, "step": 18535 }, { "epoch": 0.8613382899628252, "grad_norm": 0.6370167990622749, "learning_rate": 5.030211099278304e-07, "loss": 0.0764, "step": 18536 }, { "epoch": 0.8613847583643123, "grad_norm": 0.6974237142400352, "learning_rate": 5.026922107500942e-07, "loss": 0.0485, "step": 18537 }, { "epoch": 0.8614312267657993, "grad_norm": 0.4859553886827477, "learning_rate": 5.023634134411132e-07, "loss": 0.0501, "step": 18538 }, { "epoch": 0.8614776951672862, "grad_norm": 0.3823898411071599, "learning_rate": 5.020347180083346e-07, "loss": 0.0333, "step": 18539 }, { "epoch": 0.8615241635687733, "grad_norm": 0.5960192386902827, "learning_rate": 5.017061244592025e-07, "loss": 0.0472, "step": 18540 }, { "epoch": 0.8615706319702602, "grad_norm": 0.5014695396556884, "learning_rate": 5.013776328011621e-07, "loss": 0.0496, "step": 18541 }, { "epoch": 0.8616171003717472, "grad_norm": 0.4769055558197519, "learning_rate": 5.010492430416535e-07, "loss": 0.0619, "step": 18542 }, { "epoch": 0.8616635687732342, "grad_norm": 0.6207162172757503, "learning_rate": 5.007209551881137e-07, "loss": 0.0464, "step": 18543 }, { "epoch": 0.8617100371747212, "grad_norm": 0.5173382480408558, "learning_rate": 5.003927692479798e-07, "loss": 0.0642, "step": 18544 }, { "epoch": 0.8617565055762082, "grad_norm": 0.4678771146779433, "learning_rate": 5.000646852286855e-07, "loss": 0.0428, "step": 18545 }, { "epoch": 0.8618029739776951, "grad_norm": 3.1206936829373504, "learning_rate": 4.997367031376627e-07, "loss": 0.0801, "step": 18546 }, { "epoch": 0.8618494423791822, "grad_norm": 0.612852644739774, "learning_rate": 4.994088229823385e-07, "loss": 0.039, "step": 18547 }, { "epoch": 0.8618959107806692, "grad_norm": 0.4440029877810121, "learning_rate": 4.990810447701438e-07, "loss": 0.0536, "step": 18548 }, { "epoch": 0.8619423791821561, "grad_norm": 0.3573061223955879, "learning_rate": 4.987533685084989e-07, "loss": 0.0366, "step": 18549 }, { "epoch": 0.8619888475836431, "grad_norm": 0.399490538380496, "learning_rate": 4.984257942048271e-07, "loss": 0.0563, "step": 18550 }, { "epoch": 0.8620353159851301, "grad_norm": 0.40995942379003447, "learning_rate": 4.980983218665497e-07, "loss": 0.0297, "step": 18551 }, { "epoch": 0.8620817843866171, "grad_norm": 0.6253667395542474, "learning_rate": 4.977709515010837e-07, "loss": 0.0608, "step": 18552 }, { "epoch": 0.8621282527881041, "grad_norm": 0.5670246717336077, "learning_rate": 4.974436831158441e-07, "loss": 0.0486, "step": 18553 }, { "epoch": 0.862174721189591, "grad_norm": 0.39472340477854057, "learning_rate": 4.971165167182434e-07, "loss": 0.0519, "step": 18554 }, { "epoch": 0.8622211895910781, "grad_norm": 0.783810682425592, "learning_rate": 4.967894523156935e-07, "loss": 0.0736, "step": 18555 }, { "epoch": 0.8622676579925651, "grad_norm": 0.45195457076251216, "learning_rate": 4.964624899156018e-07, "loss": 0.0452, "step": 18556 }, { "epoch": 0.862314126394052, "grad_norm": 0.5573252861429357, "learning_rate": 4.961356295253738e-07, "loss": 0.0367, "step": 18557 }, { "epoch": 0.8623605947955391, "grad_norm": 0.6173273132526612, "learning_rate": 4.958088711524144e-07, "loss": 0.0525, "step": 18558 }, { "epoch": 0.862407063197026, "grad_norm": 0.6433735429942725, "learning_rate": 4.954822148041255e-07, "loss": 0.0757, "step": 18559 }, { "epoch": 0.862453531598513, "grad_norm": 0.4516572913822309, "learning_rate": 4.951556604879049e-07, "loss": 0.0438, "step": 18560 }, { "epoch": 0.8625, "grad_norm": 0.5392137415914567, "learning_rate": 4.948292082111494e-07, "loss": 0.0453, "step": 18561 }, { "epoch": 0.862546468401487, "grad_norm": 0.5388158225896831, "learning_rate": 4.945028579812544e-07, "loss": 0.0475, "step": 18562 }, { "epoch": 0.862592936802974, "grad_norm": 0.3340053237635658, "learning_rate": 4.941766098056117e-07, "loss": 0.0302, "step": 18563 }, { "epoch": 0.8626394052044609, "grad_norm": 0.6245886480012706, "learning_rate": 4.938504636916103e-07, "loss": 0.0679, "step": 18564 }, { "epoch": 0.862685873605948, "grad_norm": 0.4373778828870786, "learning_rate": 4.935244196466388e-07, "loss": 0.0539, "step": 18565 }, { "epoch": 0.862732342007435, "grad_norm": 0.4126743586785324, "learning_rate": 4.931984776780835e-07, "loss": 0.0517, "step": 18566 }, { "epoch": 0.8627788104089219, "grad_norm": 0.43217354081100917, "learning_rate": 4.928726377933246e-07, "loss": 0.0375, "step": 18567 }, { "epoch": 0.8628252788104089, "grad_norm": 0.5374765120671148, "learning_rate": 4.925468999997445e-07, "loss": 0.0552, "step": 18568 }, { "epoch": 0.8628717472118959, "grad_norm": 0.6031295823043922, "learning_rate": 4.922212643047213e-07, "loss": 0.0568, "step": 18569 }, { "epoch": 0.8629182156133829, "grad_norm": 0.48976243856674356, "learning_rate": 4.918957307156308e-07, "loss": 0.0661, "step": 18570 }, { "epoch": 0.8629646840148699, "grad_norm": 0.5427918830048473, "learning_rate": 4.915702992398452e-07, "loss": 0.0645, "step": 18571 }, { "epoch": 0.8630111524163568, "grad_norm": 0.4226408752777563, "learning_rate": 4.912449698847399e-07, "loss": 0.0358, "step": 18572 }, { "epoch": 0.8630576208178439, "grad_norm": 0.450027634510286, "learning_rate": 4.909197426576806e-07, "loss": 0.066, "step": 18573 }, { "epoch": 0.8631040892193309, "grad_norm": 0.6334141430810957, "learning_rate": 4.905946175660337e-07, "loss": 0.0471, "step": 18574 }, { "epoch": 0.8631505576208178, "grad_norm": 0.5597251989701584, "learning_rate": 4.902695946171654e-07, "loss": 0.0576, "step": 18575 }, { "epoch": 0.8631970260223049, "grad_norm": 0.3695180001944493, "learning_rate": 4.899446738184371e-07, "loss": 0.0589, "step": 18576 }, { "epoch": 0.8632434944237918, "grad_norm": 0.8210828396376797, "learning_rate": 4.896198551772091e-07, "loss": 0.0591, "step": 18577 }, { "epoch": 0.8632899628252788, "grad_norm": 0.848363513956569, "learning_rate": 4.892951387008387e-07, "loss": 0.0657, "step": 18578 }, { "epoch": 0.8633364312267658, "grad_norm": 0.34449270192270826, "learning_rate": 4.889705243966803e-07, "loss": 0.0485, "step": 18579 }, { "epoch": 0.8633828996282528, "grad_norm": 0.47957486878711986, "learning_rate": 4.886460122720876e-07, "loss": 0.0493, "step": 18580 }, { "epoch": 0.8634293680297398, "grad_norm": 0.532731434816082, "learning_rate": 4.8832160233441e-07, "loss": 0.0532, "step": 18581 }, { "epoch": 0.8634758364312267, "grad_norm": 0.5532124629863604, "learning_rate": 4.879972945909972e-07, "loss": 0.0466, "step": 18582 }, { "epoch": 0.8635223048327137, "grad_norm": 0.2638670333726808, "learning_rate": 4.876730890491943e-07, "loss": 0.03, "step": 18583 }, { "epoch": 0.8635687732342008, "grad_norm": 0.3915927618040603, "learning_rate": 4.873489857163455e-07, "loss": 0.0485, "step": 18584 }, { "epoch": 0.8636152416356877, "grad_norm": 0.4717917372770531, "learning_rate": 4.870249845997915e-07, "loss": 0.0567, "step": 18585 }, { "epoch": 0.8636617100371747, "grad_norm": 0.397026349382963, "learning_rate": 4.867010857068716e-07, "loss": 0.0504, "step": 18586 }, { "epoch": 0.8637081784386617, "grad_norm": 0.40266960280305764, "learning_rate": 4.86377289044922e-07, "loss": 0.0499, "step": 18587 }, { "epoch": 0.8637546468401487, "grad_norm": 0.30482871042230086, "learning_rate": 4.860535946212763e-07, "loss": 0.0494, "step": 18588 }, { "epoch": 0.8638011152416357, "grad_norm": 0.5303306095078655, "learning_rate": 4.857300024432687e-07, "loss": 0.0672, "step": 18589 }, { "epoch": 0.8638475836431226, "grad_norm": 0.41610649851015935, "learning_rate": 4.854065125182289e-07, "loss": 0.0412, "step": 18590 }, { "epoch": 0.8638940520446097, "grad_norm": 0.5138240142026891, "learning_rate": 4.850831248534805e-07, "loss": 0.0601, "step": 18591 }, { "epoch": 0.8639405204460967, "grad_norm": 0.609033620996181, "learning_rate": 4.84759839456353e-07, "loss": 0.0464, "step": 18592 }, { "epoch": 0.8639869888475836, "grad_norm": 0.39011038350613597, "learning_rate": 4.844366563341668e-07, "loss": 0.0413, "step": 18593 }, { "epoch": 0.8640334572490707, "grad_norm": 0.5998899032976012, "learning_rate": 4.841135754942427e-07, "loss": 0.0337, "step": 18594 }, { "epoch": 0.8640799256505576, "grad_norm": 0.4851979853621474, "learning_rate": 4.837905969438994e-07, "loss": 0.0566, "step": 18595 }, { "epoch": 0.8641263940520446, "grad_norm": 0.4152385183199337, "learning_rate": 4.83467720690452e-07, "loss": 0.0523, "step": 18596 }, { "epoch": 0.8641728624535316, "grad_norm": 0.3654666872993721, "learning_rate": 4.831449467412147e-07, "loss": 0.0513, "step": 18597 }, { "epoch": 0.8642193308550186, "grad_norm": 0.7291888954738063, "learning_rate": 4.828222751034972e-07, "loss": 0.0517, "step": 18598 }, { "epoch": 0.8642657992565056, "grad_norm": 0.667321234380963, "learning_rate": 4.824997057846098e-07, "loss": 0.0439, "step": 18599 }, { "epoch": 0.8643122676579925, "grad_norm": 0.5865788021995021, "learning_rate": 4.821772387918594e-07, "loss": 0.055, "step": 18600 }, { "epoch": 0.8643587360594795, "grad_norm": 0.735682545865396, "learning_rate": 4.818548741325491e-07, "loss": 0.058, "step": 18601 }, { "epoch": 0.8644052044609666, "grad_norm": 0.4305529504951534, "learning_rate": 4.815326118139813e-07, "loss": 0.0483, "step": 18602 }, { "epoch": 0.8644516728624535, "grad_norm": 0.41808406984084917, "learning_rate": 4.812104518434551e-07, "loss": 0.0331, "step": 18603 }, { "epoch": 0.8644981412639405, "grad_norm": 0.6163219135809789, "learning_rate": 4.808883942282682e-07, "loss": 0.0636, "step": 18604 }, { "epoch": 0.8645446096654275, "grad_norm": 0.6269451771212518, "learning_rate": 4.805664389757147e-07, "loss": 0.0707, "step": 18605 }, { "epoch": 0.8645910780669145, "grad_norm": 0.8894190873985244, "learning_rate": 4.802445860930894e-07, "loss": 0.0714, "step": 18606 }, { "epoch": 0.8646375464684015, "grad_norm": 0.4176097708998172, "learning_rate": 4.799228355876812e-07, "loss": 0.0369, "step": 18607 }, { "epoch": 0.8646840148698884, "grad_norm": 0.5329247587301428, "learning_rate": 4.796011874667767e-07, "loss": 0.0436, "step": 18608 }, { "epoch": 0.8647304832713755, "grad_norm": 0.5356515596550729, "learning_rate": 4.792796417376638e-07, "loss": 0.0638, "step": 18609 }, { "epoch": 0.8647769516728625, "grad_norm": 0.2913496055772136, "learning_rate": 4.789581984076247e-07, "loss": 0.0435, "step": 18610 }, { "epoch": 0.8648234200743494, "grad_norm": 0.593211404477825, "learning_rate": 4.786368574839412e-07, "loss": 0.0443, "step": 18611 }, { "epoch": 0.8648698884758365, "grad_norm": 0.7539381507779258, "learning_rate": 4.783156189738903e-07, "loss": 0.0489, "step": 18612 }, { "epoch": 0.8649163568773234, "grad_norm": 0.685226158338197, "learning_rate": 4.779944828847515e-07, "loss": 0.0511, "step": 18613 }, { "epoch": 0.8649628252788104, "grad_norm": 0.6514636448948539, "learning_rate": 4.776734492237961e-07, "loss": 0.0617, "step": 18614 }, { "epoch": 0.8650092936802974, "grad_norm": 0.5328039629538943, "learning_rate": 4.773525179982958e-07, "loss": 0.0644, "step": 18615 }, { "epoch": 0.8650557620817844, "grad_norm": 0.5065672836257149, "learning_rate": 4.770316892155219e-07, "loss": 0.0632, "step": 18616 }, { "epoch": 0.8651022304832714, "grad_norm": 0.47121160542609397, "learning_rate": 4.7671096288274045e-07, "loss": 0.0449, "step": 18617 }, { "epoch": 0.8651486988847583, "grad_norm": 0.50169410103378, "learning_rate": 4.7639033900721667e-07, "loss": 0.0569, "step": 18618 }, { "epoch": 0.8651951672862453, "grad_norm": 0.5108257567341723, "learning_rate": 4.760698175962125e-07, "loss": 0.0412, "step": 18619 }, { "epoch": 0.8652416356877324, "grad_norm": 0.5955975432497351, "learning_rate": 4.7574939865698834e-07, "loss": 0.0514, "step": 18620 }, { "epoch": 0.8652881040892193, "grad_norm": 0.38539991180811817, "learning_rate": 4.754290821968022e-07, "loss": 0.052, "step": 18621 }, { "epoch": 0.8653345724907063, "grad_norm": 0.4483565039811142, "learning_rate": 4.7510886822290825e-07, "loss": 0.0486, "step": 18622 }, { "epoch": 0.8653810408921933, "grad_norm": 0.628791134768595, "learning_rate": 4.747887567425619e-07, "loss": 0.0628, "step": 18623 }, { "epoch": 0.8654275092936803, "grad_norm": 0.5638636860122728, "learning_rate": 4.744687477630122e-07, "loss": 0.0555, "step": 18624 }, { "epoch": 0.8654739776951673, "grad_norm": 0.39724359951419197, "learning_rate": 4.7414884129150906e-07, "loss": 0.0476, "step": 18625 }, { "epoch": 0.8655204460966542, "grad_norm": 0.4558316573285011, "learning_rate": 4.7382903733529826e-07, "loss": 0.0367, "step": 18626 }, { "epoch": 0.8655669144981413, "grad_norm": 0.6715279160015746, "learning_rate": 4.73509335901623e-07, "loss": 0.0424, "step": 18627 }, { "epoch": 0.8656133828996283, "grad_norm": 0.374276388982888, "learning_rate": 4.7318973699772563e-07, "loss": 0.0345, "step": 18628 }, { "epoch": 0.8656598513011152, "grad_norm": 0.5945955042706876, "learning_rate": 4.728702406308439e-07, "loss": 0.0667, "step": 18629 }, { "epoch": 0.8657063197026023, "grad_norm": 0.3541152375306444, "learning_rate": 4.7255084680821685e-07, "loss": 0.0454, "step": 18630 }, { "epoch": 0.8657527881040892, "grad_norm": 0.4443662088104014, "learning_rate": 4.722315555370793e-07, "loss": 0.044, "step": 18631 }, { "epoch": 0.8657992565055762, "grad_norm": 0.46531454460334803, "learning_rate": 4.719123668246606e-07, "loss": 0.0568, "step": 18632 }, { "epoch": 0.8658457249070632, "grad_norm": 0.4882970489358743, "learning_rate": 4.715932806781931e-07, "loss": 0.0543, "step": 18633 }, { "epoch": 0.8658921933085502, "grad_norm": 0.4380824310095578, "learning_rate": 4.712742971049039e-07, "loss": 0.0525, "step": 18634 }, { "epoch": 0.8659386617100372, "grad_norm": 0.32406125212226705, "learning_rate": 4.7095541611201833e-07, "loss": 0.0454, "step": 18635 }, { "epoch": 0.8659851301115241, "grad_norm": 0.30061336773412345, "learning_rate": 4.7063663770675896e-07, "loss": 0.0325, "step": 18636 }, { "epoch": 0.8660315985130111, "grad_norm": 0.5630777683149866, "learning_rate": 4.703179618963477e-07, "loss": 0.0534, "step": 18637 }, { "epoch": 0.8660780669144982, "grad_norm": 0.5308277880641499, "learning_rate": 4.6999938868800167e-07, "loss": 0.0582, "step": 18638 }, { "epoch": 0.8661245353159851, "grad_norm": 0.9351432734008714, "learning_rate": 4.6968091808893613e-07, "loss": 0.0582, "step": 18639 }, { "epoch": 0.8661710037174721, "grad_norm": 0.4831793306413495, "learning_rate": 4.69362550106367e-07, "loss": 0.0523, "step": 18640 }, { "epoch": 0.866217472118959, "grad_norm": 0.3943226492579048, "learning_rate": 4.690442847475046e-07, "loss": 0.0413, "step": 18641 }, { "epoch": 0.8662639405204461, "grad_norm": 0.40221915893211846, "learning_rate": 4.687261220195577e-07, "loss": 0.0384, "step": 18642 }, { "epoch": 0.8663104089219331, "grad_norm": 0.757127446640625, "learning_rate": 4.684080619297332e-07, "loss": 0.0512, "step": 18643 }, { "epoch": 0.86635687732342, "grad_norm": 0.4305797975301548, "learning_rate": 4.680901044852354e-07, "loss": 0.0281, "step": 18644 }, { "epoch": 0.8664033457249071, "grad_norm": 0.35037811576136846, "learning_rate": 4.6777224969326684e-07, "loss": 0.04, "step": 18645 }, { "epoch": 0.8664498141263941, "grad_norm": 0.5071497046870355, "learning_rate": 4.674544975610257e-07, "loss": 0.0365, "step": 18646 }, { "epoch": 0.866496282527881, "grad_norm": 0.4865109974314533, "learning_rate": 4.6713684809571116e-07, "loss": 0.0438, "step": 18647 }, { "epoch": 0.866542750929368, "grad_norm": 0.44464344544325635, "learning_rate": 4.6681930130451856e-07, "loss": 0.0412, "step": 18648 }, { "epoch": 0.866589219330855, "grad_norm": 0.6632076973188136, "learning_rate": 4.6650185719463945e-07, "loss": 0.051, "step": 18649 }, { "epoch": 0.866635687732342, "grad_norm": 0.3840956032612466, "learning_rate": 4.661845157732647e-07, "loss": 0.056, "step": 18650 }, { "epoch": 0.866682156133829, "grad_norm": 0.3701686724741429, "learning_rate": 4.658672770475825e-07, "loss": 0.0402, "step": 18651 }, { "epoch": 0.866728624535316, "grad_norm": 0.22595864406814795, "learning_rate": 4.655501410247781e-07, "loss": 0.0286, "step": 18652 }, { "epoch": 0.866775092936803, "grad_norm": 0.49454038053601374, "learning_rate": 4.6523310771203536e-07, "loss": 0.0393, "step": 18653 }, { "epoch": 0.8668215613382899, "grad_norm": 0.4928103092336456, "learning_rate": 4.649161771165367e-07, "loss": 0.035, "step": 18654 }, { "epoch": 0.866868029739777, "grad_norm": 0.4988518209204099, "learning_rate": 4.6459934924545923e-07, "loss": 0.0542, "step": 18655 }, { "epoch": 0.866914498141264, "grad_norm": 0.8096496632604476, "learning_rate": 4.6428262410597834e-07, "loss": 0.0644, "step": 18656 }, { "epoch": 0.8669609665427509, "grad_norm": 0.34135238831033476, "learning_rate": 4.6396600170527107e-07, "loss": 0.0444, "step": 18657 }, { "epoch": 0.8670074349442379, "grad_norm": 0.5229257210671417, "learning_rate": 4.636494820505083e-07, "loss": 0.0518, "step": 18658 }, { "epoch": 0.8670539033457249, "grad_norm": 0.6596996340622484, "learning_rate": 4.633330651488588e-07, "loss": 0.0522, "step": 18659 }, { "epoch": 0.8671003717472119, "grad_norm": 0.31040114962405696, "learning_rate": 4.630167510074901e-07, "loss": 0.0196, "step": 18660 }, { "epoch": 0.8671468401486989, "grad_norm": 0.635216329637004, "learning_rate": 4.6270053963356707e-07, "loss": 0.0591, "step": 18661 }, { "epoch": 0.8671933085501858, "grad_norm": 0.3896748347633686, "learning_rate": 4.6238443103425234e-07, "loss": 0.0364, "step": 18662 }, { "epoch": 0.8672397769516729, "grad_norm": 0.33934718856242774, "learning_rate": 4.6206842521670513e-07, "loss": 0.0613, "step": 18663 }, { "epoch": 0.8672862453531599, "grad_norm": 0.8382090224885734, "learning_rate": 4.617525221880853e-07, "loss": 0.0673, "step": 18664 }, { "epoch": 0.8673327137546468, "grad_norm": 0.4437253835968119, "learning_rate": 4.6143672195554713e-07, "loss": 0.069, "step": 18665 }, { "epoch": 0.8673791821561339, "grad_norm": 0.3778016945862327, "learning_rate": 4.6112102452624373e-07, "loss": 0.0417, "step": 18666 }, { "epoch": 0.8674256505576208, "grad_norm": 0.6863447548397251, "learning_rate": 4.608054299073261e-07, "loss": 0.0532, "step": 18667 }, { "epoch": 0.8674721189591078, "grad_norm": 0.42609086340022867, "learning_rate": 4.60489938105943e-07, "loss": 0.0619, "step": 18668 }, { "epoch": 0.8675185873605948, "grad_norm": 0.38203645347839643, "learning_rate": 4.601745491292409e-07, "loss": 0.052, "step": 18669 }, { "epoch": 0.8675650557620818, "grad_norm": 0.6579810740620458, "learning_rate": 4.5985926298436236e-07, "loss": 0.0609, "step": 18670 }, { "epoch": 0.8676115241635688, "grad_norm": 0.4409495206869086, "learning_rate": 4.595440796784506e-07, "loss": 0.0445, "step": 18671 }, { "epoch": 0.8676579925650557, "grad_norm": 0.4340327787299569, "learning_rate": 4.592289992186449e-07, "loss": 0.0533, "step": 18672 }, { "epoch": 0.8677044609665427, "grad_norm": 0.8507367924437238, "learning_rate": 4.589140216120802e-07, "loss": 0.0746, "step": 18673 }, { "epoch": 0.8677509293680298, "grad_norm": 0.712468471504022, "learning_rate": 4.585991468658929e-07, "loss": 0.0659, "step": 18674 }, { "epoch": 0.8677973977695167, "grad_norm": 0.5507059316859064, "learning_rate": 4.5828437498721455e-07, "loss": 0.0399, "step": 18675 }, { "epoch": 0.8678438661710037, "grad_norm": 0.4843205670451095, "learning_rate": 4.57969705983175e-07, "loss": 0.0621, "step": 18676 }, { "epoch": 0.8678903345724907, "grad_norm": 0.36332210384456565, "learning_rate": 4.576551398609019e-07, "loss": 0.0319, "step": 18677 }, { "epoch": 0.8679368029739777, "grad_norm": 0.5752078994401028, "learning_rate": 4.5734067662752177e-07, "loss": 0.0472, "step": 18678 }, { "epoch": 0.8679832713754647, "grad_norm": 0.46026390881433854, "learning_rate": 4.5702631629015557e-07, "loss": 0.0406, "step": 18679 }, { "epoch": 0.8680297397769516, "grad_norm": 0.5994354005924518, "learning_rate": 4.5671205885592374e-07, "loss": 0.0354, "step": 18680 }, { "epoch": 0.8680762081784387, "grad_norm": 0.7126538337092118, "learning_rate": 4.5639790433194666e-07, "loss": 0.051, "step": 18681 }, { "epoch": 0.8681226765799257, "grad_norm": 0.5705844752324674, "learning_rate": 4.560838527253392e-07, "loss": 0.0602, "step": 18682 }, { "epoch": 0.8681691449814126, "grad_norm": 0.6855695836547454, "learning_rate": 4.5576990404321343e-07, "loss": 0.0726, "step": 18683 }, { "epoch": 0.8682156133828997, "grad_norm": 0.5066901691812221, "learning_rate": 4.5545605829268427e-07, "loss": 0.0496, "step": 18684 }, { "epoch": 0.8682620817843866, "grad_norm": 0.6785997263447111, "learning_rate": 4.551423154808571e-07, "loss": 0.05, "step": 18685 }, { "epoch": 0.8683085501858736, "grad_norm": 0.6134176180979277, "learning_rate": 4.548286756148401e-07, "loss": 0.0532, "step": 18686 }, { "epoch": 0.8683550185873606, "grad_norm": 0.6092761133186014, "learning_rate": 4.545151387017366e-07, "loss": 0.0468, "step": 18687 }, { "epoch": 0.8684014869888476, "grad_norm": 0.6596857908736754, "learning_rate": 4.542017047486502e-07, "loss": 0.053, "step": 18688 }, { "epoch": 0.8684479553903346, "grad_norm": 0.3993750811784712, "learning_rate": 4.5388837376267924e-07, "loss": 0.0606, "step": 18689 }, { "epoch": 0.8684944237918215, "grad_norm": 0.754191010579974, "learning_rate": 4.535751457509213e-07, "loss": 0.0535, "step": 18690 }, { "epoch": 0.8685408921933085, "grad_norm": 0.36991851611782756, "learning_rate": 4.5326202072047187e-07, "loss": 0.0513, "step": 18691 }, { "epoch": 0.8685873605947956, "grad_norm": 0.9556019607959619, "learning_rate": 4.5294899867842247e-07, "loss": 0.0633, "step": 18692 }, { "epoch": 0.8686338289962825, "grad_norm": 0.45004072082787017, "learning_rate": 4.5263607963186294e-07, "loss": 0.0439, "step": 18693 }, { "epoch": 0.8686802973977695, "grad_norm": 0.5838473574935178, "learning_rate": 4.523232635878838e-07, "loss": 0.0723, "step": 18694 }, { "epoch": 0.8687267657992565, "grad_norm": 0.49609301378490284, "learning_rate": 4.5201055055356824e-07, "loss": 0.0483, "step": 18695 }, { "epoch": 0.8687732342007435, "grad_norm": 0.6041049909748726, "learning_rate": 4.516979405360017e-07, "loss": 0.0594, "step": 18696 }, { "epoch": 0.8688197026022305, "grad_norm": 0.5673878941455454, "learning_rate": 4.5138543354226236e-07, "loss": 0.0514, "step": 18697 }, { "epoch": 0.8688661710037174, "grad_norm": 0.5733818725117895, "learning_rate": 4.510730295794308e-07, "loss": 0.0681, "step": 18698 }, { "epoch": 0.8689126394052045, "grad_norm": 0.5521964829995233, "learning_rate": 4.5076072865458287e-07, "loss": 0.0567, "step": 18699 }, { "epoch": 0.8689591078066915, "grad_norm": 0.4436286527588279, "learning_rate": 4.5044853077479134e-07, "loss": 0.0507, "step": 18700 }, { "epoch": 0.8690055762081784, "grad_norm": 0.6105967596938756, "learning_rate": 4.5013643594712997e-07, "loss": 0.0428, "step": 18701 }, { "epoch": 0.8690520446096655, "grad_norm": 0.5637577173386907, "learning_rate": 4.4982444417866753e-07, "loss": 0.0597, "step": 18702 }, { "epoch": 0.8690985130111524, "grad_norm": 0.3795064787806463, "learning_rate": 4.495125554764701e-07, "loss": 0.0358, "step": 18703 }, { "epoch": 0.8691449814126394, "grad_norm": 0.544567459874201, "learning_rate": 4.492007698476014e-07, "loss": 0.0662, "step": 18704 }, { "epoch": 0.8691914498141264, "grad_norm": 0.9288830308587774, "learning_rate": 4.488890872991253e-07, "loss": 0.0376, "step": 18705 }, { "epoch": 0.8692379182156134, "grad_norm": 0.5300694443443456, "learning_rate": 4.4857750783810217e-07, "loss": 0.0412, "step": 18706 }, { "epoch": 0.8692843866171004, "grad_norm": 0.3826478979324049, "learning_rate": 4.4826603147158753e-07, "loss": 0.0428, "step": 18707 }, { "epoch": 0.8693308550185873, "grad_norm": 0.541344509933312, "learning_rate": 4.479546582066396e-07, "loss": 0.0628, "step": 18708 }, { "epoch": 0.8693773234200743, "grad_norm": 0.4901550810968706, "learning_rate": 4.4764338805030895e-07, "loss": 0.0558, "step": 18709 }, { "epoch": 0.8694237918215614, "grad_norm": 0.5427110722053126, "learning_rate": 4.4733222100964536e-07, "loss": 0.0496, "step": 18710 }, { "epoch": 0.8694702602230483, "grad_norm": 0.4576625580801865, "learning_rate": 4.470211570916999e-07, "loss": 0.0419, "step": 18711 }, { "epoch": 0.8695167286245353, "grad_norm": 0.5724799602244548, "learning_rate": 4.46710196303517e-07, "loss": 0.0669, "step": 18712 }, { "epoch": 0.8695631970260222, "grad_norm": 0.4767928045376731, "learning_rate": 4.4639933865214147e-07, "loss": 0.0658, "step": 18713 }, { "epoch": 0.8696096654275093, "grad_norm": 0.44067265206266115, "learning_rate": 4.4608858414461165e-07, "loss": 0.0599, "step": 18714 }, { "epoch": 0.8696561338289963, "grad_norm": 0.6930216078173532, "learning_rate": 4.457779327879691e-07, "loss": 0.0582, "step": 18715 }, { "epoch": 0.8697026022304832, "grad_norm": 0.4292037050931916, "learning_rate": 4.4546738458924987e-07, "loss": 0.0378, "step": 18716 }, { "epoch": 0.8697490706319703, "grad_norm": 0.5042434042485073, "learning_rate": 4.4515693955548665e-07, "loss": 0.0449, "step": 18717 }, { "epoch": 0.8697955390334573, "grad_norm": 0.5020267361010156, "learning_rate": 4.4484659769371385e-07, "loss": 0.0519, "step": 18718 }, { "epoch": 0.8698420074349442, "grad_norm": 0.5735535192433565, "learning_rate": 4.4453635901096083e-07, "loss": 0.0524, "step": 18719 }, { "epoch": 0.8698884758364313, "grad_norm": 0.5364150268639044, "learning_rate": 4.442262235142519e-07, "loss": 0.0487, "step": 18720 }, { "epoch": 0.8699349442379182, "grad_norm": 0.5480590133741937, "learning_rate": 4.4391619121061434e-07, "loss": 0.0547, "step": 18721 }, { "epoch": 0.8699814126394052, "grad_norm": 0.30510803000658715, "learning_rate": 4.436062621070708e-07, "loss": 0.0334, "step": 18722 }, { "epoch": 0.8700278810408922, "grad_norm": 0.7043095055498406, "learning_rate": 4.432964362106407e-07, "loss": 0.0415, "step": 18723 }, { "epoch": 0.8700743494423792, "grad_norm": 0.31529683026628785, "learning_rate": 4.429867135283411e-07, "loss": 0.0327, "step": 18724 }, { "epoch": 0.8701208178438662, "grad_norm": 0.6100885570749842, "learning_rate": 4.426770940671909e-07, "loss": 0.0547, "step": 18725 }, { "epoch": 0.8701672862453531, "grad_norm": 0.6040530477162808, "learning_rate": 4.423675778341996e-07, "loss": 0.06, "step": 18726 }, { "epoch": 0.8702137546468401, "grad_norm": 0.5336937968801043, "learning_rate": 4.4205816483637864e-07, "loss": 0.0365, "step": 18727 }, { "epoch": 0.8702602230483272, "grad_norm": 0.3551558133485628, "learning_rate": 4.417488550807386e-07, "loss": 0.0532, "step": 18728 }, { "epoch": 0.8703066914498141, "grad_norm": 0.3088812317090701, "learning_rate": 4.4143964857428443e-07, "loss": 0.0475, "step": 18729 }, { "epoch": 0.8703531598513011, "grad_norm": 0.5218375020009849, "learning_rate": 4.4113054532402e-07, "loss": 0.0295, "step": 18730 }, { "epoch": 0.870399628252788, "grad_norm": 0.5040073321709387, "learning_rate": 4.408215453369463e-07, "loss": 0.0399, "step": 18731 }, { "epoch": 0.8704460966542751, "grad_norm": 0.5018440236494764, "learning_rate": 4.4051264862006395e-07, "loss": 0.0517, "step": 18732 }, { "epoch": 0.8704925650557621, "grad_norm": 0.503751421015387, "learning_rate": 4.4020385518036835e-07, "loss": 0.0452, "step": 18733 }, { "epoch": 0.870539033457249, "grad_norm": 0.6343578743443115, "learning_rate": 4.3989516502485343e-07, "loss": 0.0557, "step": 18734 }, { "epoch": 0.8705855018587361, "grad_norm": 0.44388025643223167, "learning_rate": 4.3958657816051355e-07, "loss": 0.049, "step": 18735 }, { "epoch": 0.8706319702602231, "grad_norm": 0.410551599991792, "learning_rate": 4.392780945943376e-07, "loss": 0.0498, "step": 18736 }, { "epoch": 0.87067843866171, "grad_norm": 0.4222596586246537, "learning_rate": 4.389697143333127e-07, "loss": 0.0546, "step": 18737 }, { "epoch": 0.870724907063197, "grad_norm": 0.43038888398069414, "learning_rate": 4.386614373844245e-07, "loss": 0.0341, "step": 18738 }, { "epoch": 0.870771375464684, "grad_norm": 0.39965689023728584, "learning_rate": 4.3835326375465504e-07, "loss": 0.0277, "step": 18739 }, { "epoch": 0.870817843866171, "grad_norm": 0.2675444639989822, "learning_rate": 4.380451934509855e-07, "loss": 0.0286, "step": 18740 }, { "epoch": 0.870864312267658, "grad_norm": 0.46488182582089504, "learning_rate": 4.377372264803925e-07, "loss": 0.0615, "step": 18741 }, { "epoch": 0.870910780669145, "grad_norm": 0.36825984225794894, "learning_rate": 4.374293628498538e-07, "loss": 0.0362, "step": 18742 }, { "epoch": 0.870957249070632, "grad_norm": 0.7238219140495358, "learning_rate": 4.371216025663433e-07, "loss": 0.0492, "step": 18743 }, { "epoch": 0.8710037174721189, "grad_norm": 0.4191711758692922, "learning_rate": 4.368139456368292e-07, "loss": 0.05, "step": 18744 }, { "epoch": 0.8710501858736059, "grad_norm": 0.397634160102178, "learning_rate": 4.3650639206828274e-07, "loss": 0.0539, "step": 18745 }, { "epoch": 0.871096654275093, "grad_norm": 0.6174794063976266, "learning_rate": 4.3619894186766876e-07, "loss": 0.0536, "step": 18746 }, { "epoch": 0.8711431226765799, "grad_norm": 0.44850325094962396, "learning_rate": 4.358915950419529e-07, "loss": 0.0472, "step": 18747 }, { "epoch": 0.8711895910780669, "grad_norm": 0.6044350471864677, "learning_rate": 4.355843515980951e-07, "loss": 0.0592, "step": 18748 }, { "epoch": 0.8712360594795538, "grad_norm": 0.41347195686661997, "learning_rate": 4.3527721154305703e-07, "loss": 0.051, "step": 18749 }, { "epoch": 0.8712825278810409, "grad_norm": 0.44285028289804595, "learning_rate": 4.349701748837937e-07, "loss": 0.0531, "step": 18750 }, { "epoch": 0.8713289962825279, "grad_norm": 0.5088173816887486, "learning_rate": 4.3466324162725946e-07, "loss": 0.0245, "step": 18751 }, { "epoch": 0.8713754646840148, "grad_norm": 0.4677522966769511, "learning_rate": 4.3435641178040887e-07, "loss": 0.0479, "step": 18752 }, { "epoch": 0.8714219330855019, "grad_norm": 0.4490766462827528, "learning_rate": 4.3404968535019076e-07, "loss": 0.0461, "step": 18753 }, { "epoch": 0.8714684014869889, "grad_norm": 0.4670371061815196, "learning_rate": 4.337430623435529e-07, "loss": 0.0426, "step": 18754 }, { "epoch": 0.8715148698884758, "grad_norm": 0.399100258819856, "learning_rate": 4.334365427674403e-07, "loss": 0.0435, "step": 18755 }, { "epoch": 0.8715613382899629, "grad_norm": 0.5575635627927074, "learning_rate": 4.331301266287968e-07, "loss": 0.0467, "step": 18756 }, { "epoch": 0.8716078066914498, "grad_norm": 0.6936982562885754, "learning_rate": 4.3282381393456194e-07, "loss": 0.0611, "step": 18757 }, { "epoch": 0.8716542750929368, "grad_norm": 0.4982594681280562, "learning_rate": 4.3251760469167403e-07, "loss": 0.0453, "step": 18758 }, { "epoch": 0.8717007434944238, "grad_norm": 0.5262639870723852, "learning_rate": 4.322114989070703e-07, "loss": 0.0648, "step": 18759 }, { "epoch": 0.8717472118959108, "grad_norm": 0.5690645033905579, "learning_rate": 4.3190549658768345e-07, "loss": 0.0511, "step": 18760 }, { "epoch": 0.8717936802973978, "grad_norm": 0.5083007996438257, "learning_rate": 4.3159959774044523e-07, "loss": 0.0397, "step": 18761 }, { "epoch": 0.8718401486988847, "grad_norm": 0.6147056543584698, "learning_rate": 4.31293802372284e-07, "loss": 0.0568, "step": 18762 }, { "epoch": 0.8718866171003717, "grad_norm": 0.5010583734135341, "learning_rate": 4.3098811049012645e-07, "loss": 0.0366, "step": 18763 }, { "epoch": 0.8719330855018588, "grad_norm": 0.40825995113872826, "learning_rate": 4.3068252210089754e-07, "loss": 0.0442, "step": 18764 }, { "epoch": 0.8719795539033457, "grad_norm": 0.6003645240650722, "learning_rate": 4.303770372115174e-07, "loss": 0.0483, "step": 18765 }, { "epoch": 0.8720260223048327, "grad_norm": 0.6112041099610356, "learning_rate": 4.300716558289081e-07, "loss": 0.0589, "step": 18766 }, { "epoch": 0.8720724907063196, "grad_norm": 0.4804133073776325, "learning_rate": 4.297663779599859e-07, "loss": 0.04, "step": 18767 }, { "epoch": 0.8721189591078067, "grad_norm": 0.7069068344896611, "learning_rate": 4.2946120361166357e-07, "loss": 0.0661, "step": 18768 }, { "epoch": 0.8721654275092937, "grad_norm": 0.5842803067101168, "learning_rate": 4.291561327908567e-07, "loss": 0.0521, "step": 18769 }, { "epoch": 0.8722118959107806, "grad_norm": 0.43058429356909683, "learning_rate": 4.2885116550447416e-07, "loss": 0.056, "step": 18770 }, { "epoch": 0.8722583643122677, "grad_norm": 0.4771352811815407, "learning_rate": 4.2854630175942326e-07, "loss": 0.0719, "step": 18771 }, { "epoch": 0.8723048327137547, "grad_norm": 0.3799175947819994, "learning_rate": 4.282415415626101e-07, "loss": 0.0532, "step": 18772 }, { "epoch": 0.8723513011152416, "grad_norm": 0.5217441552356298, "learning_rate": 4.279368849209381e-07, "loss": 0.0476, "step": 18773 }, { "epoch": 0.8723977695167286, "grad_norm": 0.3752452037470034, "learning_rate": 4.276323318413073e-07, "loss": 0.0404, "step": 18774 }, { "epoch": 0.8724442379182156, "grad_norm": 0.5560893691594905, "learning_rate": 4.2732788233061596e-07, "loss": 0.0542, "step": 18775 }, { "epoch": 0.8724907063197026, "grad_norm": 0.6772438437270262, "learning_rate": 4.2702353639576146e-07, "loss": 0.063, "step": 18776 }, { "epoch": 0.8725371747211896, "grad_norm": 0.45835270330657674, "learning_rate": 4.267192940436365e-07, "loss": 0.0379, "step": 18777 }, { "epoch": 0.8725836431226766, "grad_norm": 0.6573862358433109, "learning_rate": 4.2641515528113355e-07, "loss": 0.0418, "step": 18778 }, { "epoch": 0.8726301115241636, "grad_norm": 0.5506658515362093, "learning_rate": 4.2611112011514024e-07, "loss": 0.053, "step": 18779 }, { "epoch": 0.8726765799256505, "grad_norm": 0.5158192943747687, "learning_rate": 4.2580718855254446e-07, "loss": 0.0565, "step": 18780 }, { "epoch": 0.8727230483271375, "grad_norm": 0.4244442709964161, "learning_rate": 4.2550336060022957e-07, "loss": 0.0551, "step": 18781 }, { "epoch": 0.8727695167286246, "grad_norm": 0.40993108132351197, "learning_rate": 4.251996362650779e-07, "loss": 0.0528, "step": 18782 }, { "epoch": 0.8728159851301115, "grad_norm": 0.3235066533307256, "learning_rate": 4.2489601555396997e-07, "loss": 0.024, "step": 18783 }, { "epoch": 0.8728624535315985, "grad_norm": 0.6336399659211674, "learning_rate": 4.2459249847378314e-07, "loss": 0.059, "step": 18784 }, { "epoch": 0.8729089219330854, "grad_norm": 0.2739933784342913, "learning_rate": 4.242890850313902e-07, "loss": 0.0363, "step": 18785 }, { "epoch": 0.8729553903345725, "grad_norm": 0.5227918452446317, "learning_rate": 4.2398577523366627e-07, "loss": 0.0699, "step": 18786 }, { "epoch": 0.8730018587360595, "grad_norm": 0.5338070460343145, "learning_rate": 4.236825690874807e-07, "loss": 0.0462, "step": 18787 }, { "epoch": 0.8730483271375464, "grad_norm": 0.4006188949898495, "learning_rate": 4.2337946659970154e-07, "loss": 0.0546, "step": 18788 }, { "epoch": 0.8730947955390335, "grad_norm": 0.6127580300327669, "learning_rate": 4.230764677771931e-07, "loss": 0.0591, "step": 18789 }, { "epoch": 0.8731412639405205, "grad_norm": 0.4884308263450386, "learning_rate": 4.2277357262682175e-07, "loss": 0.0602, "step": 18790 }, { "epoch": 0.8731877323420074, "grad_norm": 0.8953192875176491, "learning_rate": 4.2247078115544513e-07, "loss": 0.058, "step": 18791 }, { "epoch": 0.8732342007434944, "grad_norm": 0.4487710508559904, "learning_rate": 4.2216809336992294e-07, "loss": 0.0624, "step": 18792 }, { "epoch": 0.8732806691449814, "grad_norm": 0.604291263215783, "learning_rate": 4.2186550927711177e-07, "loss": 0.067, "step": 18793 }, { "epoch": 0.8733271375464684, "grad_norm": 0.40660846112436266, "learning_rate": 4.215630288838651e-07, "loss": 0.037, "step": 18794 }, { "epoch": 0.8733736059479554, "grad_norm": 0.33732664212110924, "learning_rate": 4.2126065219703525e-07, "loss": 0.0316, "step": 18795 }, { "epoch": 0.8734200743494424, "grad_norm": 0.5215928563499584, "learning_rate": 4.2095837922347003e-07, "loss": 0.0534, "step": 18796 }, { "epoch": 0.8734665427509294, "grad_norm": 0.42044893405411493, "learning_rate": 4.2065620997001734e-07, "loss": 0.0465, "step": 18797 }, { "epoch": 0.8735130111524163, "grad_norm": 0.43930360066487345, "learning_rate": 4.203541444435211e-07, "loss": 0.0767, "step": 18798 }, { "epoch": 0.8735594795539033, "grad_norm": 0.4321317768474464, "learning_rate": 4.2005218265082246e-07, "loss": 0.036, "step": 18799 }, { "epoch": 0.8736059479553904, "grad_norm": 0.3666297419512333, "learning_rate": 4.1975032459876277e-07, "loss": 0.0501, "step": 18800 }, { "epoch": 0.8736524163568773, "grad_norm": 0.9371434478563527, "learning_rate": 4.1944857029417916e-07, "loss": 0.0575, "step": 18801 }, { "epoch": 0.8736988847583643, "grad_norm": 0.43682859679487246, "learning_rate": 4.191469197439063e-07, "loss": 0.0675, "step": 18802 }, { "epoch": 0.8737453531598512, "grad_norm": 0.4113760367262045, "learning_rate": 4.18845372954777e-07, "loss": 0.0604, "step": 18803 }, { "epoch": 0.8737918215613383, "grad_norm": 0.4096089824193966, "learning_rate": 4.185439299336219e-07, "loss": 0.0435, "step": 18804 }, { "epoch": 0.8738382899628253, "grad_norm": 0.6011365538649827, "learning_rate": 4.1824259068726834e-07, "loss": 0.059, "step": 18805 }, { "epoch": 0.8738847583643122, "grad_norm": 0.4646948999330214, "learning_rate": 4.1794135522254143e-07, "loss": 0.046, "step": 18806 }, { "epoch": 0.8739312267657993, "grad_norm": 0.5491425144666002, "learning_rate": 4.176402235462662e-07, "loss": 0.0558, "step": 18807 }, { "epoch": 0.8739776951672863, "grad_norm": 0.469141861531149, "learning_rate": 4.173391956652639e-07, "loss": 0.0438, "step": 18808 }, { "epoch": 0.8740241635687732, "grad_norm": 0.4760359260813447, "learning_rate": 4.1703827158634967e-07, "loss": 0.0468, "step": 18809 }, { "epoch": 0.8740706319702602, "grad_norm": 0.330991848065686, "learning_rate": 4.167374513163436e-07, "loss": 0.0442, "step": 18810 }, { "epoch": 0.8741171003717472, "grad_norm": 0.4249858305708189, "learning_rate": 4.1643673486205794e-07, "loss": 0.042, "step": 18811 }, { "epoch": 0.8741635687732342, "grad_norm": 0.613758333964739, "learning_rate": 4.1613612223030395e-07, "loss": 0.0563, "step": 18812 }, { "epoch": 0.8742100371747212, "grad_norm": 0.7538883167267884, "learning_rate": 4.158356134278907e-07, "loss": 0.0562, "step": 18813 }, { "epoch": 0.8742565055762082, "grad_norm": 0.4933733581464293, "learning_rate": 4.155352084616271e-07, "loss": 0.0565, "step": 18814 }, { "epoch": 0.8743029739776952, "grad_norm": 0.5922215362088757, "learning_rate": 4.152349073383155e-07, "loss": 0.0523, "step": 18815 }, { "epoch": 0.8743494423791821, "grad_norm": 0.3504933979947044, "learning_rate": 4.1493471006475716e-07, "loss": 0.0386, "step": 18816 }, { "epoch": 0.8743959107806691, "grad_norm": 0.5275097962638144, "learning_rate": 4.1463461664775496e-07, "loss": 0.0472, "step": 18817 }, { "epoch": 0.8744423791821562, "grad_norm": 0.4073920260050703, "learning_rate": 4.14334627094104e-07, "loss": 0.0448, "step": 18818 }, { "epoch": 0.8744888475836431, "grad_norm": 0.38964834646811913, "learning_rate": 4.140347414106005e-07, "loss": 0.0349, "step": 18819 }, { "epoch": 0.8745353159851301, "grad_norm": 0.45455275388820554, "learning_rate": 4.1373495960403685e-07, "loss": 0.0428, "step": 18820 }, { "epoch": 0.874581784386617, "grad_norm": 0.6849223857679296, "learning_rate": 4.134352816812032e-07, "loss": 0.0498, "step": 18821 }, { "epoch": 0.8746282527881041, "grad_norm": 0.5841210676482718, "learning_rate": 4.131357076488873e-07, "loss": 0.0533, "step": 18822 }, { "epoch": 0.8746747211895911, "grad_norm": 0.4925756341657872, "learning_rate": 4.1283623751387445e-07, "loss": 0.0453, "step": 18823 }, { "epoch": 0.874721189591078, "grad_norm": 0.42835095011702295, "learning_rate": 4.125368712829497e-07, "loss": 0.0557, "step": 18824 }, { "epoch": 0.8747676579925651, "grad_norm": 0.48829135884809344, "learning_rate": 4.1223760896289323e-07, "loss": 0.061, "step": 18825 }, { "epoch": 0.8748141263940521, "grad_norm": 0.4408053528402445, "learning_rate": 4.119384505604834e-07, "loss": 0.0363, "step": 18826 }, { "epoch": 0.874860594795539, "grad_norm": 0.3999317764512875, "learning_rate": 4.116393960824966e-07, "loss": 0.0591, "step": 18827 }, { "epoch": 0.874907063197026, "grad_norm": 0.4880192499575592, "learning_rate": 4.113404455357062e-07, "loss": 0.0525, "step": 18828 }, { "epoch": 0.874953531598513, "grad_norm": 0.5672143245868803, "learning_rate": 4.110415989268846e-07, "loss": 0.056, "step": 18829 }, { "epoch": 0.875, "grad_norm": 0.5731431420629919, "learning_rate": 4.107428562627996e-07, "loss": 0.0585, "step": 18830 }, { "epoch": 0.875046468401487, "grad_norm": 0.5439345433122912, "learning_rate": 4.10444217550221e-07, "loss": 0.0612, "step": 18831 }, { "epoch": 0.875092936802974, "grad_norm": 0.39094990668204377, "learning_rate": 4.101456827959099e-07, "loss": 0.0609, "step": 18832 }, { "epoch": 0.875139405204461, "grad_norm": 0.5682128993240971, "learning_rate": 4.098472520066293e-07, "loss": 0.0597, "step": 18833 }, { "epoch": 0.8751858736059479, "grad_norm": 0.4252434680921012, "learning_rate": 4.095489251891405e-07, "loss": 0.0323, "step": 18834 }, { "epoch": 0.8752323420074349, "grad_norm": 0.6426087592091959, "learning_rate": 4.0925070235019957e-07, "loss": 0.0665, "step": 18835 }, { "epoch": 0.875278810408922, "grad_norm": 0.3393673756061131, "learning_rate": 4.0895258349656186e-07, "loss": 0.0403, "step": 18836 }, { "epoch": 0.8753252788104089, "grad_norm": 0.34442523234844286, "learning_rate": 4.086545686349802e-07, "loss": 0.028, "step": 18837 }, { "epoch": 0.8753717472118959, "grad_norm": 0.5300083832791943, "learning_rate": 4.0835665777220477e-07, "loss": 0.0517, "step": 18838 }, { "epoch": 0.875418215613383, "grad_norm": 0.58982866269476, "learning_rate": 4.080588509149841e-07, "loss": 0.0623, "step": 18839 }, { "epoch": 0.8754646840148699, "grad_norm": 0.4529539379590649, "learning_rate": 4.077611480700622e-07, "loss": 0.0366, "step": 18840 }, { "epoch": 0.8755111524163569, "grad_norm": 0.6000075461365441, "learning_rate": 4.074635492441842e-07, "loss": 0.0552, "step": 18841 }, { "epoch": 0.8755576208178438, "grad_norm": 0.49474774404332095, "learning_rate": 4.0716605444409084e-07, "loss": 0.0539, "step": 18842 }, { "epoch": 0.8756040892193309, "grad_norm": 0.3788643021770261, "learning_rate": 4.068686636765201e-07, "loss": 0.0442, "step": 18843 }, { "epoch": 0.8756505576208179, "grad_norm": 0.6069231815391384, "learning_rate": 4.0657137694820826e-07, "loss": 0.0604, "step": 18844 }, { "epoch": 0.8756970260223048, "grad_norm": 0.25615478936299046, "learning_rate": 4.0627419426588874e-07, "loss": 0.0432, "step": 18845 }, { "epoch": 0.8757434944237918, "grad_norm": 0.7639202985536474, "learning_rate": 4.0597711563629406e-07, "loss": 0.0517, "step": 18846 }, { "epoch": 0.8757899628252788, "grad_norm": 0.6176471436100399, "learning_rate": 4.0568014106615205e-07, "loss": 0.0397, "step": 18847 }, { "epoch": 0.8758364312267658, "grad_norm": 0.4052381450101759, "learning_rate": 4.0538327056219075e-07, "loss": 0.061, "step": 18848 }, { "epoch": 0.8758828996282528, "grad_norm": 0.3792254928528745, "learning_rate": 4.0508650413113527e-07, "loss": 0.0321, "step": 18849 }, { "epoch": 0.8759293680297398, "grad_norm": 0.5967803655446251, "learning_rate": 4.047898417797047e-07, "loss": 0.0574, "step": 18850 }, { "epoch": 0.8759758364312268, "grad_norm": 0.4627890907924583, "learning_rate": 4.0449328351462147e-07, "loss": 0.0397, "step": 18851 }, { "epoch": 0.8760223048327137, "grad_norm": 0.47646242462552496, "learning_rate": 4.041968293426024e-07, "loss": 0.0306, "step": 18852 }, { "epoch": 0.8760687732342007, "grad_norm": 0.7975760909701852, "learning_rate": 4.039004792703616e-07, "loss": 0.0681, "step": 18853 }, { "epoch": 0.8761152416356878, "grad_norm": 0.4431974867413011, "learning_rate": 4.036042333046114e-07, "loss": 0.0357, "step": 18854 }, { "epoch": 0.8761617100371747, "grad_norm": 0.5246555777474504, "learning_rate": 4.0330809145206486e-07, "loss": 0.0401, "step": 18855 }, { "epoch": 0.8762081784386617, "grad_norm": 0.617415266568128, "learning_rate": 4.030120537194271e-07, "loss": 0.0504, "step": 18856 }, { "epoch": 0.8762546468401488, "grad_norm": 0.4756967722032322, "learning_rate": 4.027161201134039e-07, "loss": 0.0447, "step": 18857 }, { "epoch": 0.8763011152416357, "grad_norm": 0.34100151901006837, "learning_rate": 4.0242029064069987e-07, "loss": 0.0387, "step": 18858 }, { "epoch": 0.8763475836431227, "grad_norm": 0.48845256986921637, "learning_rate": 4.021245653080147e-07, "loss": 0.0513, "step": 18859 }, { "epoch": 0.8763940520446096, "grad_norm": 0.42742366047845654, "learning_rate": 4.01828944122048e-07, "loss": 0.0436, "step": 18860 }, { "epoch": 0.8764405204460967, "grad_norm": 0.4514013201903728, "learning_rate": 4.015334270894949e-07, "loss": 0.0537, "step": 18861 }, { "epoch": 0.8764869888475837, "grad_norm": 0.5295541365321296, "learning_rate": 4.0123801421704913e-07, "loss": 0.0419, "step": 18862 }, { "epoch": 0.8765334572490706, "grad_norm": 0.46585818256593775, "learning_rate": 4.009427055114029e-07, "loss": 0.0309, "step": 18863 }, { "epoch": 0.8765799256505576, "grad_norm": 0.5853033293874325, "learning_rate": 4.0064750097924374e-07, "loss": 0.0646, "step": 18864 }, { "epoch": 0.8766263940520446, "grad_norm": 0.6844048498516385, "learning_rate": 4.003524006272602e-07, "loss": 0.0566, "step": 18865 }, { "epoch": 0.8766728624535316, "grad_norm": 0.5212938087121364, "learning_rate": 4.0005740446213636e-07, "loss": 0.0564, "step": 18866 }, { "epoch": 0.8767193308550186, "grad_norm": 0.541171212648651, "learning_rate": 3.9976251249055297e-07, "loss": 0.0692, "step": 18867 }, { "epoch": 0.8767657992565056, "grad_norm": 0.44126245979103085, "learning_rate": 3.994677247191908e-07, "loss": 0.0493, "step": 18868 }, { "epoch": 0.8768122676579926, "grad_norm": 0.5669882097919239, "learning_rate": 3.991730411547262e-07, "loss": 0.0495, "step": 18869 }, { "epoch": 0.8768587360594795, "grad_norm": 0.5508471995803215, "learning_rate": 3.988784618038349e-07, "loss": 0.068, "step": 18870 }, { "epoch": 0.8769052044609665, "grad_norm": 0.6940914549535663, "learning_rate": 3.985839866731883e-07, "loss": 0.0528, "step": 18871 }, { "epoch": 0.8769516728624536, "grad_norm": 0.6798356720583908, "learning_rate": 3.982896157694577e-07, "loss": 0.0577, "step": 18872 }, { "epoch": 0.8769981412639405, "grad_norm": 0.4281056959130074, "learning_rate": 3.979953490993116e-07, "loss": 0.0546, "step": 18873 }, { "epoch": 0.8770446096654275, "grad_norm": 0.33761216087522716, "learning_rate": 3.977011866694125e-07, "loss": 0.0399, "step": 18874 }, { "epoch": 0.8770910780669146, "grad_norm": 0.578555266175519, "learning_rate": 3.9740712848642616e-07, "loss": 0.0414, "step": 18875 }, { "epoch": 0.8771375464684015, "grad_norm": 0.5623123806231307, "learning_rate": 3.971131745570128e-07, "loss": 0.0546, "step": 18876 }, { "epoch": 0.8771840148698885, "grad_norm": 0.5239210812957689, "learning_rate": 3.9681932488783046e-07, "loss": 0.0284, "step": 18877 }, { "epoch": 0.8772304832713754, "grad_norm": 0.7656512913762424, "learning_rate": 3.965255794855344e-07, "loss": 0.0336, "step": 18878 }, { "epoch": 0.8772769516728625, "grad_norm": 0.656117278172099, "learning_rate": 3.962319383567803e-07, "loss": 0.0574, "step": 18879 }, { "epoch": 0.8773234200743495, "grad_norm": 0.5253157198435151, "learning_rate": 3.959384015082174e-07, "loss": 0.0419, "step": 18880 }, { "epoch": 0.8773698884758364, "grad_norm": 0.6253651518151838, "learning_rate": 3.956449689464947e-07, "loss": 0.0549, "step": 18881 }, { "epoch": 0.8774163568773234, "grad_norm": 0.5108014091671574, "learning_rate": 3.953516406782604e-07, "loss": 0.0599, "step": 18882 }, { "epoch": 0.8774628252788104, "grad_norm": 0.35631386196319303, "learning_rate": 3.950584167101579e-07, "loss": 0.0317, "step": 18883 }, { "epoch": 0.8775092936802974, "grad_norm": 0.4704617787449471, "learning_rate": 3.947652970488286e-07, "loss": 0.0622, "step": 18884 }, { "epoch": 0.8775557620817844, "grad_norm": 0.598505414598377, "learning_rate": 3.944722817009122e-07, "loss": 0.038, "step": 18885 }, { "epoch": 0.8776022304832714, "grad_norm": 0.30824236010200373, "learning_rate": 3.9417937067304623e-07, "loss": 0.0534, "step": 18886 }, { "epoch": 0.8776486988847584, "grad_norm": 0.49463663262476204, "learning_rate": 3.9388656397186476e-07, "loss": 0.0441, "step": 18887 }, { "epoch": 0.8776951672862453, "grad_norm": 0.3761944814667477, "learning_rate": 3.9359386160399913e-07, "loss": 0.0384, "step": 18888 }, { "epoch": 0.8777416356877323, "grad_norm": 0.5970408309708238, "learning_rate": 3.9330126357608244e-07, "loss": 0.05, "step": 18889 }, { "epoch": 0.8777881040892194, "grad_norm": 0.5183677557360881, "learning_rate": 3.93008769894741e-07, "loss": 0.0499, "step": 18890 }, { "epoch": 0.8778345724907063, "grad_norm": 0.5634930792736614, "learning_rate": 3.927163805665979e-07, "loss": 0.0669, "step": 18891 }, { "epoch": 0.8778810408921933, "grad_norm": 0.43376211163000333, "learning_rate": 3.924240955982789e-07, "loss": 0.0427, "step": 18892 }, { "epoch": 0.8779275092936804, "grad_norm": 0.5027896837420162, "learning_rate": 3.9213191499640314e-07, "loss": 0.0445, "step": 18893 }, { "epoch": 0.8779739776951673, "grad_norm": 0.4213092819256359, "learning_rate": 3.918398387675898e-07, "loss": 0.0519, "step": 18894 }, { "epoch": 0.8780204460966543, "grad_norm": 0.5562707149891953, "learning_rate": 3.9154786691845246e-07, "loss": 0.0468, "step": 18895 }, { "epoch": 0.8780669144981412, "grad_norm": 0.5099110090466292, "learning_rate": 3.9125599945560866e-07, "loss": 0.0447, "step": 18896 }, { "epoch": 0.8781133828996283, "grad_norm": 0.6776752733052457, "learning_rate": 3.909642363856658e-07, "loss": 0.0525, "step": 18897 }, { "epoch": 0.8781598513011153, "grad_norm": 0.3595358785295574, "learning_rate": 3.906725777152337e-07, "loss": 0.0345, "step": 18898 }, { "epoch": 0.8782063197026022, "grad_norm": 0.6292108829611088, "learning_rate": 3.903810234509192e-07, "loss": 0.0618, "step": 18899 }, { "epoch": 0.8782527881040892, "grad_norm": 0.4344605904900747, "learning_rate": 3.9008957359932595e-07, "loss": 0.0598, "step": 18900 }, { "epoch": 0.8782992565055762, "grad_norm": 0.56797362008857, "learning_rate": 3.897982281670565e-07, "loss": 0.0521, "step": 18901 }, { "epoch": 0.8783457249070632, "grad_norm": 0.5465148444783615, "learning_rate": 3.895069871607088e-07, "loss": 0.0488, "step": 18902 }, { "epoch": 0.8783921933085502, "grad_norm": 0.6769160784882139, "learning_rate": 3.892158505868798e-07, "loss": 0.0579, "step": 18903 }, { "epoch": 0.8784386617100371, "grad_norm": 0.36711074361963325, "learning_rate": 3.8892481845216543e-07, "loss": 0.0527, "step": 18904 }, { "epoch": 0.8784851301115242, "grad_norm": 0.517365296163305, "learning_rate": 3.886338907631554e-07, "loss": 0.0648, "step": 18905 }, { "epoch": 0.8785315985130111, "grad_norm": 0.4002328733187341, "learning_rate": 3.883430675264427e-07, "loss": 0.0418, "step": 18906 }, { "epoch": 0.8785780669144981, "grad_norm": 0.642538412463311, "learning_rate": 3.880523487486126e-07, "loss": 0.0586, "step": 18907 }, { "epoch": 0.8786245353159852, "grad_norm": 0.36423558699143305, "learning_rate": 3.877617344362511e-07, "loss": 0.0392, "step": 18908 }, { "epoch": 0.8786710037174721, "grad_norm": 0.6457855686869255, "learning_rate": 3.8747122459594057e-07, "loss": 0.0609, "step": 18909 }, { "epoch": 0.8787174721189591, "grad_norm": 0.5415828028438108, "learning_rate": 3.871808192342613e-07, "loss": 0.0611, "step": 18910 }, { "epoch": 0.8787639405204462, "grad_norm": 0.5245902566995975, "learning_rate": 3.8689051835779145e-07, "loss": 0.0602, "step": 18911 }, { "epoch": 0.8788104089219331, "grad_norm": 0.505573065614827, "learning_rate": 3.866003219731057e-07, "loss": 0.0682, "step": 18912 }, { "epoch": 0.8788568773234201, "grad_norm": 0.2839042011074954, "learning_rate": 3.863102300867788e-07, "loss": 0.0302, "step": 18913 }, { "epoch": 0.878903345724907, "grad_norm": 0.5530528250466491, "learning_rate": 3.860202427053822e-07, "loss": 0.0535, "step": 18914 }, { "epoch": 0.8789498141263941, "grad_norm": 0.4477406210837362, "learning_rate": 3.8573035983548167e-07, "loss": 0.0465, "step": 18915 }, { "epoch": 0.8789962825278811, "grad_norm": 0.4664282386974451, "learning_rate": 3.8544058148364536e-07, "loss": 0.06, "step": 18916 }, { "epoch": 0.879042750929368, "grad_norm": 0.5099097676931491, "learning_rate": 3.851509076564369e-07, "loss": 0.0509, "step": 18917 }, { "epoch": 0.879089219330855, "grad_norm": 0.522520602014541, "learning_rate": 3.848613383604172e-07, "loss": 0.051, "step": 18918 }, { "epoch": 0.879135687732342, "grad_norm": 0.5213230168938656, "learning_rate": 3.8457187360214474e-07, "loss": 0.0643, "step": 18919 }, { "epoch": 0.879182156133829, "grad_norm": 0.4562359173235154, "learning_rate": 3.842825133881789e-07, "loss": 0.0382, "step": 18920 }, { "epoch": 0.879228624535316, "grad_norm": 0.5749885924477067, "learning_rate": 3.83993257725071e-07, "loss": 0.0489, "step": 18921 }, { "epoch": 0.879275092936803, "grad_norm": 0.559391383409657, "learning_rate": 3.837041066193731e-07, "loss": 0.0535, "step": 18922 }, { "epoch": 0.87932156133829, "grad_norm": 0.5457622988178841, "learning_rate": 3.834150600776371e-07, "loss": 0.0683, "step": 18923 }, { "epoch": 0.8793680297397769, "grad_norm": 0.6765154929481505, "learning_rate": 3.831261181064083e-07, "loss": 0.0629, "step": 18924 }, { "epoch": 0.8794144981412639, "grad_norm": 0.5797892539338575, "learning_rate": 3.8283728071223214e-07, "loss": 0.0563, "step": 18925 }, { "epoch": 0.879460966542751, "grad_norm": 0.45486929164374684, "learning_rate": 3.8254854790165166e-07, "loss": 0.0355, "step": 18926 }, { "epoch": 0.8795074349442379, "grad_norm": 0.3463076459212896, "learning_rate": 3.8225991968120545e-07, "loss": 0.0302, "step": 18927 }, { "epoch": 0.8795539033457249, "grad_norm": 0.3497521690722578, "learning_rate": 3.819713960574328e-07, "loss": 0.0402, "step": 18928 }, { "epoch": 0.879600371747212, "grad_norm": 0.379225478428391, "learning_rate": 3.816829770368674e-07, "loss": 0.0336, "step": 18929 }, { "epoch": 0.8796468401486989, "grad_norm": 0.5943940760401286, "learning_rate": 3.81394662626044e-07, "loss": 0.06, "step": 18930 }, { "epoch": 0.8796933085501859, "grad_norm": 0.5437164695217018, "learning_rate": 3.8110645283149285e-07, "loss": 0.0503, "step": 18931 }, { "epoch": 0.8797397769516728, "grad_norm": 0.5262109348525331, "learning_rate": 3.8081834765974155e-07, "loss": 0.0546, "step": 18932 }, { "epoch": 0.8797862453531599, "grad_norm": 0.664479291997493, "learning_rate": 3.8053034711731604e-07, "loss": 0.0688, "step": 18933 }, { "epoch": 0.8798327137546469, "grad_norm": 0.731349725301714, "learning_rate": 3.8024245121074055e-07, "loss": 0.0556, "step": 18934 }, { "epoch": 0.8798791821561338, "grad_norm": 0.5790885092110749, "learning_rate": 3.7995465994653536e-07, "loss": 0.0602, "step": 18935 }, { "epoch": 0.8799256505576208, "grad_norm": 0.7811672226821027, "learning_rate": 3.7966697333121915e-07, "loss": 0.0637, "step": 18936 }, { "epoch": 0.8799721189591078, "grad_norm": 0.5118631366885321, "learning_rate": 3.793793913713095e-07, "loss": 0.0566, "step": 18937 }, { "epoch": 0.8800185873605948, "grad_norm": 0.4613778704141332, "learning_rate": 3.790919140733207e-07, "loss": 0.051, "step": 18938 }, { "epoch": 0.8800650557620818, "grad_norm": 0.4519876536809332, "learning_rate": 3.7880454144376134e-07, "loss": 0.0515, "step": 18939 }, { "epoch": 0.8801115241635687, "grad_norm": 0.3338518345724544, "learning_rate": 3.7851727348914404e-07, "loss": 0.0594, "step": 18940 }, { "epoch": 0.8801579925650558, "grad_norm": 0.4481487860574614, "learning_rate": 3.7823011021597466e-07, "loss": 0.0308, "step": 18941 }, { "epoch": 0.8802044609665427, "grad_norm": 0.4757709893520886, "learning_rate": 3.7794305163075697e-07, "loss": 0.0569, "step": 18942 }, { "epoch": 0.8802509293680297, "grad_norm": 0.7516638993363264, "learning_rate": 3.7765609773999343e-07, "loss": 0.0564, "step": 18943 }, { "epoch": 0.8802973977695168, "grad_norm": 0.597342555692141, "learning_rate": 3.773692485501862e-07, "loss": 0.0654, "step": 18944 }, { "epoch": 0.8803438661710037, "grad_norm": 0.4658900444191992, "learning_rate": 3.770825040678294e-07, "loss": 0.0678, "step": 18945 }, { "epoch": 0.8803903345724907, "grad_norm": 0.4721004030355824, "learning_rate": 3.767958642994185e-07, "loss": 0.0356, "step": 18946 }, { "epoch": 0.8804368029739778, "grad_norm": 0.6499401317909929, "learning_rate": 3.765093292514482e-07, "loss": 0.06, "step": 18947 }, { "epoch": 0.8804832713754647, "grad_norm": 0.8095868843013573, "learning_rate": 3.762228989304079e-07, "loss": 0.0694, "step": 18948 }, { "epoch": 0.8805297397769517, "grad_norm": 0.44314630286371226, "learning_rate": 3.75936573342785e-07, "loss": 0.0599, "step": 18949 }, { "epoch": 0.8805762081784386, "grad_norm": 0.5199321089696225, "learning_rate": 3.756503524950655e-07, "loss": 0.0437, "step": 18950 }, { "epoch": 0.8806226765799257, "grad_norm": 0.38808799411542905, "learning_rate": 3.7536423639373263e-07, "loss": 0.0527, "step": 18951 }, { "epoch": 0.8806691449814127, "grad_norm": 0.6599909107551819, "learning_rate": 3.7507822504526725e-07, "loss": 0.0602, "step": 18952 }, { "epoch": 0.8807156133828996, "grad_norm": 0.5122612371085642, "learning_rate": 3.7479231845614696e-07, "loss": 0.0578, "step": 18953 }, { "epoch": 0.8807620817843866, "grad_norm": 0.2810937135595012, "learning_rate": 3.7450651663284875e-07, "loss": 0.0315, "step": 18954 }, { "epoch": 0.8808085501858736, "grad_norm": 0.5470821299006208, "learning_rate": 3.742208195818475e-07, "loss": 0.0445, "step": 18955 }, { "epoch": 0.8808550185873606, "grad_norm": 0.5435548425551904, "learning_rate": 3.7393522730961195e-07, "loss": 0.0478, "step": 18956 }, { "epoch": 0.8809014869888476, "grad_norm": 0.3435973078783598, "learning_rate": 3.73649739822613e-07, "loss": 0.0523, "step": 18957 }, { "epoch": 0.8809479553903345, "grad_norm": 0.5570541775794888, "learning_rate": 3.73364357127316e-07, "loss": 0.0692, "step": 18958 }, { "epoch": 0.8809944237918216, "grad_norm": 0.5642521822179476, "learning_rate": 3.730790792301864e-07, "loss": 0.0566, "step": 18959 }, { "epoch": 0.8810408921933085, "grad_norm": 0.5112238390991983, "learning_rate": 3.7279390613768396e-07, "loss": 0.0401, "step": 18960 }, { "epoch": 0.8810873605947955, "grad_norm": 0.6061864835717995, "learning_rate": 3.725088378562719e-07, "loss": 0.0675, "step": 18961 }, { "epoch": 0.8811338289962826, "grad_norm": 0.2520732405273909, "learning_rate": 3.7222387439240395e-07, "loss": 0.019, "step": 18962 }, { "epoch": 0.8811802973977695, "grad_norm": 0.6524441715942308, "learning_rate": 3.719390157525349e-07, "loss": 0.0566, "step": 18963 }, { "epoch": 0.8812267657992565, "grad_norm": 0.3021146781172497, "learning_rate": 3.7165426194311906e-07, "loss": 0.0428, "step": 18964 }, { "epoch": 0.8812732342007435, "grad_norm": 0.4813755803313033, "learning_rate": 3.713696129706057e-07, "loss": 0.0456, "step": 18965 }, { "epoch": 0.8813197026022305, "grad_norm": 0.672341313367233, "learning_rate": 3.710850688414419e-07, "loss": 0.0586, "step": 18966 }, { "epoch": 0.8813661710037175, "grad_norm": 0.5290256884277812, "learning_rate": 3.708006295620731e-07, "loss": 0.0592, "step": 18967 }, { "epoch": 0.8814126394052044, "grad_norm": 0.7066803755832234, "learning_rate": 3.7051629513894185e-07, "loss": 0.0611, "step": 18968 }, { "epoch": 0.8814591078066915, "grad_norm": 0.6161773536049354, "learning_rate": 3.702320655784891e-07, "loss": 0.0647, "step": 18969 }, { "epoch": 0.8815055762081785, "grad_norm": 0.9658501944635274, "learning_rate": 3.699479408871526e-07, "loss": 0.0446, "step": 18970 }, { "epoch": 0.8815520446096654, "grad_norm": 0.26643839950610587, "learning_rate": 3.6966392107136874e-07, "loss": 0.0374, "step": 18971 }, { "epoch": 0.8815985130111524, "grad_norm": 0.7461223910740528, "learning_rate": 3.693800061375702e-07, "loss": 0.0662, "step": 18972 }, { "epoch": 0.8816449814126394, "grad_norm": 0.5005258300649509, "learning_rate": 3.6909619609218795e-07, "loss": 0.0657, "step": 18973 }, { "epoch": 0.8816914498141264, "grad_norm": 0.5252291687046236, "learning_rate": 3.688124909416513e-07, "loss": 0.0664, "step": 18974 }, { "epoch": 0.8817379182156134, "grad_norm": 0.4589695237895289, "learning_rate": 3.685288906923862e-07, "loss": 0.0513, "step": 18975 }, { "epoch": 0.8817843866171003, "grad_norm": 0.5106537149690842, "learning_rate": 3.6824539535081585e-07, "loss": 0.0562, "step": 18976 }, { "epoch": 0.8818308550185874, "grad_norm": 0.5443938263320725, "learning_rate": 3.679620049233612e-07, "loss": 0.0433, "step": 18977 }, { "epoch": 0.8818773234200743, "grad_norm": 0.545060327833888, "learning_rate": 3.676787194164433e-07, "loss": 0.0588, "step": 18978 }, { "epoch": 0.8819237918215613, "grad_norm": 0.49218793619914114, "learning_rate": 3.673955388364786e-07, "loss": 0.0411, "step": 18979 }, { "epoch": 0.8819702602230484, "grad_norm": 0.4924752137110047, "learning_rate": 3.671124631898787e-07, "loss": 0.0423, "step": 18980 }, { "epoch": 0.8820167286245353, "grad_norm": 0.6231686262237667, "learning_rate": 3.66829492483059e-07, "loss": 0.0449, "step": 18981 }, { "epoch": 0.8820631970260223, "grad_norm": 0.4662295606665264, "learning_rate": 3.6654662672242715e-07, "loss": 0.0385, "step": 18982 }, { "epoch": 0.8821096654275093, "grad_norm": 0.3768748022217528, "learning_rate": 3.662638659143908e-07, "loss": 0.0433, "step": 18983 }, { "epoch": 0.8821561338289963, "grad_norm": 0.5914272725060595, "learning_rate": 3.6598121006535425e-07, "loss": 0.0819, "step": 18984 }, { "epoch": 0.8822026022304833, "grad_norm": 0.571107767860605, "learning_rate": 3.6569865918172187e-07, "loss": 0.0586, "step": 18985 }, { "epoch": 0.8822490706319702, "grad_norm": 0.7398022146312804, "learning_rate": 3.6541621326989183e-07, "loss": 0.0581, "step": 18986 }, { "epoch": 0.8822955390334573, "grad_norm": 0.7048735920821766, "learning_rate": 3.651338723362613e-07, "loss": 0.0481, "step": 18987 }, { "epoch": 0.8823420074349443, "grad_norm": 0.4821005358591038, "learning_rate": 3.648516363872279e-07, "loss": 0.045, "step": 18988 }, { "epoch": 0.8823884758364312, "grad_norm": 0.4484350265755148, "learning_rate": 3.645695054291831e-07, "loss": 0.0391, "step": 18989 }, { "epoch": 0.8824349442379182, "grad_norm": 0.688523311354252, "learning_rate": 3.642874794685175e-07, "loss": 0.0539, "step": 18990 }, { "epoch": 0.8824814126394052, "grad_norm": 0.5027501220990283, "learning_rate": 3.640055585116198e-07, "loss": 0.0519, "step": 18991 }, { "epoch": 0.8825278810408922, "grad_norm": 0.5171848198000171, "learning_rate": 3.637237425648754e-07, "loss": 0.0353, "step": 18992 }, { "epoch": 0.8825743494423792, "grad_norm": 0.4214358990449128, "learning_rate": 3.6344203163466764e-07, "loss": 0.0424, "step": 18993 }, { "epoch": 0.8826208178438661, "grad_norm": 0.5957221625963041, "learning_rate": 3.631604257273774e-07, "loss": 0.0507, "step": 18994 }, { "epoch": 0.8826672862453532, "grad_norm": 0.4536436785312327, "learning_rate": 3.6287892484938414e-07, "loss": 0.0391, "step": 18995 }, { "epoch": 0.8827137546468401, "grad_norm": 0.47431140214837675, "learning_rate": 3.625975290070638e-07, "loss": 0.0415, "step": 18996 }, { "epoch": 0.8827602230483271, "grad_norm": 0.424069226359649, "learning_rate": 3.6231623820679017e-07, "loss": 0.0626, "step": 18997 }, { "epoch": 0.8828066914498142, "grad_norm": 0.6181927918814383, "learning_rate": 3.6203505245493486e-07, "loss": 0.0433, "step": 18998 }, { "epoch": 0.8828531598513011, "grad_norm": 0.39577165335441633, "learning_rate": 3.6175397175786666e-07, "loss": 0.0356, "step": 18999 }, { "epoch": 0.8828996282527881, "grad_norm": 0.5891059936932888, "learning_rate": 3.614729961219532e-07, "loss": 0.0582, "step": 19000 }, { "epoch": 0.8829460966542751, "grad_norm": 0.3935199968346162, "learning_rate": 3.611921255535572e-07, "loss": 0.0619, "step": 19001 }, { "epoch": 0.8829925650557621, "grad_norm": 0.4794954694330577, "learning_rate": 3.6091136005904246e-07, "loss": 0.0532, "step": 19002 }, { "epoch": 0.8830390334572491, "grad_norm": 0.7994643850590802, "learning_rate": 3.606306996447695e-07, "loss": 0.0557, "step": 19003 }, { "epoch": 0.883085501858736, "grad_norm": 0.5137369955716703, "learning_rate": 3.6035014431709146e-07, "loss": 0.0467, "step": 19004 }, { "epoch": 0.883131970260223, "grad_norm": 0.39886760711460084, "learning_rate": 3.6006969408236725e-07, "loss": 0.0497, "step": 19005 }, { "epoch": 0.8831784386617101, "grad_norm": 0.4545941661772714, "learning_rate": 3.5978934894694783e-07, "loss": 0.0532, "step": 19006 }, { "epoch": 0.883224907063197, "grad_norm": 0.5080235004479314, "learning_rate": 3.5950910891718317e-07, "loss": 0.0591, "step": 19007 }, { "epoch": 0.883271375464684, "grad_norm": 0.524321867126712, "learning_rate": 3.592289739994215e-07, "loss": 0.0534, "step": 19008 }, { "epoch": 0.883317843866171, "grad_norm": 0.5171996916930174, "learning_rate": 3.589489442000077e-07, "loss": 0.0416, "step": 19009 }, { "epoch": 0.883364312267658, "grad_norm": 0.7734532422407846, "learning_rate": 3.586690195252851e-07, "loss": 0.0503, "step": 19010 }, { "epoch": 0.883410780669145, "grad_norm": 0.4922575232646156, "learning_rate": 3.5838919998159305e-07, "loss": 0.0525, "step": 19011 }, { "epoch": 0.883457249070632, "grad_norm": 0.4958051607871734, "learning_rate": 3.58109485575272e-07, "loss": 0.0573, "step": 19012 }, { "epoch": 0.883503717472119, "grad_norm": 0.5078024515700194, "learning_rate": 3.5782987631265643e-07, "loss": 0.0592, "step": 19013 }, { "epoch": 0.8835501858736059, "grad_norm": 0.598195792142026, "learning_rate": 3.575503722000795e-07, "loss": 0.038, "step": 19014 }, { "epoch": 0.8835966542750929, "grad_norm": 0.3282615933867946, "learning_rate": 3.572709732438734e-07, "loss": 0.0234, "step": 19015 }, { "epoch": 0.88364312267658, "grad_norm": 0.559851805795731, "learning_rate": 3.5699167945036585e-07, "loss": 0.0582, "step": 19016 }, { "epoch": 0.8836895910780669, "grad_norm": 0.5419595469335216, "learning_rate": 3.5671249082588344e-07, "loss": 0.0339, "step": 19017 }, { "epoch": 0.8837360594795539, "grad_norm": 0.6917583278547198, "learning_rate": 3.5643340737674945e-07, "loss": 0.0401, "step": 19018 }, { "epoch": 0.883782527881041, "grad_norm": 0.41723246541406933, "learning_rate": 3.561544291092867e-07, "loss": 0.0445, "step": 19019 }, { "epoch": 0.8838289962825279, "grad_norm": 0.4736658192641719, "learning_rate": 3.558755560298144e-07, "loss": 0.0449, "step": 19020 }, { "epoch": 0.8838754646840149, "grad_norm": 0.35042230466319, "learning_rate": 3.5559678814464705e-07, "loss": 0.0404, "step": 19021 }, { "epoch": 0.8839219330855018, "grad_norm": 0.43838190904678026, "learning_rate": 3.553181254601018e-07, "loss": 0.0503, "step": 19022 }, { "epoch": 0.8839684014869889, "grad_norm": 0.40851061594079535, "learning_rate": 3.550395679824886e-07, "loss": 0.0543, "step": 19023 }, { "epoch": 0.8840148698884759, "grad_norm": 0.3948360219278578, "learning_rate": 3.5476111571811856e-07, "loss": 0.0444, "step": 19024 }, { "epoch": 0.8840613382899628, "grad_norm": 0.4662335369631534, "learning_rate": 3.5448276867329767e-07, "loss": 0.0681, "step": 19025 }, { "epoch": 0.8841078066914498, "grad_norm": 0.6561296539047269, "learning_rate": 3.542045268543326e-07, "loss": 0.0696, "step": 19026 }, { "epoch": 0.8841542750929368, "grad_norm": 0.31943946106169857, "learning_rate": 3.5392639026752384e-07, "loss": 0.0364, "step": 19027 }, { "epoch": 0.8842007434944238, "grad_norm": 0.5310803333150078, "learning_rate": 3.536483589191714e-07, "loss": 0.0634, "step": 19028 }, { "epoch": 0.8842472118959108, "grad_norm": 0.5463043123873997, "learning_rate": 3.533704328155746e-07, "loss": 0.0629, "step": 19029 }, { "epoch": 0.8842936802973977, "grad_norm": 0.6075493744259353, "learning_rate": 3.530926119630279e-07, "loss": 0.0424, "step": 19030 }, { "epoch": 0.8843401486988848, "grad_norm": 0.6142745662043614, "learning_rate": 3.528148963678246e-07, "loss": 0.0382, "step": 19031 }, { "epoch": 0.8843866171003717, "grad_norm": 0.48694575774841997, "learning_rate": 3.525372860362547e-07, "loss": 0.0432, "step": 19032 }, { "epoch": 0.8844330855018587, "grad_norm": 0.6163693831570347, "learning_rate": 3.52259780974607e-07, "loss": 0.0476, "step": 19033 }, { "epoch": 0.8844795539033458, "grad_norm": 0.4343825886721707, "learning_rate": 3.519823811891665e-07, "loss": 0.05, "step": 19034 }, { "epoch": 0.8845260223048327, "grad_norm": 0.6848945293955356, "learning_rate": 3.517050866862165e-07, "loss": 0.0425, "step": 19035 }, { "epoch": 0.8845724907063197, "grad_norm": 0.5683535686827672, "learning_rate": 3.514278974720392e-07, "loss": 0.0672, "step": 19036 }, { "epoch": 0.8846189591078067, "grad_norm": 0.46522092512162605, "learning_rate": 3.5115081355291293e-07, "loss": 0.0538, "step": 19037 }, { "epoch": 0.8846654275092937, "grad_norm": 0.5260002663120912, "learning_rate": 3.5087383493511315e-07, "loss": 0.0422, "step": 19038 }, { "epoch": 0.8847118959107807, "grad_norm": 0.3099836021205964, "learning_rate": 3.505969616249144e-07, "loss": 0.0452, "step": 19039 }, { "epoch": 0.8847583643122676, "grad_norm": 0.560364508484197, "learning_rate": 3.503201936285877e-07, "loss": 0.0665, "step": 19040 }, { "epoch": 0.8848048327137547, "grad_norm": 0.590727957532183, "learning_rate": 3.5004353095240196e-07, "loss": 0.0482, "step": 19041 }, { "epoch": 0.8848513011152417, "grad_norm": 0.5728724509880305, "learning_rate": 3.497669736026238e-07, "loss": 0.0472, "step": 19042 }, { "epoch": 0.8848977695167286, "grad_norm": 0.45734459877435324, "learning_rate": 3.4949052158551875e-07, "loss": 0.0498, "step": 19043 }, { "epoch": 0.8849442379182156, "grad_norm": 0.4160728520468593, "learning_rate": 3.492141749073491e-07, "loss": 0.0624, "step": 19044 }, { "epoch": 0.8849907063197026, "grad_norm": 0.5219560860333189, "learning_rate": 3.489379335743709e-07, "loss": 0.0652, "step": 19045 }, { "epoch": 0.8850371747211896, "grad_norm": 0.4605720298360219, "learning_rate": 3.4866179759284525e-07, "loss": 0.0459, "step": 19046 }, { "epoch": 0.8850836431226766, "grad_norm": 0.23358846451338672, "learning_rate": 3.483857669690249e-07, "loss": 0.0307, "step": 19047 }, { "epoch": 0.8851301115241635, "grad_norm": 0.3710867207814996, "learning_rate": 3.4810984170916274e-07, "loss": 0.0309, "step": 19048 }, { "epoch": 0.8851765799256506, "grad_norm": 0.6732926547826544, "learning_rate": 3.4783402181950754e-07, "loss": 0.0533, "step": 19049 }, { "epoch": 0.8852230483271375, "grad_norm": 0.4665663027235142, "learning_rate": 3.475583073063105e-07, "loss": 0.0343, "step": 19050 }, { "epoch": 0.8852695167286245, "grad_norm": 0.3190696133735081, "learning_rate": 3.472826981758126e-07, "loss": 0.0345, "step": 19051 }, { "epoch": 0.8853159851301116, "grad_norm": 0.3763035814745551, "learning_rate": 3.4700719443425846e-07, "loss": 0.0502, "step": 19052 }, { "epoch": 0.8853624535315985, "grad_norm": 0.3237614270508545, "learning_rate": 3.467317960878891e-07, "loss": 0.0456, "step": 19053 }, { "epoch": 0.8854089219330855, "grad_norm": 0.5881239945751011, "learning_rate": 3.4645650314294175e-07, "loss": 0.047, "step": 19054 }, { "epoch": 0.8854553903345725, "grad_norm": 0.3429836112959309, "learning_rate": 3.461813156056526e-07, "loss": 0.0376, "step": 19055 }, { "epoch": 0.8855018587360595, "grad_norm": 0.44570856353946975, "learning_rate": 3.459062334822549e-07, "loss": 0.0352, "step": 19056 }, { "epoch": 0.8855483271375465, "grad_norm": 0.36539825830013983, "learning_rate": 3.4563125677897936e-07, "loss": 0.0483, "step": 19057 }, { "epoch": 0.8855947955390334, "grad_norm": 0.46000552855452637, "learning_rate": 3.453563855020542e-07, "loss": 0.0479, "step": 19058 }, { "epoch": 0.8856412639405205, "grad_norm": 0.61788665656503, "learning_rate": 3.4508161965770506e-07, "loss": 0.0573, "step": 19059 }, { "epoch": 0.8856877323420075, "grad_norm": 3.434929401202805, "learning_rate": 3.448069592521575e-07, "loss": 0.0684, "step": 19060 }, { "epoch": 0.8857342007434944, "grad_norm": 0.3099061162260047, "learning_rate": 3.445324042916315e-07, "loss": 0.0471, "step": 19061 }, { "epoch": 0.8857806691449814, "grad_norm": 0.7684155328160315, "learning_rate": 3.442579547823466e-07, "loss": 0.0439, "step": 19062 }, { "epoch": 0.8858271375464684, "grad_norm": 0.3475063016920423, "learning_rate": 3.439836107305189e-07, "loss": 0.0473, "step": 19063 }, { "epoch": 0.8858736059479554, "grad_norm": 0.8359709477802759, "learning_rate": 3.437093721423629e-07, "loss": 0.0754, "step": 19064 }, { "epoch": 0.8859200743494424, "grad_norm": 0.4061415635620285, "learning_rate": 3.4343523902409025e-07, "loss": 0.0432, "step": 19065 }, { "epoch": 0.8859665427509293, "grad_norm": 0.3878265298094518, "learning_rate": 3.431612113819094e-07, "loss": 0.0415, "step": 19066 }, { "epoch": 0.8860130111524164, "grad_norm": 0.5534113700429519, "learning_rate": 3.428872892220303e-07, "loss": 0.0431, "step": 19067 }, { "epoch": 0.8860594795539033, "grad_norm": 0.3169578613347819, "learning_rate": 3.426134725506547e-07, "loss": 0.0538, "step": 19068 }, { "epoch": 0.8861059479553903, "grad_norm": 0.845759382967882, "learning_rate": 3.4233976137398486e-07, "loss": 0.0613, "step": 19069 }, { "epoch": 0.8861524163568774, "grad_norm": 0.6513599387867414, "learning_rate": 3.4206615569822246e-07, "loss": 0.0528, "step": 19070 }, { "epoch": 0.8861988847583643, "grad_norm": 0.419540510292524, "learning_rate": 3.4179265552956366e-07, "loss": 0.0578, "step": 19071 }, { "epoch": 0.8862453531598513, "grad_norm": 0.42432566646011527, "learning_rate": 3.4151926087420404e-07, "loss": 0.0295, "step": 19072 }, { "epoch": 0.8862918215613383, "grad_norm": 0.4938356624566684, "learning_rate": 3.412459717383365e-07, "loss": 0.056, "step": 19073 }, { "epoch": 0.8863382899628253, "grad_norm": 0.4759449793529151, "learning_rate": 3.409727881281505e-07, "loss": 0.0605, "step": 19074 }, { "epoch": 0.8863847583643123, "grad_norm": 0.3171198207907561, "learning_rate": 3.4069971004983495e-07, "loss": 0.0386, "step": 19075 }, { "epoch": 0.8864312267657992, "grad_norm": 0.5638824883875752, "learning_rate": 3.4042673750957323e-07, "loss": 0.067, "step": 19076 }, { "epoch": 0.8864776951672863, "grad_norm": 0.5352260790806811, "learning_rate": 3.401538705135515e-07, "loss": 0.047, "step": 19077 }, { "epoch": 0.8865241635687733, "grad_norm": 0.6067157397814766, "learning_rate": 3.398811090679488e-07, "loss": 0.0501, "step": 19078 }, { "epoch": 0.8865706319702602, "grad_norm": 0.5438227074890137, "learning_rate": 3.396084531789434e-07, "loss": 0.0516, "step": 19079 }, { "epoch": 0.8866171003717472, "grad_norm": 0.5025549169527609, "learning_rate": 3.3933590285271154e-07, "loss": 0.06, "step": 19080 }, { "epoch": 0.8866635687732342, "grad_norm": 0.40818765330252793, "learning_rate": 3.3906345809542717e-07, "loss": 0.0541, "step": 19081 }, { "epoch": 0.8867100371747212, "grad_norm": 0.5457685750495226, "learning_rate": 3.387911189132609e-07, "loss": 0.0524, "step": 19082 }, { "epoch": 0.8867565055762082, "grad_norm": 0.6635152860550378, "learning_rate": 3.3851888531238053e-07, "loss": 0.0774, "step": 19083 }, { "epoch": 0.8868029739776951, "grad_norm": 0.7042275643290599, "learning_rate": 3.382467572989545e-07, "loss": 0.0507, "step": 19084 }, { "epoch": 0.8868494423791822, "grad_norm": 0.43156241137113505, "learning_rate": 3.379747348791462e-07, "loss": 0.0298, "step": 19085 }, { "epoch": 0.8868959107806691, "grad_norm": 0.393170063376812, "learning_rate": 3.3770281805911577e-07, "loss": 0.0473, "step": 19086 }, { "epoch": 0.8869423791821561, "grad_norm": 0.5398549587063044, "learning_rate": 3.374310068450243e-07, "loss": 0.046, "step": 19087 }, { "epoch": 0.8869888475836432, "grad_norm": 0.5829027299809109, "learning_rate": 3.371593012430274e-07, "loss": 0.0586, "step": 19088 }, { "epoch": 0.8870353159851301, "grad_norm": 0.3865719550601705, "learning_rate": 3.368877012592803e-07, "loss": 0.0433, "step": 19089 }, { "epoch": 0.8870817843866171, "grad_norm": 0.5768800475057548, "learning_rate": 3.3661620689993344e-07, "loss": 0.0759, "step": 19090 }, { "epoch": 0.8871282527881041, "grad_norm": 0.5651241038790272, "learning_rate": 3.3634481817113926e-07, "loss": 0.0617, "step": 19091 }, { "epoch": 0.8871747211895911, "grad_norm": 0.5058506352273404, "learning_rate": 3.360735350790428e-07, "loss": 0.0632, "step": 19092 }, { "epoch": 0.8872211895910781, "grad_norm": 0.5565992642307568, "learning_rate": 3.3580235762978864e-07, "loss": 0.0344, "step": 19093 }, { "epoch": 0.887267657992565, "grad_norm": 0.6533015252446921, "learning_rate": 3.3553128582952064e-07, "loss": 0.0493, "step": 19094 }, { "epoch": 0.887314126394052, "grad_norm": 0.3604930726783199, "learning_rate": 3.352603196843784e-07, "loss": 0.0393, "step": 19095 }, { "epoch": 0.8873605947955391, "grad_norm": 0.28681679235990004, "learning_rate": 3.3498945920049985e-07, "loss": 0.0436, "step": 19096 }, { "epoch": 0.887407063197026, "grad_norm": 0.36418526440543986, "learning_rate": 3.347187043840194e-07, "loss": 0.0404, "step": 19097 }, { "epoch": 0.887453531598513, "grad_norm": 0.43914871163787483, "learning_rate": 3.344480552410706e-07, "loss": 0.0453, "step": 19098 }, { "epoch": 0.8875, "grad_norm": 0.7057209181503701, "learning_rate": 3.3417751177778347e-07, "loss": 0.0492, "step": 19099 }, { "epoch": 0.887546468401487, "grad_norm": 0.49887560307082823, "learning_rate": 3.339070740002859e-07, "loss": 0.0316, "step": 19100 }, { "epoch": 0.887592936802974, "grad_norm": 0.7001548169054983, "learning_rate": 3.3363674191470463e-07, "loss": 0.0527, "step": 19101 }, { "epoch": 0.8876394052044609, "grad_norm": 0.4031479960446494, "learning_rate": 3.333665155271626e-07, "loss": 0.0584, "step": 19102 }, { "epoch": 0.887685873605948, "grad_norm": 0.5295695522285822, "learning_rate": 3.3309639484378043e-07, "loss": 0.0373, "step": 19103 }, { "epoch": 0.8877323420074349, "grad_norm": 0.7188204116409798, "learning_rate": 3.328263798706771e-07, "loss": 0.0546, "step": 19104 }, { "epoch": 0.8877788104089219, "grad_norm": 0.5068996683112114, "learning_rate": 3.3255647061396777e-07, "loss": 0.0512, "step": 19105 }, { "epoch": 0.887825278810409, "grad_norm": 0.3215314904691274, "learning_rate": 3.32286667079767e-07, "loss": 0.0572, "step": 19106 }, { "epoch": 0.8878717472118959, "grad_norm": 0.20558971578841728, "learning_rate": 3.3201696927418535e-07, "loss": 0.0209, "step": 19107 }, { "epoch": 0.8879182156133829, "grad_norm": 0.23135561613918176, "learning_rate": 3.3174737720333307e-07, "loss": 0.0267, "step": 19108 }, { "epoch": 0.8879646840148699, "grad_norm": 1.0117291054209396, "learning_rate": 3.314778908733163e-07, "loss": 0.0573, "step": 19109 }, { "epoch": 0.8880111524163569, "grad_norm": 0.28186523548565934, "learning_rate": 3.312085102902379e-07, "loss": 0.0287, "step": 19110 }, { "epoch": 0.8880576208178439, "grad_norm": 0.5141383554532076, "learning_rate": 3.309392354602009e-07, "loss": 0.05, "step": 19111 }, { "epoch": 0.8881040892193308, "grad_norm": 0.5993457344037192, "learning_rate": 3.306700663893042e-07, "loss": 0.0613, "step": 19112 }, { "epoch": 0.8881505576208178, "grad_norm": 0.35611296838705125, "learning_rate": 3.304010030836452e-07, "loss": 0.0557, "step": 19113 }, { "epoch": 0.8881970260223049, "grad_norm": 0.8072968864564142, "learning_rate": 3.3013204554931734e-07, "loss": 0.0755, "step": 19114 }, { "epoch": 0.8882434944237918, "grad_norm": 0.5001365746276981, "learning_rate": 3.2986319379241516e-07, "loss": 0.0283, "step": 19115 }, { "epoch": 0.8882899628252788, "grad_norm": 0.6245834976808539, "learning_rate": 3.2959444781902603e-07, "loss": 0.0557, "step": 19116 }, { "epoch": 0.8883364312267658, "grad_norm": 0.4928058848549433, "learning_rate": 3.2932580763523737e-07, "loss": 0.073, "step": 19117 }, { "epoch": 0.8883828996282528, "grad_norm": 0.5338483632871556, "learning_rate": 3.2905727324713586e-07, "loss": 0.0515, "step": 19118 }, { "epoch": 0.8884293680297398, "grad_norm": 0.27417234211817787, "learning_rate": 3.2878884466080286e-07, "loss": 0.0296, "step": 19119 }, { "epoch": 0.8884758364312267, "grad_norm": 0.48375826745829986, "learning_rate": 3.28520521882319e-07, "loss": 0.0497, "step": 19120 }, { "epoch": 0.8885223048327138, "grad_norm": 0.4482111048199462, "learning_rate": 3.2825230491776214e-07, "loss": 0.0428, "step": 19121 }, { "epoch": 0.8885687732342007, "grad_norm": 0.5188324766730439, "learning_rate": 3.2798419377320753e-07, "loss": 0.0491, "step": 19122 }, { "epoch": 0.8886152416356877, "grad_norm": 0.4283799230747232, "learning_rate": 3.2771618845472806e-07, "loss": 0.0471, "step": 19123 }, { "epoch": 0.8886617100371748, "grad_norm": 0.3718167225540522, "learning_rate": 3.274482889683933e-07, "loss": 0.0532, "step": 19124 }, { "epoch": 0.8887081784386617, "grad_norm": 0.44050140036220226, "learning_rate": 3.2718049532027395e-07, "loss": 0.0568, "step": 19125 }, { "epoch": 0.8887546468401487, "grad_norm": 0.3825506736982117, "learning_rate": 3.2691280751643464e-07, "loss": 0.0421, "step": 19126 }, { "epoch": 0.8888011152416357, "grad_norm": 0.40155977560353484, "learning_rate": 3.2664522556293764e-07, "loss": 0.0606, "step": 19127 }, { "epoch": 0.8888475836431227, "grad_norm": 0.4427548687558929, "learning_rate": 3.263777494658449e-07, "loss": 0.04, "step": 19128 }, { "epoch": 0.8888940520446097, "grad_norm": 0.4427548687558929, "learning_rate": 3.263777494658449e-07, "loss": 0.0437, "step": 19129 }, { "epoch": 0.8889405204460966, "grad_norm": 0.4148721351221621, "learning_rate": 3.261103792312154e-07, "loss": 0.0662, "step": 19130 }, { "epoch": 0.8889869888475836, "grad_norm": 0.5407536926903376, "learning_rate": 3.258431148651053e-07, "loss": 0.0604, "step": 19131 }, { "epoch": 0.8890334572490707, "grad_norm": 0.5808361559382609, "learning_rate": 3.255759563735672e-07, "loss": 0.046, "step": 19132 }, { "epoch": 0.8890799256505576, "grad_norm": 0.4761117042543741, "learning_rate": 3.2530890376265447e-07, "loss": 0.0661, "step": 19133 }, { "epoch": 0.8891263940520446, "grad_norm": 0.4492247131214222, "learning_rate": 3.25041957038415e-07, "loss": 0.0518, "step": 19134 }, { "epoch": 0.8891728624535316, "grad_norm": 0.4492247131214222, "learning_rate": 3.25041957038415e-07, "loss": 0.0514, "step": 19135 }, { "epoch": 0.8892193308550186, "grad_norm": 0.3870504597270098, "learning_rate": 3.247751162068946e-07, "loss": 0.0322, "step": 19136 }, { "epoch": 0.8892657992565056, "grad_norm": 0.31532497078438804, "learning_rate": 3.2450838127413897e-07, "loss": 0.0376, "step": 19137 }, { "epoch": 0.8893122676579925, "grad_norm": 0.6593467741809617, "learning_rate": 3.242417522461894e-07, "loss": 0.0659, "step": 19138 }, { "epoch": 0.8893587360594796, "grad_norm": 0.551033120778452, "learning_rate": 3.239752291290849e-07, "loss": 0.0524, "step": 19139 }, { "epoch": 0.8894052044609665, "grad_norm": 0.32923558108939155, "learning_rate": 3.237088119288634e-07, "loss": 0.0391, "step": 19140 }, { "epoch": 0.8894516728624535, "grad_norm": 0.759631901103732, "learning_rate": 3.2344250065155903e-07, "loss": 0.058, "step": 19141 }, { "epoch": 0.8894981412639406, "grad_norm": 0.47470445414792783, "learning_rate": 3.231762953032036e-07, "loss": 0.0458, "step": 19142 }, { "epoch": 0.8895446096654275, "grad_norm": 0.37364362168336207, "learning_rate": 3.229101958898273e-07, "loss": 0.0383, "step": 19143 }, { "epoch": 0.8895910780669145, "grad_norm": 0.47032253580902705, "learning_rate": 3.226442024174575e-07, "loss": 0.0436, "step": 19144 }, { "epoch": 0.8896375464684015, "grad_norm": 0.2836178714395322, "learning_rate": 3.2237831489212e-07, "loss": 0.0358, "step": 19145 }, { "epoch": 0.8896840148698885, "grad_norm": 0.5361942553071372, "learning_rate": 3.221125333198366e-07, "loss": 0.0587, "step": 19146 }, { "epoch": 0.8897304832713755, "grad_norm": 0.7150985500490636, "learning_rate": 3.218468577066275e-07, "loss": 0.0548, "step": 19147 }, { "epoch": 0.8897769516728624, "grad_norm": 0.5091223813338989, "learning_rate": 3.215812880585112e-07, "loss": 0.048, "step": 19148 }, { "epoch": 0.8898234200743494, "grad_norm": 0.4374142078482095, "learning_rate": 3.2131582438150233e-07, "loss": 0.0488, "step": 19149 }, { "epoch": 0.8898698884758365, "grad_norm": 0.6398720290888825, "learning_rate": 3.210504666816133e-07, "loss": 0.0568, "step": 19150 }, { "epoch": 0.8899163568773234, "grad_norm": 0.38423708606151435, "learning_rate": 3.2078521496485714e-07, "loss": 0.0417, "step": 19151 }, { "epoch": 0.8899628252788104, "grad_norm": 0.5909714265895848, "learning_rate": 3.2052006923724064e-07, "loss": 0.0434, "step": 19152 }, { "epoch": 0.8900092936802974, "grad_norm": 0.81185915114839, "learning_rate": 3.202550295047685e-07, "loss": 0.0539, "step": 19153 }, { "epoch": 0.8900557620817844, "grad_norm": 0.47631532230674817, "learning_rate": 3.1999009577344587e-07, "loss": 0.0403, "step": 19154 }, { "epoch": 0.8901022304832714, "grad_norm": 0.5345256889585569, "learning_rate": 3.197252680492735e-07, "loss": 0.0471, "step": 19155 }, { "epoch": 0.8901486988847583, "grad_norm": 0.4098297542610948, "learning_rate": 3.1946054633824884e-07, "loss": 0.0438, "step": 19156 }, { "epoch": 0.8901951672862454, "grad_norm": 0.4909891133326458, "learning_rate": 3.191959306463699e-07, "loss": 0.0543, "step": 19157 }, { "epoch": 0.8902416356877323, "grad_norm": 0.41144741074005936, "learning_rate": 3.1893142097963016e-07, "loss": 0.0477, "step": 19158 }, { "epoch": 0.8902881040892193, "grad_norm": 0.7803726850669148, "learning_rate": 3.1866701734401927e-07, "loss": 0.0533, "step": 19159 }, { "epoch": 0.8903345724907064, "grad_norm": 0.6783222326369218, "learning_rate": 3.1840271974552696e-07, "loss": 0.0485, "step": 19160 }, { "epoch": 0.8903810408921933, "grad_norm": 0.5281903588751224, "learning_rate": 3.1813852819014166e-07, "loss": 0.0559, "step": 19161 }, { "epoch": 0.8904275092936803, "grad_norm": 0.731772297186198, "learning_rate": 3.1787444268384596e-07, "loss": 0.0656, "step": 19162 }, { "epoch": 0.8904739776951673, "grad_norm": 0.4855727528858284, "learning_rate": 3.17610463232621e-07, "loss": 0.0544, "step": 19163 }, { "epoch": 0.8905204460966543, "grad_norm": 0.6936478188030165, "learning_rate": 3.1734658984244884e-07, "loss": 0.0491, "step": 19164 }, { "epoch": 0.8905669144981413, "grad_norm": 0.5640020476427525, "learning_rate": 3.17082822519304e-07, "loss": 0.0628, "step": 19165 }, { "epoch": 0.8906133828996282, "grad_norm": 0.30960004823535286, "learning_rate": 3.168191612691618e-07, "loss": 0.043, "step": 19166 }, { "epoch": 0.8906598513011152, "grad_norm": 0.3228300917083603, "learning_rate": 3.1655560609799405e-07, "loss": 0.0356, "step": 19167 }, { "epoch": 0.8907063197026023, "grad_norm": 0.5522325593939942, "learning_rate": 3.162921570117716e-07, "loss": 0.0764, "step": 19168 }, { "epoch": 0.8907527881040892, "grad_norm": 0.5631851761242762, "learning_rate": 3.1602881401646133e-07, "loss": 0.0476, "step": 19169 }, { "epoch": 0.8907992565055762, "grad_norm": 0.739444148498418, "learning_rate": 3.1576557711802845e-07, "loss": 0.0777, "step": 19170 }, { "epoch": 0.8908457249070632, "grad_norm": 0.43659206989920235, "learning_rate": 3.1550244632243486e-07, "loss": 0.0568, "step": 19171 }, { "epoch": 0.8908921933085502, "grad_norm": 0.4360953975791069, "learning_rate": 3.1523942163564134e-07, "loss": 0.0509, "step": 19172 }, { "epoch": 0.8909386617100372, "grad_norm": 0.5623397100711466, "learning_rate": 3.149765030636054e-07, "loss": 0.0543, "step": 19173 }, { "epoch": 0.8909851301115241, "grad_norm": 0.5469357595129536, "learning_rate": 3.147136906122827e-07, "loss": 0.0533, "step": 19174 }, { "epoch": 0.8910315985130112, "grad_norm": 0.5713107408053076, "learning_rate": 3.144509842876259e-07, "loss": 0.0623, "step": 19175 }, { "epoch": 0.8910780669144981, "grad_norm": 0.5578739605843558, "learning_rate": 3.141883840955873e-07, "loss": 0.049, "step": 19176 }, { "epoch": 0.8911245353159851, "grad_norm": 0.503189363584218, "learning_rate": 3.139258900421116e-07, "loss": 0.0729, "step": 19177 }, { "epoch": 0.8911710037174722, "grad_norm": 0.2547136402316196, "learning_rate": 3.1366350213314687e-07, "loss": 0.0455, "step": 19178 }, { "epoch": 0.8912174721189591, "grad_norm": 0.6196869349001647, "learning_rate": 3.134012203746367e-07, "loss": 0.0382, "step": 19179 }, { "epoch": 0.8912639405204461, "grad_norm": 0.26409822470652566, "learning_rate": 3.1313904477252024e-07, "loss": 0.0385, "step": 19180 }, { "epoch": 0.8913104089219331, "grad_norm": 0.6464061369473039, "learning_rate": 3.128769753327382e-07, "loss": 0.0522, "step": 19181 }, { "epoch": 0.8913568773234201, "grad_norm": 0.5666052787178877, "learning_rate": 3.1261501206122647e-07, "loss": 0.0372, "step": 19182 }, { "epoch": 0.8914033457249071, "grad_norm": 0.5346961669666279, "learning_rate": 3.123531549639175e-07, "loss": 0.0453, "step": 19183 }, { "epoch": 0.891449814126394, "grad_norm": 0.4166622683599718, "learning_rate": 3.120914040467421e-07, "loss": 0.027, "step": 19184 }, { "epoch": 0.891496282527881, "grad_norm": 0.41272305562905887, "learning_rate": 3.1182975931563167e-07, "loss": 0.0449, "step": 19185 }, { "epoch": 0.8915427509293681, "grad_norm": 0.4081897004193976, "learning_rate": 3.115682207765108e-07, "loss": 0.0495, "step": 19186 }, { "epoch": 0.891589219330855, "grad_norm": 0.6744807274928745, "learning_rate": 3.1130678843530494e-07, "loss": 0.065, "step": 19187 }, { "epoch": 0.891635687732342, "grad_norm": 0.6030071250839528, "learning_rate": 3.1104546229793476e-07, "loss": 0.0616, "step": 19188 }, { "epoch": 0.891682156133829, "grad_norm": 0.5874438785501527, "learning_rate": 3.1078424237031943e-07, "loss": 0.0587, "step": 19189 }, { "epoch": 0.891728624535316, "grad_norm": 0.588623113650873, "learning_rate": 3.1052312865837597e-07, "loss": 0.0532, "step": 19190 }, { "epoch": 0.891775092936803, "grad_norm": 0.4620528469434937, "learning_rate": 3.1026212116802015e-07, "loss": 0.0705, "step": 19191 }, { "epoch": 0.8918215613382899, "grad_norm": 0.5312380793358079, "learning_rate": 3.100012199051627e-07, "loss": 0.0566, "step": 19192 }, { "epoch": 0.891868029739777, "grad_norm": 0.6930986220474511, "learning_rate": 3.0974042487571467e-07, "loss": 0.0568, "step": 19193 }, { "epoch": 0.8919144981412639, "grad_norm": 0.7554968627258871, "learning_rate": 3.094797360855811e-07, "loss": 0.0465, "step": 19194 }, { "epoch": 0.8919609665427509, "grad_norm": 0.7487508221232161, "learning_rate": 3.0921915354066855e-07, "loss": 0.0435, "step": 19195 }, { "epoch": 0.892007434944238, "grad_norm": 0.4538204410511119, "learning_rate": 3.0895867724687945e-07, "loss": 0.0581, "step": 19196 }, { "epoch": 0.8920539033457249, "grad_norm": 0.6214017971049602, "learning_rate": 3.0869830721011293e-07, "loss": 0.052, "step": 19197 }, { "epoch": 0.8921003717472119, "grad_norm": 0.5855850374927717, "learning_rate": 3.0843804343626773e-07, "loss": 0.0619, "step": 19198 }, { "epoch": 0.8921468401486989, "grad_norm": 0.36253427821957257, "learning_rate": 3.0817788593123956e-07, "loss": 0.0396, "step": 19199 }, { "epoch": 0.8921933085501859, "grad_norm": 0.4335196091101469, "learning_rate": 3.0791783470091876e-07, "loss": 0.0528, "step": 19200 }, { "epoch": 0.8922397769516729, "grad_norm": 0.7528961719426245, "learning_rate": 3.076578897511978e-07, "loss": 0.0593, "step": 19201 }, { "epoch": 0.8922862453531598, "grad_norm": 0.6326107394419656, "learning_rate": 3.0739805108796483e-07, "loss": 0.0438, "step": 19202 }, { "epoch": 0.8923327137546468, "grad_norm": 0.44543752748822557, "learning_rate": 3.071383187171045e-07, "loss": 0.0538, "step": 19203 }, { "epoch": 0.8923791821561339, "grad_norm": 0.43806550030912034, "learning_rate": 3.0687869264449945e-07, "loss": 0.0525, "step": 19204 }, { "epoch": 0.8924256505576208, "grad_norm": 0.6722337794679115, "learning_rate": 3.066191728760332e-07, "loss": 0.0674, "step": 19205 }, { "epoch": 0.8924721189591078, "grad_norm": 0.34304556373009276, "learning_rate": 3.0635975941758155e-07, "loss": 0.0465, "step": 19206 }, { "epoch": 0.8925185873605948, "grad_norm": 0.7375769266521053, "learning_rate": 3.0610045227502103e-07, "loss": 0.0557, "step": 19207 }, { "epoch": 0.8925650557620818, "grad_norm": 0.533016778339565, "learning_rate": 3.0584125145422584e-07, "loss": 0.0455, "step": 19208 }, { "epoch": 0.8926115241635688, "grad_norm": 0.49445572978317237, "learning_rate": 3.0558215696106677e-07, "loss": 0.048, "step": 19209 }, { "epoch": 0.8926579925650557, "grad_norm": 0.5187969023456666, "learning_rate": 3.0532316880141243e-07, "loss": 0.0525, "step": 19210 }, { "epoch": 0.8927044609665428, "grad_norm": 0.49597319417158775, "learning_rate": 3.050642869811299e-07, "loss": 0.0686, "step": 19211 }, { "epoch": 0.8927509293680297, "grad_norm": 0.3338970796927673, "learning_rate": 3.048055115060822e-07, "loss": 0.0343, "step": 19212 }, { "epoch": 0.8927973977695167, "grad_norm": 0.5476041551275848, "learning_rate": 3.045468423821313e-07, "loss": 0.0599, "step": 19213 }, { "epoch": 0.8928438661710038, "grad_norm": 0.4310238549850458, "learning_rate": 3.0428827961513585e-07, "loss": 0.0451, "step": 19214 }, { "epoch": 0.8928903345724907, "grad_norm": 0.5798180382456265, "learning_rate": 3.0402982321095345e-07, "loss": 0.0552, "step": 19215 }, { "epoch": 0.8929368029739777, "grad_norm": 0.7897446602907635, "learning_rate": 3.0377147317543823e-07, "loss": 0.0668, "step": 19216 }, { "epoch": 0.8929832713754647, "grad_norm": 0.5614699183212478, "learning_rate": 3.035132295144416e-07, "loss": 0.0648, "step": 19217 }, { "epoch": 0.8930297397769517, "grad_norm": 0.6022406548358169, "learning_rate": 3.0325509223381345e-07, "loss": 0.0506, "step": 19218 }, { "epoch": 0.8930762081784387, "grad_norm": 0.48571246287358505, "learning_rate": 3.029970613394007e-07, "loss": 0.0711, "step": 19219 }, { "epoch": 0.8931226765799256, "grad_norm": 0.7334532800497309, "learning_rate": 3.027391368370475e-07, "loss": 0.0488, "step": 19220 }, { "epoch": 0.8931691449814126, "grad_norm": 0.35419869207118304, "learning_rate": 3.024813187325965e-07, "loss": 0.0376, "step": 19221 }, { "epoch": 0.8932156133828997, "grad_norm": 0.4306698551875739, "learning_rate": 3.022236070318885e-07, "loss": 0.0453, "step": 19222 }, { "epoch": 0.8932620817843866, "grad_norm": 0.6234476881219003, "learning_rate": 3.0196600174076054e-07, "loss": 0.0442, "step": 19223 }, { "epoch": 0.8933085501858736, "grad_norm": 0.46591174143944536, "learning_rate": 3.017085028650457e-07, "loss": 0.0581, "step": 19224 }, { "epoch": 0.8933550185873605, "grad_norm": 0.34487156146695647, "learning_rate": 3.0145111041057883e-07, "loss": 0.0221, "step": 19225 }, { "epoch": 0.8934014869888476, "grad_norm": 0.5291104643037107, "learning_rate": 3.0119382438318967e-07, "loss": 0.0513, "step": 19226 }, { "epoch": 0.8934479553903346, "grad_norm": 0.44678835597445127, "learning_rate": 3.009366447887058e-07, "loss": 0.0385, "step": 19227 }, { "epoch": 0.8934944237918215, "grad_norm": 0.3794192238938592, "learning_rate": 3.0067957163295135e-07, "loss": 0.0516, "step": 19228 }, { "epoch": 0.8935408921933086, "grad_norm": 0.42550278364681865, "learning_rate": 3.004226049217529e-07, "loss": 0.045, "step": 19229 }, { "epoch": 0.8935873605947955, "grad_norm": 0.2918974186839619, "learning_rate": 3.001657446609274e-07, "loss": 0.0289, "step": 19230 }, { "epoch": 0.8936338289962825, "grad_norm": 0.4397384885020731, "learning_rate": 2.9990899085629354e-07, "loss": 0.0521, "step": 19231 }, { "epoch": 0.8936802973977696, "grad_norm": 0.44531799822802126, "learning_rate": 2.9965234351366835e-07, "loss": 0.0516, "step": 19232 }, { "epoch": 0.8937267657992565, "grad_norm": 0.23394556747848583, "learning_rate": 2.993958026388649e-07, "loss": 0.0429, "step": 19233 }, { "epoch": 0.8937732342007435, "grad_norm": 0.7282672964449588, "learning_rate": 2.9913936823769365e-07, "loss": 0.058, "step": 19234 }, { "epoch": 0.8938197026022305, "grad_norm": 0.5135825709376687, "learning_rate": 2.9888304031596373e-07, "loss": 0.0637, "step": 19235 }, { "epoch": 0.8938661710037175, "grad_norm": 0.3300719639599903, "learning_rate": 2.986268188794805e-07, "loss": 0.0423, "step": 19236 }, { "epoch": 0.8939126394052045, "grad_norm": 0.4381297630707477, "learning_rate": 2.9837070393404774e-07, "loss": 0.0514, "step": 19237 }, { "epoch": 0.8939591078066914, "grad_norm": 0.4075062310364672, "learning_rate": 2.9811469548546634e-07, "loss": 0.0405, "step": 19238 }, { "epoch": 0.8940055762081784, "grad_norm": 0.7064329435314236, "learning_rate": 2.978587935395366e-07, "loss": 0.0589, "step": 19239 }, { "epoch": 0.8940520446096655, "grad_norm": 0.4070098080850313, "learning_rate": 2.9760299810205396e-07, "loss": 0.0507, "step": 19240 }, { "epoch": 0.8940985130111524, "grad_norm": 0.6156818412085214, "learning_rate": 2.973473091788121e-07, "loss": 0.0505, "step": 19241 }, { "epoch": 0.8941449814126394, "grad_norm": 0.3446755996761111, "learning_rate": 2.9709172677560407e-07, "loss": 0.0427, "step": 19242 }, { "epoch": 0.8941914498141263, "grad_norm": 0.4823079223358442, "learning_rate": 2.9683625089821757e-07, "loss": 0.0526, "step": 19243 }, { "epoch": 0.8942379182156134, "grad_norm": 0.5362017379595685, "learning_rate": 2.965808815524401e-07, "loss": 0.0591, "step": 19244 }, { "epoch": 0.8942843866171004, "grad_norm": 0.6101337351662702, "learning_rate": 2.9632561874405496e-07, "loss": 0.0615, "step": 19245 }, { "epoch": 0.8943308550185873, "grad_norm": 0.9907121784462264, "learning_rate": 2.9607046247884685e-07, "loss": 0.0758, "step": 19246 }, { "epoch": 0.8943773234200744, "grad_norm": 0.6755076260833787, "learning_rate": 2.9581541276259286e-07, "loss": 0.0641, "step": 19247 }, { "epoch": 0.8944237918215613, "grad_norm": 0.7712131578462217, "learning_rate": 2.9556046960106997e-07, "loss": 0.0495, "step": 19248 }, { "epoch": 0.8944702602230483, "grad_norm": 0.5667778713549899, "learning_rate": 2.953056330000542e-07, "loss": 0.0626, "step": 19249 }, { "epoch": 0.8945167286245354, "grad_norm": 0.4277956222890122, "learning_rate": 2.950509029653181e-07, "loss": 0.0555, "step": 19250 }, { "epoch": 0.8945631970260223, "grad_norm": 0.35944830220146634, "learning_rate": 2.9479627950263043e-07, "loss": 0.0478, "step": 19251 }, { "epoch": 0.8946096654275093, "grad_norm": 0.5042978243573959, "learning_rate": 2.9454176261775937e-07, "loss": 0.0441, "step": 19252 }, { "epoch": 0.8946561338289963, "grad_norm": 0.4242687135183894, "learning_rate": 2.9428735231646967e-07, "loss": 0.0548, "step": 19253 }, { "epoch": 0.8947026022304833, "grad_norm": 0.4348781499178014, "learning_rate": 2.9403304860452454e-07, "loss": 0.0467, "step": 19254 }, { "epoch": 0.8947490706319703, "grad_norm": 0.5751868842050625, "learning_rate": 2.937788514876827e-07, "loss": 0.0521, "step": 19255 }, { "epoch": 0.8947955390334572, "grad_norm": 0.73923134310881, "learning_rate": 2.9352476097170344e-07, "loss": 0.0516, "step": 19256 }, { "epoch": 0.8948420074349442, "grad_norm": 0.4229445186995742, "learning_rate": 2.932707770623422e-07, "loss": 0.0528, "step": 19257 }, { "epoch": 0.8948884758364313, "grad_norm": 0.5409727066556774, "learning_rate": 2.9301689976535206e-07, "loss": 0.0654, "step": 19258 }, { "epoch": 0.8949349442379182, "grad_norm": 0.637265353760179, "learning_rate": 2.927631290864824e-07, "loss": 0.0494, "step": 19259 }, { "epoch": 0.8949814126394052, "grad_norm": 0.332680629356376, "learning_rate": 2.9250946503148245e-07, "loss": 0.0553, "step": 19260 }, { "epoch": 0.8950278810408921, "grad_norm": 0.5762615026812609, "learning_rate": 2.922559076060977e-07, "loss": 0.0479, "step": 19261 }, { "epoch": 0.8950743494423792, "grad_norm": 0.7023891260582923, "learning_rate": 2.9200245681607065e-07, "loss": 0.054, "step": 19262 }, { "epoch": 0.8951208178438662, "grad_norm": 0.3335123215565597, "learning_rate": 2.91749112667144e-07, "loss": 0.0381, "step": 19263 }, { "epoch": 0.8951672862453531, "grad_norm": 0.4424605653624334, "learning_rate": 2.9149587516505597e-07, "loss": 0.0581, "step": 19264 }, { "epoch": 0.8952137546468402, "grad_norm": 0.4788194425826569, "learning_rate": 2.912427443155402e-07, "loss": 0.045, "step": 19265 }, { "epoch": 0.8952602230483271, "grad_norm": 0.5969943135369805, "learning_rate": 2.909897201243328e-07, "loss": 0.0529, "step": 19266 }, { "epoch": 0.8953066914498141, "grad_norm": 0.4754355778506021, "learning_rate": 2.9073680259716465e-07, "loss": 0.0352, "step": 19267 }, { "epoch": 0.8953531598513012, "grad_norm": 0.6905313199577676, "learning_rate": 2.904839917397645e-07, "loss": 0.0591, "step": 19268 }, { "epoch": 0.8953996282527881, "grad_norm": 0.6112263857810438, "learning_rate": 2.902312875578578e-07, "loss": 0.0422, "step": 19269 }, { "epoch": 0.8954460966542751, "grad_norm": 0.5210126199642082, "learning_rate": 2.899786900571705e-07, "loss": 0.07, "step": 19270 }, { "epoch": 0.8954925650557621, "grad_norm": 0.6468782702674004, "learning_rate": 2.897261992434225e-07, "loss": 0.0607, "step": 19271 }, { "epoch": 0.8955390334572491, "grad_norm": 0.406371313045409, "learning_rate": 2.894738151223331e-07, "loss": 0.0412, "step": 19272 }, { "epoch": 0.8955855018587361, "grad_norm": 0.5304615788842864, "learning_rate": 2.892215376996199e-07, "loss": 0.0471, "step": 19273 }, { "epoch": 0.895631970260223, "grad_norm": 0.5913870405662718, "learning_rate": 2.889693669809973e-07, "loss": 0.0355, "step": 19274 }, { "epoch": 0.89567843866171, "grad_norm": 0.4455610369107356, "learning_rate": 2.8871730297217626e-07, "loss": 0.0439, "step": 19275 }, { "epoch": 0.8957249070631971, "grad_norm": 0.6394466312040662, "learning_rate": 2.884653456788672e-07, "loss": 0.0533, "step": 19276 }, { "epoch": 0.895771375464684, "grad_norm": 0.31495605198485266, "learning_rate": 2.8821349510677664e-07, "loss": 0.0474, "step": 19277 }, { "epoch": 0.895817843866171, "grad_norm": 0.6508684877915174, "learning_rate": 2.879617512616101e-07, "loss": 0.0639, "step": 19278 }, { "epoch": 0.895864312267658, "grad_norm": 0.46787597114596075, "learning_rate": 2.877101141490679e-07, "loss": 0.0551, "step": 19279 }, { "epoch": 0.895910780669145, "grad_norm": 0.468391312672638, "learning_rate": 2.874585837748517e-07, "loss": 0.0356, "step": 19280 }, { "epoch": 0.895957249070632, "grad_norm": 0.6440764481711286, "learning_rate": 2.872071601446591e-07, "loss": 0.0724, "step": 19281 }, { "epoch": 0.8960037174721189, "grad_norm": 0.3741480657337416, "learning_rate": 2.869558432641839e-07, "loss": 0.0396, "step": 19282 }, { "epoch": 0.896050185873606, "grad_norm": 0.5670970453470043, "learning_rate": 2.867046331391199e-07, "loss": 0.0734, "step": 19283 }, { "epoch": 0.8960966542750929, "grad_norm": 0.47772829562624525, "learning_rate": 2.864535297751558e-07, "loss": 0.0488, "step": 19284 }, { "epoch": 0.8961431226765799, "grad_norm": 0.35916255569337, "learning_rate": 2.862025331779805e-07, "loss": 0.0501, "step": 19285 }, { "epoch": 0.896189591078067, "grad_norm": 0.3630728705679009, "learning_rate": 2.859516433532783e-07, "loss": 0.0362, "step": 19286 }, { "epoch": 0.8962360594795539, "grad_norm": 0.5548821090372577, "learning_rate": 2.8570086030673405e-07, "loss": 0.0472, "step": 19287 }, { "epoch": 0.8962825278810409, "grad_norm": 0.807160156497613, "learning_rate": 2.854501840440271e-07, "loss": 0.0543, "step": 19288 }, { "epoch": 0.8963289962825279, "grad_norm": 0.5469379727400423, "learning_rate": 2.851996145708341e-07, "loss": 0.0577, "step": 19289 }, { "epoch": 0.8963754646840149, "grad_norm": 0.43522188295321307, "learning_rate": 2.8494915189283325e-07, "loss": 0.0628, "step": 19290 }, { "epoch": 0.8964219330855019, "grad_norm": 0.4522774532107526, "learning_rate": 2.846987960156961e-07, "loss": 0.0607, "step": 19291 }, { "epoch": 0.8964684014869888, "grad_norm": 0.5443352830072334, "learning_rate": 2.8444854694509417e-07, "loss": 0.0469, "step": 19292 }, { "epoch": 0.8965148698884758, "grad_norm": 0.42079282284108804, "learning_rate": 2.841984046866947e-07, "loss": 0.0362, "step": 19293 }, { "epoch": 0.8965613382899629, "grad_norm": 0.5527797420628573, "learning_rate": 2.839483692461664e-07, "loss": 0.0431, "step": 19294 }, { "epoch": 0.8966078066914498, "grad_norm": 0.4948907796548795, "learning_rate": 2.836984406291704e-07, "loss": 0.036, "step": 19295 }, { "epoch": 0.8966542750929368, "grad_norm": 0.7069129234045838, "learning_rate": 2.8344861884136765e-07, "loss": 0.0703, "step": 19296 }, { "epoch": 0.8967007434944237, "grad_norm": 0.4649526401152232, "learning_rate": 2.8319890388841865e-07, "loss": 0.0496, "step": 19297 }, { "epoch": 0.8967472118959108, "grad_norm": 0.4603690976207041, "learning_rate": 2.829492957759783e-07, "loss": 0.0567, "step": 19298 }, { "epoch": 0.8967936802973978, "grad_norm": 0.5394292878943193, "learning_rate": 2.8269979450970155e-07, "loss": 0.0607, "step": 19299 }, { "epoch": 0.8968401486988847, "grad_norm": 0.5413860655741379, "learning_rate": 2.824504000952394e-07, "loss": 0.0408, "step": 19300 }, { "epoch": 0.8968866171003718, "grad_norm": 0.4482774393548446, "learning_rate": 2.8220111253824066e-07, "loss": 0.0474, "step": 19301 }, { "epoch": 0.8969330855018587, "grad_norm": 0.452123955637979, "learning_rate": 2.81951931844352e-07, "loss": 0.0476, "step": 19302 }, { "epoch": 0.8969795539033457, "grad_norm": 0.4539897459372743, "learning_rate": 2.8170285801921715e-07, "loss": 0.0475, "step": 19303 }, { "epoch": 0.8970260223048327, "grad_norm": 0.6462329379715568, "learning_rate": 2.8145389106847886e-07, "loss": 0.0548, "step": 19304 }, { "epoch": 0.8970724907063197, "grad_norm": 0.44441304252129615, "learning_rate": 2.8120503099777706e-07, "loss": 0.0302, "step": 19305 }, { "epoch": 0.8971189591078067, "grad_norm": 0.6567238943356972, "learning_rate": 2.809562778127467e-07, "loss": 0.0362, "step": 19306 }, { "epoch": 0.8971654275092937, "grad_norm": 0.6333460451047507, "learning_rate": 2.807076315190238e-07, "loss": 0.0573, "step": 19307 }, { "epoch": 0.8972118959107807, "grad_norm": 0.6626190716556493, "learning_rate": 2.8045909212224e-07, "loss": 0.0549, "step": 19308 }, { "epoch": 0.8972583643122677, "grad_norm": 0.42424444782103393, "learning_rate": 2.8021065962802464e-07, "loss": 0.0418, "step": 19309 }, { "epoch": 0.8973048327137546, "grad_norm": 0.3864133920690581, "learning_rate": 2.799623340420049e-07, "loss": 0.0324, "step": 19310 }, { "epoch": 0.8973513011152416, "grad_norm": 0.555339443001113, "learning_rate": 2.797141153698074e-07, "loss": 0.0639, "step": 19311 }, { "epoch": 0.8973977695167287, "grad_norm": 0.613440336004834, "learning_rate": 2.7946600361705266e-07, "loss": 0.0572, "step": 19312 }, { "epoch": 0.8974442379182156, "grad_norm": 0.6567047667302465, "learning_rate": 2.792179987893606e-07, "loss": 0.0634, "step": 19313 }, { "epoch": 0.8974907063197026, "grad_norm": 0.5861478767618333, "learning_rate": 2.7897010089234954e-07, "loss": 0.0566, "step": 19314 }, { "epoch": 0.8975371747211895, "grad_norm": 0.8137928689561774, "learning_rate": 2.78722309931635e-07, "loss": 0.0668, "step": 19315 }, { "epoch": 0.8975836431226766, "grad_norm": 0.508611664394296, "learning_rate": 2.78474625912829e-07, "loss": 0.0591, "step": 19316 }, { "epoch": 0.8976301115241636, "grad_norm": 0.3680300462711451, "learning_rate": 2.782270488415423e-07, "loss": 0.0446, "step": 19317 }, { "epoch": 0.8976765799256505, "grad_norm": 0.43625306021760174, "learning_rate": 2.7797957872338245e-07, "loss": 0.0529, "step": 19318 }, { "epoch": 0.8977230483271376, "grad_norm": 0.6628686730108839, "learning_rate": 2.7773221556395446e-07, "loss": 0.0583, "step": 19319 }, { "epoch": 0.8977695167286245, "grad_norm": 0.44384000273665364, "learning_rate": 2.774849593688617e-07, "loss": 0.0372, "step": 19320 }, { "epoch": 0.8978159851301115, "grad_norm": 0.4591096873843, "learning_rate": 2.7723781014370566e-07, "loss": 0.0497, "step": 19321 }, { "epoch": 0.8978624535315985, "grad_norm": 0.6236654805381584, "learning_rate": 2.7699076789408364e-07, "loss": 0.0362, "step": 19322 }, { "epoch": 0.8979089219330855, "grad_norm": 0.7160391615193639, "learning_rate": 2.767438326255917e-07, "loss": 0.0372, "step": 19323 }, { "epoch": 0.8979553903345725, "grad_norm": 0.7317359360653458, "learning_rate": 2.764970043438231e-07, "loss": 0.0511, "step": 19324 }, { "epoch": 0.8980018587360595, "grad_norm": 0.4605695183605792, "learning_rate": 2.762502830543684e-07, "loss": 0.0587, "step": 19325 }, { "epoch": 0.8980483271375465, "grad_norm": 0.6635721618677585, "learning_rate": 2.76003668762817e-07, "loss": 0.0729, "step": 19326 }, { "epoch": 0.8980947955390335, "grad_norm": 0.49222764717093936, "learning_rate": 2.757571614747534e-07, "loss": 0.06, "step": 19327 }, { "epoch": 0.8981412639405204, "grad_norm": 0.3465693810704418, "learning_rate": 2.7551076119576304e-07, "loss": 0.0607, "step": 19328 }, { "epoch": 0.8981877323420074, "grad_norm": 0.7317479378415999, "learning_rate": 2.75264467931427e-07, "loss": 0.0531, "step": 19329 }, { "epoch": 0.8982342007434945, "grad_norm": 0.5629001601026795, "learning_rate": 2.750182816873226e-07, "loss": 0.0573, "step": 19330 }, { "epoch": 0.8982806691449814, "grad_norm": 0.3784840816256647, "learning_rate": 2.7477220246902693e-07, "loss": 0.0507, "step": 19331 }, { "epoch": 0.8983271375464684, "grad_norm": 0.3743116566507363, "learning_rate": 2.7452623028211443e-07, "loss": 0.0447, "step": 19332 }, { "epoch": 0.8983736059479553, "grad_norm": 0.47504423313280436, "learning_rate": 2.742803651321563e-07, "loss": 0.0688, "step": 19333 }, { "epoch": 0.8984200743494424, "grad_norm": 0.44978611034972404, "learning_rate": 2.740346070247213e-07, "loss": 0.0454, "step": 19334 }, { "epoch": 0.8984665427509294, "grad_norm": 0.41106492302986863, "learning_rate": 2.7378895596537725e-07, "loss": 0.0662, "step": 19335 }, { "epoch": 0.8985130111524163, "grad_norm": 0.47176871621225036, "learning_rate": 2.7354341195968694e-07, "loss": 0.0683, "step": 19336 }, { "epoch": 0.8985594795539034, "grad_norm": 0.6705517103921874, "learning_rate": 2.7329797501321255e-07, "loss": 0.0703, "step": 19337 }, { "epoch": 0.8986059479553903, "grad_norm": 0.4640583327335151, "learning_rate": 2.730526451315141e-07, "loss": 0.0299, "step": 19338 }, { "epoch": 0.8986524163568773, "grad_norm": 0.681289360096702, "learning_rate": 2.728074223201488e-07, "loss": 0.0482, "step": 19339 }, { "epoch": 0.8986988847583643, "grad_norm": 0.382146596886767, "learning_rate": 2.7256230658467e-07, "loss": 0.0476, "step": 19340 }, { "epoch": 0.8987453531598513, "grad_norm": 0.7299038326888698, "learning_rate": 2.72317297930631e-07, "loss": 0.0727, "step": 19341 }, { "epoch": 0.8987918215613383, "grad_norm": 0.39579940640456557, "learning_rate": 2.720723963635807e-07, "loss": 0.0538, "step": 19342 }, { "epoch": 0.8988382899628253, "grad_norm": 0.5155989855531816, "learning_rate": 2.71827601889067e-07, "loss": 0.0601, "step": 19343 }, { "epoch": 0.8988847583643123, "grad_norm": 0.5332823702387908, "learning_rate": 2.7158291451263306e-07, "loss": 0.0416, "step": 19344 }, { "epoch": 0.8989312267657993, "grad_norm": 0.37286105247468615, "learning_rate": 2.71338334239824e-07, "loss": 0.0435, "step": 19345 }, { "epoch": 0.8989776951672862, "grad_norm": 0.4973435182215916, "learning_rate": 2.710938610761782e-07, "loss": 0.0405, "step": 19346 }, { "epoch": 0.8990241635687732, "grad_norm": 0.5257403431736662, "learning_rate": 2.708494950272339e-07, "loss": 0.0514, "step": 19347 }, { "epoch": 0.8990706319702603, "grad_norm": 0.5300343596980722, "learning_rate": 2.7060523609852564e-07, "loss": 0.0459, "step": 19348 }, { "epoch": 0.8991171003717472, "grad_norm": 0.48686102552336274, "learning_rate": 2.703610842955862e-07, "loss": 0.0533, "step": 19349 }, { "epoch": 0.8991635687732342, "grad_norm": 0.49684528208174866, "learning_rate": 2.701170396239461e-07, "loss": 0.0392, "step": 19350 }, { "epoch": 0.8992100371747211, "grad_norm": 0.48305134281509043, "learning_rate": 2.6987310208913264e-07, "loss": 0.0635, "step": 19351 }, { "epoch": 0.8992565055762082, "grad_norm": 0.5780280029254191, "learning_rate": 2.6962927169667254e-07, "loss": 0.0441, "step": 19352 }, { "epoch": 0.8993029739776952, "grad_norm": 0.4815189846569545, "learning_rate": 2.6938554845208854e-07, "loss": 0.058, "step": 19353 }, { "epoch": 0.8993494423791821, "grad_norm": 0.6228714538756607, "learning_rate": 2.6914193236089957e-07, "loss": 0.0614, "step": 19354 }, { "epoch": 0.8993959107806692, "grad_norm": 0.6618921893359568, "learning_rate": 2.688984234286257e-07, "loss": 0.0559, "step": 19355 }, { "epoch": 0.8994423791821561, "grad_norm": 0.41262767371884346, "learning_rate": 2.6865502166078196e-07, "loss": 0.0595, "step": 19356 }, { "epoch": 0.8994888475836431, "grad_norm": 0.3806847326239466, "learning_rate": 2.684117270628811e-07, "loss": 0.047, "step": 19357 }, { "epoch": 0.8995353159851301, "grad_norm": 0.41429760974648294, "learning_rate": 2.681685396404343e-07, "loss": 0.0463, "step": 19358 }, { "epoch": 0.8995817843866171, "grad_norm": 0.571182507866855, "learning_rate": 2.6792545939895165e-07, "loss": 0.0587, "step": 19359 }, { "epoch": 0.8996282527881041, "grad_norm": 0.6041709676328737, "learning_rate": 2.6768248634393703e-07, "loss": 0.0729, "step": 19360 }, { "epoch": 0.8996747211895911, "grad_norm": 0.5087847903174029, "learning_rate": 2.6743962048089376e-07, "loss": 0.0477, "step": 19361 }, { "epoch": 0.899721189591078, "grad_norm": 0.5492195793308343, "learning_rate": 2.671968618153248e-07, "loss": 0.0524, "step": 19362 }, { "epoch": 0.8997676579925651, "grad_norm": 0.5292189253958406, "learning_rate": 2.669542103527284e-07, "loss": 0.0613, "step": 19363 }, { "epoch": 0.899814126394052, "grad_norm": 0.4534238549609554, "learning_rate": 2.6671166609860024e-07, "loss": 0.038, "step": 19364 }, { "epoch": 0.899860594795539, "grad_norm": 0.4914944226212158, "learning_rate": 2.664692290584348e-07, "loss": 0.0435, "step": 19365 }, { "epoch": 0.8999070631970261, "grad_norm": 0.3267346749910893, "learning_rate": 2.662268992377232e-07, "loss": 0.0441, "step": 19366 }, { "epoch": 0.899953531598513, "grad_norm": 0.4886440969514756, "learning_rate": 2.659846766419544e-07, "loss": 0.064, "step": 19367 }, { "epoch": 0.9, "grad_norm": 0.39291070330596073, "learning_rate": 2.6574256127661457e-07, "loss": 0.0486, "step": 19368 }, { "epoch": 0.9000464684014869, "grad_norm": 0.4085572526123646, "learning_rate": 2.655005531471894e-07, "loss": 0.0499, "step": 19369 }, { "epoch": 0.900092936802974, "grad_norm": 0.49900434825663964, "learning_rate": 2.652586522591599e-07, "loss": 0.0461, "step": 19370 }, { "epoch": 0.900139405204461, "grad_norm": 0.34901706599223314, "learning_rate": 2.650168586180041e-07, "loss": 0.0605, "step": 19371 }, { "epoch": 0.9001858736059479, "grad_norm": 0.6144752993820721, "learning_rate": 2.6477517222920026e-07, "loss": 0.0639, "step": 19372 }, { "epoch": 0.900232342007435, "grad_norm": 0.6437351908111978, "learning_rate": 2.645335930982229e-07, "loss": 0.0449, "step": 19373 }, { "epoch": 0.9002788104089219, "grad_norm": 0.5555152330738444, "learning_rate": 2.6429212123054327e-07, "loss": 0.0552, "step": 19374 }, { "epoch": 0.9003252788104089, "grad_norm": 0.39847255873754633, "learning_rate": 2.640507566316308e-07, "loss": 0.0469, "step": 19375 }, { "epoch": 0.900371747211896, "grad_norm": 0.6277076494993097, "learning_rate": 2.638094993069551e-07, "loss": 0.0481, "step": 19376 }, { "epoch": 0.9004182156133829, "grad_norm": 0.5699975905235016, "learning_rate": 2.6356834926197783e-07, "loss": 0.0657, "step": 19377 }, { "epoch": 0.9004646840148699, "grad_norm": 0.41747489037407165, "learning_rate": 2.633273065021613e-07, "loss": 0.0437, "step": 19378 }, { "epoch": 0.9005111524163569, "grad_norm": 0.635333027757721, "learning_rate": 2.6308637103296787e-07, "loss": 0.0622, "step": 19379 }, { "epoch": 0.9005576208178439, "grad_norm": 0.6299619080101936, "learning_rate": 2.6284554285985367e-07, "loss": 0.0416, "step": 19380 }, { "epoch": 0.9006040892193309, "grad_norm": 0.4098108357976082, "learning_rate": 2.626048219882732e-07, "loss": 0.0447, "step": 19381 }, { "epoch": 0.9006505576208178, "grad_norm": 0.6385345569248254, "learning_rate": 2.6236420842367994e-07, "loss": 0.0431, "step": 19382 }, { "epoch": 0.9006970260223048, "grad_norm": 0.5059333351025899, "learning_rate": 2.621237021715234e-07, "loss": 0.0492, "step": 19383 }, { "epoch": 0.9007434944237919, "grad_norm": 0.7817623991337853, "learning_rate": 2.618833032372514e-07, "loss": 0.0636, "step": 19384 }, { "epoch": 0.9007899628252788, "grad_norm": 0.47877892879712636, "learning_rate": 2.616430116263091e-07, "loss": 0.0475, "step": 19385 }, { "epoch": 0.9008364312267658, "grad_norm": 0.7269522477676722, "learning_rate": 2.614028273441399e-07, "loss": 0.0607, "step": 19386 }, { "epoch": 0.9008828996282527, "grad_norm": 0.6270668941306864, "learning_rate": 2.6116275039618443e-07, "loss": 0.041, "step": 19387 }, { "epoch": 0.9009293680297398, "grad_norm": 0.6255516486289404, "learning_rate": 2.6092278078788004e-07, "loss": 0.0384, "step": 19388 }, { "epoch": 0.9009758364312268, "grad_norm": 0.47084106006541143, "learning_rate": 2.6068291852466234e-07, "loss": 0.0691, "step": 19389 }, { "epoch": 0.9010223048327137, "grad_norm": 0.5746639541483591, "learning_rate": 2.604431636119642e-07, "loss": 0.0434, "step": 19390 }, { "epoch": 0.9010687732342008, "grad_norm": 0.7366442918793169, "learning_rate": 2.6020351605521745e-07, "loss": 0.0501, "step": 19391 }, { "epoch": 0.9011152416356877, "grad_norm": 0.44205059000317, "learning_rate": 2.599639758598482e-07, "loss": 0.0522, "step": 19392 }, { "epoch": 0.9011617100371747, "grad_norm": 0.3867414128052488, "learning_rate": 2.59724543031285e-07, "loss": 0.0457, "step": 19393 }, { "epoch": 0.9012081784386617, "grad_norm": 0.4137806808878588, "learning_rate": 2.5948521757495015e-07, "loss": 0.0515, "step": 19394 }, { "epoch": 0.9012546468401487, "grad_norm": 0.3934748706353648, "learning_rate": 2.5924599949626314e-07, "loss": 0.0399, "step": 19395 }, { "epoch": 0.9013011152416357, "grad_norm": 0.4226902263029469, "learning_rate": 2.5900688880064474e-07, "loss": 0.0578, "step": 19396 }, { "epoch": 0.9013475836431227, "grad_norm": 0.35532586188832427, "learning_rate": 2.5876788549350997e-07, "loss": 0.0391, "step": 19397 }, { "epoch": 0.9013940520446097, "grad_norm": 0.5486922050920221, "learning_rate": 2.5852898958027285e-07, "loss": 0.0526, "step": 19398 }, { "epoch": 0.9014405204460967, "grad_norm": 0.39324170356964794, "learning_rate": 2.582902010663435e-07, "loss": 0.0544, "step": 19399 }, { "epoch": 0.9014869888475836, "grad_norm": 0.5693401024028665, "learning_rate": 2.5805151995713263e-07, "loss": 0.0556, "step": 19400 }, { "epoch": 0.9015334572490706, "grad_norm": 0.5217443651972956, "learning_rate": 2.578129462580453e-07, "loss": 0.0606, "step": 19401 }, { "epoch": 0.9015799256505577, "grad_norm": 0.5837129490832405, "learning_rate": 2.575744799744856e-07, "loss": 0.0632, "step": 19402 }, { "epoch": 0.9016263940520446, "grad_norm": 0.45291796278176255, "learning_rate": 2.5733612111185526e-07, "loss": 0.0478, "step": 19403 }, { "epoch": 0.9016728624535316, "grad_norm": 0.46605352184284204, "learning_rate": 2.570978696755538e-07, "loss": 0.0342, "step": 19404 }, { "epoch": 0.9017193308550185, "grad_norm": 0.5359017562683136, "learning_rate": 2.5685972567097704e-07, "loss": 0.0532, "step": 19405 }, { "epoch": 0.9017657992565056, "grad_norm": 0.3762432937595123, "learning_rate": 2.5662168910352004e-07, "loss": 0.0544, "step": 19406 }, { "epoch": 0.9018122676579926, "grad_norm": 0.7226358701716177, "learning_rate": 2.5638375997857346e-07, "loss": 0.0498, "step": 19407 }, { "epoch": 0.9018587360594795, "grad_norm": 0.6090693979596831, "learning_rate": 2.5614593830152743e-07, "loss": 0.0645, "step": 19408 }, { "epoch": 0.9019052044609666, "grad_norm": 0.48722253627519196, "learning_rate": 2.5590822407776825e-07, "loss": 0.063, "step": 19409 }, { "epoch": 0.9019516728624535, "grad_norm": 0.4658092572349532, "learning_rate": 2.556706173126816e-07, "loss": 0.05, "step": 19410 }, { "epoch": 0.9019981412639405, "grad_norm": 0.5449340415313466, "learning_rate": 2.554331180116487e-07, "loss": 0.0572, "step": 19411 }, { "epoch": 0.9020446096654275, "grad_norm": 0.3669007167191673, "learning_rate": 2.5519572618004915e-07, "loss": 0.0417, "step": 19412 }, { "epoch": 0.9020910780669145, "grad_norm": 0.5491993447144229, "learning_rate": 2.5495844182326034e-07, "loss": 0.0815, "step": 19413 }, { "epoch": 0.9021375464684015, "grad_norm": 0.32630500530450385, "learning_rate": 2.547212649466568e-07, "loss": 0.0214, "step": 19414 }, { "epoch": 0.9021840148698885, "grad_norm": 0.4208674052988143, "learning_rate": 2.54484195555611e-07, "loss": 0.0447, "step": 19415 }, { "epoch": 0.9022304832713754, "grad_norm": 0.756671579083072, "learning_rate": 2.5424723365549186e-07, "loss": 0.0598, "step": 19416 }, { "epoch": 0.9022769516728625, "grad_norm": 0.39602685101515395, "learning_rate": 2.5401037925166903e-07, "loss": 0.0615, "step": 19417 }, { "epoch": 0.9023234200743494, "grad_norm": 0.6785337941213224, "learning_rate": 2.5377363234950604e-07, "loss": 0.0506, "step": 19418 }, { "epoch": 0.9023698884758364, "grad_norm": 0.47387300126034404, "learning_rate": 2.5353699295436465e-07, "loss": 0.0579, "step": 19419 }, { "epoch": 0.9024163568773235, "grad_norm": 0.6527065387790054, "learning_rate": 2.533004610716067e-07, "loss": 0.0441, "step": 19420 }, { "epoch": 0.9024628252788104, "grad_norm": 0.41062280262849593, "learning_rate": 2.5306403670658895e-07, "loss": 0.0365, "step": 19421 }, { "epoch": 0.9025092936802974, "grad_norm": 0.3876010012918178, "learning_rate": 2.528277198646667e-07, "loss": 0.0559, "step": 19422 }, { "epoch": 0.9025557620817843, "grad_norm": 0.5014853324932175, "learning_rate": 2.525915105511928e-07, "loss": 0.066, "step": 19423 }, { "epoch": 0.9026022304832714, "grad_norm": 0.4177164071874877, "learning_rate": 2.52355408771518e-07, "loss": 0.0552, "step": 19424 }, { "epoch": 0.9026486988847584, "grad_norm": 0.32346031296847205, "learning_rate": 2.521194145309902e-07, "loss": 0.034, "step": 19425 }, { "epoch": 0.9026951672862453, "grad_norm": 0.5721748281959355, "learning_rate": 2.518835278349535e-07, "loss": 0.0541, "step": 19426 }, { "epoch": 0.9027416356877324, "grad_norm": 0.7016295955805939, "learning_rate": 2.516477486887531e-07, "loss": 0.0439, "step": 19427 }, { "epoch": 0.9027881040892193, "grad_norm": 0.44672677073722294, "learning_rate": 2.514120770977291e-07, "loss": 0.0338, "step": 19428 }, { "epoch": 0.9028345724907063, "grad_norm": 0.49291295393564116, "learning_rate": 2.511765130672189e-07, "loss": 0.0638, "step": 19429 }, { "epoch": 0.9028810408921933, "grad_norm": 0.7443123052336165, "learning_rate": 2.5094105660255887e-07, "loss": 0.0424, "step": 19430 }, { "epoch": 0.9029275092936803, "grad_norm": 0.5581493120466763, "learning_rate": 2.507057077090824e-07, "loss": 0.054, "step": 19431 }, { "epoch": 0.9029739776951673, "grad_norm": 0.7573558697433193, "learning_rate": 2.504704663921198e-07, "loss": 0.0428, "step": 19432 }, { "epoch": 0.9030204460966543, "grad_norm": 0.6819057443620486, "learning_rate": 2.502353326569995e-07, "loss": 0.0614, "step": 19433 }, { "epoch": 0.9030669144981412, "grad_norm": 0.2598248350444523, "learning_rate": 2.5000030650904837e-07, "loss": 0.0274, "step": 19434 }, { "epoch": 0.9031133828996283, "grad_norm": 0.5497436136767367, "learning_rate": 2.4976538795359053e-07, "loss": 0.0532, "step": 19435 }, { "epoch": 0.9031598513011152, "grad_norm": 0.40625604196347154, "learning_rate": 2.4953057699594495e-07, "loss": 0.0313, "step": 19436 }, { "epoch": 0.9032063197026022, "grad_norm": 0.34697350206795197, "learning_rate": 2.4929587364143195e-07, "loss": 0.0399, "step": 19437 }, { "epoch": 0.9032527881040893, "grad_norm": 0.570849986873449, "learning_rate": 2.490612778953677e-07, "loss": 0.0761, "step": 19438 }, { "epoch": 0.9032992565055762, "grad_norm": 0.5015910588253741, "learning_rate": 2.488267897630653e-07, "loss": 0.0479, "step": 19439 }, { "epoch": 0.9033457249070632, "grad_norm": 0.36269340855377047, "learning_rate": 2.4859240924983653e-07, "loss": 0.0535, "step": 19440 }, { "epoch": 0.9033921933085501, "grad_norm": 0.7677117468891044, "learning_rate": 2.483581363609916e-07, "loss": 0.0616, "step": 19441 }, { "epoch": 0.9034386617100372, "grad_norm": 0.23368393646812283, "learning_rate": 2.4812397110183515e-07, "loss": 0.0316, "step": 19442 }, { "epoch": 0.9034851301115242, "grad_norm": 0.5126521217923177, "learning_rate": 2.4788991347767135e-07, "loss": 0.0534, "step": 19443 }, { "epoch": 0.9035315985130111, "grad_norm": 0.2822655738117126, "learning_rate": 2.4765596349380306e-07, "loss": 0.0279, "step": 19444 }, { "epoch": 0.9035780669144982, "grad_norm": 0.4862302036512264, "learning_rate": 2.4742212115552943e-07, "loss": 0.0372, "step": 19445 }, { "epoch": 0.9036245353159851, "grad_norm": 0.3997575320871633, "learning_rate": 2.4718838646814627e-07, "loss": 0.068, "step": 19446 }, { "epoch": 0.9036710037174721, "grad_norm": 0.3930394502838623, "learning_rate": 2.469547594369487e-07, "loss": 0.0414, "step": 19447 }, { "epoch": 0.9037174721189591, "grad_norm": 0.44941452539461385, "learning_rate": 2.467212400672281e-07, "loss": 0.0611, "step": 19448 }, { "epoch": 0.9037639405204461, "grad_norm": 0.36653292774370977, "learning_rate": 2.464878283642741e-07, "loss": 0.0497, "step": 19449 }, { "epoch": 0.9038104089219331, "grad_norm": 0.34540540657847174, "learning_rate": 2.4625452433337303e-07, "loss": 0.0393, "step": 19450 }, { "epoch": 0.9038568773234201, "grad_norm": 0.4952348258828836, "learning_rate": 2.4602132797981126e-07, "loss": 0.0557, "step": 19451 }, { "epoch": 0.903903345724907, "grad_norm": 0.6874780869320443, "learning_rate": 2.457882393088695e-07, "loss": 0.0676, "step": 19452 }, { "epoch": 0.9039498141263941, "grad_norm": 0.6563113341015119, "learning_rate": 2.4555525832582805e-07, "loss": 0.0621, "step": 19453 }, { "epoch": 0.903996282527881, "grad_norm": 0.6097789408224643, "learning_rate": 2.453223850359643e-07, "loss": 0.0402, "step": 19454 }, { "epoch": 0.904042750929368, "grad_norm": 0.48480799118701307, "learning_rate": 2.4508961944455235e-07, "loss": 0.0416, "step": 19455 }, { "epoch": 0.9040892193308551, "grad_norm": 0.5026264197156785, "learning_rate": 2.448569615568647e-07, "loss": 0.0286, "step": 19456 }, { "epoch": 0.904135687732342, "grad_norm": 0.34030849453230905, "learning_rate": 2.446244113781715e-07, "loss": 0.0409, "step": 19457 }, { "epoch": 0.904182156133829, "grad_norm": 0.7381948303386302, "learning_rate": 2.443919689137403e-07, "loss": 0.0444, "step": 19458 }, { "epoch": 0.9042286245353159, "grad_norm": 0.3255765257883542, "learning_rate": 2.4415963416883736e-07, "loss": 0.0425, "step": 19459 }, { "epoch": 0.904275092936803, "grad_norm": 0.5357013630914497, "learning_rate": 2.43927407148723e-07, "loss": 0.0538, "step": 19460 }, { "epoch": 0.90432156133829, "grad_norm": 0.359392504383353, "learning_rate": 2.4369528785865847e-07, "loss": 0.0405, "step": 19461 }, { "epoch": 0.9043680297397769, "grad_norm": 0.6015218334363408, "learning_rate": 2.434632763039024e-07, "loss": 0.0797, "step": 19462 }, { "epoch": 0.904414498141264, "grad_norm": 0.6550849325878919, "learning_rate": 2.4323137248970896e-07, "loss": 0.0307, "step": 19463 }, { "epoch": 0.9044609665427509, "grad_norm": 0.4494217920183978, "learning_rate": 2.4299957642133054e-07, "loss": 0.054, "step": 19464 }, { "epoch": 0.9045074349442379, "grad_norm": 0.5024136568916828, "learning_rate": 2.4276788810401963e-07, "loss": 0.0348, "step": 19465 }, { "epoch": 0.9045539033457249, "grad_norm": 0.7662034693957076, "learning_rate": 2.425363075430226e-07, "loss": 0.0441, "step": 19466 }, { "epoch": 0.9046003717472119, "grad_norm": 0.5800428601996129, "learning_rate": 2.4230483474358413e-07, "loss": 0.0636, "step": 19467 }, { "epoch": 0.9046468401486989, "grad_norm": 0.3449014562833272, "learning_rate": 2.4207346971095004e-07, "loss": 0.0349, "step": 19468 }, { "epoch": 0.9046933085501859, "grad_norm": 0.6287926403910505, "learning_rate": 2.4184221245035887e-07, "loss": 0.0588, "step": 19469 }, { "epoch": 0.9047397769516728, "grad_norm": 0.5638404158993887, "learning_rate": 2.4161106296704983e-07, "loss": 0.0579, "step": 19470 }, { "epoch": 0.9047862453531599, "grad_norm": 3.6720851764973665, "learning_rate": 2.413800212662576e-07, "loss": 0.0428, "step": 19471 }, { "epoch": 0.9048327137546468, "grad_norm": 0.34763372513384383, "learning_rate": 2.4114908735321685e-07, "loss": 0.0328, "step": 19472 }, { "epoch": 0.9048791821561338, "grad_norm": 0.7020689098852942, "learning_rate": 2.4091826123315784e-07, "loss": 0.0456, "step": 19473 }, { "epoch": 0.9049256505576209, "grad_norm": 0.30726083062391885, "learning_rate": 2.406875429113087e-07, "loss": 0.0477, "step": 19474 }, { "epoch": 0.9049721189591078, "grad_norm": 0.2538377971261476, "learning_rate": 2.404569323928957e-07, "loss": 0.0277, "step": 19475 }, { "epoch": 0.9050185873605948, "grad_norm": 0.5305244183213041, "learning_rate": 2.40226429683143e-07, "loss": 0.0484, "step": 19476 }, { "epoch": 0.9050650557620817, "grad_norm": 0.48814496200631785, "learning_rate": 2.3999603478727095e-07, "loss": 0.0363, "step": 19477 }, { "epoch": 0.9051115241635688, "grad_norm": 0.6106105316681883, "learning_rate": 2.397657477104992e-07, "loss": 0.0478, "step": 19478 }, { "epoch": 0.9051579925650558, "grad_norm": 0.5142132398770132, "learning_rate": 2.395355684580425e-07, "loss": 0.0642, "step": 19479 }, { "epoch": 0.9052044609665427, "grad_norm": 0.7960018123941784, "learning_rate": 2.393054970351161e-07, "loss": 0.0511, "step": 19480 }, { "epoch": 0.9052509293680298, "grad_norm": 0.4466482561014893, "learning_rate": 2.3907553344692967e-07, "loss": 0.0547, "step": 19481 }, { "epoch": 0.9052973977695167, "grad_norm": 0.3719971766706546, "learning_rate": 2.388456776986947e-07, "loss": 0.0398, "step": 19482 }, { "epoch": 0.9053438661710037, "grad_norm": 0.5252232661882645, "learning_rate": 2.386159297956153e-07, "loss": 0.0524, "step": 19483 }, { "epoch": 0.9053903345724907, "grad_norm": 0.6767216386654769, "learning_rate": 2.383862897428957e-07, "loss": 0.0519, "step": 19484 }, { "epoch": 0.9054368029739777, "grad_norm": 0.6589301960200885, "learning_rate": 2.3815675754573885e-07, "loss": 0.0485, "step": 19485 }, { "epoch": 0.9054832713754647, "grad_norm": 0.5642540561399622, "learning_rate": 2.3792733320934348e-07, "loss": 0.0557, "step": 19486 }, { "epoch": 0.9055297397769517, "grad_norm": 0.9772038293036885, "learning_rate": 2.376980167389059e-07, "loss": 0.0557, "step": 19487 }, { "epoch": 0.9055762081784386, "grad_norm": 0.6933688791998632, "learning_rate": 2.3746880813962037e-07, "loss": 0.067, "step": 19488 }, { "epoch": 0.9056226765799257, "grad_norm": 0.6889994149527108, "learning_rate": 2.372397074166788e-07, "loss": 0.0552, "step": 19489 }, { "epoch": 0.9056691449814126, "grad_norm": 0.7823971349381943, "learning_rate": 2.3701071457527037e-07, "loss": 0.0631, "step": 19490 }, { "epoch": 0.9057156133828996, "grad_norm": 0.32897876280458566, "learning_rate": 2.3678182962058205e-07, "loss": 0.0447, "step": 19491 }, { "epoch": 0.9057620817843867, "grad_norm": 0.39803171047360353, "learning_rate": 2.3655305255779858e-07, "loss": 0.0637, "step": 19492 }, { "epoch": 0.9058085501858736, "grad_norm": 0.3827047095051008, "learning_rate": 2.3632438339210196e-07, "loss": 0.0375, "step": 19493 }, { "epoch": 0.9058550185873606, "grad_norm": 0.5726024256256413, "learning_rate": 2.360958221286719e-07, "loss": 0.0499, "step": 19494 }, { "epoch": 0.9059014869888475, "grad_norm": 0.44815236119839813, "learning_rate": 2.358673687726859e-07, "loss": 0.0455, "step": 19495 }, { "epoch": 0.9059479553903346, "grad_norm": 0.4538515880114246, "learning_rate": 2.3563902332931766e-07, "loss": 0.0587, "step": 19496 }, { "epoch": 0.9059944237918216, "grad_norm": 0.5678127851072517, "learning_rate": 2.3541078580373967e-07, "loss": 0.0537, "step": 19497 }, { "epoch": 0.9060408921933085, "grad_norm": 0.3887008079049204, "learning_rate": 2.351826562011217e-07, "loss": 0.0507, "step": 19498 }, { "epoch": 0.9060873605947956, "grad_norm": 0.5292520787225135, "learning_rate": 2.3495463452663236e-07, "loss": 0.0538, "step": 19499 }, { "epoch": 0.9061338289962825, "grad_norm": 0.31530356172098956, "learning_rate": 2.347267207854359e-07, "loss": 0.0274, "step": 19500 }, { "epoch": 0.9061802973977695, "grad_norm": 0.8852772797115924, "learning_rate": 2.3449891498269318e-07, "loss": 0.0603, "step": 19501 }, { "epoch": 0.9062267657992565, "grad_norm": 0.4198764996066736, "learning_rate": 2.3427121712356672e-07, "loss": 0.0568, "step": 19502 }, { "epoch": 0.9062732342007435, "grad_norm": 0.4522017207678236, "learning_rate": 2.3404362721321293e-07, "loss": 0.0309, "step": 19503 }, { "epoch": 0.9063197026022305, "grad_norm": 0.6018677252304789, "learning_rate": 2.3381614525678664e-07, "loss": 0.0385, "step": 19504 }, { "epoch": 0.9063661710037175, "grad_norm": 0.4037478321066326, "learning_rate": 2.335887712594409e-07, "loss": 0.034, "step": 19505 }, { "epoch": 0.9064126394052044, "grad_norm": 0.8877762449841131, "learning_rate": 2.3336150522632662e-07, "loss": 0.0406, "step": 19506 }, { "epoch": 0.9064591078066915, "grad_norm": 0.550866116699938, "learning_rate": 2.331343471625913e-07, "loss": 0.04, "step": 19507 }, { "epoch": 0.9065055762081784, "grad_norm": 0.687148047705302, "learning_rate": 2.3290729707337868e-07, "loss": 0.042, "step": 19508 }, { "epoch": 0.9065520446096654, "grad_norm": 0.22538884228976339, "learning_rate": 2.32680354963834e-07, "loss": 0.0262, "step": 19509 }, { "epoch": 0.9065985130111525, "grad_norm": 0.7434024529332978, "learning_rate": 2.3245352083909657e-07, "loss": 0.0479, "step": 19510 }, { "epoch": 0.9066449814126394, "grad_norm": 0.43119410114073337, "learning_rate": 2.3222679470430498e-07, "loss": 0.047, "step": 19511 }, { "epoch": 0.9066914498141264, "grad_norm": 0.7126720481575168, "learning_rate": 2.3200017656459462e-07, "loss": 0.0583, "step": 19512 }, { "epoch": 0.9067379182156133, "grad_norm": 0.4353504695822595, "learning_rate": 2.3177366642509802e-07, "loss": 0.0503, "step": 19513 }, { "epoch": 0.9067843866171004, "grad_norm": 0.3481095908174838, "learning_rate": 2.3154726429094664e-07, "loss": 0.0534, "step": 19514 }, { "epoch": 0.9068308550185874, "grad_norm": 0.39802194683843445, "learning_rate": 2.3132097016726752e-07, "loss": 0.032, "step": 19515 }, { "epoch": 0.9068773234200743, "grad_norm": 0.6277728820669572, "learning_rate": 2.3109478405918818e-07, "loss": 0.0391, "step": 19516 }, { "epoch": 0.9069237918215614, "grad_norm": 0.5515207684058377, "learning_rate": 2.308687059718312e-07, "loss": 0.0497, "step": 19517 }, { "epoch": 0.9069702602230483, "grad_norm": 0.32658569204847565, "learning_rate": 2.306427359103175e-07, "loss": 0.0592, "step": 19518 }, { "epoch": 0.9070167286245353, "grad_norm": 0.5108003274972148, "learning_rate": 2.304168738797652e-07, "loss": 0.0482, "step": 19519 }, { "epoch": 0.9070631970260223, "grad_norm": 0.4325121322471639, "learning_rate": 2.3019111988529076e-07, "loss": 0.0394, "step": 19520 }, { "epoch": 0.9071096654275093, "grad_norm": 0.4051599197544436, "learning_rate": 2.2996547393200785e-07, "loss": 0.0451, "step": 19521 }, { "epoch": 0.9071561338289963, "grad_norm": 0.5893397285214342, "learning_rate": 2.297399360250263e-07, "loss": 0.0599, "step": 19522 }, { "epoch": 0.9072026022304833, "grad_norm": 0.5672947219229888, "learning_rate": 2.2951450616945647e-07, "loss": 0.0233, "step": 19523 }, { "epoch": 0.9072490706319702, "grad_norm": 0.6438566959200916, "learning_rate": 2.2928918437040482e-07, "loss": 0.0592, "step": 19524 }, { "epoch": 0.9072955390334573, "grad_norm": 0.4729240732967492, "learning_rate": 2.290639706329728e-07, "loss": 0.0579, "step": 19525 }, { "epoch": 0.9073420074349442, "grad_norm": 0.457026335676729, "learning_rate": 2.2883886496226414e-07, "loss": 0.031, "step": 19526 }, { "epoch": 0.9073884758364312, "grad_norm": 0.5559260376525417, "learning_rate": 2.2861386736337643e-07, "loss": 0.0663, "step": 19527 }, { "epoch": 0.9074349442379183, "grad_norm": 0.4475445919488699, "learning_rate": 2.2838897784140612e-07, "loss": 0.0589, "step": 19528 }, { "epoch": 0.9074814126394052, "grad_norm": 0.3795400691818813, "learning_rate": 2.2816419640144748e-07, "loss": 0.0555, "step": 19529 }, { "epoch": 0.9075278810408922, "grad_norm": 0.5133743423523466, "learning_rate": 2.2793952304859256e-07, "loss": 0.0328, "step": 19530 }, { "epoch": 0.9075743494423791, "grad_norm": 0.33916436226814445, "learning_rate": 2.2771495778793006e-07, "loss": 0.0504, "step": 19531 }, { "epoch": 0.9076208178438662, "grad_norm": 0.559500301990987, "learning_rate": 2.2749050062454537e-07, "loss": 0.0498, "step": 19532 }, { "epoch": 0.9076672862453532, "grad_norm": 0.39076719418422895, "learning_rate": 2.2726615156352494e-07, "loss": 0.0488, "step": 19533 }, { "epoch": 0.9077137546468401, "grad_norm": 0.5104768016125673, "learning_rate": 2.2704191060994919e-07, "loss": 0.0509, "step": 19534 }, { "epoch": 0.9077602230483272, "grad_norm": 0.8687949003977883, "learning_rate": 2.2681777776889736e-07, "loss": 0.0453, "step": 19535 }, { "epoch": 0.9078066914498141, "grad_norm": 0.46515530895709867, "learning_rate": 2.265937530454465e-07, "loss": 0.0381, "step": 19536 }, { "epoch": 0.9078531598513011, "grad_norm": 0.4770875150003507, "learning_rate": 2.263698364446715e-07, "loss": 0.0485, "step": 19537 }, { "epoch": 0.9078996282527881, "grad_norm": 0.3582267589088023, "learning_rate": 2.261460279716432e-07, "loss": 0.0394, "step": 19538 }, { "epoch": 0.9079460966542751, "grad_norm": 0.6666428297246314, "learning_rate": 2.259223276314315e-07, "loss": 0.0543, "step": 19539 }, { "epoch": 0.9079925650557621, "grad_norm": 0.4137324309766738, "learning_rate": 2.256987354291046e-07, "loss": 0.0496, "step": 19540 }, { "epoch": 0.9080390334572491, "grad_norm": 0.4983910561421786, "learning_rate": 2.254752513697267e-07, "loss": 0.0434, "step": 19541 }, { "epoch": 0.908085501858736, "grad_norm": 0.5987922694020128, "learning_rate": 2.2525187545835825e-07, "loss": 0.0487, "step": 19542 }, { "epoch": 0.9081319702602231, "grad_norm": 0.42395003006842047, "learning_rate": 2.2502860770006018e-07, "loss": 0.0546, "step": 19543 }, { "epoch": 0.90817843866171, "grad_norm": 0.8020897386701236, "learning_rate": 2.2480544809989014e-07, "loss": 0.0662, "step": 19544 }, { "epoch": 0.908224907063197, "grad_norm": 0.584350339737306, "learning_rate": 2.2458239666290294e-07, "loss": 0.0515, "step": 19545 }, { "epoch": 0.9082713754646841, "grad_norm": 0.43038782840631107, "learning_rate": 2.2435945339414956e-07, "loss": 0.0475, "step": 19546 }, { "epoch": 0.908317843866171, "grad_norm": 0.5000619658159801, "learning_rate": 2.241366182986815e-07, "loss": 0.0597, "step": 19547 }, { "epoch": 0.908364312267658, "grad_norm": 0.49717627557663246, "learning_rate": 2.2391389138154584e-07, "loss": 0.0468, "step": 19548 }, { "epoch": 0.9084107806691449, "grad_norm": 0.47160129033124615, "learning_rate": 2.2369127264778578e-07, "loss": 0.054, "step": 19549 }, { "epoch": 0.908457249070632, "grad_norm": 0.388566190863005, "learning_rate": 2.2346876210244671e-07, "loss": 0.0542, "step": 19550 }, { "epoch": 0.908503717472119, "grad_norm": 0.7923581319676676, "learning_rate": 2.2324635975056684e-07, "loss": 0.0524, "step": 19551 }, { "epoch": 0.9085501858736059, "grad_norm": 0.36037022915443023, "learning_rate": 2.2302406559718493e-07, "loss": 0.0426, "step": 19552 }, { "epoch": 0.908596654275093, "grad_norm": 0.4265459053532413, "learning_rate": 2.2280187964733524e-07, "loss": 0.0609, "step": 19553 }, { "epoch": 0.9086431226765799, "grad_norm": 0.6188373767149187, "learning_rate": 2.22579801906051e-07, "loss": 0.0615, "step": 19554 }, { "epoch": 0.9086895910780669, "grad_norm": 0.4552019322680953, "learning_rate": 2.2235783237836206e-07, "loss": 0.0403, "step": 19555 }, { "epoch": 0.9087360594795539, "grad_norm": 0.48649978064094507, "learning_rate": 2.2213597106929608e-07, "loss": 0.0506, "step": 19556 }, { "epoch": 0.9087825278810409, "grad_norm": 0.4062214857764037, "learning_rate": 2.2191421798388012e-07, "loss": 0.0343, "step": 19557 }, { "epoch": 0.9088289962825279, "grad_norm": 0.38651958008773296, "learning_rate": 2.216925731271352e-07, "loss": 0.0495, "step": 19558 }, { "epoch": 0.9088754646840149, "grad_norm": 0.49697878391459166, "learning_rate": 2.2147103650408343e-07, "loss": 0.0642, "step": 19559 }, { "epoch": 0.9089219330855018, "grad_norm": 0.5494866308598891, "learning_rate": 2.2124960811974127e-07, "loss": 0.0498, "step": 19560 }, { "epoch": 0.9089684014869889, "grad_norm": 0.5798791457139032, "learning_rate": 2.2102828797912535e-07, "loss": 0.0369, "step": 19561 }, { "epoch": 0.9090148698884758, "grad_norm": 0.43792694315031894, "learning_rate": 2.2080707608724827e-07, "loss": 0.0432, "step": 19562 }, { "epoch": 0.9090613382899628, "grad_norm": 0.6327167687443487, "learning_rate": 2.205859724491205e-07, "loss": 0.0425, "step": 19563 }, { "epoch": 0.9091078066914499, "grad_norm": 0.571533632006556, "learning_rate": 2.2036497706975136e-07, "loss": 0.0575, "step": 19564 }, { "epoch": 0.9091542750929368, "grad_norm": 0.6091589690281152, "learning_rate": 2.2014408995414682e-07, "loss": 0.0639, "step": 19565 }, { "epoch": 0.9092007434944238, "grad_norm": 0.6766726736827792, "learning_rate": 2.1992331110730737e-07, "loss": 0.0435, "step": 19566 }, { "epoch": 0.9092472118959107, "grad_norm": 0.5723246240246701, "learning_rate": 2.197026405342373e-07, "loss": 0.038, "step": 19567 }, { "epoch": 0.9092936802973978, "grad_norm": 0.523453777405366, "learning_rate": 2.1948207823993317e-07, "loss": 0.0635, "step": 19568 }, { "epoch": 0.9093401486988848, "grad_norm": 0.7379016306297596, "learning_rate": 2.19261624229391e-07, "loss": 0.076, "step": 19569 }, { "epoch": 0.9093866171003717, "grad_norm": 0.799479639373279, "learning_rate": 2.190412785076046e-07, "loss": 0.0728, "step": 19570 }, { "epoch": 0.9094330855018588, "grad_norm": 0.28410541940540873, "learning_rate": 2.1882104107956603e-07, "loss": 0.0316, "step": 19571 }, { "epoch": 0.9094795539033457, "grad_norm": 0.5105066081136771, "learning_rate": 2.18600911950263e-07, "loss": 0.0293, "step": 19572 }, { "epoch": 0.9095260223048327, "grad_norm": 0.45265312750929093, "learning_rate": 2.1838089112468042e-07, "loss": 0.0499, "step": 19573 }, { "epoch": 0.9095724907063197, "grad_norm": 0.5439028222785002, "learning_rate": 2.1816097860780373e-07, "loss": 0.0632, "step": 19574 }, { "epoch": 0.9096189591078067, "grad_norm": 0.3485150754385095, "learning_rate": 2.1794117440461393e-07, "loss": 0.0406, "step": 19575 }, { "epoch": 0.9096654275092937, "grad_norm": 0.3645006969863268, "learning_rate": 2.1772147852008986e-07, "loss": 0.0633, "step": 19576 }, { "epoch": 0.9097118959107807, "grad_norm": 0.6907176985546614, "learning_rate": 2.1750189095920693e-07, "loss": 0.0456, "step": 19577 }, { "epoch": 0.9097583643122676, "grad_norm": 0.4893439127831797, "learning_rate": 2.172824117269401e-07, "loss": 0.0753, "step": 19578 }, { "epoch": 0.9098048327137547, "grad_norm": 0.8301072538371562, "learning_rate": 2.1706304082826034e-07, "loss": 0.0539, "step": 19579 }, { "epoch": 0.9098513011152416, "grad_norm": 0.5458940538842444, "learning_rate": 2.1684377826813595e-07, "loss": 0.041, "step": 19580 }, { "epoch": 0.9098977695167286, "grad_norm": 0.5104763695414037, "learning_rate": 2.166246240515346e-07, "loss": 0.0337, "step": 19581 }, { "epoch": 0.9099442379182157, "grad_norm": 0.4290845269398825, "learning_rate": 2.164055781834201e-07, "loss": 0.0442, "step": 19582 }, { "epoch": 0.9099907063197026, "grad_norm": 0.5437039107003995, "learning_rate": 2.1618664066875405e-07, "loss": 0.0602, "step": 19583 }, { "epoch": 0.9100371747211896, "grad_norm": 0.6458064427930701, "learning_rate": 2.1596781151249524e-07, "loss": 0.0599, "step": 19584 }, { "epoch": 0.9100836431226765, "grad_norm": 0.6014029970226951, "learning_rate": 2.1574909071960083e-07, "loss": 0.0589, "step": 19585 }, { "epoch": 0.9101301115241636, "grad_norm": 0.6134190431743376, "learning_rate": 2.155304782950246e-07, "loss": 0.0468, "step": 19586 }, { "epoch": 0.9101765799256506, "grad_norm": 0.8525198417034885, "learning_rate": 2.153119742437182e-07, "loss": 0.0594, "step": 19587 }, { "epoch": 0.9102230483271375, "grad_norm": 0.5363153855241265, "learning_rate": 2.1509357857063152e-07, "loss": 0.0566, "step": 19588 }, { "epoch": 0.9102695167286246, "grad_norm": 0.37047487619310693, "learning_rate": 2.1487529128071227e-07, "loss": 0.028, "step": 19589 }, { "epoch": 0.9103159851301115, "grad_norm": 0.4556519562365936, "learning_rate": 2.1465711237890264e-07, "loss": 0.0429, "step": 19590 }, { "epoch": 0.9103624535315985, "grad_norm": 0.6121827781635965, "learning_rate": 2.144390418701464e-07, "loss": 0.0496, "step": 19591 }, { "epoch": 0.9104089219330855, "grad_norm": 0.3875168242297135, "learning_rate": 2.14221079759383e-07, "loss": 0.0382, "step": 19592 }, { "epoch": 0.9104553903345725, "grad_norm": 0.5099219279084811, "learning_rate": 2.14003226051549e-07, "loss": 0.0338, "step": 19593 }, { "epoch": 0.9105018587360595, "grad_norm": 0.6583182426529793, "learning_rate": 2.1378548075157824e-07, "loss": 0.0375, "step": 19594 }, { "epoch": 0.9105483271375465, "grad_norm": 0.5890140293859473, "learning_rate": 2.135678438644051e-07, "loss": 0.0607, "step": 19595 }, { "epoch": 0.9105947955390334, "grad_norm": 0.38008720158890674, "learning_rate": 2.1335031539495677e-07, "loss": 0.0444, "step": 19596 }, { "epoch": 0.9106412639405205, "grad_norm": 0.42930752038937614, "learning_rate": 2.131328953481615e-07, "loss": 0.0378, "step": 19597 }, { "epoch": 0.9106877323420074, "grad_norm": 0.6555497836751791, "learning_rate": 2.1291558372894482e-07, "loss": 0.0533, "step": 19598 }, { "epoch": 0.9107342007434944, "grad_norm": 0.5956315807855608, "learning_rate": 2.126983805422289e-07, "loss": 0.0506, "step": 19599 }, { "epoch": 0.9107806691449815, "grad_norm": 0.5298212177645932, "learning_rate": 2.124812857929326e-07, "loss": 0.0268, "step": 19600 }, { "epoch": 0.9108271375464684, "grad_norm": 0.7290065934811893, "learning_rate": 2.1226429948597417e-07, "loss": 0.0603, "step": 19601 }, { "epoch": 0.9108736059479554, "grad_norm": 0.6174290797479267, "learning_rate": 2.1204742162626856e-07, "loss": 0.0553, "step": 19602 }, { "epoch": 0.9109200743494423, "grad_norm": 0.48027886589507174, "learning_rate": 2.1183065221872802e-07, "loss": 0.0282, "step": 19603 }, { "epoch": 0.9109665427509294, "grad_norm": 0.579116254962248, "learning_rate": 2.1161399126826187e-07, "loss": 0.0463, "step": 19604 }, { "epoch": 0.9110130111524164, "grad_norm": 0.5631529628422984, "learning_rate": 2.1139743877977903e-07, "loss": 0.0481, "step": 19605 }, { "epoch": 0.9110594795539033, "grad_norm": 0.3075082261319833, "learning_rate": 2.1118099475818498e-07, "loss": 0.0366, "step": 19606 }, { "epoch": 0.9111059479553903, "grad_norm": 0.36959026501476183, "learning_rate": 2.1096465920837973e-07, "loss": 0.0384, "step": 19607 }, { "epoch": 0.9111524163568773, "grad_norm": 0.4625557814471314, "learning_rate": 2.107484321352665e-07, "loss": 0.03, "step": 19608 }, { "epoch": 0.9111988847583643, "grad_norm": 0.5677367573610174, "learning_rate": 2.1053231354374205e-07, "loss": 0.0536, "step": 19609 }, { "epoch": 0.9112453531598513, "grad_norm": 0.5030957782344706, "learning_rate": 2.1031630343870124e-07, "loss": 0.0455, "step": 19610 }, { "epoch": 0.9112918215613383, "grad_norm": 0.5475115923768984, "learning_rate": 2.1010040182503633e-07, "loss": 0.0498, "step": 19611 }, { "epoch": 0.9113382899628253, "grad_norm": 0.5138838065657553, "learning_rate": 2.0988460870764005e-07, "loss": 0.0346, "step": 19612 }, { "epoch": 0.9113847583643123, "grad_norm": 0.4701501470100438, "learning_rate": 2.096689240913985e-07, "loss": 0.0341, "step": 19613 }, { "epoch": 0.9114312267657992, "grad_norm": 0.5468455044181951, "learning_rate": 2.0945334798119664e-07, "loss": 0.0372, "step": 19614 }, { "epoch": 0.9114776951672863, "grad_norm": 0.4634681643354719, "learning_rate": 2.0923788038191895e-07, "loss": 0.0379, "step": 19615 }, { "epoch": 0.9115241635687732, "grad_norm": 0.5254341590116196, "learning_rate": 2.090225212984459e-07, "loss": 0.0473, "step": 19616 }, { "epoch": 0.9115706319702602, "grad_norm": 0.7232362820776564, "learning_rate": 2.0880727073565476e-07, "loss": 0.0561, "step": 19617 }, { "epoch": 0.9116171003717473, "grad_norm": 0.7535285886382858, "learning_rate": 2.085921286984216e-07, "loss": 0.058, "step": 19618 }, { "epoch": 0.9116635687732342, "grad_norm": 0.4193925644809596, "learning_rate": 2.0837709519161975e-07, "loss": 0.0545, "step": 19619 }, { "epoch": 0.9117100371747212, "grad_norm": 0.46705622119052254, "learning_rate": 2.0816217022011976e-07, "loss": 0.0512, "step": 19620 }, { "epoch": 0.9117565055762081, "grad_norm": 0.48639500123368695, "learning_rate": 2.0794735378878939e-07, "loss": 0.0471, "step": 19621 }, { "epoch": 0.9118029739776952, "grad_norm": 0.35400451373726266, "learning_rate": 2.0773264590249532e-07, "loss": 0.0606, "step": 19622 }, { "epoch": 0.9118494423791822, "grad_norm": 0.6385628833048272, "learning_rate": 2.0751804656610087e-07, "loss": 0.0501, "step": 19623 }, { "epoch": 0.9118959107806691, "grad_norm": 0.573141394486185, "learning_rate": 2.0730355578446658e-07, "loss": 0.0507, "step": 19624 }, { "epoch": 0.9119423791821561, "grad_norm": 0.5134587042841819, "learning_rate": 2.0708917356245084e-07, "loss": 0.0529, "step": 19625 }, { "epoch": 0.9119888475836431, "grad_norm": 0.6057612520512045, "learning_rate": 2.0687489990491027e-07, "loss": 0.0531, "step": 19626 }, { "epoch": 0.9120353159851301, "grad_norm": 0.4134094547635257, "learning_rate": 2.0666073481669714e-07, "loss": 0.0433, "step": 19627 }, { "epoch": 0.9120817843866171, "grad_norm": 0.3742341988034519, "learning_rate": 2.0644667830266308e-07, "loss": 0.0602, "step": 19628 }, { "epoch": 0.912128252788104, "grad_norm": 0.5632741600413734, "learning_rate": 2.0623273036765755e-07, "loss": 0.0799, "step": 19629 }, { "epoch": 0.9121747211895911, "grad_norm": 0.29245240341673207, "learning_rate": 2.060188910165267e-07, "loss": 0.0268, "step": 19630 }, { "epoch": 0.9122211895910781, "grad_norm": 0.5145031576298076, "learning_rate": 2.0580516025411225e-07, "loss": 0.0524, "step": 19631 }, { "epoch": 0.912267657992565, "grad_norm": 0.4034586086930752, "learning_rate": 2.0559153808525688e-07, "loss": 0.0364, "step": 19632 }, { "epoch": 0.9123141263940521, "grad_norm": 0.515063043530625, "learning_rate": 2.053780245147996e-07, "loss": 0.0699, "step": 19633 }, { "epoch": 0.912360594795539, "grad_norm": 0.4581105808868287, "learning_rate": 2.0516461954757537e-07, "loss": 0.0419, "step": 19634 }, { "epoch": 0.912407063197026, "grad_norm": 0.43242090044197284, "learning_rate": 2.0495132318842036e-07, "loss": 0.0454, "step": 19635 }, { "epoch": 0.9124535315985131, "grad_norm": 0.5203698726775176, "learning_rate": 2.047381354421646e-07, "loss": 0.0562, "step": 19636 }, { "epoch": 0.9125, "grad_norm": 0.4452823969265812, "learning_rate": 2.045250563136364e-07, "loss": 0.0342, "step": 19637 }, { "epoch": 0.912546468401487, "grad_norm": 0.6959811499414096, "learning_rate": 2.0431208580766193e-07, "loss": 0.0508, "step": 19638 }, { "epoch": 0.9125929368029739, "grad_norm": 0.7826659470712147, "learning_rate": 2.0409922392906734e-07, "loss": 0.0801, "step": 19639 }, { "epoch": 0.912639405204461, "grad_norm": 0.5497968651665625, "learning_rate": 2.0388647068267265e-07, "loss": 0.0483, "step": 19640 }, { "epoch": 0.912685873605948, "grad_norm": 0.4112489711318065, "learning_rate": 2.0367382607329678e-07, "loss": 0.0605, "step": 19641 }, { "epoch": 0.9127323420074349, "grad_norm": 0.5924005797543206, "learning_rate": 2.0346129010575756e-07, "loss": 0.0411, "step": 19642 }, { "epoch": 0.912778810408922, "grad_norm": 0.516235703630949, "learning_rate": 2.0324886278486832e-07, "loss": 0.0626, "step": 19643 }, { "epoch": 0.9128252788104089, "grad_norm": 0.5852657867443967, "learning_rate": 2.0303654411544027e-07, "loss": 0.0514, "step": 19644 }, { "epoch": 0.9128717472118959, "grad_norm": 0.41943785795251753, "learning_rate": 2.0282433410228287e-07, "loss": 0.0297, "step": 19645 }, { "epoch": 0.9129182156133829, "grad_norm": 0.43937210220453715, "learning_rate": 2.026122327502039e-07, "loss": 0.0642, "step": 19646 }, { "epoch": 0.9129646840148699, "grad_norm": 0.5600237405315405, "learning_rate": 2.0240024006400682e-07, "loss": 0.0513, "step": 19647 }, { "epoch": 0.9130111524163569, "grad_norm": 0.7829402476930221, "learning_rate": 2.0218835604849386e-07, "loss": 0.0633, "step": 19648 }, { "epoch": 0.9130576208178439, "grad_norm": 0.5229136404914387, "learning_rate": 2.019765807084645e-07, "loss": 0.0533, "step": 19649 }, { "epoch": 0.9131040892193308, "grad_norm": 0.3814412218463221, "learning_rate": 2.017649140487149e-07, "loss": 0.0452, "step": 19650 }, { "epoch": 0.9131505576208179, "grad_norm": 0.621712045059708, "learning_rate": 2.0155335607403957e-07, "loss": 0.0645, "step": 19651 }, { "epoch": 0.9131970260223048, "grad_norm": 0.43942219601119603, "learning_rate": 2.0134190678923193e-07, "loss": 0.0596, "step": 19652 }, { "epoch": 0.9132434944237918, "grad_norm": 0.5168427392125362, "learning_rate": 2.0113056619908034e-07, "loss": 0.0506, "step": 19653 }, { "epoch": 0.9132899628252789, "grad_norm": 0.48259317610405345, "learning_rate": 2.0091933430837264e-07, "loss": 0.0461, "step": 19654 }, { "epoch": 0.9133364312267658, "grad_norm": 0.4363762820962566, "learning_rate": 2.0070821112189164e-07, "loss": 0.0424, "step": 19655 }, { "epoch": 0.9133828996282528, "grad_norm": 0.3568737623579276, "learning_rate": 2.0049719664442192e-07, "loss": 0.0454, "step": 19656 }, { "epoch": 0.9134293680297397, "grad_norm": 0.7200798811217574, "learning_rate": 2.002862908807418e-07, "loss": 0.0444, "step": 19657 }, { "epoch": 0.9134758364312268, "grad_norm": 0.5200431387302992, "learning_rate": 2.0007549383562808e-07, "loss": 0.0496, "step": 19658 }, { "epoch": 0.9135223048327138, "grad_norm": 0.6123034943062866, "learning_rate": 1.99864805513858e-07, "loss": 0.0533, "step": 19659 }, { "epoch": 0.9135687732342007, "grad_norm": 0.6718228490426923, "learning_rate": 1.9965422592020113e-07, "loss": 0.0637, "step": 19660 }, { "epoch": 0.9136152416356877, "grad_norm": 0.6059544522682118, "learning_rate": 1.9944375505942747e-07, "loss": 0.0606, "step": 19661 }, { "epoch": 0.9136617100371747, "grad_norm": 0.5444788088986874, "learning_rate": 1.9923339293630605e-07, "loss": 0.0441, "step": 19662 }, { "epoch": 0.9137081784386617, "grad_norm": 0.49385004531874205, "learning_rate": 1.9902313955560083e-07, "loss": 0.0549, "step": 19663 }, { "epoch": 0.9137546468401487, "grad_norm": 0.6983247260792446, "learning_rate": 1.9881299492207517e-07, "loss": 0.0698, "step": 19664 }, { "epoch": 0.9138011152416357, "grad_norm": 0.7235138196469633, "learning_rate": 1.9860295904048754e-07, "loss": 0.0518, "step": 19665 }, { "epoch": 0.9138475836431227, "grad_norm": 0.6702166546752536, "learning_rate": 1.9839303191559632e-07, "loss": 0.0537, "step": 19666 }, { "epoch": 0.9138940520446097, "grad_norm": 0.7623802747110916, "learning_rate": 1.981832135521572e-07, "loss": 0.0638, "step": 19667 }, { "epoch": 0.9139405204460966, "grad_norm": 0.5344856819432509, "learning_rate": 1.9797350395492077e-07, "loss": 0.0514, "step": 19668 }, { "epoch": 0.9139869888475837, "grad_norm": 0.5083199281544952, "learning_rate": 1.9776390312863936e-07, "loss": 0.0426, "step": 19669 }, { "epoch": 0.9140334572490706, "grad_norm": 0.5816854328931904, "learning_rate": 1.9755441107805973e-07, "loss": 0.0529, "step": 19670 }, { "epoch": 0.9140799256505576, "grad_norm": 0.8684766253275834, "learning_rate": 1.9734502780792807e-07, "loss": 0.0481, "step": 19671 }, { "epoch": 0.9141263940520447, "grad_norm": 0.480076557421037, "learning_rate": 1.9713575332298507e-07, "loss": 0.0327, "step": 19672 }, { "epoch": 0.9141728624535316, "grad_norm": 0.3264506736860177, "learning_rate": 1.9692658762797245e-07, "loss": 0.0382, "step": 19673 }, { "epoch": 0.9142193308550186, "grad_norm": 0.606390661627216, "learning_rate": 1.9671753072762755e-07, "loss": 0.0614, "step": 19674 }, { "epoch": 0.9142657992565055, "grad_norm": 0.6653758863919, "learning_rate": 1.9650858262668603e-07, "loss": 0.0629, "step": 19675 }, { "epoch": 0.9143122676579926, "grad_norm": 1.0831044128903615, "learning_rate": 1.9629974332988076e-07, "loss": 0.0715, "step": 19676 }, { "epoch": 0.9143587360594796, "grad_norm": 0.5172741296167501, "learning_rate": 1.96091012841943e-07, "loss": 0.0543, "step": 19677 }, { "epoch": 0.9144052044609665, "grad_norm": 0.5876914560865361, "learning_rate": 1.9588239116759833e-07, "loss": 0.0577, "step": 19678 }, { "epoch": 0.9144516728624535, "grad_norm": 0.5475030184525294, "learning_rate": 1.956738783115747e-07, "loss": 0.0366, "step": 19679 }, { "epoch": 0.9144981412639405, "grad_norm": 0.48893079114403215, "learning_rate": 1.9546547427859386e-07, "loss": 0.0532, "step": 19680 }, { "epoch": 0.9145446096654275, "grad_norm": 0.6234671908222377, "learning_rate": 1.9525717907337704e-07, "loss": 0.0343, "step": 19681 }, { "epoch": 0.9145910780669145, "grad_norm": 0.5040508811332938, "learning_rate": 1.9504899270064105e-07, "loss": 0.06, "step": 19682 }, { "epoch": 0.9146375464684015, "grad_norm": 0.5910035672101553, "learning_rate": 1.9484091516510372e-07, "loss": 0.0489, "step": 19683 }, { "epoch": 0.9146840148698885, "grad_norm": 0.45663003452251, "learning_rate": 1.946329464714769e-07, "loss": 0.05, "step": 19684 }, { "epoch": 0.9147304832713755, "grad_norm": 0.6620271637409197, "learning_rate": 1.9442508662447013e-07, "loss": 0.035, "step": 19685 }, { "epoch": 0.9147769516728624, "grad_norm": 0.5499810904777237, "learning_rate": 1.9421733562879408e-07, "loss": 0.0428, "step": 19686 }, { "epoch": 0.9148234200743495, "grad_norm": 0.7497738589120513, "learning_rate": 1.9400969348915277e-07, "loss": 0.0662, "step": 19687 }, { "epoch": 0.9148698884758364, "grad_norm": 0.5614655431287345, "learning_rate": 1.9380216021025078e-07, "loss": 0.0624, "step": 19688 }, { "epoch": 0.9149163568773234, "grad_norm": 0.36933711287480375, "learning_rate": 1.9359473579678767e-07, "loss": 0.0406, "step": 19689 }, { "epoch": 0.9149628252788105, "grad_norm": 0.4261738951974301, "learning_rate": 1.9338742025346301e-07, "loss": 0.0478, "step": 19690 }, { "epoch": 0.9150092936802974, "grad_norm": 0.5442249471312279, "learning_rate": 1.9318021358497196e-07, "loss": 0.0517, "step": 19691 }, { "epoch": 0.9150557620817844, "grad_norm": 0.3543258750294499, "learning_rate": 1.9297311579600743e-07, "loss": 0.0284, "step": 19692 }, { "epoch": 0.9151022304832713, "grad_norm": 0.5368361139870476, "learning_rate": 1.9276612689126173e-07, "loss": 0.0513, "step": 19693 }, { "epoch": 0.9151486988847584, "grad_norm": 0.3319521163210147, "learning_rate": 1.9255924687542283e-07, "loss": 0.039, "step": 19694 }, { "epoch": 0.9151951672862454, "grad_norm": 0.5245801242400959, "learning_rate": 1.9235247575317694e-07, "loss": 0.0499, "step": 19695 }, { "epoch": 0.9152416356877323, "grad_norm": 0.3430820670341223, "learning_rate": 1.9214581352920704e-07, "loss": 0.0359, "step": 19696 }, { "epoch": 0.9152881040892193, "grad_norm": 0.5423219614699161, "learning_rate": 1.9193926020819487e-07, "loss": 0.0624, "step": 19697 }, { "epoch": 0.9153345724907063, "grad_norm": 0.6321854769310428, "learning_rate": 1.9173281579481896e-07, "loss": 0.0595, "step": 19698 }, { "epoch": 0.9153810408921933, "grad_norm": 0.4974561636598797, "learning_rate": 1.9152648029375443e-07, "loss": 0.0464, "step": 19699 }, { "epoch": 0.9154275092936803, "grad_norm": 0.4659237035784461, "learning_rate": 1.9132025370967699e-07, "loss": 0.041, "step": 19700 }, { "epoch": 0.9154739776951673, "grad_norm": 0.7504795356506584, "learning_rate": 1.9111413604725736e-07, "loss": 0.0613, "step": 19701 }, { "epoch": 0.9155204460966543, "grad_norm": 0.21346123025783925, "learning_rate": 1.9090812731116236e-07, "loss": 0.0238, "step": 19702 }, { "epoch": 0.9155669144981413, "grad_norm": 0.5602296129023154, "learning_rate": 1.9070222750606048e-07, "loss": 0.0375, "step": 19703 }, { "epoch": 0.9156133828996282, "grad_norm": 0.4759868339603018, "learning_rate": 1.9049643663661464e-07, "loss": 0.0538, "step": 19704 }, { "epoch": 0.9156598513011153, "grad_norm": 0.6285355848406721, "learning_rate": 1.902907547074867e-07, "loss": 0.0505, "step": 19705 }, { "epoch": 0.9157063197026022, "grad_norm": 0.7174212307233001, "learning_rate": 1.9008518172333458e-07, "loss": 0.0658, "step": 19706 }, { "epoch": 0.9157527881040892, "grad_norm": 0.5203375714603363, "learning_rate": 1.8987971768881676e-07, "loss": 0.0509, "step": 19707 }, { "epoch": 0.9157992565055763, "grad_norm": 0.4088786437915525, "learning_rate": 1.8967436260858507e-07, "loss": 0.0476, "step": 19708 }, { "epoch": 0.9158457249070632, "grad_norm": 0.467799110057009, "learning_rate": 1.894691164872914e-07, "loss": 0.0401, "step": 19709 }, { "epoch": 0.9158921933085502, "grad_norm": 0.33428786165938285, "learning_rate": 1.8926397932958584e-07, "loss": 0.0341, "step": 19710 }, { "epoch": 0.9159386617100371, "grad_norm": 0.32921265244329884, "learning_rate": 1.8905895114011419e-07, "loss": 0.0316, "step": 19711 }, { "epoch": 0.9159851301115242, "grad_norm": 0.37464989139152605, "learning_rate": 1.8885403192352104e-07, "loss": 0.0387, "step": 19712 }, { "epoch": 0.9160315985130112, "grad_norm": 0.4622668382661012, "learning_rate": 1.886492216844471e-07, "loss": 0.0497, "step": 19713 }, { "epoch": 0.9160780669144981, "grad_norm": 0.8329572401974731, "learning_rate": 1.884445204275326e-07, "loss": 0.0667, "step": 19714 }, { "epoch": 0.9161245353159851, "grad_norm": 0.6525768792861526, "learning_rate": 1.8823992815741377e-07, "loss": 0.0597, "step": 19715 }, { "epoch": 0.9161710037174721, "grad_norm": 0.5592435084019103, "learning_rate": 1.8803544487872415e-07, "loss": 0.0574, "step": 19716 }, { "epoch": 0.9162174721189591, "grad_norm": 0.47100168357523825, "learning_rate": 1.878310705960973e-07, "loss": 0.0457, "step": 19717 }, { "epoch": 0.9162639405204461, "grad_norm": 0.49959302047574156, "learning_rate": 1.876268053141611e-07, "loss": 0.0331, "step": 19718 }, { "epoch": 0.916310408921933, "grad_norm": 0.4695482527883517, "learning_rate": 1.8742264903754192e-07, "loss": 0.0571, "step": 19719 }, { "epoch": 0.9163568773234201, "grad_norm": 0.5256955142721017, "learning_rate": 1.8721860177086604e-07, "loss": 0.046, "step": 19720 }, { "epoch": 0.9164033457249071, "grad_norm": 0.44924856104308475, "learning_rate": 1.8701466351875363e-07, "loss": 0.0409, "step": 19721 }, { "epoch": 0.916449814126394, "grad_norm": 0.3652229272970688, "learning_rate": 1.8681083428582436e-07, "loss": 0.0388, "step": 19722 }, { "epoch": 0.9164962825278811, "grad_norm": 0.3781607664981118, "learning_rate": 1.8660711407669562e-07, "loss": 0.0409, "step": 19723 }, { "epoch": 0.916542750929368, "grad_norm": 0.5232129798856195, "learning_rate": 1.8640350289598264e-07, "loss": 0.0405, "step": 19724 }, { "epoch": 0.916589219330855, "grad_norm": 0.6388485259127636, "learning_rate": 1.8620000074829557e-07, "loss": 0.0458, "step": 19725 }, { "epoch": 0.916635687732342, "grad_norm": 0.46157413731852676, "learning_rate": 1.8599660763824467e-07, "loss": 0.0459, "step": 19726 }, { "epoch": 0.916682156133829, "grad_norm": 0.9315090792355256, "learning_rate": 1.857933235704379e-07, "loss": 0.0571, "step": 19727 }, { "epoch": 0.916728624535316, "grad_norm": 0.42041053591293526, "learning_rate": 1.8559014854947876e-07, "loss": 0.0306, "step": 19728 }, { "epoch": 0.9167750929368029, "grad_norm": 0.5922750010741288, "learning_rate": 1.8538708257997029e-07, "loss": 0.0724, "step": 19729 }, { "epoch": 0.91682156133829, "grad_norm": 0.7478015882348477, "learning_rate": 1.8518412566651157e-07, "loss": 0.046, "step": 19730 }, { "epoch": 0.916868029739777, "grad_norm": 0.4441662450216132, "learning_rate": 1.849812778137e-07, "loss": 0.0572, "step": 19731 }, { "epoch": 0.9169144981412639, "grad_norm": 0.5574828427263643, "learning_rate": 1.847785390261303e-07, "loss": 0.0592, "step": 19732 }, { "epoch": 0.9169609665427509, "grad_norm": 0.5684368750660288, "learning_rate": 1.8457590930839375e-07, "loss": 0.0463, "step": 19733 }, { "epoch": 0.9170074349442379, "grad_norm": 0.4263761668822575, "learning_rate": 1.843733886650817e-07, "loss": 0.0369, "step": 19734 }, { "epoch": 0.9170539033457249, "grad_norm": 0.320555115883145, "learning_rate": 1.8417097710078102e-07, "loss": 0.0386, "step": 19735 }, { "epoch": 0.9171003717472119, "grad_norm": 0.5058868733168861, "learning_rate": 1.8396867462007585e-07, "loss": 0.0401, "step": 19736 }, { "epoch": 0.9171468401486988, "grad_norm": 0.4231560994275183, "learning_rate": 1.8376648122754913e-07, "loss": 0.0509, "step": 19737 }, { "epoch": 0.9171933085501859, "grad_norm": 0.5099283905028815, "learning_rate": 1.8356439692778062e-07, "loss": 0.0332, "step": 19738 }, { "epoch": 0.9172397769516729, "grad_norm": 0.5404751099136196, "learning_rate": 1.8336242172534768e-07, "loss": 0.045, "step": 19739 }, { "epoch": 0.9172862453531598, "grad_norm": 0.431052223362609, "learning_rate": 1.8316055562482448e-07, "loss": 0.0475, "step": 19740 }, { "epoch": 0.9173327137546469, "grad_norm": 0.5316756620971916, "learning_rate": 1.8295879863078514e-07, "loss": 0.0412, "step": 19741 }, { "epoch": 0.9173791821561338, "grad_norm": 0.27124729420960725, "learning_rate": 1.8275715074779988e-07, "loss": 0.0364, "step": 19742 }, { "epoch": 0.9174256505576208, "grad_norm": 0.504572138191754, "learning_rate": 1.8255561198043336e-07, "loss": 0.0641, "step": 19743 }, { "epoch": 0.9174721189591079, "grad_norm": 0.4311859240519358, "learning_rate": 1.8235418233325364e-07, "loss": 0.0522, "step": 19744 }, { "epoch": 0.9175185873605948, "grad_norm": 0.53883422907511, "learning_rate": 1.8215286181082147e-07, "loss": 0.0642, "step": 19745 }, { "epoch": 0.9175650557620818, "grad_norm": 0.4633154195013544, "learning_rate": 1.8195165041769825e-07, "loss": 0.0361, "step": 19746 }, { "epoch": 0.9176115241635687, "grad_norm": 0.5522885337591953, "learning_rate": 1.817505481584403e-07, "loss": 0.037, "step": 19747 }, { "epoch": 0.9176579925650558, "grad_norm": 0.8861822797709751, "learning_rate": 1.8154955503760507e-07, "loss": 0.0615, "step": 19748 }, { "epoch": 0.9177044609665428, "grad_norm": 0.4740943245382761, "learning_rate": 1.813486710597423e-07, "loss": 0.0205, "step": 19749 }, { "epoch": 0.9177509293680297, "grad_norm": 0.5700001215556545, "learning_rate": 1.811478962294039e-07, "loss": 0.066, "step": 19750 }, { "epoch": 0.9177973977695167, "grad_norm": 0.5386712352859665, "learning_rate": 1.8094723055113784e-07, "loss": 0.0606, "step": 19751 }, { "epoch": 0.9178438661710037, "grad_norm": 0.6441036080979892, "learning_rate": 1.8074667402948888e-07, "loss": 0.0592, "step": 19752 }, { "epoch": 0.9178903345724907, "grad_norm": 0.43017327817168943, "learning_rate": 1.8054622666899947e-07, "loss": 0.0399, "step": 19753 }, { "epoch": 0.9179368029739777, "grad_norm": 0.26430042517595337, "learning_rate": 1.8034588847421098e-07, "loss": 0.0296, "step": 19754 }, { "epoch": 0.9179832713754646, "grad_norm": 0.5062705307812398, "learning_rate": 1.8014565944966034e-07, "loss": 0.0544, "step": 19755 }, { "epoch": 0.9180297397769517, "grad_norm": 0.5927240195203743, "learning_rate": 1.799455395998839e-07, "loss": 0.0565, "step": 19756 }, { "epoch": 0.9180762081784387, "grad_norm": 0.8197514890004295, "learning_rate": 1.797455289294131e-07, "loss": 0.0546, "step": 19757 }, { "epoch": 0.9181226765799256, "grad_norm": 0.45655809790654384, "learning_rate": 1.795456274427798e-07, "loss": 0.0488, "step": 19758 }, { "epoch": 0.9181691449814127, "grad_norm": 0.5592588298805835, "learning_rate": 1.7934583514451097e-07, "loss": 0.0451, "step": 19759 }, { "epoch": 0.9182156133828996, "grad_norm": 0.49868543750807426, "learning_rate": 1.7914615203913353e-07, "loss": 0.0431, "step": 19760 }, { "epoch": 0.9182620817843866, "grad_norm": 0.3425717554205546, "learning_rate": 1.789465781311689e-07, "loss": 0.0521, "step": 19761 }, { "epoch": 0.9183085501858737, "grad_norm": 0.4275380477703217, "learning_rate": 1.7874711342513896e-07, "loss": 0.0407, "step": 19762 }, { "epoch": 0.9183550185873606, "grad_norm": 0.364044618862063, "learning_rate": 1.7854775792556067e-07, "loss": 0.0426, "step": 19763 }, { "epoch": 0.9184014869888476, "grad_norm": 0.6284739889680682, "learning_rate": 1.7834851163694988e-07, "loss": 0.0548, "step": 19764 }, { "epoch": 0.9184479553903345, "grad_norm": 0.6506579411177434, "learning_rate": 1.781493745638202e-07, "loss": 0.0705, "step": 19765 }, { "epoch": 0.9184944237918216, "grad_norm": 0.6067781215599707, "learning_rate": 1.77950346710683e-07, "loss": 0.042, "step": 19766 }, { "epoch": 0.9185408921933086, "grad_norm": 0.33146238420162333, "learning_rate": 1.777514280820447e-07, "loss": 0.0543, "step": 19767 }, { "epoch": 0.9185873605947955, "grad_norm": 0.3968849620651197, "learning_rate": 1.775526186824117e-07, "loss": 0.0538, "step": 19768 }, { "epoch": 0.9186338289962825, "grad_norm": 0.6556732700225377, "learning_rate": 1.7735391851628814e-07, "loss": 0.0601, "step": 19769 }, { "epoch": 0.9186802973977695, "grad_norm": 0.458306821551538, "learning_rate": 1.7715532758817377e-07, "loss": 0.0462, "step": 19770 }, { "epoch": 0.9187267657992565, "grad_norm": 0.5688301440873728, "learning_rate": 1.7695684590256668e-07, "loss": 0.0419, "step": 19771 }, { "epoch": 0.9187732342007435, "grad_norm": 0.6366960241711105, "learning_rate": 1.767584734639638e-07, "loss": 0.067, "step": 19772 }, { "epoch": 0.9188197026022304, "grad_norm": 0.49636967428704054, "learning_rate": 1.765602102768582e-07, "loss": 0.0519, "step": 19773 }, { "epoch": 0.9188661710037175, "grad_norm": 0.30783147097240166, "learning_rate": 1.763620563457391e-07, "loss": 0.0372, "step": 19774 }, { "epoch": 0.9189126394052045, "grad_norm": 0.4587387829196236, "learning_rate": 1.7616401167509733e-07, "loss": 0.037, "step": 19775 }, { "epoch": 0.9189591078066914, "grad_norm": 0.5928592480386463, "learning_rate": 1.7596607626941708e-07, "loss": 0.0615, "step": 19776 }, { "epoch": 0.9190055762081785, "grad_norm": 0.6224975822633333, "learning_rate": 1.7576825013318312e-07, "loss": 0.0654, "step": 19777 }, { "epoch": 0.9190520446096654, "grad_norm": 0.5326415977807246, "learning_rate": 1.7557053327087513e-07, "loss": 0.0536, "step": 19778 }, { "epoch": 0.9190985130111524, "grad_norm": 0.5771932264654078, "learning_rate": 1.753729256869724e-07, "loss": 0.0423, "step": 19779 }, { "epoch": 0.9191449814126395, "grad_norm": 0.5890997380854562, "learning_rate": 1.7517542738595071e-07, "loss": 0.0567, "step": 19780 }, { "epoch": 0.9191914498141264, "grad_norm": 0.4619486417779873, "learning_rate": 1.749780383722832e-07, "loss": 0.0335, "step": 19781 }, { "epoch": 0.9192379182156134, "grad_norm": 0.5416445261258527, "learning_rate": 1.7478075865044187e-07, "loss": 0.0427, "step": 19782 }, { "epoch": 0.9192843866171003, "grad_norm": 0.5112545134100585, "learning_rate": 1.745835882248953e-07, "loss": 0.0604, "step": 19783 }, { "epoch": 0.9193308550185874, "grad_norm": 0.6655737482256648, "learning_rate": 1.7438652710010772e-07, "loss": 0.0478, "step": 19784 }, { "epoch": 0.9193773234200744, "grad_norm": 0.887099677403478, "learning_rate": 1.7418957528054503e-07, "loss": 0.0484, "step": 19785 }, { "epoch": 0.9194237918215613, "grad_norm": 0.6969628714120544, "learning_rate": 1.7399273277066752e-07, "loss": 0.0486, "step": 19786 }, { "epoch": 0.9194702602230483, "grad_norm": 0.6242262176166812, "learning_rate": 1.7379599957493388e-07, "loss": 0.0595, "step": 19787 }, { "epoch": 0.9195167286245353, "grad_norm": 0.47557233032973695, "learning_rate": 1.7359937569779993e-07, "loss": 0.0377, "step": 19788 }, { "epoch": 0.9195631970260223, "grad_norm": 0.5946264067198532, "learning_rate": 1.734028611437205e-07, "loss": 0.0571, "step": 19789 }, { "epoch": 0.9196096654275093, "grad_norm": 0.6694406186663754, "learning_rate": 1.7320645591714592e-07, "loss": 0.0528, "step": 19790 }, { "epoch": 0.9196561338289962, "grad_norm": 0.7631284748624424, "learning_rate": 1.7301016002252479e-07, "loss": 0.0506, "step": 19791 }, { "epoch": 0.9197026022304833, "grad_norm": 0.45096002609965197, "learning_rate": 1.7281397346430416e-07, "loss": 0.0453, "step": 19792 }, { "epoch": 0.9197490706319703, "grad_norm": 0.3946738325998189, "learning_rate": 1.7261789624692772e-07, "loss": 0.0326, "step": 19793 }, { "epoch": 0.9197955390334572, "grad_norm": 0.6592055575364234, "learning_rate": 1.7242192837483628e-07, "loss": 0.0566, "step": 19794 }, { "epoch": 0.9198420074349443, "grad_norm": 0.4848053438670047, "learning_rate": 1.722260698524697e-07, "loss": 0.0668, "step": 19795 }, { "epoch": 0.9198884758364312, "grad_norm": 0.2328552188534495, "learning_rate": 1.720303206842633e-07, "loss": 0.0217, "step": 19796 }, { "epoch": 0.9199349442379182, "grad_norm": 0.5322664232717362, "learning_rate": 1.7183468087465184e-07, "loss": 0.0603, "step": 19797 }, { "epoch": 0.9199814126394052, "grad_norm": 0.46895759293633416, "learning_rate": 1.7163915042806568e-07, "loss": 0.0501, "step": 19798 }, { "epoch": 0.9200278810408922, "grad_norm": 0.4464524480284625, "learning_rate": 1.714437293489346e-07, "loss": 0.041, "step": 19799 }, { "epoch": 0.9200743494423792, "grad_norm": 0.4402899111105897, "learning_rate": 1.7124841764168564e-07, "loss": 0.0608, "step": 19800 }, { "epoch": 0.9201208178438661, "grad_norm": 0.5266694188213382, "learning_rate": 1.7105321531074193e-07, "loss": 0.0509, "step": 19801 }, { "epoch": 0.9201672862453532, "grad_norm": 0.5229080455225336, "learning_rate": 1.7085812236052547e-07, "loss": 0.0568, "step": 19802 }, { "epoch": 0.9202137546468402, "grad_norm": 0.6400606219430889, "learning_rate": 1.7066313879545494e-07, "loss": 0.0573, "step": 19803 }, { "epoch": 0.9202602230483271, "grad_norm": 0.48493640231157253, "learning_rate": 1.7046826461994738e-07, "loss": 0.0449, "step": 19804 }, { "epoch": 0.9203066914498141, "grad_norm": 0.5245500849797461, "learning_rate": 1.702734998384159e-07, "loss": 0.0219, "step": 19805 }, { "epoch": 0.9203531598513011, "grad_norm": 0.5173137988743683, "learning_rate": 1.7007884445527313e-07, "loss": 0.0427, "step": 19806 }, { "epoch": 0.9203996282527881, "grad_norm": 0.5031843514114439, "learning_rate": 1.6988429847492882e-07, "loss": 0.0467, "step": 19807 }, { "epoch": 0.9204460966542751, "grad_norm": 0.45463848028809156, "learning_rate": 1.6968986190178728e-07, "loss": 0.0662, "step": 19808 }, { "epoch": 0.920492565055762, "grad_norm": 0.5823336016211046, "learning_rate": 1.6949553474025494e-07, "loss": 0.0569, "step": 19809 }, { "epoch": 0.9205390334572491, "grad_norm": 0.7743541773482989, "learning_rate": 1.6930131699473328e-07, "loss": 0.0542, "step": 19810 }, { "epoch": 0.9205855018587361, "grad_norm": 0.6401544155354674, "learning_rate": 1.691072086696205e-07, "loss": 0.0409, "step": 19811 }, { "epoch": 0.920631970260223, "grad_norm": 0.5092675476900916, "learning_rate": 1.6891320976931303e-07, "loss": 0.0462, "step": 19812 }, { "epoch": 0.9206784386617101, "grad_norm": 0.4784447837452311, "learning_rate": 1.6871932029820736e-07, "loss": 0.055, "step": 19813 }, { "epoch": 0.920724907063197, "grad_norm": 0.42076489526220723, "learning_rate": 1.6852554026069335e-07, "loss": 0.0605, "step": 19814 }, { "epoch": 0.920771375464684, "grad_norm": 0.6296271956781129, "learning_rate": 1.6833186966116078e-07, "loss": 0.0476, "step": 19815 }, { "epoch": 0.920817843866171, "grad_norm": 0.4048386846855459, "learning_rate": 1.681383085039967e-07, "loss": 0.0524, "step": 19816 }, { "epoch": 0.920864312267658, "grad_norm": 0.7065611216414001, "learning_rate": 1.679448567935854e-07, "loss": 0.0425, "step": 19817 }, { "epoch": 0.920910780669145, "grad_norm": 0.44539928050762884, "learning_rate": 1.6775151453430837e-07, "loss": 0.0499, "step": 19818 }, { "epoch": 0.9209572490706319, "grad_norm": 0.5974099026085438, "learning_rate": 1.67558281730546e-07, "loss": 0.0733, "step": 19819 }, { "epoch": 0.921003717472119, "grad_norm": 0.41228701476144036, "learning_rate": 1.6736515838667423e-07, "loss": 0.0425, "step": 19820 }, { "epoch": 0.921050185873606, "grad_norm": 0.6036098814215655, "learning_rate": 1.6717214450706841e-07, "loss": 0.0414, "step": 19821 }, { "epoch": 0.9210966542750929, "grad_norm": 0.5385147160251073, "learning_rate": 1.66979240096099e-07, "loss": 0.0577, "step": 19822 }, { "epoch": 0.9211431226765799, "grad_norm": 0.4654298770977157, "learning_rate": 1.6678644515813692e-07, "loss": 0.06, "step": 19823 }, { "epoch": 0.9211895910780669, "grad_norm": 0.5067537440636739, "learning_rate": 1.6659375969754864e-07, "loss": 0.0425, "step": 19824 }, { "epoch": 0.9212360594795539, "grad_norm": 0.6945624974337429, "learning_rate": 1.6640118371869908e-07, "loss": 0.0513, "step": 19825 }, { "epoch": 0.9212825278810409, "grad_norm": 0.4169098109525946, "learning_rate": 1.6620871722595023e-07, "loss": 0.0398, "step": 19826 }, { "epoch": 0.9213289962825278, "grad_norm": 0.5514373571185904, "learning_rate": 1.6601636022366085e-07, "loss": 0.0479, "step": 19827 }, { "epoch": 0.9213754646840149, "grad_norm": 0.6121819081097523, "learning_rate": 1.6582411271618915e-07, "loss": 0.0502, "step": 19828 }, { "epoch": 0.9214219330855019, "grad_norm": 0.4982960218249099, "learning_rate": 1.6563197470788827e-07, "loss": 0.0313, "step": 19829 }, { "epoch": 0.9214684014869888, "grad_norm": 0.5795283919337848, "learning_rate": 1.6543994620311143e-07, "loss": 0.0613, "step": 19830 }, { "epoch": 0.9215148698884759, "grad_norm": 0.43951850419493566, "learning_rate": 1.6524802720620902e-07, "loss": 0.0489, "step": 19831 }, { "epoch": 0.9215613382899628, "grad_norm": 0.3922986948775059, "learning_rate": 1.6505621772152647e-07, "loss": 0.035, "step": 19832 }, { "epoch": 0.9216078066914498, "grad_norm": 0.6006919172601513, "learning_rate": 1.6486451775340916e-07, "loss": 0.0535, "step": 19833 }, { "epoch": 0.9216542750929368, "grad_norm": 0.45704503088297943, "learning_rate": 1.6467292730619977e-07, "loss": 0.0602, "step": 19834 }, { "epoch": 0.9217007434944238, "grad_norm": 0.511624540717954, "learning_rate": 1.6448144638423813e-07, "loss": 0.0353, "step": 19835 }, { "epoch": 0.9217472118959108, "grad_norm": 0.40197195178281414, "learning_rate": 1.6429007499186023e-07, "loss": 0.041, "step": 19836 }, { "epoch": 0.9217936802973977, "grad_norm": 0.45838061117386353, "learning_rate": 1.6409881313340203e-07, "loss": 0.058, "step": 19837 }, { "epoch": 0.9218401486988848, "grad_norm": 0.645577863622804, "learning_rate": 1.6390766081319565e-07, "loss": 0.0516, "step": 19838 }, { "epoch": 0.9218866171003718, "grad_norm": 0.6412565012565207, "learning_rate": 1.6371661803556927e-07, "loss": 0.0663, "step": 19839 }, { "epoch": 0.9219330855018587, "grad_norm": 0.6079705840326436, "learning_rate": 1.6352568480485277e-07, "loss": 0.0625, "step": 19840 }, { "epoch": 0.9219795539033457, "grad_norm": 0.45926484711687504, "learning_rate": 1.633348611253699e-07, "loss": 0.0568, "step": 19841 }, { "epoch": 0.9220260223048327, "grad_norm": 0.4449548897141588, "learning_rate": 1.6314414700144276e-07, "loss": 0.0435, "step": 19842 }, { "epoch": 0.9220724907063197, "grad_norm": 0.5568862120852313, "learning_rate": 1.6295354243739124e-07, "loss": 0.0555, "step": 19843 }, { "epoch": 0.9221189591078067, "grad_norm": 0.5476040054273384, "learning_rate": 1.6276304743753301e-07, "loss": 0.0668, "step": 19844 }, { "epoch": 0.9221654275092936, "grad_norm": 0.36684043308460196, "learning_rate": 1.6257266200618294e-07, "loss": 0.0369, "step": 19845 }, { "epoch": 0.9222118959107807, "grad_norm": 0.6012724576860199, "learning_rate": 1.6238238614765256e-07, "loss": 0.0437, "step": 19846 }, { "epoch": 0.9222583643122677, "grad_norm": 0.5540536870380874, "learning_rate": 1.6219221986625345e-07, "loss": 0.0562, "step": 19847 }, { "epoch": 0.9223048327137546, "grad_norm": 0.4586494091795114, "learning_rate": 1.620021631662927e-07, "loss": 0.0464, "step": 19848 }, { "epoch": 0.9223513011152417, "grad_norm": 0.570632127228779, "learning_rate": 1.6181221605207353e-07, "loss": 0.052, "step": 19849 }, { "epoch": 0.9223977695167286, "grad_norm": 0.33953797401008606, "learning_rate": 1.6162237852790085e-07, "loss": 0.0405, "step": 19850 }, { "epoch": 0.9224442379182156, "grad_norm": 0.5108521262142682, "learning_rate": 1.614326505980729e-07, "loss": 0.0379, "step": 19851 }, { "epoch": 0.9224907063197026, "grad_norm": 0.5091747550184549, "learning_rate": 1.6124303226688843e-07, "loss": 0.0513, "step": 19852 }, { "epoch": 0.9225371747211896, "grad_norm": 0.3823311164111897, "learning_rate": 1.6105352353864122e-07, "loss": 0.0457, "step": 19853 }, { "epoch": 0.9225836431226766, "grad_norm": 0.52796101504777, "learning_rate": 1.608641244176262e-07, "loss": 0.0486, "step": 19854 }, { "epoch": 0.9226301115241635, "grad_norm": 0.66877811915559, "learning_rate": 1.6067483490813053e-07, "loss": 0.0505, "step": 19855 }, { "epoch": 0.9226765799256506, "grad_norm": 0.5575441437712388, "learning_rate": 1.604856550144429e-07, "loss": 0.0398, "step": 19856 }, { "epoch": 0.9227230483271376, "grad_norm": 0.9259498966451225, "learning_rate": 1.6029658474084942e-07, "loss": 0.0661, "step": 19857 }, { "epoch": 0.9227695167286245, "grad_norm": 0.4026109351381032, "learning_rate": 1.6010762409163216e-07, "loss": 0.065, "step": 19858 }, { "epoch": 0.9228159851301115, "grad_norm": 0.4887748843540144, "learning_rate": 1.599187730710705e-07, "loss": 0.0526, "step": 19859 }, { "epoch": 0.9228624535315985, "grad_norm": 0.5255441607525098, "learning_rate": 1.5973003168344325e-07, "loss": 0.0552, "step": 19860 }, { "epoch": 0.9229089219330855, "grad_norm": 0.3154776137087872, "learning_rate": 1.5954139993302474e-07, "loss": 0.0433, "step": 19861 }, { "epoch": 0.9229553903345725, "grad_norm": 0.4636536444161431, "learning_rate": 1.5935287782408827e-07, "loss": 0.0555, "step": 19862 }, { "epoch": 0.9230018587360594, "grad_norm": 0.49309515635246753, "learning_rate": 1.5916446536090314e-07, "loss": 0.0416, "step": 19863 }, { "epoch": 0.9230483271375465, "grad_norm": 0.3876269872451951, "learning_rate": 1.5897616254773818e-07, "loss": 0.0604, "step": 19864 }, { "epoch": 0.9230947955390335, "grad_norm": 0.4004584277765206, "learning_rate": 1.5878796938885832e-07, "loss": 0.0418, "step": 19865 }, { "epoch": 0.9231412639405204, "grad_norm": 0.3707944098985552, "learning_rate": 1.5859988588852627e-07, "loss": 0.0402, "step": 19866 }, { "epoch": 0.9231877323420075, "grad_norm": 0.751568518743735, "learning_rate": 1.5841191205100192e-07, "loss": 0.0532, "step": 19867 }, { "epoch": 0.9232342007434944, "grad_norm": 0.6080603753490312, "learning_rate": 1.5822404788054412e-07, "loss": 0.0579, "step": 19868 }, { "epoch": 0.9232806691449814, "grad_norm": 0.5880666022620604, "learning_rate": 1.5803629338140724e-07, "loss": 0.0613, "step": 19869 }, { "epoch": 0.9233271375464684, "grad_norm": 0.5346211704038801, "learning_rate": 1.5784864855784342e-07, "loss": 0.0481, "step": 19870 }, { "epoch": 0.9233736059479554, "grad_norm": 0.5684955562033901, "learning_rate": 1.5766111341410484e-07, "loss": 0.0521, "step": 19871 }, { "epoch": 0.9234200743494424, "grad_norm": 0.6000972679712376, "learning_rate": 1.574736879544392e-07, "loss": 0.0635, "step": 19872 }, { "epoch": 0.9234665427509293, "grad_norm": 0.30619057765610386, "learning_rate": 1.572863721830903e-07, "loss": 0.0257, "step": 19873 }, { "epoch": 0.9235130111524164, "grad_norm": 0.36104500776908727, "learning_rate": 1.5709916610430199e-07, "loss": 0.06, "step": 19874 }, { "epoch": 0.9235594795539034, "grad_norm": 0.4429002671639721, "learning_rate": 1.5691206972231477e-07, "loss": 0.0546, "step": 19875 }, { "epoch": 0.9236059479553903, "grad_norm": 0.5972978414211625, "learning_rate": 1.5672508304136634e-07, "loss": 0.0359, "step": 19876 }, { "epoch": 0.9236524163568773, "grad_norm": 0.4115664232713274, "learning_rate": 1.565382060656917e-07, "loss": 0.0495, "step": 19877 }, { "epoch": 0.9236988847583643, "grad_norm": 0.37815438473473145, "learning_rate": 1.5635143879952575e-07, "loss": 0.0353, "step": 19878 }, { "epoch": 0.9237453531598513, "grad_norm": 0.5349873500830851, "learning_rate": 1.5616478124709676e-07, "loss": 0.0694, "step": 19879 }, { "epoch": 0.9237918215613383, "grad_norm": 0.45922772030862824, "learning_rate": 1.5597823341263307e-07, "loss": 0.0458, "step": 19880 }, { "epoch": 0.9238382899628252, "grad_norm": 0.7403764432870995, "learning_rate": 1.5579179530036127e-07, "loss": 0.0545, "step": 19881 }, { "epoch": 0.9238847583643123, "grad_norm": 0.6113889744230921, "learning_rate": 1.556054669145035e-07, "loss": 0.0389, "step": 19882 }, { "epoch": 0.9239312267657993, "grad_norm": 0.5420691860876532, "learning_rate": 1.5541924825928088e-07, "loss": 0.0616, "step": 19883 }, { "epoch": 0.9239776951672862, "grad_norm": 0.6456894709640234, "learning_rate": 1.552331393389106e-07, "loss": 0.0631, "step": 19884 }, { "epoch": 0.9240241635687733, "grad_norm": 0.4061543949275602, "learning_rate": 1.5504714015760925e-07, "loss": 0.053, "step": 19885 }, { "epoch": 0.9240706319702602, "grad_norm": 0.6649117548049138, "learning_rate": 1.54861250719589e-07, "loss": 0.0623, "step": 19886 }, { "epoch": 0.9241171003717472, "grad_norm": 0.3943475886189428, "learning_rate": 1.5467547102906044e-07, "loss": 0.0445, "step": 19887 }, { "epoch": 0.9241635687732342, "grad_norm": 0.4995785950919905, "learning_rate": 1.5448980109023293e-07, "loss": 0.0555, "step": 19888 }, { "epoch": 0.9242100371747212, "grad_norm": 0.476457651385953, "learning_rate": 1.5430424090731088e-07, "loss": 0.0383, "step": 19889 }, { "epoch": 0.9242565055762082, "grad_norm": 0.5185702216697062, "learning_rate": 1.5411879048449762e-07, "loss": 0.0555, "step": 19890 }, { "epoch": 0.9243029739776951, "grad_norm": 0.5540346328895639, "learning_rate": 1.5393344982599423e-07, "loss": 0.0406, "step": 19891 }, { "epoch": 0.9243494423791822, "grad_norm": 0.6255751101175735, "learning_rate": 1.5374821893599844e-07, "loss": 0.0595, "step": 19892 }, { "epoch": 0.9243959107806692, "grad_norm": 0.39012942192087136, "learning_rate": 1.5356309781870638e-07, "loss": 0.0425, "step": 19893 }, { "epoch": 0.9244423791821561, "grad_norm": 0.4449469437208127, "learning_rate": 1.533780864783102e-07, "loss": 0.0574, "step": 19894 }, { "epoch": 0.9244888475836431, "grad_norm": 0.6499739504501023, "learning_rate": 1.5319318491900215e-07, "loss": 0.0578, "step": 19895 }, { "epoch": 0.9245353159851301, "grad_norm": 0.5027304471030868, "learning_rate": 1.530083931449694e-07, "loss": 0.0553, "step": 19896 }, { "epoch": 0.9245817843866171, "grad_norm": 0.49722352087264754, "learning_rate": 1.5282371116039752e-07, "loss": 0.0559, "step": 19897 }, { "epoch": 0.9246282527881041, "grad_norm": 0.621798796973066, "learning_rate": 1.5263913896947036e-07, "loss": 0.058, "step": 19898 }, { "epoch": 0.924674721189591, "grad_norm": 0.379450306525756, "learning_rate": 1.5245467657636848e-07, "loss": 0.0468, "step": 19899 }, { "epoch": 0.9247211895910781, "grad_norm": 0.5185331270416532, "learning_rate": 1.5227032398527019e-07, "loss": 0.0487, "step": 19900 }, { "epoch": 0.9247676579925651, "grad_norm": 0.38184314519547125, "learning_rate": 1.5208608120035162e-07, "loss": 0.0519, "step": 19901 }, { "epoch": 0.924814126394052, "grad_norm": 0.5703877253649697, "learning_rate": 1.5190194822578497e-07, "loss": 0.0691, "step": 19902 }, { "epoch": 0.9248605947955391, "grad_norm": 0.48403018251260416, "learning_rate": 1.517179250657419e-07, "loss": 0.0424, "step": 19903 }, { "epoch": 0.924907063197026, "grad_norm": 0.7642534643972727, "learning_rate": 1.515340117243902e-07, "loss": 0.0493, "step": 19904 }, { "epoch": 0.924953531598513, "grad_norm": 0.5378720555218329, "learning_rate": 1.5135020820589653e-07, "loss": 0.0562, "step": 19905 }, { "epoch": 0.925, "grad_norm": 0.6126564400059112, "learning_rate": 1.511665145144242e-07, "loss": 0.054, "step": 19906 }, { "epoch": 0.925046468401487, "grad_norm": 0.49098041121674824, "learning_rate": 1.5098293065413271e-07, "loss": 0.0317, "step": 19907 }, { "epoch": 0.925092936802974, "grad_norm": 0.49184557592556316, "learning_rate": 1.5079945662918204e-07, "loss": 0.0518, "step": 19908 }, { "epoch": 0.9251394052044609, "grad_norm": 0.6214364563516765, "learning_rate": 1.506160924437272e-07, "loss": 0.0424, "step": 19909 }, { "epoch": 0.925185873605948, "grad_norm": 0.3035265744118337, "learning_rate": 1.5043283810192154e-07, "loss": 0.0352, "step": 19910 }, { "epoch": 0.925232342007435, "grad_norm": 0.6760553237422539, "learning_rate": 1.5024969360791564e-07, "loss": 0.0533, "step": 19911 }, { "epoch": 0.9252788104089219, "grad_norm": 0.713740319781525, "learning_rate": 1.500666589658595e-07, "loss": 0.0761, "step": 19912 }, { "epoch": 0.9253252788104089, "grad_norm": 0.48050488508141587, "learning_rate": 1.498837341798981e-07, "loss": 0.0497, "step": 19913 }, { "epoch": 0.9253717472118959, "grad_norm": 0.6080856319431931, "learning_rate": 1.4970091925417372e-07, "loss": 0.0572, "step": 19914 }, { "epoch": 0.9254182156133829, "grad_norm": 0.3051961009911518, "learning_rate": 1.4951821419282865e-07, "loss": 0.0362, "step": 19915 }, { "epoch": 0.9254646840148699, "grad_norm": 0.4980898456674446, "learning_rate": 1.493356190000017e-07, "loss": 0.0587, "step": 19916 }, { "epoch": 0.9255111524163568, "grad_norm": 0.7230128364452444, "learning_rate": 1.4915313367982797e-07, "loss": 0.0548, "step": 19917 }, { "epoch": 0.9255576208178439, "grad_norm": 0.41127713158344104, "learning_rate": 1.4897075823644026e-07, "loss": 0.0627, "step": 19918 }, { "epoch": 0.9256040892193309, "grad_norm": 0.4938840059738675, "learning_rate": 1.487884926739719e-07, "loss": 0.0541, "step": 19919 }, { "epoch": 0.9256505576208178, "grad_norm": 0.45928249452464964, "learning_rate": 1.486063369965496e-07, "loss": 0.0479, "step": 19920 }, { "epoch": 0.9256970260223049, "grad_norm": 0.5891737007438552, "learning_rate": 1.4842429120829893e-07, "loss": 0.0516, "step": 19921 }, { "epoch": 0.9257434944237918, "grad_norm": 0.5104441313580036, "learning_rate": 1.4824235531334497e-07, "loss": 0.0431, "step": 19922 }, { "epoch": 0.9257899628252788, "grad_norm": 0.40039619757019546, "learning_rate": 1.4806052931580771e-07, "loss": 0.0336, "step": 19923 }, { "epoch": 0.9258364312267658, "grad_norm": 0.2532283470530639, "learning_rate": 1.4787881321980667e-07, "loss": 0.0324, "step": 19924 }, { "epoch": 0.9258828996282528, "grad_norm": 0.5276497837414132, "learning_rate": 1.4769720702945688e-07, "loss": 0.0448, "step": 19925 }, { "epoch": 0.9259293680297398, "grad_norm": 0.5889244257549259, "learning_rate": 1.4751571074887227e-07, "loss": 0.0475, "step": 19926 }, { "epoch": 0.9259758364312267, "grad_norm": 0.5983460385725998, "learning_rate": 1.47334324382164e-07, "loss": 0.0733, "step": 19927 }, { "epoch": 0.9260223048327137, "grad_norm": 0.7468601728658684, "learning_rate": 1.4715304793343987e-07, "loss": 0.0586, "step": 19928 }, { "epoch": 0.9260687732342008, "grad_norm": 0.5560017135902755, "learning_rate": 1.4697188140680774e-07, "loss": 0.0375, "step": 19929 }, { "epoch": 0.9261152416356877, "grad_norm": 0.4283104221825392, "learning_rate": 1.4679082480636986e-07, "loss": 0.0453, "step": 19930 }, { "epoch": 0.9261617100371747, "grad_norm": 0.685328077032672, "learning_rate": 1.4660987813622796e-07, "loss": 0.0382, "step": 19931 }, { "epoch": 0.9262081784386617, "grad_norm": 0.7628713410185584, "learning_rate": 1.4642904140047986e-07, "loss": 0.0656, "step": 19932 }, { "epoch": 0.9262546468401487, "grad_norm": 0.6733306796504945, "learning_rate": 1.4624831460322285e-07, "loss": 0.0617, "step": 19933 }, { "epoch": 0.9263011152416357, "grad_norm": 0.6186157541934895, "learning_rate": 1.460676977485498e-07, "loss": 0.0601, "step": 19934 }, { "epoch": 0.9263475836431226, "grad_norm": 0.4021907948309475, "learning_rate": 1.458871908405518e-07, "loss": 0.044, "step": 19935 }, { "epoch": 0.9263940520446097, "grad_norm": 0.504793262751563, "learning_rate": 1.4570679388331787e-07, "loss": 0.0335, "step": 19936 }, { "epoch": 0.9264405204460967, "grad_norm": 0.38547881156863717, "learning_rate": 1.455265068809353e-07, "loss": 0.033, "step": 19937 }, { "epoch": 0.9264869888475836, "grad_norm": 0.4578221817457948, "learning_rate": 1.4534632983748465e-07, "loss": 0.0548, "step": 19938 }, { "epoch": 0.9265334572490707, "grad_norm": 0.40315880703496293, "learning_rate": 1.4516626275704993e-07, "loss": 0.0404, "step": 19939 }, { "epoch": 0.9265799256505576, "grad_norm": 0.6269242122337728, "learning_rate": 1.449863056437095e-07, "loss": 0.0476, "step": 19940 }, { "epoch": 0.9266263940520446, "grad_norm": 0.5204331820762182, "learning_rate": 1.448064585015385e-07, "loss": 0.0456, "step": 19941 }, { "epoch": 0.9266728624535316, "grad_norm": 0.37401741408196043, "learning_rate": 1.4462672133461087e-07, "loss": 0.047, "step": 19942 }, { "epoch": 0.9267193308550186, "grad_norm": 0.67715871859458, "learning_rate": 1.4444709414699886e-07, "loss": 0.0493, "step": 19943 }, { "epoch": 0.9267657992565056, "grad_norm": 0.4465591732033528, "learning_rate": 1.4426757694277038e-07, "loss": 0.0444, "step": 19944 }, { "epoch": 0.9268122676579925, "grad_norm": 0.8017696923714097, "learning_rate": 1.440881697259916e-07, "loss": 0.0604, "step": 19945 }, { "epoch": 0.9268587360594795, "grad_norm": 0.5875155605736835, "learning_rate": 1.439088725007265e-07, "loss": 0.0526, "step": 19946 }, { "epoch": 0.9269052044609666, "grad_norm": 0.4902743661047836, "learning_rate": 1.437296852710368e-07, "loss": 0.0545, "step": 19947 }, { "epoch": 0.9269516728624535, "grad_norm": 0.4972587678175572, "learning_rate": 1.4355060804098043e-07, "loss": 0.0438, "step": 19948 }, { "epoch": 0.9269981412639405, "grad_norm": 0.5407764393173987, "learning_rate": 1.4337164081461463e-07, "loss": 0.0501, "step": 19949 }, { "epoch": 0.9270446096654275, "grad_norm": 0.46199462695941546, "learning_rate": 1.4319278359599232e-07, "loss": 0.0572, "step": 19950 }, { "epoch": 0.9270910780669145, "grad_norm": 0.6459038800606484, "learning_rate": 1.4301403638916522e-07, "loss": 0.0477, "step": 19951 }, { "epoch": 0.9271375464684015, "grad_norm": 0.5304888875937962, "learning_rate": 1.428353991981818e-07, "loss": 0.0526, "step": 19952 }, { "epoch": 0.9271840148698884, "grad_norm": 0.25077414742031406, "learning_rate": 1.4265687202708877e-07, "loss": 0.0205, "step": 19953 }, { "epoch": 0.9272304832713755, "grad_norm": 0.4535926302149467, "learning_rate": 1.4247845487993072e-07, "loss": 0.0386, "step": 19954 }, { "epoch": 0.9272769516728625, "grad_norm": 0.7066277905174142, "learning_rate": 1.4230014776074663e-07, "loss": 0.0525, "step": 19955 }, { "epoch": 0.9273234200743494, "grad_norm": 0.5154631566971986, "learning_rate": 1.421219506735777e-07, "loss": 0.0233, "step": 19956 }, { "epoch": 0.9273698884758365, "grad_norm": 0.5673324536754014, "learning_rate": 1.419438636224596e-07, "loss": 0.0493, "step": 19957 }, { "epoch": 0.9274163568773234, "grad_norm": 0.5251832480893356, "learning_rate": 1.4176588661142577e-07, "loss": 0.0527, "step": 19958 }, { "epoch": 0.9274628252788104, "grad_norm": 0.4053741749502662, "learning_rate": 1.4158801964450742e-07, "loss": 0.0425, "step": 19959 }, { "epoch": 0.9275092936802974, "grad_norm": 0.4024312444888134, "learning_rate": 1.4141026272573467e-07, "loss": 0.0324, "step": 19960 }, { "epoch": 0.9275557620817844, "grad_norm": 0.5544811725694714, "learning_rate": 1.412326158591326e-07, "loss": 0.0535, "step": 19961 }, { "epoch": 0.9276022304832714, "grad_norm": 0.5961107407400895, "learning_rate": 1.4105507904872473e-07, "loss": 0.0401, "step": 19962 }, { "epoch": 0.9276486988847583, "grad_norm": 0.3631405240980065, "learning_rate": 1.4087765229853388e-07, "loss": 0.0651, "step": 19963 }, { "epoch": 0.9276951672862453, "grad_norm": 0.9354750293109162, "learning_rate": 1.4070033561257858e-07, "loss": 0.0632, "step": 19964 }, { "epoch": 0.9277416356877324, "grad_norm": 0.5402254073201647, "learning_rate": 1.4052312899487496e-07, "loss": 0.0591, "step": 19965 }, { "epoch": 0.9277881040892193, "grad_norm": 0.5389171875786296, "learning_rate": 1.4034603244943713e-07, "loss": 0.0507, "step": 19966 }, { "epoch": 0.9278345724907063, "grad_norm": 0.49813469163420304, "learning_rate": 1.4016904598027625e-07, "loss": 0.0492, "step": 19967 }, { "epoch": 0.9278810408921933, "grad_norm": 0.7276312556203142, "learning_rate": 1.3999216959140082e-07, "loss": 0.0501, "step": 19968 }, { "epoch": 0.9279275092936803, "grad_norm": 0.4234479774094563, "learning_rate": 1.3981540328681818e-07, "loss": 0.0555, "step": 19969 }, { "epoch": 0.9279739776951673, "grad_norm": 0.7169732308639416, "learning_rate": 1.3963874707053183e-07, "loss": 0.0259, "step": 19970 }, { "epoch": 0.9280204460966542, "grad_norm": 0.4083688328108695, "learning_rate": 1.394622009465435e-07, "loss": 0.04, "step": 19971 }, { "epoch": 0.9280669144981413, "grad_norm": 0.6405815842948563, "learning_rate": 1.3928576491885172e-07, "loss": 0.0469, "step": 19972 }, { "epoch": 0.9281133828996283, "grad_norm": 0.6152801344164981, "learning_rate": 1.3910943899145324e-07, "loss": 0.0423, "step": 19973 }, { "epoch": 0.9281598513011152, "grad_norm": 0.33392153641144945, "learning_rate": 1.389332231683421e-07, "loss": 0.0589, "step": 19974 }, { "epoch": 0.9282063197026023, "grad_norm": 0.49145873170637905, "learning_rate": 1.3875711745350962e-07, "loss": 0.0479, "step": 19975 }, { "epoch": 0.9282527881040892, "grad_norm": 0.35676297374883287, "learning_rate": 1.3858112185094418e-07, "loss": 0.0274, "step": 19976 }, { "epoch": 0.9282992565055762, "grad_norm": 0.6364079225324434, "learning_rate": 1.3840523636463377e-07, "loss": 0.0577, "step": 19977 }, { "epoch": 0.9283457249070632, "grad_norm": 0.6464889074098611, "learning_rate": 1.3822946099856184e-07, "loss": 0.0511, "step": 19978 }, { "epoch": 0.9283921933085502, "grad_norm": 0.7809722480316938, "learning_rate": 1.3805379575670797e-07, "loss": 0.0672, "step": 19979 }, { "epoch": 0.9284386617100372, "grad_norm": 0.4663900007199169, "learning_rate": 1.37878240643054e-07, "loss": 0.0677, "step": 19980 }, { "epoch": 0.9284851301115241, "grad_norm": 0.8742111062458028, "learning_rate": 1.3770279566157452e-07, "loss": 0.0604, "step": 19981 }, { "epoch": 0.9285315985130111, "grad_norm": 0.49220298095326037, "learning_rate": 1.375274608162447e-07, "loss": 0.0621, "step": 19982 }, { "epoch": 0.9285780669144982, "grad_norm": 0.42789359471356914, "learning_rate": 1.3735223611103464e-07, "loss": 0.0477, "step": 19983 }, { "epoch": 0.9286245353159851, "grad_norm": 0.6667285447200743, "learning_rate": 1.3717712154991568e-07, "loss": 0.0508, "step": 19984 }, { "epoch": 0.9286710037174721, "grad_norm": 0.359078944651611, "learning_rate": 1.370021171368524e-07, "loss": 0.0484, "step": 19985 }, { "epoch": 0.928717472118959, "grad_norm": 0.5929313945605987, "learning_rate": 1.368272228758083e-07, "loss": 0.0478, "step": 19986 }, { "epoch": 0.9287639405204461, "grad_norm": 0.5035716432517937, "learning_rate": 1.366524387707474e-07, "loss": 0.0467, "step": 19987 }, { "epoch": 0.9288104089219331, "grad_norm": 0.48048575615335204, "learning_rate": 1.3647776482562658e-07, "loss": 0.0434, "step": 19988 }, { "epoch": 0.92885687732342, "grad_norm": 0.629174729043586, "learning_rate": 1.3630320104440376e-07, "loss": 0.0504, "step": 19989 }, { "epoch": 0.9289033457249071, "grad_norm": 0.32756197710478246, "learning_rate": 1.3612874743103188e-07, "loss": 0.0431, "step": 19990 }, { "epoch": 0.9289498141263941, "grad_norm": 0.3854392635043449, "learning_rate": 1.3595440398946334e-07, "loss": 0.0465, "step": 19991 }, { "epoch": 0.928996282527881, "grad_norm": 0.4792856410006351, "learning_rate": 1.357801707236467e-07, "loss": 0.0663, "step": 19992 }, { "epoch": 0.929042750929368, "grad_norm": 0.5463890809461949, "learning_rate": 1.3560604763752815e-07, "loss": 0.0436, "step": 19993 }, { "epoch": 0.929089219330855, "grad_norm": 0.6098489245010098, "learning_rate": 1.3543203473505295e-07, "loss": 0.048, "step": 19994 }, { "epoch": 0.929135687732342, "grad_norm": 0.6785983438107797, "learning_rate": 1.352581320201618e-07, "loss": 0.0608, "step": 19995 }, { "epoch": 0.929182156133829, "grad_norm": 0.5382966110095191, "learning_rate": 1.3508433949679433e-07, "loss": 0.0458, "step": 19996 }, { "epoch": 0.929228624535316, "grad_norm": 0.2916007879335923, "learning_rate": 1.3491065716888628e-07, "loss": 0.0399, "step": 19997 }, { "epoch": 0.929275092936803, "grad_norm": 0.5205132618076249, "learning_rate": 1.3473708504037286e-07, "loss": 0.0465, "step": 19998 }, { "epoch": 0.9293215613382899, "grad_norm": 0.3081066423054546, "learning_rate": 1.3456362311518478e-07, "loss": 0.0357, "step": 19999 }, { "epoch": 0.929368029739777, "grad_norm": 0.4139792838729402, "learning_rate": 1.343902713972506e-07, "loss": 0.0525, "step": 20000 }, { "epoch": 0.929414498141264, "grad_norm": 0.38358529256562085, "learning_rate": 1.3421702989049878e-07, "loss": 0.0194, "step": 20001 }, { "epoch": 0.9294609665427509, "grad_norm": 0.5382668171165803, "learning_rate": 1.340438985988529e-07, "loss": 0.0533, "step": 20002 }, { "epoch": 0.9295074349442379, "grad_norm": 0.5595883095557912, "learning_rate": 1.338708775262326e-07, "loss": 0.059, "step": 20003 }, { "epoch": 0.9295539033457249, "grad_norm": 0.8826652419068098, "learning_rate": 1.3369796667655922e-07, "loss": 0.0539, "step": 20004 }, { "epoch": 0.9296003717472119, "grad_norm": 0.652288255316176, "learning_rate": 1.3352516605374844e-07, "loss": 0.0473, "step": 20005 }, { "epoch": 0.9296468401486989, "grad_norm": 0.8050299581019739, "learning_rate": 1.3335247566171439e-07, "loss": 0.0705, "step": 20006 }, { "epoch": 0.9296933085501858, "grad_norm": 0.6148442581893209, "learning_rate": 1.3317989550436838e-07, "loss": 0.0412, "step": 20007 }, { "epoch": 0.9297397769516729, "grad_norm": 0.717358826705494, "learning_rate": 1.330074255856212e-07, "loss": 0.0583, "step": 20008 }, { "epoch": 0.9297862453531599, "grad_norm": 0.562538333258078, "learning_rate": 1.32835065909378e-07, "loss": 0.0501, "step": 20009 }, { "epoch": 0.9298327137546468, "grad_norm": 0.4747659291001869, "learning_rate": 1.3266281647954238e-07, "loss": 0.0483, "step": 20010 }, { "epoch": 0.9298791821561339, "grad_norm": 0.35995329527802006, "learning_rate": 1.3249067730001675e-07, "loss": 0.048, "step": 20011 }, { "epoch": 0.9299256505576208, "grad_norm": 0.3201649909493338, "learning_rate": 1.3231864837470078e-07, "loss": 0.0427, "step": 20012 }, { "epoch": 0.9299721189591078, "grad_norm": 0.4349333867927877, "learning_rate": 1.3214672970749075e-07, "loss": 0.0414, "step": 20013 }, { "epoch": 0.9300185873605948, "grad_norm": 0.41505891164573994, "learning_rate": 1.319749213022803e-07, "loss": 0.0648, "step": 20014 }, { "epoch": 0.9300650557620818, "grad_norm": 0.31386465612607134, "learning_rate": 1.3180322316296125e-07, "loss": 0.0533, "step": 20015 }, { "epoch": 0.9301115241635688, "grad_norm": 0.7194784512323186, "learning_rate": 1.3163163529342272e-07, "loss": 0.0618, "step": 20016 }, { "epoch": 0.9301579925650557, "grad_norm": 0.3708153305108931, "learning_rate": 1.3146015769755162e-07, "loss": 0.0604, "step": 20017 }, { "epoch": 0.9302044609665427, "grad_norm": 0.6063025763470034, "learning_rate": 1.3128879037923204e-07, "loss": 0.0512, "step": 20018 }, { "epoch": 0.9302509293680298, "grad_norm": 0.25015658709275995, "learning_rate": 1.3111753334234644e-07, "loss": 0.0365, "step": 20019 }, { "epoch": 0.9302973977695167, "grad_norm": 0.7262508663795147, "learning_rate": 1.3094638659077174e-07, "loss": 0.0714, "step": 20020 }, { "epoch": 0.9303438661710037, "grad_norm": 0.5250555965914692, "learning_rate": 1.3077535012838648e-07, "loss": 0.0852, "step": 20021 }, { "epoch": 0.9303903345724907, "grad_norm": 0.4005233444178426, "learning_rate": 1.306044239590648e-07, "loss": 0.0563, "step": 20022 }, { "epoch": 0.9304368029739777, "grad_norm": 0.5317477846665318, "learning_rate": 1.3043360808667749e-07, "loss": 0.0543, "step": 20023 }, { "epoch": 0.9304832713754647, "grad_norm": 0.7726426239050753, "learning_rate": 1.3026290251509365e-07, "loss": 0.0477, "step": 20024 }, { "epoch": 0.9305297397769516, "grad_norm": 0.5476229515663572, "learning_rate": 1.3009230724818134e-07, "loss": 0.0453, "step": 20025 }, { "epoch": 0.9305762081784387, "grad_norm": 0.5164496362841035, "learning_rate": 1.2992182228980299e-07, "loss": 0.0712, "step": 20026 }, { "epoch": 0.9306226765799257, "grad_norm": 0.6511272899784238, "learning_rate": 1.2975144764382108e-07, "loss": 0.052, "step": 20027 }, { "epoch": 0.9306691449814126, "grad_norm": 0.32099130550621946, "learning_rate": 1.2958118331409475e-07, "loss": 0.0526, "step": 20028 }, { "epoch": 0.9307156133828997, "grad_norm": 0.48314852404659575, "learning_rate": 1.2941102930448091e-07, "loss": 0.0531, "step": 20029 }, { "epoch": 0.9307620817843866, "grad_norm": 0.666453404625045, "learning_rate": 1.2924098561883368e-07, "loss": 0.0554, "step": 20030 }, { "epoch": 0.9308085501858736, "grad_norm": 0.6240926444691844, "learning_rate": 1.2907105226100392e-07, "loss": 0.0435, "step": 20031 }, { "epoch": 0.9308550185873606, "grad_norm": 0.5573271322890241, "learning_rate": 1.2890122923484182e-07, "loss": 0.0376, "step": 20032 }, { "epoch": 0.9309014869888476, "grad_norm": 0.38378315638781846, "learning_rate": 1.2873151654419324e-07, "loss": 0.0397, "step": 20033 }, { "epoch": 0.9309479553903346, "grad_norm": 0.6966063325691876, "learning_rate": 1.2856191419290288e-07, "loss": 0.0547, "step": 20034 }, { "epoch": 0.9309944237918215, "grad_norm": 0.36630005618281886, "learning_rate": 1.2839242218481208e-07, "loss": 0.0431, "step": 20035 }, { "epoch": 0.9310408921933085, "grad_norm": 0.7141071254261898, "learning_rate": 1.2822304052376056e-07, "loss": 0.0419, "step": 20036 }, { "epoch": 0.9310873605947956, "grad_norm": 0.35099324295479895, "learning_rate": 1.280537692135847e-07, "loss": 0.0415, "step": 20037 }, { "epoch": 0.9311338289962825, "grad_norm": 0.33250627320518833, "learning_rate": 1.2788460825811811e-07, "loss": 0.0322, "step": 20038 }, { "epoch": 0.9311802973977695, "grad_norm": 0.3495918868628534, "learning_rate": 1.2771555766119327e-07, "loss": 0.0423, "step": 20039 }, { "epoch": 0.9312267657992565, "grad_norm": 0.5916914984920909, "learning_rate": 1.275466174266393e-07, "loss": 0.0334, "step": 20040 }, { "epoch": 0.9312732342007435, "grad_norm": 0.36758302340177157, "learning_rate": 1.2737778755828157e-07, "loss": 0.0597, "step": 20041 }, { "epoch": 0.9313197026022305, "grad_norm": 0.8156814116701987, "learning_rate": 1.2720906805994637e-07, "loss": 0.054, "step": 20042 }, { "epoch": 0.9313661710037174, "grad_norm": 0.5079635017030645, "learning_rate": 1.2704045893545403e-07, "loss": 0.0394, "step": 20043 }, { "epoch": 0.9314126394052045, "grad_norm": 0.4188698972256886, "learning_rate": 1.268719601886237e-07, "loss": 0.0415, "step": 20044 }, { "epoch": 0.9314591078066915, "grad_norm": 0.5615501827042169, "learning_rate": 1.267035718232723e-07, "loss": 0.0587, "step": 20045 }, { "epoch": 0.9315055762081784, "grad_norm": 0.6129506998319801, "learning_rate": 1.2653529384321462e-07, "loss": 0.0587, "step": 20046 }, { "epoch": 0.9315520446096655, "grad_norm": 0.6435689775989227, "learning_rate": 1.2636712625226144e-07, "loss": 0.0548, "step": 20047 }, { "epoch": 0.9315985130111524, "grad_norm": 0.574738846040072, "learning_rate": 1.2619906905422196e-07, "loss": 0.0479, "step": 20048 }, { "epoch": 0.9316449814126394, "grad_norm": 0.44979610048453217, "learning_rate": 1.260311222529037e-07, "loss": 0.0512, "step": 20049 }, { "epoch": 0.9316914498141264, "grad_norm": 0.7575586733702401, "learning_rate": 1.2586328585211027e-07, "loss": 0.0733, "step": 20050 }, { "epoch": 0.9317379182156134, "grad_norm": 0.7239049826942464, "learning_rate": 1.256955598556425e-07, "loss": 0.0638, "step": 20051 }, { "epoch": 0.9317843866171004, "grad_norm": 0.532594754842482, "learning_rate": 1.255279442673013e-07, "loss": 0.0473, "step": 20052 }, { "epoch": 0.9318308550185873, "grad_norm": 0.676343457872099, "learning_rate": 1.253604390908819e-07, "loss": 0.0343, "step": 20053 }, { "epoch": 0.9318773234200743, "grad_norm": 0.38854315159703123, "learning_rate": 1.2519304433017964e-07, "loss": 0.0458, "step": 20054 }, { "epoch": 0.9319237918215614, "grad_norm": 0.5333921621044927, "learning_rate": 1.2502575998898537e-07, "loss": 0.0582, "step": 20055 }, { "epoch": 0.9319702602230483, "grad_norm": 0.5123243381452447, "learning_rate": 1.2485858607108825e-07, "loss": 0.0674, "step": 20056 }, { "epoch": 0.9320167286245353, "grad_norm": 0.6315253199033075, "learning_rate": 1.2469152258027528e-07, "loss": 0.0692, "step": 20057 }, { "epoch": 0.9320631970260222, "grad_norm": 0.5270447570032141, "learning_rate": 1.2452456952032954e-07, "loss": 0.055, "step": 20058 }, { "epoch": 0.9321096654275093, "grad_norm": 0.37518701064466164, "learning_rate": 1.2435772689503467e-07, "loss": 0.0569, "step": 20059 }, { "epoch": 0.9321561338289963, "grad_norm": 0.4149790573491003, "learning_rate": 1.2419099470816876e-07, "loss": 0.0333, "step": 20060 }, { "epoch": 0.9322026022304832, "grad_norm": 0.7538202956523733, "learning_rate": 1.2402437296350876e-07, "loss": 0.0553, "step": 20061 }, { "epoch": 0.9322490706319703, "grad_norm": 0.43821491536324836, "learning_rate": 1.2385786166482782e-07, "loss": 0.0501, "step": 20062 }, { "epoch": 0.9322955390334573, "grad_norm": 0.3394917797318644, "learning_rate": 1.23691460815899e-07, "loss": 0.0544, "step": 20063 }, { "epoch": 0.9323420074349442, "grad_norm": 0.5462400566355315, "learning_rate": 1.2352517042049095e-07, "loss": 0.0286, "step": 20064 }, { "epoch": 0.9323884758364313, "grad_norm": 0.44784960578949856, "learning_rate": 1.23358990482369e-07, "loss": 0.0462, "step": 20065 }, { "epoch": 0.9324349442379182, "grad_norm": 0.5648830139327168, "learning_rate": 1.2319292100529956e-07, "loss": 0.0556, "step": 20066 }, { "epoch": 0.9324814126394052, "grad_norm": 0.6261510068464108, "learning_rate": 1.2302696199304353e-07, "loss": 0.0527, "step": 20067 }, { "epoch": 0.9325278810408922, "grad_norm": 0.4462650147097766, "learning_rate": 1.2286111344935846e-07, "loss": 0.0491, "step": 20068 }, { "epoch": 0.9325743494423792, "grad_norm": 0.4873126751916508, "learning_rate": 1.2269537537800302e-07, "loss": 0.042, "step": 20069 }, { "epoch": 0.9326208178438662, "grad_norm": 0.7072261892593377, "learning_rate": 1.2252974778273085e-07, "loss": 0.0595, "step": 20070 }, { "epoch": 0.9326672862453531, "grad_norm": 0.4668104280279857, "learning_rate": 1.2236423066729286e-07, "loss": 0.0454, "step": 20071 }, { "epoch": 0.9327137546468401, "grad_norm": 0.5566687327639168, "learning_rate": 1.2219882403543882e-07, "loss": 0.0575, "step": 20072 }, { "epoch": 0.9327602230483272, "grad_norm": 0.319604280434506, "learning_rate": 1.2203352789091515e-07, "loss": 0.0429, "step": 20073 }, { "epoch": 0.9328066914498141, "grad_norm": 0.5876233616231311, "learning_rate": 1.2186834223746612e-07, "loss": 0.0397, "step": 20074 }, { "epoch": 0.9328531598513011, "grad_norm": 0.5423111252017295, "learning_rate": 1.2170326707883317e-07, "loss": 0.0678, "step": 20075 }, { "epoch": 0.932899628252788, "grad_norm": 0.9383868759981672, "learning_rate": 1.2153830241875554e-07, "loss": 0.05, "step": 20076 }, { "epoch": 0.9329460966542751, "grad_norm": 0.4191511252755411, "learning_rate": 1.2137344826097018e-07, "loss": 0.0584, "step": 20077 }, { "epoch": 0.9329925650557621, "grad_norm": 0.4707128085441032, "learning_rate": 1.212087046092103e-07, "loss": 0.0419, "step": 20078 }, { "epoch": 0.933039033457249, "grad_norm": 0.5298308982608075, "learning_rate": 1.2104407146720898e-07, "loss": 0.0589, "step": 20079 }, { "epoch": 0.9330855018587361, "grad_norm": 0.6212002188937417, "learning_rate": 1.2087954883869378e-07, "loss": 0.0564, "step": 20080 }, { "epoch": 0.9331319702602231, "grad_norm": 0.5999033084850168, "learning_rate": 1.2071513672739232e-07, "loss": 0.0563, "step": 20081 }, { "epoch": 0.93317843866171, "grad_norm": 0.6693543502442351, "learning_rate": 1.205508351370277e-07, "loss": 0.0542, "step": 20082 }, { "epoch": 0.933224907063197, "grad_norm": 0.5957998194973755, "learning_rate": 1.203866440713225e-07, "loss": 0.0527, "step": 20083 }, { "epoch": 0.933271375464684, "grad_norm": 0.5219282865798692, "learning_rate": 1.2022256353399597e-07, "loss": 0.0543, "step": 20084 }, { "epoch": 0.933317843866171, "grad_norm": 0.662354844525262, "learning_rate": 1.2005859352876348e-07, "loss": 0.0548, "step": 20085 }, { "epoch": 0.933364312267658, "grad_norm": 0.7699184202160205, "learning_rate": 1.1989473405934037e-07, "loss": 0.0431, "step": 20086 }, { "epoch": 0.933410780669145, "grad_norm": 0.42294871398450146, "learning_rate": 1.1973098512943815e-07, "loss": 0.0491, "step": 20087 }, { "epoch": 0.933457249070632, "grad_norm": 0.44212354167590495, "learning_rate": 1.1956734674276492e-07, "loss": 0.0491, "step": 20088 }, { "epoch": 0.9335037174721189, "grad_norm": 0.7248858513518363, "learning_rate": 1.1940381890302776e-07, "loss": 0.0757, "step": 20089 }, { "epoch": 0.9335501858736059, "grad_norm": 0.6456948034108165, "learning_rate": 1.1924040161393147e-07, "loss": 0.0603, "step": 20090 }, { "epoch": 0.933596654275093, "grad_norm": 0.3868902132610916, "learning_rate": 1.190770948791764e-07, "loss": 0.0606, "step": 20091 }, { "epoch": 0.9336431226765799, "grad_norm": 0.34022781114342515, "learning_rate": 1.189138987024624e-07, "loss": 0.0318, "step": 20092 }, { "epoch": 0.9336895910780669, "grad_norm": 0.4917500419286108, "learning_rate": 1.1875081308748593e-07, "loss": 0.0598, "step": 20093 }, { "epoch": 0.9337360594795538, "grad_norm": 0.753950264165538, "learning_rate": 1.1858783803794072e-07, "loss": 0.0678, "step": 20094 }, { "epoch": 0.9337825278810409, "grad_norm": 0.570926899147312, "learning_rate": 1.1842497355751825e-07, "loss": 0.0495, "step": 20095 }, { "epoch": 0.9338289962825279, "grad_norm": 0.49562361021049506, "learning_rate": 1.1826221964990892e-07, "loss": 0.0365, "step": 20096 }, { "epoch": 0.9338754646840148, "grad_norm": 0.46638591734401297, "learning_rate": 1.180995763187981e-07, "loss": 0.0408, "step": 20097 }, { "epoch": 0.9339219330855019, "grad_norm": 0.5421421948798585, "learning_rate": 1.179370435678695e-07, "loss": 0.0616, "step": 20098 }, { "epoch": 0.9339684014869889, "grad_norm": 0.5914507541456714, "learning_rate": 1.1777462140080464e-07, "loss": 0.0599, "step": 20099 }, { "epoch": 0.9340148698884758, "grad_norm": 0.39440099042971855, "learning_rate": 1.176123098212839e-07, "loss": 0.0338, "step": 20100 }, { "epoch": 0.9340613382899629, "grad_norm": 0.5364819022908612, "learning_rate": 1.1745010883298269e-07, "loss": 0.0424, "step": 20101 }, { "epoch": 0.9341078066914498, "grad_norm": 0.7446170617869527, "learning_rate": 1.1728801843957526e-07, "loss": 0.066, "step": 20102 }, { "epoch": 0.9341542750929368, "grad_norm": 0.479393317922392, "learning_rate": 1.1712603864473315e-07, "loss": 0.0426, "step": 20103 }, { "epoch": 0.9342007434944238, "grad_norm": 0.28664931206434136, "learning_rate": 1.1696416945212563e-07, "loss": 0.0254, "step": 20104 }, { "epoch": 0.9342472118959108, "grad_norm": 0.5657721949155861, "learning_rate": 1.1680241086541921e-07, "loss": 0.0594, "step": 20105 }, { "epoch": 0.9342936802973978, "grad_norm": 0.5697161326561352, "learning_rate": 1.1664076288827653e-07, "loss": 0.0477, "step": 20106 }, { "epoch": 0.9343401486988847, "grad_norm": 0.584612739424701, "learning_rate": 1.1647922552436131e-07, "loss": 0.0455, "step": 20107 }, { "epoch": 0.9343866171003717, "grad_norm": 0.40231126774275366, "learning_rate": 1.1631779877733118e-07, "loss": 0.0436, "step": 20108 }, { "epoch": 0.9344330855018588, "grad_norm": 0.5943935947620659, "learning_rate": 1.1615648265084267e-07, "loss": 0.0549, "step": 20109 }, { "epoch": 0.9344795539033457, "grad_norm": 0.5551813497276706, "learning_rate": 1.1599527714855008e-07, "loss": 0.0653, "step": 20110 }, { "epoch": 0.9345260223048327, "grad_norm": 0.4192851479515673, "learning_rate": 1.1583418227410492e-07, "loss": 0.0406, "step": 20111 }, { "epoch": 0.9345724907063196, "grad_norm": 0.6510388058123465, "learning_rate": 1.156731980311554e-07, "loss": 0.0572, "step": 20112 }, { "epoch": 0.9346189591078067, "grad_norm": 0.5055064750466918, "learning_rate": 1.1551232442334914e-07, "loss": 0.0526, "step": 20113 }, { "epoch": 0.9346654275092937, "grad_norm": 0.6602102121746858, "learning_rate": 1.1535156145433047e-07, "loss": 0.053, "step": 20114 }, { "epoch": 0.9347118959107806, "grad_norm": 0.35678035992961776, "learning_rate": 1.1519090912773867e-07, "loss": 0.0608, "step": 20115 }, { "epoch": 0.9347583643122677, "grad_norm": 0.5543474965128055, "learning_rate": 1.1503036744721419e-07, "loss": 0.059, "step": 20116 }, { "epoch": 0.9348048327137547, "grad_norm": 0.35229792196782217, "learning_rate": 1.14869936416393e-07, "loss": 0.0239, "step": 20117 }, { "epoch": 0.9348513011152416, "grad_norm": 0.6546348876457726, "learning_rate": 1.1470961603890995e-07, "loss": 0.0539, "step": 20118 }, { "epoch": 0.9348977695167286, "grad_norm": 0.31017677661009574, "learning_rate": 1.1454940631839495e-07, "loss": 0.0343, "step": 20119 }, { "epoch": 0.9349442379182156, "grad_norm": 0.6550300295038507, "learning_rate": 1.1438930725847841e-07, "loss": 0.0647, "step": 20120 }, { "epoch": 0.9349907063197026, "grad_norm": 0.4103618599984554, "learning_rate": 1.1422931886278632e-07, "loss": 0.0346, "step": 20121 }, { "epoch": 0.9350371747211896, "grad_norm": 0.486523118340298, "learning_rate": 1.1406944113494189e-07, "loss": 0.041, "step": 20122 }, { "epoch": 0.9350836431226766, "grad_norm": 0.5994784916707796, "learning_rate": 1.1390967407856612e-07, "loss": 0.0485, "step": 20123 }, { "epoch": 0.9351301115241636, "grad_norm": 0.6195078949680114, "learning_rate": 1.1375001769728e-07, "loss": 0.0461, "step": 20124 }, { "epoch": 0.9351765799256505, "grad_norm": 0.41520865520516664, "learning_rate": 1.1359047199469841e-07, "loss": 0.0523, "step": 20125 }, { "epoch": 0.9352230483271375, "grad_norm": 0.5264847162155666, "learning_rate": 1.1343103697443514e-07, "loss": 0.039, "step": 20126 }, { "epoch": 0.9352695167286246, "grad_norm": 0.4388702836657229, "learning_rate": 1.1327171264010228e-07, "loss": 0.0369, "step": 20127 }, { "epoch": 0.9353159851301115, "grad_norm": 0.7582903020909179, "learning_rate": 1.1311249899530807e-07, "loss": 0.0426, "step": 20128 }, { "epoch": 0.9353624535315985, "grad_norm": 0.44991267912822364, "learning_rate": 1.1295339604365851e-07, "loss": 0.0442, "step": 20129 }, { "epoch": 0.9354089219330854, "grad_norm": 0.5797129198463848, "learning_rate": 1.1279440378875906e-07, "loss": 0.0404, "step": 20130 }, { "epoch": 0.9354553903345725, "grad_norm": 0.570380498243317, "learning_rate": 1.1263552223421015e-07, "loss": 0.0427, "step": 20131 }, { "epoch": 0.9355018587360595, "grad_norm": 0.6109085507448777, "learning_rate": 1.1247675138361003e-07, "loss": 0.0463, "step": 20132 }, { "epoch": 0.9355483271375464, "grad_norm": 0.4967862355507007, "learning_rate": 1.1231809124055527e-07, "loss": 0.0515, "step": 20133 }, { "epoch": 0.9355947955390335, "grad_norm": 0.4557513948277488, "learning_rate": 1.121595418086402e-07, "loss": 0.0436, "step": 20134 }, { "epoch": 0.9356412639405205, "grad_norm": 0.37337469554655905, "learning_rate": 1.1200110309145584e-07, "loss": 0.0541, "step": 20135 }, { "epoch": 0.9356877323420074, "grad_norm": 0.39971786626689165, "learning_rate": 1.1184277509259045e-07, "loss": 0.0592, "step": 20136 }, { "epoch": 0.9357342007434944, "grad_norm": 0.45132849506298883, "learning_rate": 1.1168455781563225e-07, "loss": 0.0327, "step": 20137 }, { "epoch": 0.9357806691449814, "grad_norm": 0.771081951576693, "learning_rate": 1.1152645126416284e-07, "loss": 0.0393, "step": 20138 }, { "epoch": 0.9358271375464684, "grad_norm": 0.5722177985297023, "learning_rate": 1.1136845544176433e-07, "loss": 0.0484, "step": 20139 }, { "epoch": 0.9358736059479554, "grad_norm": 0.45233209733428315, "learning_rate": 1.1121057035201554e-07, "loss": 0.0425, "step": 20140 }, { "epoch": 0.9359200743494424, "grad_norm": 0.515665820958324, "learning_rate": 1.110527959984925e-07, "loss": 0.0531, "step": 20141 }, { "epoch": 0.9359665427509294, "grad_norm": 0.46371641736606267, "learning_rate": 1.1089513238476956e-07, "loss": 0.0407, "step": 20142 }, { "epoch": 0.9360130111524163, "grad_norm": 0.8339444945626053, "learning_rate": 1.1073757951441778e-07, "loss": 0.0467, "step": 20143 }, { "epoch": 0.9360594795539033, "grad_norm": 0.46687904814436915, "learning_rate": 1.1058013739100537e-07, "loss": 0.0477, "step": 20144 }, { "epoch": 0.9361059479553904, "grad_norm": 0.5895134850190502, "learning_rate": 1.104228060180984e-07, "loss": 0.0585, "step": 20145 }, { "epoch": 0.9361524163568773, "grad_norm": 0.5198749547137685, "learning_rate": 1.1026558539926124e-07, "loss": 0.0668, "step": 20146 }, { "epoch": 0.9361988847583643, "grad_norm": 0.5012613327420251, "learning_rate": 1.1010847553805548e-07, "loss": 0.0472, "step": 20147 }, { "epoch": 0.9362453531598512, "grad_norm": 0.7085552015692577, "learning_rate": 1.0995147643803882e-07, "loss": 0.0751, "step": 20148 }, { "epoch": 0.9362918215613383, "grad_norm": 0.6040526042279726, "learning_rate": 1.0979458810276844e-07, "loss": 0.0424, "step": 20149 }, { "epoch": 0.9363382899628253, "grad_norm": 0.49837333238119547, "learning_rate": 1.0963781053579648e-07, "loss": 0.0287, "step": 20150 }, { "epoch": 0.9363847583643122, "grad_norm": 0.5612281510682586, "learning_rate": 1.0948114374067564e-07, "loss": 0.0694, "step": 20151 }, { "epoch": 0.9364312267657993, "grad_norm": 0.5162181787130916, "learning_rate": 1.0932458772095423e-07, "loss": 0.0522, "step": 20152 }, { "epoch": 0.9364776951672863, "grad_norm": 0.6243701894882329, "learning_rate": 1.0916814248017771e-07, "loss": 0.0668, "step": 20153 }, { "epoch": 0.9365241635687732, "grad_norm": 0.4050366038547122, "learning_rate": 1.0901180802189104e-07, "loss": 0.0429, "step": 20154 }, { "epoch": 0.9365706319702602, "grad_norm": 0.5878503070060096, "learning_rate": 1.0885558434963472e-07, "loss": 0.0549, "step": 20155 }, { "epoch": 0.9366171003717472, "grad_norm": 0.30404850634961333, "learning_rate": 1.0869947146694649e-07, "loss": 0.0383, "step": 20156 }, { "epoch": 0.9366635687732342, "grad_norm": 0.41873233933028847, "learning_rate": 1.0854346937736349e-07, "loss": 0.0503, "step": 20157 }, { "epoch": 0.9367100371747212, "grad_norm": 0.6159840358954964, "learning_rate": 1.0838757808441902e-07, "loss": 0.0638, "step": 20158 }, { "epoch": 0.9367565055762082, "grad_norm": 0.6926646346633566, "learning_rate": 1.0823179759164471e-07, "loss": 0.0666, "step": 20159 }, { "epoch": 0.9368029739776952, "grad_norm": 0.7865250549194752, "learning_rate": 1.0807612790256771e-07, "loss": 0.059, "step": 20160 }, { "epoch": 0.9368494423791821, "grad_norm": 0.5975956822752095, "learning_rate": 1.0792056902071635e-07, "loss": 0.0447, "step": 20161 }, { "epoch": 0.9368959107806691, "grad_norm": 0.4662962446571517, "learning_rate": 1.0776512094961222e-07, "loss": 0.0495, "step": 20162 }, { "epoch": 0.9369423791821562, "grad_norm": 0.3766800590703695, "learning_rate": 1.0760978369277642e-07, "loss": 0.0458, "step": 20163 }, { "epoch": 0.9369888475836431, "grad_norm": 0.5464853835645158, "learning_rate": 1.0745455725372889e-07, "loss": 0.0523, "step": 20164 }, { "epoch": 0.9370353159851301, "grad_norm": 0.39929211243977397, "learning_rate": 1.0729944163598516e-07, "loss": 0.0451, "step": 20165 }, { "epoch": 0.937081784386617, "grad_norm": 0.4507011223288622, "learning_rate": 1.0714443684305853e-07, "loss": 0.0487, "step": 20166 }, { "epoch": 0.9371282527881041, "grad_norm": 0.516588569771525, "learning_rate": 1.0698954287845952e-07, "loss": 0.0512, "step": 20167 }, { "epoch": 0.9371747211895911, "grad_norm": 0.5027386296847668, "learning_rate": 1.0683475974569757e-07, "loss": 0.0438, "step": 20168 }, { "epoch": 0.937221189591078, "grad_norm": 0.4865296730609416, "learning_rate": 1.0668008744827818e-07, "loss": 0.0559, "step": 20169 }, { "epoch": 0.9372676579925651, "grad_norm": 0.3201158840181613, "learning_rate": 1.0652552598970467e-07, "loss": 0.0457, "step": 20170 }, { "epoch": 0.9373141263940521, "grad_norm": 0.45418836985565764, "learning_rate": 1.0637107537347868e-07, "loss": 0.039, "step": 20171 }, { "epoch": 0.937360594795539, "grad_norm": 1.1008492924657922, "learning_rate": 1.0621673560309798e-07, "loss": 0.0461, "step": 20172 }, { "epoch": 0.937407063197026, "grad_norm": 0.4624373677368476, "learning_rate": 1.060625066820592e-07, "loss": 0.0686, "step": 20173 }, { "epoch": 0.937453531598513, "grad_norm": 0.5882158875945805, "learning_rate": 1.059083886138551e-07, "loss": 0.0653, "step": 20174 }, { "epoch": 0.9375, "grad_norm": 0.3709041011922713, "learning_rate": 1.057543814019768e-07, "loss": 0.0561, "step": 20175 }, { "epoch": 0.937546468401487, "grad_norm": 0.4409334300320358, "learning_rate": 1.0560048504991316e-07, "loss": 0.0452, "step": 20176 }, { "epoch": 0.937592936802974, "grad_norm": 0.6740650985646338, "learning_rate": 1.0544669956114972e-07, "loss": 0.0471, "step": 20177 }, { "epoch": 0.937639405204461, "grad_norm": 0.565535504731979, "learning_rate": 1.0529302493916981e-07, "loss": 0.0547, "step": 20178 }, { "epoch": 0.9376858736059479, "grad_norm": 0.24811437473461742, "learning_rate": 1.051394611874551e-07, "loss": 0.0328, "step": 20179 }, { "epoch": 0.9377323420074349, "grad_norm": 0.3271314942233248, "learning_rate": 1.0498600830948224e-07, "loss": 0.0465, "step": 20180 }, { "epoch": 0.937778810408922, "grad_norm": 0.34677898733082013, "learning_rate": 1.0483266630872902e-07, "loss": 0.0447, "step": 20181 }, { "epoch": 0.9378252788104089, "grad_norm": 0.7030950224236148, "learning_rate": 1.0467943518866818e-07, "loss": 0.0491, "step": 20182 }, { "epoch": 0.9378717472118959, "grad_norm": 0.4493533002245193, "learning_rate": 1.0452631495276977e-07, "loss": 0.0446, "step": 20183 }, { "epoch": 0.937918215613383, "grad_norm": 0.7268885258747048, "learning_rate": 1.0437330560450265e-07, "loss": 0.0653, "step": 20184 }, { "epoch": 0.9379646840148699, "grad_norm": 0.5732774522788437, "learning_rate": 1.042204071473335e-07, "loss": 0.0554, "step": 20185 }, { "epoch": 0.9380111524163569, "grad_norm": 0.5037556252006542, "learning_rate": 1.04067619584724e-07, "loss": 0.0329, "step": 20186 }, { "epoch": 0.9380576208178438, "grad_norm": 0.7948820276609013, "learning_rate": 1.0391494292013582e-07, "loss": 0.0482, "step": 20187 }, { "epoch": 0.9381040892193309, "grad_norm": 0.4412621156203391, "learning_rate": 1.0376237715702786e-07, "loss": 0.0516, "step": 20188 }, { "epoch": 0.9381505576208179, "grad_norm": 0.671877675128752, "learning_rate": 1.0360992229885458e-07, "loss": 0.0523, "step": 20189 }, { "epoch": 0.9381970260223048, "grad_norm": 0.37908535143278166, "learning_rate": 1.0345757834907045e-07, "loss": 0.037, "step": 20190 }, { "epoch": 0.9382434944237918, "grad_norm": 0.4553552443353575, "learning_rate": 1.0330534531112546e-07, "loss": 0.0654, "step": 20191 }, { "epoch": 0.9382899628252788, "grad_norm": 0.39215891186522833, "learning_rate": 1.0315322318846855e-07, "loss": 0.0422, "step": 20192 }, { "epoch": 0.9383364312267658, "grad_norm": 0.5903147560109725, "learning_rate": 1.0300121198454416e-07, "loss": 0.0381, "step": 20193 }, { "epoch": 0.9383828996282528, "grad_norm": 0.5064891735697168, "learning_rate": 1.0284931170279622e-07, "loss": 0.055, "step": 20194 }, { "epoch": 0.9384293680297398, "grad_norm": 0.4936102908929961, "learning_rate": 1.0269752234666642e-07, "loss": 0.0433, "step": 20195 }, { "epoch": 0.9384758364312268, "grad_norm": 0.34100234321656053, "learning_rate": 1.02545843919592e-07, "loss": 0.0495, "step": 20196 }, { "epoch": 0.9385223048327137, "grad_norm": 0.32142573803832414, "learning_rate": 1.0239427642500799e-07, "loss": 0.0443, "step": 20197 }, { "epoch": 0.9385687732342007, "grad_norm": 0.4626236457443223, "learning_rate": 1.0224281986634832e-07, "loss": 0.055, "step": 20198 }, { "epoch": 0.9386152416356878, "grad_norm": 0.39688567215431336, "learning_rate": 1.0209147424704413e-07, "loss": 0.0472, "step": 20199 }, { "epoch": 0.9386617100371747, "grad_norm": 0.4833114179026219, "learning_rate": 1.0194023957052269e-07, "loss": 0.0587, "step": 20200 }, { "epoch": 0.9387081784386617, "grad_norm": 0.30084669167274336, "learning_rate": 1.0178911584020956e-07, "loss": 0.0347, "step": 20201 }, { "epoch": 0.9387546468401488, "grad_norm": 0.37552435637629206, "learning_rate": 1.0163810305952925e-07, "loss": 0.0289, "step": 20202 }, { "epoch": 0.9388011152416357, "grad_norm": 0.6495974941098737, "learning_rate": 1.014872012319007e-07, "loss": 0.0485, "step": 20203 }, { "epoch": 0.9388475836431227, "grad_norm": 0.5087804671920915, "learning_rate": 1.0133641036074226e-07, "loss": 0.0514, "step": 20204 }, { "epoch": 0.9388940520446096, "grad_norm": 0.6094228907565727, "learning_rate": 1.0118573044947011e-07, "loss": 0.0362, "step": 20205 }, { "epoch": 0.9389405204460967, "grad_norm": 0.529407130372839, "learning_rate": 1.0103516150149762e-07, "loss": 0.0459, "step": 20206 }, { "epoch": 0.9389869888475837, "grad_norm": 0.5347383809686801, "learning_rate": 1.0088470352023427e-07, "loss": 0.068, "step": 20207 }, { "epoch": 0.9390334572490706, "grad_norm": 1.374982822581465, "learning_rate": 1.0073435650908902e-07, "loss": 0.0532, "step": 20208 }, { "epoch": 0.9390799256505576, "grad_norm": 0.5942232715267032, "learning_rate": 1.0058412047146637e-07, "loss": 0.057, "step": 20209 }, { "epoch": 0.9391263940520446, "grad_norm": 0.46019894383392485, "learning_rate": 1.0043399541077026e-07, "loss": 0.0391, "step": 20210 }, { "epoch": 0.9391728624535316, "grad_norm": 0.560620653904312, "learning_rate": 1.0028398133040074e-07, "loss": 0.0453, "step": 20211 }, { "epoch": 0.9392193308550186, "grad_norm": 0.47320165456180147, "learning_rate": 1.001340782337562e-07, "loss": 0.0601, "step": 20212 }, { "epoch": 0.9392657992565056, "grad_norm": 0.6803747677015581, "learning_rate": 9.998428612423172e-08, "loss": 0.0416, "step": 20213 }, { "epoch": 0.9393122676579926, "grad_norm": 0.34637594525032017, "learning_rate": 9.983460500522014e-08, "loss": 0.0511, "step": 20214 }, { "epoch": 0.9393587360594795, "grad_norm": 0.6306524051262204, "learning_rate": 9.968503488011206e-08, "loss": 0.0504, "step": 20215 }, { "epoch": 0.9394052044609665, "grad_norm": 0.2605452903925822, "learning_rate": 9.953557575229589e-08, "loss": 0.0376, "step": 20216 }, { "epoch": 0.9394516728624536, "grad_norm": 0.6181809430059316, "learning_rate": 9.938622762515615e-08, "loss": 0.0574, "step": 20217 }, { "epoch": 0.9394981412639405, "grad_norm": 0.5098749570068707, "learning_rate": 9.923699050207569e-08, "loss": 0.0424, "step": 20218 }, { "epoch": 0.9395446096654275, "grad_norm": 0.42725765225377127, "learning_rate": 9.908786438643625e-08, "loss": 0.047, "step": 20219 }, { "epoch": 0.9395910780669146, "grad_norm": 0.4103196820365561, "learning_rate": 9.893884928161457e-08, "loss": 0.0451, "step": 20220 }, { "epoch": 0.9396375464684015, "grad_norm": 0.5354131462080763, "learning_rate": 9.878994519098573e-08, "loss": 0.0594, "step": 20221 }, { "epoch": 0.9396840148698885, "grad_norm": 0.7242892299603666, "learning_rate": 9.864115211792369e-08, "loss": 0.0568, "step": 20222 }, { "epoch": 0.9397304832713754, "grad_norm": 0.5211514824905932, "learning_rate": 9.849247006579799e-08, "loss": 0.0537, "step": 20223 }, { "epoch": 0.9397769516728625, "grad_norm": 0.5405853994612158, "learning_rate": 9.834389903797647e-08, "loss": 0.0556, "step": 20224 }, { "epoch": 0.9398234200743495, "grad_norm": 0.45240373799007105, "learning_rate": 9.819543903782425e-08, "loss": 0.0563, "step": 20225 }, { "epoch": 0.9398698884758364, "grad_norm": 0.5082957181457394, "learning_rate": 9.804709006870472e-08, "loss": 0.0504, "step": 20226 }, { "epoch": 0.9399163568773234, "grad_norm": 0.529333091596775, "learning_rate": 9.789885213397798e-08, "loss": 0.047, "step": 20227 }, { "epoch": 0.9399628252788104, "grad_norm": 0.49341398793265245, "learning_rate": 9.775072523700135e-08, "loss": 0.0564, "step": 20228 }, { "epoch": 0.9400092936802974, "grad_norm": 0.6036954773593022, "learning_rate": 9.760270938113048e-08, "loss": 0.0516, "step": 20229 }, { "epoch": 0.9400557620817844, "grad_norm": 0.7026270887445126, "learning_rate": 9.745480456971768e-08, "loss": 0.0593, "step": 20230 }, { "epoch": 0.9401022304832714, "grad_norm": 1.1875473307618378, "learning_rate": 9.73070108061136e-08, "loss": 0.0493, "step": 20231 }, { "epoch": 0.9401486988847584, "grad_norm": 0.40219413140409505, "learning_rate": 9.715932809366557e-08, "loss": 0.0382, "step": 20232 }, { "epoch": 0.9401951672862453, "grad_norm": 0.4596277230209454, "learning_rate": 9.701175643571925e-08, "loss": 0.0522, "step": 20233 }, { "epoch": 0.9402416356877323, "grad_norm": 0.7167292215202675, "learning_rate": 9.686429583561641e-08, "loss": 0.0758, "step": 20234 }, { "epoch": 0.9402881040892194, "grad_norm": 0.45535656652838036, "learning_rate": 9.67169462966977e-08, "loss": 0.0516, "step": 20235 }, { "epoch": 0.9403345724907063, "grad_norm": 0.3029938423073965, "learning_rate": 9.656970782230102e-08, "loss": 0.0325, "step": 20236 }, { "epoch": 0.9403810408921933, "grad_norm": 0.4872154642741504, "learning_rate": 9.642258041576146e-08, "loss": 0.0456, "step": 20237 }, { "epoch": 0.9404275092936804, "grad_norm": 0.5658090237381801, "learning_rate": 9.627556408041139e-08, "loss": 0.0485, "step": 20238 }, { "epoch": 0.9404739776951673, "grad_norm": 0.6634000658611924, "learning_rate": 9.612865881958034e-08, "loss": 0.0653, "step": 20239 }, { "epoch": 0.9405204460966543, "grad_norm": 0.4037237061106797, "learning_rate": 9.598186463659676e-08, "loss": 0.0342, "step": 20240 }, { "epoch": 0.9405669144981412, "grad_norm": 0.5307638740400662, "learning_rate": 9.583518153478577e-08, "loss": 0.0471, "step": 20241 }, { "epoch": 0.9406133828996283, "grad_norm": 0.514630273666311, "learning_rate": 9.568860951746862e-08, "loss": 0.0594, "step": 20242 }, { "epoch": 0.9406598513011153, "grad_norm": 0.41551940012772637, "learning_rate": 9.554214858796651e-08, "loss": 0.0431, "step": 20243 }, { "epoch": 0.9407063197026022, "grad_norm": 0.575408675104462, "learning_rate": 9.539579874959681e-08, "loss": 0.0596, "step": 20244 }, { "epoch": 0.9407527881040892, "grad_norm": 0.3387355127098399, "learning_rate": 9.524956000567353e-08, "loss": 0.0403, "step": 20245 }, { "epoch": 0.9407992565055762, "grad_norm": 0.6970895869093756, "learning_rate": 9.51034323595107e-08, "loss": 0.0557, "step": 20246 }, { "epoch": 0.9408457249070632, "grad_norm": 0.65066412507733, "learning_rate": 9.495741581441731e-08, "loss": 0.0675, "step": 20247 }, { "epoch": 0.9408921933085502, "grad_norm": 0.41346313544365754, "learning_rate": 9.481151037370073e-08, "loss": 0.0421, "step": 20248 }, { "epoch": 0.9409386617100371, "grad_norm": 0.4057123663526446, "learning_rate": 9.46657160406661e-08, "loss": 0.0287, "step": 20249 }, { "epoch": 0.9409851301115242, "grad_norm": 0.7714169391595698, "learning_rate": 9.452003281861577e-08, "loss": 0.0439, "step": 20250 }, { "epoch": 0.9410315985130111, "grad_norm": 0.3561019005478534, "learning_rate": 9.437446071084933e-08, "loss": 0.0558, "step": 20251 }, { "epoch": 0.9410780669144981, "grad_norm": 0.44199633382407794, "learning_rate": 9.422899972066468e-08, "loss": 0.04, "step": 20252 }, { "epoch": 0.9411245353159852, "grad_norm": 0.8215837240721473, "learning_rate": 9.408364985135642e-08, "loss": 0.0563, "step": 20253 }, { "epoch": 0.9411710037174721, "grad_norm": 0.3949292965448251, "learning_rate": 9.393841110621692e-08, "loss": 0.0315, "step": 20254 }, { "epoch": 0.9412174721189591, "grad_norm": 0.27303903719304995, "learning_rate": 9.379328348853577e-08, "loss": 0.0463, "step": 20255 }, { "epoch": 0.9412639405204462, "grad_norm": 0.48508846886586365, "learning_rate": 9.36482670016009e-08, "loss": 0.0595, "step": 20256 }, { "epoch": 0.9413104089219331, "grad_norm": 0.5502682138569235, "learning_rate": 9.350336164869633e-08, "loss": 0.051, "step": 20257 }, { "epoch": 0.9413568773234201, "grad_norm": 0.76969282831234, "learning_rate": 9.335856743310445e-08, "loss": 0.065, "step": 20258 }, { "epoch": 0.941403345724907, "grad_norm": 0.5593307523002413, "learning_rate": 9.321388435810485e-08, "loss": 0.0468, "step": 20259 }, { "epoch": 0.9414498141263941, "grad_norm": 0.3361176134011326, "learning_rate": 9.306931242697548e-08, "loss": 0.034, "step": 20260 }, { "epoch": 0.9414962825278811, "grad_norm": 0.4246857950480115, "learning_rate": 9.292485164299148e-08, "loss": 0.0506, "step": 20261 }, { "epoch": 0.941542750929368, "grad_norm": 0.5388760457556045, "learning_rate": 9.278050200942301e-08, "loss": 0.0355, "step": 20262 }, { "epoch": 0.941589219330855, "grad_norm": 0.4452396621912812, "learning_rate": 9.263626352954136e-08, "loss": 0.0503, "step": 20263 }, { "epoch": 0.941635687732342, "grad_norm": 0.4232587595141487, "learning_rate": 9.249213620661334e-08, "loss": 0.0402, "step": 20264 }, { "epoch": 0.941682156133829, "grad_norm": 0.5765855765988875, "learning_rate": 9.234812004390303e-08, "loss": 0.0513, "step": 20265 }, { "epoch": 0.941728624535316, "grad_norm": 0.5810940458155528, "learning_rate": 9.22042150446728e-08, "loss": 0.0476, "step": 20266 }, { "epoch": 0.941775092936803, "grad_norm": 0.5812887508669293, "learning_rate": 9.206042121218395e-08, "loss": 0.045, "step": 20267 }, { "epoch": 0.94182156133829, "grad_norm": 0.7380783667179366, "learning_rate": 9.191673854969107e-08, "loss": 0.0682, "step": 20268 }, { "epoch": 0.9418680297397769, "grad_norm": 0.5385668641933972, "learning_rate": 9.177316706044936e-08, "loss": 0.0437, "step": 20269 }, { "epoch": 0.9419144981412639, "grad_norm": 0.6930408641635545, "learning_rate": 9.162970674771177e-08, "loss": 0.044, "step": 20270 }, { "epoch": 0.941960966542751, "grad_norm": 0.5849201470750346, "learning_rate": 9.148635761472679e-08, "loss": 0.0392, "step": 20271 }, { "epoch": 0.9420074349442379, "grad_norm": 0.543139276079247, "learning_rate": 9.134311966474241e-08, "loss": 0.0578, "step": 20272 }, { "epoch": 0.9420539033457249, "grad_norm": 0.540271654006255, "learning_rate": 9.119999290100268e-08, "loss": 0.062, "step": 20273 }, { "epoch": 0.942100371747212, "grad_norm": 0.6341539477558417, "learning_rate": 9.105697732674945e-08, "loss": 0.0681, "step": 20274 }, { "epoch": 0.9421468401486989, "grad_norm": 0.3583908250656133, "learning_rate": 9.091407294522237e-08, "loss": 0.0438, "step": 20275 }, { "epoch": 0.9421933085501859, "grad_norm": 0.42305028764512764, "learning_rate": 9.077127975965716e-08, "loss": 0.0502, "step": 20276 }, { "epoch": 0.9422397769516728, "grad_norm": 0.6909545237076437, "learning_rate": 9.06285977732907e-08, "loss": 0.054, "step": 20277 }, { "epoch": 0.9422862453531599, "grad_norm": 0.579518690527406, "learning_rate": 9.048602698935317e-08, "loss": 0.0539, "step": 20278 }, { "epoch": 0.9423327137546469, "grad_norm": 0.44849994391250897, "learning_rate": 9.034356741107419e-08, "loss": 0.0317, "step": 20279 }, { "epoch": 0.9423791821561338, "grad_norm": 0.7301123484214507, "learning_rate": 9.020121904168067e-08, "loss": 0.0618, "step": 20280 }, { "epoch": 0.9424256505576208, "grad_norm": 0.5565339749031187, "learning_rate": 9.005898188439721e-08, "loss": 0.0589, "step": 20281 }, { "epoch": 0.9424721189591078, "grad_norm": 0.3713741231380309, "learning_rate": 8.991685594244515e-08, "loss": 0.0385, "step": 20282 }, { "epoch": 0.9425185873605948, "grad_norm": 0.44869547523477543, "learning_rate": 8.977484121904357e-08, "loss": 0.0485, "step": 20283 }, { "epoch": 0.9425650557620818, "grad_norm": 1.1027449203716408, "learning_rate": 8.963293771741044e-08, "loss": 0.0752, "step": 20284 }, { "epoch": 0.9426115241635687, "grad_norm": 0.4434590505375998, "learning_rate": 8.949114544075988e-08, "loss": 0.0508, "step": 20285 }, { "epoch": 0.9426579925650558, "grad_norm": 0.5989530814535781, "learning_rate": 8.934946439230208e-08, "loss": 0.0627, "step": 20286 }, { "epoch": 0.9427044609665427, "grad_norm": 0.6537990516307427, "learning_rate": 8.920789457524781e-08, "loss": 0.0596, "step": 20287 }, { "epoch": 0.9427509293680297, "grad_norm": 0.6080824464074309, "learning_rate": 8.906643599280285e-08, "loss": 0.0587, "step": 20288 }, { "epoch": 0.9427973977695168, "grad_norm": 0.30841786803093885, "learning_rate": 8.89250886481724e-08, "loss": 0.0238, "step": 20289 }, { "epoch": 0.9428438661710037, "grad_norm": 0.44964217852226546, "learning_rate": 8.878385254455668e-08, "loss": 0.0639, "step": 20290 }, { "epoch": 0.9428903345724907, "grad_norm": 0.4444285624168363, "learning_rate": 8.864272768515646e-08, "loss": 0.0466, "step": 20291 }, { "epoch": 0.9429368029739778, "grad_norm": 0.5228241021563249, "learning_rate": 8.850171407316754e-08, "loss": 0.0515, "step": 20292 }, { "epoch": 0.9429832713754647, "grad_norm": 0.6393472775693547, "learning_rate": 8.836081171178346e-08, "loss": 0.0489, "step": 20293 }, { "epoch": 0.9430297397769517, "grad_norm": 0.5428520376748327, "learning_rate": 8.822002060419666e-08, "loss": 0.0648, "step": 20294 }, { "epoch": 0.9430762081784386, "grad_norm": 0.4651140140940081, "learning_rate": 8.80793407535957e-08, "loss": 0.0426, "step": 20295 }, { "epoch": 0.9431226765799257, "grad_norm": 0.611230743390413, "learning_rate": 8.793877216316804e-08, "loss": 0.0394, "step": 20296 }, { "epoch": 0.9431691449814127, "grad_norm": 0.44254105676469546, "learning_rate": 8.779831483609668e-08, "loss": 0.0494, "step": 20297 }, { "epoch": 0.9432156133828996, "grad_norm": 0.5002323754207287, "learning_rate": 8.765796877556354e-08, "loss": 0.0373, "step": 20298 }, { "epoch": 0.9432620817843866, "grad_norm": 0.6002871276139397, "learning_rate": 8.751773398474716e-08, "loss": 0.0506, "step": 20299 }, { "epoch": 0.9433085501858736, "grad_norm": 0.43512875155458514, "learning_rate": 8.737761046682447e-08, "loss": 0.0488, "step": 20300 }, { "epoch": 0.9433550185873606, "grad_norm": 0.558313164792673, "learning_rate": 8.723759822496958e-08, "loss": 0.0628, "step": 20301 }, { "epoch": 0.9434014869888476, "grad_norm": 0.3466133679023866, "learning_rate": 8.709769726235385e-08, "loss": 0.0575, "step": 20302 }, { "epoch": 0.9434479553903345, "grad_norm": 0.467138091279981, "learning_rate": 8.695790758214584e-08, "loss": 0.0459, "step": 20303 }, { "epoch": 0.9434944237918216, "grad_norm": 0.4832618406917957, "learning_rate": 8.681822918751248e-08, "loss": 0.0577, "step": 20304 }, { "epoch": 0.9435408921933085, "grad_norm": 0.31397396359450996, "learning_rate": 8.66786620816168e-08, "loss": 0.0374, "step": 20305 }, { "epoch": 0.9435873605947955, "grad_norm": 0.5081580753921994, "learning_rate": 8.653920626762069e-08, "loss": 0.0675, "step": 20306 }, { "epoch": 0.9436338289962826, "grad_norm": 0.4407800424301722, "learning_rate": 8.63998617486822e-08, "loss": 0.0582, "step": 20307 }, { "epoch": 0.9436802973977695, "grad_norm": 0.3787456351948978, "learning_rate": 8.62606285279599e-08, "loss": 0.0567, "step": 20308 }, { "epoch": 0.9437267657992565, "grad_norm": 0.498850615369305, "learning_rate": 8.612150660860519e-08, "loss": 0.0425, "step": 20309 }, { "epoch": 0.9437732342007435, "grad_norm": 0.40014277521683955, "learning_rate": 8.59824959937694e-08, "loss": 0.0311, "step": 20310 }, { "epoch": 0.9438197026022305, "grad_norm": 0.41095062205589294, "learning_rate": 8.58435966866028e-08, "loss": 0.0304, "step": 20311 }, { "epoch": 0.9438661710037175, "grad_norm": 0.6116003832626847, "learning_rate": 8.570480869025066e-08, "loss": 0.058, "step": 20312 }, { "epoch": 0.9439126394052044, "grad_norm": 0.6668937484953839, "learning_rate": 8.556613200785658e-08, "loss": 0.0641, "step": 20313 }, { "epoch": 0.9439591078066915, "grad_norm": 0.6210554521616583, "learning_rate": 8.542756664256191e-08, "loss": 0.0586, "step": 20314 }, { "epoch": 0.9440055762081785, "grad_norm": 0.5252731899760493, "learning_rate": 8.528911259750583e-08, "loss": 0.0514, "step": 20315 }, { "epoch": 0.9440520446096654, "grad_norm": 0.6756485518053683, "learning_rate": 8.51507698758236e-08, "loss": 0.0634, "step": 20316 }, { "epoch": 0.9440985130111524, "grad_norm": 0.8378617328770935, "learning_rate": 8.501253848064939e-08, "loss": 0.0604, "step": 20317 }, { "epoch": 0.9441449814126394, "grad_norm": 0.615871208862159, "learning_rate": 8.487441841511402e-08, "loss": 0.0622, "step": 20318 }, { "epoch": 0.9441914498141264, "grad_norm": 0.5310979262195085, "learning_rate": 8.473640968234609e-08, "loss": 0.0599, "step": 20319 }, { "epoch": 0.9442379182156134, "grad_norm": 0.6191953668184084, "learning_rate": 8.4598512285472e-08, "loss": 0.0743, "step": 20320 }, { "epoch": 0.9442843866171003, "grad_norm": 0.5090088794424474, "learning_rate": 8.44607262276148e-08, "loss": 0.051, "step": 20321 }, { "epoch": 0.9443308550185874, "grad_norm": 0.5088429537797834, "learning_rate": 8.432305151189536e-08, "loss": 0.0306, "step": 20322 }, { "epoch": 0.9443773234200743, "grad_norm": 0.6361123166062495, "learning_rate": 8.418548814143279e-08, "loss": 0.0637, "step": 20323 }, { "epoch": 0.9444237918215613, "grad_norm": 0.36275413045966515, "learning_rate": 8.404803611934243e-08, "loss": 0.0487, "step": 20324 }, { "epoch": 0.9444702602230484, "grad_norm": 0.41272327828977284, "learning_rate": 8.391069544873842e-08, "loss": 0.0605, "step": 20325 }, { "epoch": 0.9445167286245353, "grad_norm": 0.4214201820252714, "learning_rate": 8.377346613273107e-08, "loss": 0.0451, "step": 20326 }, { "epoch": 0.9445631970260223, "grad_norm": 0.36828369260674326, "learning_rate": 8.363634817442845e-08, "loss": 0.068, "step": 20327 }, { "epoch": 0.9446096654275093, "grad_norm": 0.531917844040565, "learning_rate": 8.34993415769375e-08, "loss": 0.0482, "step": 20328 }, { "epoch": 0.9446561338289963, "grad_norm": 0.4708439889104284, "learning_rate": 8.336244634336133e-08, "loss": 0.0671, "step": 20329 }, { "epoch": 0.9447026022304833, "grad_norm": 0.2919449224865676, "learning_rate": 8.322566247679964e-08, "loss": 0.0264, "step": 20330 }, { "epoch": 0.9447490706319702, "grad_norm": 0.41338252508443435, "learning_rate": 8.308898998035164e-08, "loss": 0.0313, "step": 20331 }, { "epoch": 0.9447955390334573, "grad_norm": 0.6525768372743358, "learning_rate": 8.295242885711374e-08, "loss": 0.0533, "step": 20332 }, { "epoch": 0.9448420074349443, "grad_norm": 0.4352749698625926, "learning_rate": 8.281597911017791e-08, "loss": 0.0448, "step": 20333 }, { "epoch": 0.9448884758364312, "grad_norm": 0.37106582255136294, "learning_rate": 8.267964074263501e-08, "loss": 0.0604, "step": 20334 }, { "epoch": 0.9449349442379182, "grad_norm": 0.39801863361699835, "learning_rate": 8.254341375757368e-08, "loss": 0.0435, "step": 20335 }, { "epoch": 0.9449814126394052, "grad_norm": 0.3421575050477457, "learning_rate": 8.24072981580798e-08, "loss": 0.0426, "step": 20336 }, { "epoch": 0.9450278810408922, "grad_norm": 0.49531637354825797, "learning_rate": 8.227129394723643e-08, "loss": 0.0679, "step": 20337 }, { "epoch": 0.9450743494423792, "grad_norm": 0.5739400005511129, "learning_rate": 8.213540112812391e-08, "loss": 0.0583, "step": 20338 }, { "epoch": 0.9451208178438661, "grad_norm": 0.478674711400803, "learning_rate": 8.199961970382087e-08, "loss": 0.039, "step": 20339 }, { "epoch": 0.9451672862453532, "grad_norm": 0.436868541657608, "learning_rate": 8.186394967740208e-08, "loss": 0.0432, "step": 20340 }, { "epoch": 0.9452137546468401, "grad_norm": 0.436868541657608, "learning_rate": 8.186394967740208e-08, "loss": 0.0517, "step": 20341 }, { "epoch": 0.9452602230483271, "grad_norm": 0.42244367413257844, "learning_rate": 8.172839105194119e-08, "loss": 0.0534, "step": 20342 }, { "epoch": 0.9453066914498142, "grad_norm": 0.564116723941561, "learning_rate": 8.159294383050853e-08, "loss": 0.035, "step": 20343 }, { "epoch": 0.9453531598513011, "grad_norm": 0.486731916431971, "learning_rate": 8.14576080161722e-08, "loss": 0.0536, "step": 20344 }, { "epoch": 0.9453996282527881, "grad_norm": 0.4828427296552215, "learning_rate": 8.132238361199806e-08, "loss": 0.0587, "step": 20345 }, { "epoch": 0.9454460966542751, "grad_norm": 0.6774716128792739, "learning_rate": 8.11872706210487e-08, "loss": 0.0513, "step": 20346 }, { "epoch": 0.9454925650557621, "grad_norm": 0.7400858118649759, "learning_rate": 8.105226904638441e-08, "loss": 0.0651, "step": 20347 }, { "epoch": 0.9455390334572491, "grad_norm": 0.491702546880885, "learning_rate": 8.091737889106388e-08, "loss": 0.0558, "step": 20348 }, { "epoch": 0.945585501858736, "grad_norm": 0.37887040532232763, "learning_rate": 8.078260015814077e-08, "loss": 0.0567, "step": 20349 }, { "epoch": 0.945631970260223, "grad_norm": 0.5034115952259061, "learning_rate": 8.064793285066986e-08, "loss": 0.0543, "step": 20350 }, { "epoch": 0.9456784386617101, "grad_norm": 0.37882716098273894, "learning_rate": 8.051337697170147e-08, "loss": 0.0251, "step": 20351 }, { "epoch": 0.945724907063197, "grad_norm": 0.37972296293669255, "learning_rate": 8.037893252428152e-08, "loss": 0.044, "step": 20352 }, { "epoch": 0.945771375464684, "grad_norm": 0.42289155942683077, "learning_rate": 8.024459951145703e-08, "loss": 0.0494, "step": 20353 }, { "epoch": 0.945817843866171, "grad_norm": 0.4205611505845044, "learning_rate": 8.011037793627053e-08, "loss": 0.0525, "step": 20354 }, { "epoch": 0.945864312267658, "grad_norm": 0.678300120741786, "learning_rate": 7.997626780176237e-08, "loss": 0.0516, "step": 20355 }, { "epoch": 0.945910780669145, "grad_norm": 0.5380845784466343, "learning_rate": 7.984226911096904e-08, "loss": 0.0653, "step": 20356 }, { "epoch": 0.945957249070632, "grad_norm": 0.48815868841739357, "learning_rate": 7.970838186692753e-08, "loss": 0.0533, "step": 20357 }, { "epoch": 0.946003717472119, "grad_norm": 0.4753456442394452, "learning_rate": 7.957460607266931e-08, "loss": 0.061, "step": 20358 }, { "epoch": 0.9460501858736059, "grad_norm": 0.6958723054870074, "learning_rate": 7.944094173122474e-08, "loss": 0.0454, "step": 20359 }, { "epoch": 0.9460966542750929, "grad_norm": 0.4183545487865203, "learning_rate": 7.930738884562195e-08, "loss": 0.0594, "step": 20360 }, { "epoch": 0.94614312267658, "grad_norm": 0.34152588430212105, "learning_rate": 7.917394741888573e-08, "loss": 0.0543, "step": 20361 }, { "epoch": 0.9461895910780669, "grad_norm": 0.6242004299166157, "learning_rate": 7.904061745403813e-08, "loss": 0.0336, "step": 20362 }, { "epoch": 0.9462360594795539, "grad_norm": 0.6330860191866658, "learning_rate": 7.89073989541006e-08, "loss": 0.0655, "step": 20363 }, { "epoch": 0.946282527881041, "grad_norm": 0.6330860191866658, "learning_rate": 7.89073989541006e-08, "loss": 0.045, "step": 20364 }, { "epoch": 0.9463289962825279, "grad_norm": 0.6582708713741268, "learning_rate": 7.877429192208908e-08, "loss": 0.0659, "step": 20365 }, { "epoch": 0.9463754646840149, "grad_norm": 0.45478010653528117, "learning_rate": 7.864129636102002e-08, "loss": 0.04, "step": 20366 }, { "epoch": 0.9464219330855018, "grad_norm": 0.7546838493974833, "learning_rate": 7.850841227390437e-08, "loss": 0.0586, "step": 20367 }, { "epoch": 0.9464684014869889, "grad_norm": 0.3014842038168352, "learning_rate": 7.837563966375361e-08, "loss": 0.0536, "step": 20368 }, { "epoch": 0.9465148698884759, "grad_norm": 0.555688234671053, "learning_rate": 7.824297853357477e-08, "loss": 0.0439, "step": 20369 }, { "epoch": 0.9465613382899628, "grad_norm": 0.47144812853844914, "learning_rate": 7.81104288863721e-08, "loss": 0.0708, "step": 20370 }, { "epoch": 0.9466078066914498, "grad_norm": 0.5117880078298612, "learning_rate": 7.797799072514822e-08, "loss": 0.0605, "step": 20371 }, { "epoch": 0.9466542750929368, "grad_norm": 0.3388442696847684, "learning_rate": 7.78456640529035e-08, "loss": 0.0399, "step": 20372 }, { "epoch": 0.9467007434944238, "grad_norm": 0.6902171810444597, "learning_rate": 7.7713448872635e-08, "loss": 0.0718, "step": 20373 }, { "epoch": 0.9467472118959108, "grad_norm": 0.4584804276520715, "learning_rate": 7.758134518733752e-08, "loss": 0.0426, "step": 20374 }, { "epoch": 0.9467936802973977, "grad_norm": 0.500479131950513, "learning_rate": 7.744935300000311e-08, "loss": 0.0439, "step": 20375 }, { "epoch": 0.9468401486988848, "grad_norm": 0.48855592114000546, "learning_rate": 7.731747231362219e-08, "loss": 0.0572, "step": 20376 }, { "epoch": 0.9468866171003717, "grad_norm": 0.5633681670195982, "learning_rate": 7.718570313118068e-08, "loss": 0.0587, "step": 20377 }, { "epoch": 0.9469330855018587, "grad_norm": 0.5053690240983915, "learning_rate": 7.705404545566508e-08, "loss": 0.0606, "step": 20378 }, { "epoch": 0.9469795539033458, "grad_norm": 0.4917137478520023, "learning_rate": 7.69224992900558e-08, "loss": 0.0596, "step": 20379 }, { "epoch": 0.9470260223048327, "grad_norm": 0.42362475960238316, "learning_rate": 7.679106463733432e-08, "loss": 0.0455, "step": 20380 }, { "epoch": 0.9470724907063197, "grad_norm": 0.4836660633859205, "learning_rate": 7.665974150047606e-08, "loss": 0.0357, "step": 20381 }, { "epoch": 0.9471189591078067, "grad_norm": 0.49462134492930065, "learning_rate": 7.652852988245696e-08, "loss": 0.0351, "step": 20382 }, { "epoch": 0.9471654275092937, "grad_norm": 0.4373344355295988, "learning_rate": 7.639742978624798e-08, "loss": 0.053, "step": 20383 }, { "epoch": 0.9472118959107807, "grad_norm": 0.426859908054815, "learning_rate": 7.626644121481896e-08, "loss": 0.0402, "step": 20384 }, { "epoch": 0.9472583643122676, "grad_norm": 0.6317701970530394, "learning_rate": 7.61355641711381e-08, "loss": 0.0591, "step": 20385 }, { "epoch": 0.9473048327137547, "grad_norm": 0.6641105945097131, "learning_rate": 7.600479865816856e-08, "loss": 0.0552, "step": 20386 }, { "epoch": 0.9473513011152417, "grad_norm": 0.37570412752228904, "learning_rate": 7.587414467887244e-08, "loss": 0.0533, "step": 20387 }, { "epoch": 0.9473977695167286, "grad_norm": 0.6442046950955198, "learning_rate": 7.574360223621014e-08, "loss": 0.0576, "step": 20388 }, { "epoch": 0.9474442379182156, "grad_norm": 0.6191461010074876, "learning_rate": 7.561317133313705e-08, "loss": 0.0481, "step": 20389 }, { "epoch": 0.9474907063197026, "grad_norm": 0.6352818298959421, "learning_rate": 7.548285197260919e-08, "loss": 0.0743, "step": 20390 }, { "epoch": 0.9475371747211896, "grad_norm": 0.6855625651719423, "learning_rate": 7.535264415757693e-08, "loss": 0.0646, "step": 20391 }, { "epoch": 0.9475836431226766, "grad_norm": 0.5133539608214296, "learning_rate": 7.522254789099127e-08, "loss": 0.0512, "step": 20392 }, { "epoch": 0.9476301115241635, "grad_norm": 0.46167268398869654, "learning_rate": 7.509256317579817e-08, "loss": 0.0577, "step": 20393 }, { "epoch": 0.9476765799256506, "grad_norm": 0.7666439041266874, "learning_rate": 7.496269001494084e-08, "loss": 0.0665, "step": 20394 }, { "epoch": 0.9477230483271375, "grad_norm": 0.40468591090911143, "learning_rate": 7.483292841136303e-08, "loss": 0.0341, "step": 20395 }, { "epoch": 0.9477695167286245, "grad_norm": 0.5221168351727321, "learning_rate": 7.470327836800295e-08, "loss": 0.0506, "step": 20396 }, { "epoch": 0.9478159851301116, "grad_norm": 0.5130694739456365, "learning_rate": 7.457373988779715e-08, "loss": 0.0537, "step": 20397 }, { "epoch": 0.9478624535315985, "grad_norm": 0.5641003557120302, "learning_rate": 7.444431297367993e-08, "loss": 0.0599, "step": 20398 }, { "epoch": 0.9479089219330855, "grad_norm": 0.4086451964591797, "learning_rate": 7.43149976285834e-08, "loss": 0.0386, "step": 20399 }, { "epoch": 0.9479553903345725, "grad_norm": 0.28251552296323307, "learning_rate": 7.418579385543634e-08, "loss": 0.0314, "step": 20400 }, { "epoch": 0.9480018587360595, "grad_norm": 0.47201935591769767, "learning_rate": 7.405670165716472e-08, "loss": 0.0596, "step": 20401 }, { "epoch": 0.9480483271375465, "grad_norm": 0.5353801397076999, "learning_rate": 7.392772103669399e-08, "loss": 0.0351, "step": 20402 }, { "epoch": 0.9480947955390334, "grad_norm": 0.40854400173967076, "learning_rate": 7.379885199694458e-08, "loss": 0.0415, "step": 20403 }, { "epoch": 0.9481412639405205, "grad_norm": 0.3799179635837273, "learning_rate": 7.367009454083641e-08, "loss": 0.0416, "step": 20404 }, { "epoch": 0.9481877323420075, "grad_norm": 0.5502881560610074, "learning_rate": 7.354144867128488e-08, "loss": 0.0569, "step": 20405 }, { "epoch": 0.9482342007434944, "grad_norm": 0.6530204596937289, "learning_rate": 7.341291439120491e-08, "loss": 0.0509, "step": 20406 }, { "epoch": 0.9482806691449814, "grad_norm": 0.41698666285914343, "learning_rate": 7.328449170350749e-08, "loss": 0.041, "step": 20407 }, { "epoch": 0.9483271375464684, "grad_norm": 0.29770989451809726, "learning_rate": 7.315618061110141e-08, "loss": 0.0254, "step": 20408 }, { "epoch": 0.9483736059479554, "grad_norm": 0.46869045734314674, "learning_rate": 7.302798111689324e-08, "loss": 0.0448, "step": 20409 }, { "epoch": 0.9484200743494424, "grad_norm": 0.6483169213650223, "learning_rate": 7.289989322378732e-08, "loss": 0.0395, "step": 20410 }, { "epoch": 0.9484665427509293, "grad_norm": 0.5546377188119194, "learning_rate": 7.27719169346841e-08, "loss": 0.0684, "step": 20411 }, { "epoch": 0.9485130111524164, "grad_norm": 0.4350110443205905, "learning_rate": 7.264405225248295e-08, "loss": 0.0328, "step": 20412 }, { "epoch": 0.9485594795539033, "grad_norm": 0.5753682545375943, "learning_rate": 7.251629918007985e-08, "loss": 0.0379, "step": 20413 }, { "epoch": 0.9486059479553903, "grad_norm": 0.7493263062486025, "learning_rate": 7.238865772036863e-08, "loss": 0.0522, "step": 20414 }, { "epoch": 0.9486524163568774, "grad_norm": 0.44426600968345525, "learning_rate": 7.226112787624028e-08, "loss": 0.0388, "step": 20415 }, { "epoch": 0.9486988847583643, "grad_norm": 0.345352805556263, "learning_rate": 7.213370965058419e-08, "loss": 0.0346, "step": 20416 }, { "epoch": 0.9487453531598513, "grad_norm": 0.4757402458848778, "learning_rate": 7.200640304628636e-08, "loss": 0.0635, "step": 20417 }, { "epoch": 0.9487918215613383, "grad_norm": 0.6878559199762422, "learning_rate": 7.18792080662295e-08, "loss": 0.051, "step": 20418 }, { "epoch": 0.9488382899628253, "grad_norm": 0.4496471492643479, "learning_rate": 7.175212471329573e-08, "loss": 0.0514, "step": 20419 }, { "epoch": 0.9488847583643123, "grad_norm": 0.4931154067487178, "learning_rate": 7.162515299036277e-08, "loss": 0.0477, "step": 20420 }, { "epoch": 0.9489312267657992, "grad_norm": 0.4014131847733309, "learning_rate": 7.149829290030774e-08, "loss": 0.0324, "step": 20421 }, { "epoch": 0.9489776951672863, "grad_norm": 0.6108097901490649, "learning_rate": 7.13715444460028e-08, "loss": 0.0345, "step": 20422 }, { "epoch": 0.9490241635687733, "grad_norm": 0.5732182889231918, "learning_rate": 7.12449076303201e-08, "loss": 0.0551, "step": 20423 }, { "epoch": 0.9490706319702602, "grad_norm": 0.3461823951516391, "learning_rate": 7.11183824561279e-08, "loss": 0.0402, "step": 20424 }, { "epoch": 0.9491171003717472, "grad_norm": 0.49640954583618135, "learning_rate": 7.099196892629112e-08, "loss": 0.0464, "step": 20425 }, { "epoch": 0.9491635687732342, "grad_norm": 0.8373581848227147, "learning_rate": 7.086566704367415e-08, "loss": 0.0707, "step": 20426 }, { "epoch": 0.9492100371747212, "grad_norm": 0.5798124404367418, "learning_rate": 7.073947681113802e-08, "loss": 0.0394, "step": 20427 }, { "epoch": 0.9492565055762082, "grad_norm": 0.4123154018972962, "learning_rate": 7.061339823154045e-08, "loss": 0.0364, "step": 20428 }, { "epoch": 0.9493029739776951, "grad_norm": 0.4531949213740491, "learning_rate": 7.048743130773806e-08, "loss": 0.0441, "step": 20429 }, { "epoch": 0.9493494423791822, "grad_norm": 0.7220527921232044, "learning_rate": 7.036157604258298e-08, "loss": 0.0631, "step": 20430 }, { "epoch": 0.9493959107806691, "grad_norm": 0.34681845346279566, "learning_rate": 7.023583243892684e-08, "loss": 0.0236, "step": 20431 }, { "epoch": 0.9494423791821561, "grad_norm": 0.5797454977176992, "learning_rate": 7.011020049961681e-08, "loss": 0.056, "step": 20432 }, { "epoch": 0.9494888475836432, "grad_norm": 0.3524128785267495, "learning_rate": 6.998468022750004e-08, "loss": 0.04, "step": 20433 }, { "epoch": 0.9495353159851301, "grad_norm": 0.5505152741624932, "learning_rate": 6.985927162541928e-08, "loss": 0.0564, "step": 20434 }, { "epoch": 0.9495817843866171, "grad_norm": 0.29371904169814017, "learning_rate": 6.973397469621446e-08, "loss": 0.0398, "step": 20435 }, { "epoch": 0.9496282527881041, "grad_norm": 0.4847456755011546, "learning_rate": 6.960878944272387e-08, "loss": 0.0386, "step": 20436 }, { "epoch": 0.9496747211895911, "grad_norm": 0.4573192497302995, "learning_rate": 6.948371586778357e-08, "loss": 0.0653, "step": 20437 }, { "epoch": 0.9497211895910781, "grad_norm": 0.5915485571273308, "learning_rate": 6.935875397422687e-08, "loss": 0.0515, "step": 20438 }, { "epoch": 0.949767657992565, "grad_norm": 0.556054479169948, "learning_rate": 6.923390376488315e-08, "loss": 0.0533, "step": 20439 }, { "epoch": 0.949814126394052, "grad_norm": 0.6287530923517498, "learning_rate": 6.910916524258127e-08, "loss": 0.0624, "step": 20440 }, { "epoch": 0.9498605947955391, "grad_norm": 0.5130809802348133, "learning_rate": 6.898453841014674e-08, "loss": 0.0596, "step": 20441 }, { "epoch": 0.949907063197026, "grad_norm": 0.49718886450235034, "learning_rate": 6.886002327040176e-08, "loss": 0.0478, "step": 20442 }, { "epoch": 0.949953531598513, "grad_norm": 0.5468792582062264, "learning_rate": 6.873561982616738e-08, "loss": 0.0507, "step": 20443 }, { "epoch": 0.95, "grad_norm": 0.4330726967867374, "learning_rate": 6.861132808026139e-08, "loss": 0.0426, "step": 20444 }, { "epoch": 0.950046468401487, "grad_norm": 0.6113706941047645, "learning_rate": 6.848714803549872e-08, "loss": 0.0365, "step": 20445 }, { "epoch": 0.950092936802974, "grad_norm": 0.4710829894704834, "learning_rate": 6.836307969469269e-08, "loss": 0.0378, "step": 20446 }, { "epoch": 0.9501394052044609, "grad_norm": 0.49808913596569887, "learning_rate": 6.823912306065328e-08, "loss": 0.0514, "step": 20447 }, { "epoch": 0.950185873605948, "grad_norm": 0.7371283887531068, "learning_rate": 6.811527813618879e-08, "loss": 0.0603, "step": 20448 }, { "epoch": 0.9502323420074349, "grad_norm": 0.3941840181230369, "learning_rate": 6.799154492410309e-08, "loss": 0.0493, "step": 20449 }, { "epoch": 0.9502788104089219, "grad_norm": 0.4313881976753505, "learning_rate": 6.786792342720006e-08, "loss": 0.064, "step": 20450 }, { "epoch": 0.950325278810409, "grad_norm": 0.4794016764089246, "learning_rate": 6.774441364827966e-08, "loss": 0.0436, "step": 20451 }, { "epoch": 0.9503717472118959, "grad_norm": 0.3667654515237336, "learning_rate": 6.762101559013967e-08, "loss": 0.0402, "step": 20452 }, { "epoch": 0.9504182156133829, "grad_norm": 0.5947356594011594, "learning_rate": 6.749772925557452e-08, "loss": 0.0465, "step": 20453 }, { "epoch": 0.9504646840148699, "grad_norm": 0.46397216414256, "learning_rate": 6.737455464737752e-08, "loss": 0.0499, "step": 20454 }, { "epoch": 0.9505111524163569, "grad_norm": 0.6058721821131361, "learning_rate": 6.725149176833811e-08, "loss": 0.0507, "step": 20455 }, { "epoch": 0.9505576208178439, "grad_norm": 0.5811889243456702, "learning_rate": 6.712854062124352e-08, "loss": 0.066, "step": 20456 }, { "epoch": 0.9506040892193308, "grad_norm": 0.7185142267257963, "learning_rate": 6.700570120887983e-08, "loss": 0.0709, "step": 20457 }, { "epoch": 0.9506505576208178, "grad_norm": 0.3410813230299496, "learning_rate": 6.688297353402928e-08, "loss": 0.0529, "step": 20458 }, { "epoch": 0.9506970260223049, "grad_norm": 0.7493991616833106, "learning_rate": 6.676035759947076e-08, "loss": 0.0577, "step": 20459 }, { "epoch": 0.9507434944237918, "grad_norm": 0.5367674836043991, "learning_rate": 6.663785340798256e-08, "loss": 0.0517, "step": 20460 }, { "epoch": 0.9507899628252788, "grad_norm": 0.4990545337035324, "learning_rate": 6.651546096233918e-08, "loss": 0.0391, "step": 20461 }, { "epoch": 0.9508364312267658, "grad_norm": 0.43710886924381376, "learning_rate": 6.639318026531283e-08, "loss": 0.046, "step": 20462 }, { "epoch": 0.9508828996282528, "grad_norm": 0.5615429773401903, "learning_rate": 6.62710113196735e-08, "loss": 0.0536, "step": 20463 }, { "epoch": 0.9509293680297398, "grad_norm": 0.8646644297349388, "learning_rate": 6.6148954128189e-08, "loss": 0.0596, "step": 20464 }, { "epoch": 0.9509758364312267, "grad_norm": 0.38206753010073313, "learning_rate": 6.602700869362321e-08, "loss": 0.0453, "step": 20465 }, { "epoch": 0.9510223048327138, "grad_norm": 0.5666918324289673, "learning_rate": 6.590517501873783e-08, "loss": 0.0505, "step": 20466 }, { "epoch": 0.9510687732342007, "grad_norm": 0.48153334557184546, "learning_rate": 6.578345310629397e-08, "loss": 0.052, "step": 20467 }, { "epoch": 0.9511152416356877, "grad_norm": 0.45603509291408395, "learning_rate": 6.566184295904777e-08, "loss": 0.0614, "step": 20468 }, { "epoch": 0.9511617100371748, "grad_norm": 0.3796850347564767, "learning_rate": 6.554034457975422e-08, "loss": 0.0545, "step": 20469 }, { "epoch": 0.9512081784386617, "grad_norm": 0.4435266694322326, "learning_rate": 6.54189579711656e-08, "loss": 0.0446, "step": 20470 }, { "epoch": 0.9512546468401487, "grad_norm": 0.5386056569801377, "learning_rate": 6.52976831360308e-08, "loss": 0.0599, "step": 20471 }, { "epoch": 0.9513011152416357, "grad_norm": 0.4142366942695205, "learning_rate": 6.517652007709762e-08, "loss": 0.0448, "step": 20472 }, { "epoch": 0.9513475836431227, "grad_norm": 0.4635441039502452, "learning_rate": 6.505546879710945e-08, "loss": 0.0607, "step": 20473 }, { "epoch": 0.9513940520446097, "grad_norm": 0.7434671493731824, "learning_rate": 6.493452929880906e-08, "loss": 0.0541, "step": 20474 }, { "epoch": 0.9514405204460966, "grad_norm": 0.4490540253914154, "learning_rate": 6.481370158493538e-08, "loss": 0.0526, "step": 20475 }, { "epoch": 0.9514869888475836, "grad_norm": 0.5498640974856989, "learning_rate": 6.469298565822624e-08, "loss": 0.044, "step": 20476 }, { "epoch": 0.9515334572490707, "grad_norm": 0.3046476449952064, "learning_rate": 6.457238152141499e-08, "loss": 0.0262, "step": 20477 }, { "epoch": 0.9515799256505576, "grad_norm": 0.5828570047270911, "learning_rate": 6.445188917723333e-08, "loss": 0.043, "step": 20478 }, { "epoch": 0.9516263940520446, "grad_norm": 0.26264227791284644, "learning_rate": 6.433150862841131e-08, "loss": 0.0325, "step": 20479 }, { "epoch": 0.9516728624535316, "grad_norm": 0.46752829507687554, "learning_rate": 6.421123987767508e-08, "loss": 0.0371, "step": 20480 }, { "epoch": 0.9517193308550186, "grad_norm": 0.5025614316403354, "learning_rate": 6.409108292774912e-08, "loss": 0.0415, "step": 20481 }, { "epoch": 0.9517657992565056, "grad_norm": 0.3744842193210664, "learning_rate": 6.397103778135572e-08, "loss": 0.0485, "step": 20482 }, { "epoch": 0.9518122676579925, "grad_norm": 0.8822954079009078, "learning_rate": 6.385110444121267e-08, "loss": 0.0366, "step": 20483 }, { "epoch": 0.9518587360594796, "grad_norm": 0.4637250637623583, "learning_rate": 6.373128291003783e-08, "loss": 0.0362, "step": 20484 }, { "epoch": 0.9519052044609665, "grad_norm": 0.7932895166969091, "learning_rate": 6.361157319054456e-08, "loss": 0.0623, "step": 20485 }, { "epoch": 0.9519516728624535, "grad_norm": 0.3763867342010649, "learning_rate": 6.349197528544404e-08, "loss": 0.0172, "step": 20486 }, { "epoch": 0.9519981412639406, "grad_norm": 0.4520752991602518, "learning_rate": 6.337248919744687e-08, "loss": 0.0461, "step": 20487 }, { "epoch": 0.9520446096654275, "grad_norm": 0.47395027111528887, "learning_rate": 6.325311492925757e-08, "loss": 0.0461, "step": 20488 }, { "epoch": 0.9520910780669145, "grad_norm": 0.5686108824880214, "learning_rate": 6.313385248358173e-08, "loss": 0.0602, "step": 20489 }, { "epoch": 0.9521375464684015, "grad_norm": 0.529139035740765, "learning_rate": 6.301470186311943e-08, "loss": 0.0563, "step": 20490 }, { "epoch": 0.9521840148698885, "grad_norm": 0.5736001310805486, "learning_rate": 6.289566307057071e-08, "loss": 0.0563, "step": 20491 }, { "epoch": 0.9522304832713755, "grad_norm": 0.23682603866107443, "learning_rate": 6.277673610863178e-08, "loss": 0.0346, "step": 20492 }, { "epoch": 0.9522769516728624, "grad_norm": 0.35265760273562435, "learning_rate": 6.265792097999546e-08, "loss": 0.0493, "step": 20493 }, { "epoch": 0.9523234200743494, "grad_norm": 0.5299983215897335, "learning_rate": 6.253921768735405e-08, "loss": 0.0559, "step": 20494 }, { "epoch": 0.9523698884758365, "grad_norm": 0.6041468993793138, "learning_rate": 6.242062623339595e-08, "loss": 0.0503, "step": 20495 }, { "epoch": 0.9524163568773234, "grad_norm": 0.5877079950671444, "learning_rate": 6.23021466208068e-08, "loss": 0.0597, "step": 20496 }, { "epoch": 0.9524628252788104, "grad_norm": 0.36463589768546883, "learning_rate": 6.218377885227112e-08, "loss": 0.044, "step": 20497 }, { "epoch": 0.9525092936802974, "grad_norm": 0.359023027612865, "learning_rate": 6.206552293047008e-08, "loss": 0.0444, "step": 20498 }, { "epoch": 0.9525557620817844, "grad_norm": 0.33015216517802465, "learning_rate": 6.194737885808211e-08, "loss": 0.0467, "step": 20499 }, { "epoch": 0.9526022304832714, "grad_norm": 0.3072828155317242, "learning_rate": 6.182934663778284e-08, "loss": 0.0391, "step": 20500 }, { "epoch": 0.9526486988847583, "grad_norm": 0.5566981159204398, "learning_rate": 6.17114262722468e-08, "loss": 0.0665, "step": 20501 }, { "epoch": 0.9526951672862454, "grad_norm": 1.1763294439689251, "learning_rate": 6.159361776414352e-08, "loss": 0.0657, "step": 20502 }, { "epoch": 0.9527416356877323, "grad_norm": 0.47267020976539237, "learning_rate": 6.147592111614309e-08, "loss": 0.0341, "step": 20503 }, { "epoch": 0.9527881040892193, "grad_norm": 0.5356257142128367, "learning_rate": 6.135833633091004e-08, "loss": 0.0554, "step": 20504 }, { "epoch": 0.9528345724907064, "grad_norm": 0.49729417286845273, "learning_rate": 6.124086341110946e-08, "loss": 0.0526, "step": 20505 }, { "epoch": 0.9528810408921933, "grad_norm": 0.2186158559292251, "learning_rate": 6.11235023594009e-08, "loss": 0.0175, "step": 20506 }, { "epoch": 0.9529275092936803, "grad_norm": 0.5994607063268632, "learning_rate": 6.100625317844278e-08, "loss": 0.0286, "step": 20507 }, { "epoch": 0.9529739776951673, "grad_norm": 0.4046669836855353, "learning_rate": 6.08891158708913e-08, "loss": 0.0479, "step": 20508 }, { "epoch": 0.9530204460966543, "grad_norm": 0.644338975356748, "learning_rate": 6.077209043939991e-08, "loss": 0.0646, "step": 20509 }, { "epoch": 0.9530669144981413, "grad_norm": 0.5229780580138579, "learning_rate": 6.065517688661926e-08, "loss": 0.0454, "step": 20510 }, { "epoch": 0.9531133828996282, "grad_norm": 0.5920136576695791, "learning_rate": 6.053837521519668e-08, "loss": 0.049, "step": 20511 }, { "epoch": 0.9531598513011152, "grad_norm": 0.5589594672966847, "learning_rate": 6.042168542777949e-08, "loss": 0.0497, "step": 20512 }, { "epoch": 0.9532063197026023, "grad_norm": 0.5955310707598482, "learning_rate": 6.030510752700946e-08, "loss": 0.071, "step": 20513 }, { "epoch": 0.9532527881040892, "grad_norm": 0.6996195616039028, "learning_rate": 6.018864151552729e-08, "loss": 0.0587, "step": 20514 }, { "epoch": 0.9532992565055762, "grad_norm": 0.582027426096848, "learning_rate": 6.007228739597193e-08, "loss": 0.0585, "step": 20515 }, { "epoch": 0.9533457249070632, "grad_norm": 0.38562572219866464, "learning_rate": 5.995604517097853e-08, "loss": 0.0224, "step": 20516 }, { "epoch": 0.9533921933085502, "grad_norm": 0.3639195495575424, "learning_rate": 5.983991484317997e-08, "loss": 0.0504, "step": 20517 }, { "epoch": 0.9534386617100372, "grad_norm": 0.5405955546383042, "learning_rate": 5.972389641520637e-08, "loss": 0.0494, "step": 20518 }, { "epoch": 0.9534851301115241, "grad_norm": 0.5628243014693262, "learning_rate": 5.960798988968675e-08, "loss": 0.0584, "step": 20519 }, { "epoch": 0.9535315985130112, "grad_norm": 0.5138204747927085, "learning_rate": 5.94921952692451e-08, "loss": 0.0467, "step": 20520 }, { "epoch": 0.9535780669144981, "grad_norm": 0.44033352141044574, "learning_rate": 5.937651255650545e-08, "loss": 0.0423, "step": 20521 }, { "epoch": 0.9536245353159851, "grad_norm": 0.47093440911541495, "learning_rate": 5.926094175408736e-08, "loss": 0.0433, "step": 20522 }, { "epoch": 0.9536710037174722, "grad_norm": 0.5240160728242658, "learning_rate": 5.914548286460986e-08, "loss": 0.0562, "step": 20523 }, { "epoch": 0.9537174721189591, "grad_norm": 0.4879921797806463, "learning_rate": 5.903013589068585e-08, "loss": 0.0359, "step": 20524 }, { "epoch": 0.9537639405204461, "grad_norm": 0.3706587102164578, "learning_rate": 5.891490083493046e-08, "loss": 0.0333, "step": 20525 }, { "epoch": 0.9538104089219331, "grad_norm": 0.6660186844494503, "learning_rate": 5.879977769995326e-08, "loss": 0.0554, "step": 20526 }, { "epoch": 0.9538568773234201, "grad_norm": 0.4866142721213045, "learning_rate": 5.868476648836108e-08, "loss": 0.0475, "step": 20527 }, { "epoch": 0.9539033457249071, "grad_norm": 0.4134990957588916, "learning_rate": 5.856986720275959e-08, "loss": 0.0489, "step": 20528 }, { "epoch": 0.953949814126394, "grad_norm": 0.7094968320907219, "learning_rate": 5.8455079845751716e-08, "loss": 0.0679, "step": 20529 }, { "epoch": 0.953996282527881, "grad_norm": 0.3075144613528892, "learning_rate": 5.8340404419937046e-08, "loss": 0.0253, "step": 20530 }, { "epoch": 0.9540427509293681, "grad_norm": 0.36567846580270663, "learning_rate": 5.8225840927912944e-08, "loss": 0.0375, "step": 20531 }, { "epoch": 0.954089219330855, "grad_norm": 0.7075388957611444, "learning_rate": 5.8111389372274565e-08, "loss": 0.0528, "step": 20532 }, { "epoch": 0.954135687732342, "grad_norm": 0.5811510511832136, "learning_rate": 5.799704975561482e-08, "loss": 0.0635, "step": 20533 }, { "epoch": 0.954182156133829, "grad_norm": 0.5634640699250497, "learning_rate": 5.788282208052332e-08, "loss": 0.0505, "step": 20534 }, { "epoch": 0.954228624535316, "grad_norm": 0.4191978356456921, "learning_rate": 5.776870634958742e-08, "loss": 0.0502, "step": 20535 }, { "epoch": 0.954275092936803, "grad_norm": 0.41341308613601735, "learning_rate": 5.765470256539174e-08, "loss": 0.0548, "step": 20536 }, { "epoch": 0.9543215613382899, "grad_norm": 0.5712874527368929, "learning_rate": 5.7540810730518646e-08, "loss": 0.0547, "step": 20537 }, { "epoch": 0.954368029739777, "grad_norm": 0.5543376546058263, "learning_rate": 5.742703084754775e-08, "loss": 0.047, "step": 20538 }, { "epoch": 0.9544144981412639, "grad_norm": 0.4310115743119177, "learning_rate": 5.731336291905698e-08, "loss": 0.0403, "step": 20539 }, { "epoch": 0.9544609665427509, "grad_norm": 0.6505054657866215, "learning_rate": 5.719980694762095e-08, "loss": 0.0449, "step": 20540 }, { "epoch": 0.954507434944238, "grad_norm": 0.6086227277123094, "learning_rate": 5.7086362935810936e-08, "loss": 0.0514, "step": 20541 }, { "epoch": 0.9545539033457249, "grad_norm": 0.7458087543385623, "learning_rate": 5.697303088619765e-08, "loss": 0.0689, "step": 20542 }, { "epoch": 0.9546003717472119, "grad_norm": 0.4775767066907244, "learning_rate": 5.685981080134739e-08, "loss": 0.0483, "step": 20543 }, { "epoch": 0.9546468401486989, "grad_norm": 0.47349947280525295, "learning_rate": 5.674670268382476e-08, "loss": 0.0556, "step": 20544 }, { "epoch": 0.9546933085501859, "grad_norm": 0.31936062137915594, "learning_rate": 5.663370653619216e-08, "loss": 0.0568, "step": 20545 }, { "epoch": 0.9547397769516729, "grad_norm": 0.511364971075736, "learning_rate": 5.65208223610092e-08, "loss": 0.0613, "step": 20546 }, { "epoch": 0.9547862453531598, "grad_norm": 0.5946402396846158, "learning_rate": 5.6408050160832175e-08, "loss": 0.0342, "step": 20547 }, { "epoch": 0.9548327137546468, "grad_norm": 0.38186135864452225, "learning_rate": 5.6295389938215704e-08, "loss": 0.0544, "step": 20548 }, { "epoch": 0.9548791821561339, "grad_norm": 0.583587064778229, "learning_rate": 5.61828416957122e-08, "loss": 0.0588, "step": 20549 }, { "epoch": 0.9549256505576208, "grad_norm": 0.5269606925601789, "learning_rate": 5.6070405435870726e-08, "loss": 0.0595, "step": 20550 }, { "epoch": 0.9549721189591078, "grad_norm": 0.47474387490750225, "learning_rate": 5.595808116123758e-08, "loss": 0.0502, "step": 20551 }, { "epoch": 0.9550185873605948, "grad_norm": 0.5430102328490028, "learning_rate": 5.584586887435739e-08, "loss": 0.063, "step": 20552 }, { "epoch": 0.9550650557620818, "grad_norm": 0.4616752864521566, "learning_rate": 5.573376857777202e-08, "loss": 0.049, "step": 20553 }, { "epoch": 0.9551115241635688, "grad_norm": 0.40652850766028187, "learning_rate": 5.562178027402054e-08, "loss": 0.0279, "step": 20554 }, { "epoch": 0.9551579925650557, "grad_norm": 0.3978413127080249, "learning_rate": 5.550990396563927e-08, "loss": 0.0474, "step": 20555 }, { "epoch": 0.9552044609665428, "grad_norm": 0.6302818598023506, "learning_rate": 5.539813965516339e-08, "loss": 0.0391, "step": 20556 }, { "epoch": 0.9552509293680297, "grad_norm": 0.4198287728026104, "learning_rate": 5.5286487345123094e-08, "loss": 0.033, "step": 20557 }, { "epoch": 0.9552973977695167, "grad_norm": 0.38296171865009426, "learning_rate": 5.5174947038048597e-08, "loss": 0.0485, "step": 20558 }, { "epoch": 0.9553438661710038, "grad_norm": 0.3742269181613421, "learning_rate": 5.506351873646565e-08, "loss": 0.0467, "step": 20559 }, { "epoch": 0.9553903345724907, "grad_norm": 0.8701960310037878, "learning_rate": 5.495220244289834e-08, "loss": 0.0641, "step": 20560 }, { "epoch": 0.9554368029739777, "grad_norm": 0.3911736181946679, "learning_rate": 5.4840998159867987e-08, "loss": 0.0502, "step": 20561 }, { "epoch": 0.9554832713754647, "grad_norm": 0.6688268040884504, "learning_rate": 5.4729905889894244e-08, "loss": 0.0526, "step": 20562 }, { "epoch": 0.9555297397769517, "grad_norm": 0.6377428553802343, "learning_rate": 5.461892563549231e-08, "loss": 0.0569, "step": 20563 }, { "epoch": 0.9555762081784387, "grad_norm": 0.5152229276639652, "learning_rate": 5.4508057399177414e-08, "loss": 0.0506, "step": 20564 }, { "epoch": 0.9556226765799256, "grad_norm": 0.4977047728208778, "learning_rate": 5.439730118345976e-08, "loss": 0.0501, "step": 20565 }, { "epoch": 0.9556691449814126, "grad_norm": 0.5767639293002316, "learning_rate": 5.4286656990847897e-08, "loss": 0.0534, "step": 20566 }, { "epoch": 0.9557156133828997, "grad_norm": 0.7187013889265479, "learning_rate": 5.4176124823849266e-08, "loss": 0.0501, "step": 20567 }, { "epoch": 0.9557620817843866, "grad_norm": 0.4881946030553755, "learning_rate": 5.40657046849663e-08, "loss": 0.0575, "step": 20568 }, { "epoch": 0.9558085501858736, "grad_norm": 0.5778823928181711, "learning_rate": 5.395539657670035e-08, "loss": 0.0512, "step": 20569 }, { "epoch": 0.9558550185873605, "grad_norm": 0.7325518260424645, "learning_rate": 5.3845200501551066e-08, "loss": 0.0449, "step": 20570 }, { "epoch": 0.9559014869888476, "grad_norm": 0.6683816246657609, "learning_rate": 5.373511646201368e-08, "loss": 0.0514, "step": 20571 }, { "epoch": 0.9559479553903346, "grad_norm": 0.46243214977115, "learning_rate": 5.3625144460581204e-08, "loss": 0.0581, "step": 20572 }, { "epoch": 0.9559944237918215, "grad_norm": 0.6809339395745234, "learning_rate": 5.351528449974497e-08, "loss": 0.0488, "step": 20573 }, { "epoch": 0.9560408921933086, "grad_norm": 0.5086747772296332, "learning_rate": 5.34055365819941e-08, "loss": 0.0593, "step": 20574 }, { "epoch": 0.9560873605947955, "grad_norm": 0.5746562810661497, "learning_rate": 5.329590070981383e-08, "loss": 0.0557, "step": 20575 }, { "epoch": 0.9561338289962825, "grad_norm": 0.5547775988270908, "learning_rate": 5.3186376885688286e-08, "loss": 0.0678, "step": 20576 }, { "epoch": 0.9561802973977696, "grad_norm": 0.8902162446759918, "learning_rate": 5.3076965112097146e-08, "loss": 0.0604, "step": 20577 }, { "epoch": 0.9562267657992565, "grad_norm": 0.6141534815816911, "learning_rate": 5.296766539151954e-08, "loss": 0.0507, "step": 20578 }, { "epoch": 0.9562732342007435, "grad_norm": 0.4069736117684519, "learning_rate": 5.2858477726430714e-08, "loss": 0.04, "step": 20579 }, { "epoch": 0.9563197026022305, "grad_norm": 0.7050746092076021, "learning_rate": 5.274940211930424e-08, "loss": 0.0693, "step": 20580 }, { "epoch": 0.9563661710037175, "grad_norm": 0.28073027475188445, "learning_rate": 5.2640438572610916e-08, "loss": 0.0381, "step": 20581 }, { "epoch": 0.9564126394052045, "grad_norm": 0.48364113405386344, "learning_rate": 5.2531587088818784e-08, "loss": 0.063, "step": 20582 }, { "epoch": 0.9564591078066914, "grad_norm": 0.49780491701258506, "learning_rate": 5.2422847670393073e-08, "loss": 0.0599, "step": 20583 }, { "epoch": 0.9565055762081784, "grad_norm": 0.5611841104012175, "learning_rate": 5.2314220319796824e-08, "loss": 0.0407, "step": 20584 }, { "epoch": 0.9565520446096655, "grad_norm": 0.45628638163795043, "learning_rate": 5.220570503949085e-08, "loss": 0.0529, "step": 20585 }, { "epoch": 0.9565985130111524, "grad_norm": 0.4223143528293642, "learning_rate": 5.2097301831933175e-08, "loss": 0.0499, "step": 20586 }, { "epoch": 0.9566449814126394, "grad_norm": 0.43709255202242164, "learning_rate": 5.1989010699579625e-08, "loss": 0.044, "step": 20587 }, { "epoch": 0.9566914498141263, "grad_norm": 0.34361529572592486, "learning_rate": 5.188083164488267e-08, "loss": 0.039, "step": 20588 }, { "epoch": 0.9567379182156134, "grad_norm": 0.43002439888740546, "learning_rate": 5.177276467029202e-08, "loss": 0.0373, "step": 20589 }, { "epoch": 0.9567843866171004, "grad_norm": 0.3425220529575337, "learning_rate": 5.166480977825683e-08, "loss": 0.0382, "step": 20590 }, { "epoch": 0.9568308550185873, "grad_norm": 0.549685405384813, "learning_rate": 5.155696697122126e-08, "loss": 0.0599, "step": 20591 }, { "epoch": 0.9568773234200744, "grad_norm": 0.6065099262304634, "learning_rate": 5.1449236251628344e-08, "loss": 0.0481, "step": 20592 }, { "epoch": 0.9569237918215613, "grad_norm": 0.6013524311372292, "learning_rate": 5.134161762191892e-08, "loss": 0.0509, "step": 20593 }, { "epoch": 0.9569702602230483, "grad_norm": 0.5716453540958829, "learning_rate": 5.1234111084530467e-08, "loss": 0.0501, "step": 20594 }, { "epoch": 0.9570167286245354, "grad_norm": 0.367128695148557, "learning_rate": 5.112671664189772e-08, "loss": 0.0477, "step": 20595 }, { "epoch": 0.9570631970260223, "grad_norm": 0.5958735157088587, "learning_rate": 5.1019434296453175e-08, "loss": 0.0471, "step": 20596 }, { "epoch": 0.9571096654275093, "grad_norm": 0.3630190733898325, "learning_rate": 5.0912264050627104e-08, "loss": 0.0463, "step": 20597 }, { "epoch": 0.9571561338289963, "grad_norm": 0.6055745230953722, "learning_rate": 5.080520590684701e-08, "loss": 0.054, "step": 20598 }, { "epoch": 0.9572026022304833, "grad_norm": 0.5268531522005488, "learning_rate": 5.069825986753818e-08, "loss": 0.0529, "step": 20599 }, { "epoch": 0.9572490706319703, "grad_norm": 0.6029247767970015, "learning_rate": 5.0591425935123116e-08, "loss": 0.044, "step": 20600 }, { "epoch": 0.9572955390334572, "grad_norm": 0.5086741728932888, "learning_rate": 5.0484704112021e-08, "loss": 0.043, "step": 20601 }, { "epoch": 0.9573420074349442, "grad_norm": 0.6019100879281102, "learning_rate": 5.0378094400649337e-08, "loss": 0.05, "step": 20602 }, { "epoch": 0.9573884758364313, "grad_norm": 0.4232237237585272, "learning_rate": 5.027159680342397e-08, "loss": 0.0507, "step": 20603 }, { "epoch": 0.9574349442379182, "grad_norm": 0.4112687455137729, "learning_rate": 5.016521132275576e-08, "loss": 0.039, "step": 20604 }, { "epoch": 0.9574814126394052, "grad_norm": 0.4984097456952506, "learning_rate": 5.005893796105554e-08, "loss": 0.0436, "step": 20605 }, { "epoch": 0.9575278810408921, "grad_norm": 0.4570074942839304, "learning_rate": 4.995277672072973e-08, "loss": 0.0398, "step": 20606 }, { "epoch": 0.9575743494423792, "grad_norm": 0.5190624819785529, "learning_rate": 4.984672760418363e-08, "loss": 0.0506, "step": 20607 }, { "epoch": 0.9576208178438662, "grad_norm": 0.5219267763998561, "learning_rate": 4.97407906138192e-08, "loss": 0.045, "step": 20608 }, { "epoch": 0.9576672862453531, "grad_norm": 0.6627167240057433, "learning_rate": 4.9634965752035634e-08, "loss": 0.0739, "step": 20609 }, { "epoch": 0.9577137546468402, "grad_norm": 0.4473085347089659, "learning_rate": 4.952925302123046e-08, "loss": 0.0571, "step": 20610 }, { "epoch": 0.9577602230483271, "grad_norm": 0.5503442239189366, "learning_rate": 4.9423652423797875e-08, "loss": 0.0601, "step": 20611 }, { "epoch": 0.9578066914498141, "grad_norm": 0.6586593664745252, "learning_rate": 4.931816396212985e-08, "loss": 0.0614, "step": 20612 }, { "epoch": 0.9578531598513012, "grad_norm": 0.46824030344018286, "learning_rate": 4.921278763861614e-08, "loss": 0.0309, "step": 20613 }, { "epoch": 0.9578996282527881, "grad_norm": 0.6668127068521263, "learning_rate": 4.910752345564318e-08, "loss": 0.0569, "step": 20614 }, { "epoch": 0.9579460966542751, "grad_norm": 0.41324849716523193, "learning_rate": 4.9002371415595716e-08, "loss": 0.0391, "step": 20615 }, { "epoch": 0.9579925650557621, "grad_norm": 1.0850475957427441, "learning_rate": 4.8897331520855184e-08, "loss": 0.0788, "step": 20616 }, { "epoch": 0.9580390334572491, "grad_norm": 0.4815427446172018, "learning_rate": 4.879240377380135e-08, "loss": 0.033, "step": 20617 }, { "epoch": 0.9580855018587361, "grad_norm": 0.7383096061442131, "learning_rate": 4.8687588176810074e-08, "loss": 0.0559, "step": 20618 }, { "epoch": 0.958131970260223, "grad_norm": 0.40986974568269197, "learning_rate": 4.8582884732256696e-08, "loss": 0.0407, "step": 20619 }, { "epoch": 0.95817843866171, "grad_norm": 0.3991714675997886, "learning_rate": 4.8478293442511535e-08, "loss": 0.0366, "step": 20620 }, { "epoch": 0.9582249070631971, "grad_norm": 0.6223159275345788, "learning_rate": 4.837381430994492e-08, "loss": 0.0657, "step": 20621 }, { "epoch": 0.958271375464684, "grad_norm": 0.340942465574544, "learning_rate": 4.826944733692329e-08, "loss": 0.0384, "step": 20622 }, { "epoch": 0.958317843866171, "grad_norm": 0.38194720255358033, "learning_rate": 4.8165192525809754e-08, "loss": 0.0351, "step": 20623 }, { "epoch": 0.958364312267658, "grad_norm": 0.7331543347742675, "learning_rate": 4.8061049878966316e-08, "loss": 0.0474, "step": 20624 }, { "epoch": 0.958410780669145, "grad_norm": 0.39428109338824946, "learning_rate": 4.795701939875219e-08, "loss": 0.0308, "step": 20625 }, { "epoch": 0.958457249070632, "grad_norm": 0.5129138765825503, "learning_rate": 4.7853101087522724e-08, "loss": 0.0397, "step": 20626 }, { "epoch": 0.9585037174721189, "grad_norm": 0.42328904527942085, "learning_rate": 4.774929494763325e-08, "loss": 0.055, "step": 20627 }, { "epoch": 0.958550185873606, "grad_norm": 0.5250056320153023, "learning_rate": 4.7645600981434114e-08, "loss": 0.0565, "step": 20628 }, { "epoch": 0.9585966542750929, "grad_norm": 0.33793543732797776, "learning_rate": 4.75420191912751e-08, "loss": 0.0282, "step": 20629 }, { "epoch": 0.9586431226765799, "grad_norm": 0.45690987903549063, "learning_rate": 4.7438549579501e-08, "loss": 0.0377, "step": 20630 }, { "epoch": 0.958689591078067, "grad_norm": 0.2701133094693387, "learning_rate": 4.733519214845661e-08, "loss": 0.0284, "step": 20631 }, { "epoch": 0.9587360594795539, "grad_norm": 0.3857200974138103, "learning_rate": 4.723194690048227e-08, "loss": 0.0556, "step": 20632 }, { "epoch": 0.9587825278810409, "grad_norm": 0.38197512479877516, "learning_rate": 4.712881383791723e-08, "loss": 0.0286, "step": 20633 }, { "epoch": 0.9588289962825279, "grad_norm": 0.5982370268706628, "learning_rate": 4.702579296309739e-08, "loss": 0.0558, "step": 20634 }, { "epoch": 0.9588754646840149, "grad_norm": 0.7861286901919109, "learning_rate": 4.692288427835645e-08, "loss": 0.0521, "step": 20635 }, { "epoch": 0.9589219330855019, "grad_norm": 0.5248138325946833, "learning_rate": 4.682008778602476e-08, "loss": 0.0616, "step": 20636 }, { "epoch": 0.9589684014869888, "grad_norm": 0.5069050284181322, "learning_rate": 4.671740348843157e-08, "loss": 0.04, "step": 20637 }, { "epoch": 0.9590148698884758, "grad_norm": 0.43799459124016377, "learning_rate": 4.661483138790224e-08, "loss": 0.0364, "step": 20638 }, { "epoch": 0.9590613382899629, "grad_norm": 0.4987681494931765, "learning_rate": 4.651237148676047e-08, "loss": 0.059, "step": 20639 }, { "epoch": 0.9591078066914498, "grad_norm": 0.7014481602165074, "learning_rate": 4.6410023787326065e-08, "loss": 0.0464, "step": 20640 }, { "epoch": 0.9591542750929368, "grad_norm": 0.4751604744602146, "learning_rate": 4.6307788291919396e-08, "loss": 0.0576, "step": 20641 }, { "epoch": 0.9592007434944237, "grad_norm": 0.9959224892231656, "learning_rate": 4.6205665002854725e-08, "loss": 0.0701, "step": 20642 }, { "epoch": 0.9592472118959108, "grad_norm": 0.28876632105675204, "learning_rate": 4.61036539224452e-08, "loss": 0.0276, "step": 20643 }, { "epoch": 0.9592936802973978, "grad_norm": 0.41255846115204564, "learning_rate": 4.600175505300175e-08, "loss": 0.0489, "step": 20644 }, { "epoch": 0.9593401486988847, "grad_norm": 0.39833466259549627, "learning_rate": 4.5899968396833087e-08, "loss": 0.0403, "step": 20645 }, { "epoch": 0.9593866171003718, "grad_norm": 0.4407259136112741, "learning_rate": 4.579829395624347e-08, "loss": 0.0488, "step": 20646 }, { "epoch": 0.9594330855018587, "grad_norm": 0.49691556754707666, "learning_rate": 4.569673173353717e-08, "loss": 0.0504, "step": 20647 }, { "epoch": 0.9594795539033457, "grad_norm": 0.4696874921297957, "learning_rate": 4.559528173101457e-08, "loss": 0.0715, "step": 20648 }, { "epoch": 0.9595260223048327, "grad_norm": 0.7910045685384516, "learning_rate": 4.549394395097273e-08, "loss": 0.0421, "step": 20649 }, { "epoch": 0.9595724907063197, "grad_norm": 0.4898856183611734, "learning_rate": 4.539271839570702e-08, "loss": 0.0617, "step": 20650 }, { "epoch": 0.9596189591078067, "grad_norm": 0.36072754815088565, "learning_rate": 4.529160506751173e-08, "loss": 0.0415, "step": 20651 }, { "epoch": 0.9596654275092937, "grad_norm": 0.40484811962049316, "learning_rate": 4.5190603968676136e-08, "loss": 0.0626, "step": 20652 }, { "epoch": 0.9597118959107807, "grad_norm": 0.5738333156515619, "learning_rate": 4.5089715101488405e-08, "loss": 0.0557, "step": 20653 }, { "epoch": 0.9597583643122677, "grad_norm": 0.4096972950738873, "learning_rate": 4.498893846823338e-08, "loss": 0.0667, "step": 20654 }, { "epoch": 0.9598048327137546, "grad_norm": 0.5589842471578063, "learning_rate": 4.4888274071193674e-08, "loss": 0.0661, "step": 20655 }, { "epoch": 0.9598513011152416, "grad_norm": 0.4246485141655998, "learning_rate": 4.478772191265024e-08, "loss": 0.0508, "step": 20656 }, { "epoch": 0.9598977695167287, "grad_norm": 0.5538015630173804, "learning_rate": 4.468728199487959e-08, "loss": 0.0678, "step": 20657 }, { "epoch": 0.9599442379182156, "grad_norm": 0.3353581997891049, "learning_rate": 4.4586954320157695e-08, "loss": 0.0346, "step": 20658 }, { "epoch": 0.9599907063197026, "grad_norm": 0.45597239705711967, "learning_rate": 4.448673889075716e-08, "loss": 0.0468, "step": 20659 }, { "epoch": 0.9600371747211895, "grad_norm": 0.3302020857757838, "learning_rate": 4.438663570894674e-08, "loss": 0.0298, "step": 20660 }, { "epoch": 0.9600836431226766, "grad_norm": 0.5727560546241358, "learning_rate": 4.428664477699518e-08, "loss": 0.0488, "step": 20661 }, { "epoch": 0.9601301115241636, "grad_norm": 0.7606815536087934, "learning_rate": 4.4186766097166765e-08, "loss": 0.0535, "step": 20662 }, { "epoch": 0.9601765799256505, "grad_norm": 0.2641517144276675, "learning_rate": 4.408699967172414e-08, "loss": 0.0403, "step": 20663 }, { "epoch": 0.9602230483271376, "grad_norm": 0.4049857767190108, "learning_rate": 4.398734550292716e-08, "loss": 0.0497, "step": 20664 }, { "epoch": 0.9602695167286245, "grad_norm": 0.3516739325559594, "learning_rate": 4.388780359303291e-08, "loss": 0.0336, "step": 20665 }, { "epoch": 0.9603159851301115, "grad_norm": 0.48768484085715796, "learning_rate": 4.3788373944295694e-08, "loss": 0.0498, "step": 20666 }, { "epoch": 0.9603624535315985, "grad_norm": 0.49392909233988347, "learning_rate": 4.368905655896816e-08, "loss": 0.052, "step": 20667 }, { "epoch": 0.9604089219330855, "grad_norm": 0.4621673811902458, "learning_rate": 4.358985143930017e-08, "loss": 0.0397, "step": 20668 }, { "epoch": 0.9604553903345725, "grad_norm": 0.5741286347263029, "learning_rate": 4.349075858753826e-08, "loss": 0.0392, "step": 20669 }, { "epoch": 0.9605018587360595, "grad_norm": 0.25744413598154137, "learning_rate": 4.3391778005927865e-08, "loss": 0.043, "step": 20670 }, { "epoch": 0.9605483271375465, "grad_norm": 0.6379143745698571, "learning_rate": 4.329290969670996e-08, "loss": 0.0631, "step": 20671 }, { "epoch": 0.9605947955390335, "grad_norm": 0.9391189734368184, "learning_rate": 4.3194153662124426e-08, "loss": 0.0684, "step": 20672 }, { "epoch": 0.9606412639405204, "grad_norm": 0.36369718955189945, "learning_rate": 4.309550990440892e-08, "loss": 0.0425, "step": 20673 }, { "epoch": 0.9606877323420074, "grad_norm": 0.31261539487636353, "learning_rate": 4.299697842579609e-08, "loss": 0.0268, "step": 20674 }, { "epoch": 0.9607342007434945, "grad_norm": 0.5350916617245807, "learning_rate": 4.289855922851971e-08, "loss": 0.0462, "step": 20675 }, { "epoch": 0.9607806691449814, "grad_norm": 0.6281515354976037, "learning_rate": 4.2800252314808e-08, "loss": 0.0511, "step": 20676 }, { "epoch": 0.9608271375464684, "grad_norm": 0.4341966056318096, "learning_rate": 4.270205768688751e-08, "loss": 0.0394, "step": 20677 }, { "epoch": 0.9608736059479553, "grad_norm": 0.3326610781096745, "learning_rate": 4.260397534698368e-08, "loss": 0.0338, "step": 20678 }, { "epoch": 0.9609200743494424, "grad_norm": 0.36908331423752827, "learning_rate": 4.2506005297316964e-08, "loss": 0.0357, "step": 20679 }, { "epoch": 0.9609665427509294, "grad_norm": 0.37555200068233274, "learning_rate": 4.240814754010669e-08, "loss": 0.0397, "step": 20680 }, { "epoch": 0.9610130111524163, "grad_norm": 0.6223743404484469, "learning_rate": 4.2310402077569427e-08, "loss": 0.0539, "step": 20681 }, { "epoch": 0.9610594795539034, "grad_norm": 0.4792039800515349, "learning_rate": 4.221276891192061e-08, "loss": 0.0381, "step": 20682 }, { "epoch": 0.9611059479553903, "grad_norm": 0.4130832080949408, "learning_rate": 4.2115248045369596e-08, "loss": 0.0408, "step": 20683 }, { "epoch": 0.9611524163568773, "grad_norm": 0.3967490002598756, "learning_rate": 4.201783948012628e-08, "loss": 0.0403, "step": 20684 }, { "epoch": 0.9611988847583643, "grad_norm": 0.3687556709703044, "learning_rate": 4.1920543218397227e-08, "loss": 0.0559, "step": 20685 }, { "epoch": 0.9612453531598513, "grad_norm": 0.5186956429219902, "learning_rate": 4.182335926238623e-08, "loss": 0.06, "step": 20686 }, { "epoch": 0.9612918215613383, "grad_norm": 0.45115620946889506, "learning_rate": 4.172628761429487e-08, "loss": 0.0486, "step": 20687 }, { "epoch": 0.9613382899628253, "grad_norm": 0.6274456382735025, "learning_rate": 4.162932827632082e-08, "loss": 0.064, "step": 20688 }, { "epoch": 0.9613847583643123, "grad_norm": 0.5406476038524801, "learning_rate": 4.153248125066178e-08, "loss": 0.0515, "step": 20689 }, { "epoch": 0.9614312267657993, "grad_norm": 0.48491568931965673, "learning_rate": 4.1435746539510436e-08, "loss": 0.0409, "step": 20690 }, { "epoch": 0.9614776951672862, "grad_norm": 0.4766885009066819, "learning_rate": 4.133912414505781e-08, "loss": 0.0464, "step": 20691 }, { "epoch": 0.9615241635687732, "grad_norm": 0.4615232261125331, "learning_rate": 4.124261406949326e-08, "loss": 0.0568, "step": 20692 }, { "epoch": 0.9615706319702603, "grad_norm": 0.4975071479645124, "learning_rate": 4.114621631500282e-08, "loss": 0.0422, "step": 20693 }, { "epoch": 0.9616171003717472, "grad_norm": 0.5141024602287899, "learning_rate": 4.104993088376974e-08, "loss": 0.0524, "step": 20694 }, { "epoch": 0.9616635687732342, "grad_norm": 0.5112993620887726, "learning_rate": 4.09537577779745e-08, "loss": 0.0434, "step": 20695 }, { "epoch": 0.9617100371747211, "grad_norm": 0.45958106723266573, "learning_rate": 4.085769699979591e-08, "loss": 0.0555, "step": 20696 }, { "epoch": 0.9617565055762082, "grad_norm": 0.514749769965309, "learning_rate": 4.076174855141002e-08, "loss": 0.062, "step": 20697 }, { "epoch": 0.9618029739776952, "grad_norm": 0.33453179141786027, "learning_rate": 4.066591243499007e-08, "loss": 0.0422, "step": 20698 }, { "epoch": 0.9618494423791821, "grad_norm": 0.7484857681765152, "learning_rate": 4.057018865270657e-08, "loss": 0.0396, "step": 20699 }, { "epoch": 0.9618959107806692, "grad_norm": 0.572045185964579, "learning_rate": 4.047457720672887e-08, "loss": 0.0451, "step": 20700 }, { "epoch": 0.9619423791821561, "grad_norm": 0.5894110970068053, "learning_rate": 4.0379078099220816e-08, "loss": 0.0564, "step": 20701 }, { "epoch": 0.9619888475836431, "grad_norm": 0.5898234386057531, "learning_rate": 4.028369133234677e-08, "loss": 0.0733, "step": 20702 }, { "epoch": 0.9620353159851301, "grad_norm": 0.556761693458248, "learning_rate": 4.0188416908267245e-08, "loss": 0.0666, "step": 20703 }, { "epoch": 0.9620817843866171, "grad_norm": 0.4266427953763347, "learning_rate": 4.00932548291405e-08, "loss": 0.0604, "step": 20704 }, { "epoch": 0.9621282527881041, "grad_norm": 0.6149951813136607, "learning_rate": 3.999820509712094e-08, "loss": 0.0499, "step": 20705 }, { "epoch": 0.9621747211895911, "grad_norm": 0.8551499498681306, "learning_rate": 3.9903267714362924e-08, "loss": 0.0743, "step": 20706 }, { "epoch": 0.962221189591078, "grad_norm": 0.46615449704466105, "learning_rate": 3.980844268301642e-08, "loss": 0.0524, "step": 20707 }, { "epoch": 0.9622676579925651, "grad_norm": 0.4766736639741905, "learning_rate": 3.97137300052286e-08, "loss": 0.0541, "step": 20708 }, { "epoch": 0.962314126394052, "grad_norm": 0.5436679305185224, "learning_rate": 3.961912968314607e-08, "loss": 0.0367, "step": 20709 }, { "epoch": 0.962360594795539, "grad_norm": 0.3419861949452574, "learning_rate": 3.952464171891046e-08, "loss": 0.04, "step": 20710 }, { "epoch": 0.9624070631970261, "grad_norm": 0.8108534161171121, "learning_rate": 3.943026611466283e-08, "loss": 0.0677, "step": 20711 }, { "epoch": 0.962453531598513, "grad_norm": 0.5091612562330431, "learning_rate": 3.9336002872540355e-08, "loss": 0.057, "step": 20712 }, { "epoch": 0.9625, "grad_norm": 2.361206052652823, "learning_rate": 3.9241851994678556e-08, "loss": 0.0772, "step": 20713 }, { "epoch": 0.9625464684014869, "grad_norm": 0.3202855221503105, "learning_rate": 3.914781348321017e-08, "loss": 0.0471, "step": 20714 }, { "epoch": 0.962592936802974, "grad_norm": 0.42516010704038, "learning_rate": 3.905388734026461e-08, "loss": 0.0432, "step": 20715 }, { "epoch": 0.962639405204461, "grad_norm": 0.5223217101305555, "learning_rate": 3.8960073567969604e-08, "loss": 0.058, "step": 20716 }, { "epoch": 0.9626858736059479, "grad_norm": 0.3828465305396167, "learning_rate": 3.8866372168450685e-08, "loss": 0.0642, "step": 20717 }, { "epoch": 0.962732342007435, "grad_norm": 0.5659850503229857, "learning_rate": 3.877278314383004e-08, "loss": 0.0393, "step": 20718 }, { "epoch": 0.9627788104089219, "grad_norm": 0.35242705785829076, "learning_rate": 3.8679306496227645e-08, "loss": 0.0456, "step": 20719 }, { "epoch": 0.9628252788104089, "grad_norm": 0.5278831409019389, "learning_rate": 3.858594222776013e-08, "loss": 0.0537, "step": 20720 }, { "epoch": 0.962871747211896, "grad_norm": 0.4177542155336928, "learning_rate": 3.849269034054304e-08, "loss": 0.0565, "step": 20721 }, { "epoch": 0.9629182156133829, "grad_norm": 0.6337472459761728, "learning_rate": 3.8399550836688003e-08, "loss": 0.0604, "step": 20722 }, { "epoch": 0.9629646840148699, "grad_norm": 0.4925372718959045, "learning_rate": 3.830652371830612e-08, "loss": 0.0581, "step": 20723 }, { "epoch": 0.9630111524163569, "grad_norm": 0.5849867345956843, "learning_rate": 3.821360898750293e-08, "loss": 0.0638, "step": 20724 }, { "epoch": 0.9630576208178439, "grad_norm": 0.44359893361385366, "learning_rate": 3.81208066463834e-08, "loss": 0.048, "step": 20725 }, { "epoch": 0.9631040892193309, "grad_norm": 0.483819654057361, "learning_rate": 3.802811669705031e-08, "loss": 0.0571, "step": 20726 }, { "epoch": 0.9631505576208178, "grad_norm": 0.27727146117078355, "learning_rate": 3.7935539141602526e-08, "loss": 0.0276, "step": 20727 }, { "epoch": 0.9631970260223048, "grad_norm": 0.6039149831264776, "learning_rate": 3.784307398213727e-08, "loss": 0.0506, "step": 20728 }, { "epoch": 0.9632434944237919, "grad_norm": 0.5417127094590362, "learning_rate": 3.775072122074897e-08, "loss": 0.0484, "step": 20729 }, { "epoch": 0.9632899628252788, "grad_norm": 0.3040876312816815, "learning_rate": 3.765848085952983e-08, "loss": 0.0343, "step": 20730 }, { "epoch": 0.9633364312267658, "grad_norm": 0.5403504310556915, "learning_rate": 3.756635290056876e-08, "loss": 0.058, "step": 20731 }, { "epoch": 0.9633828996282527, "grad_norm": 0.2482496954221185, "learning_rate": 3.747433734595185e-08, "loss": 0.0271, "step": 20732 }, { "epoch": 0.9634293680297398, "grad_norm": 0.36599523766597086, "learning_rate": 3.7382434197765216e-08, "loss": 0.042, "step": 20733 }, { "epoch": 0.9634758364312268, "grad_norm": 0.584903619020793, "learning_rate": 3.729064345808886e-08, "loss": 0.055, "step": 20734 }, { "epoch": 0.9635223048327137, "grad_norm": 0.4475978164498257, "learning_rate": 3.7198965129003336e-08, "loss": 0.0399, "step": 20735 }, { "epoch": 0.9635687732342008, "grad_norm": 0.48612718652374104, "learning_rate": 3.710739921258366e-08, "loss": 0.0504, "step": 20736 }, { "epoch": 0.9636152416356877, "grad_norm": 0.6457291926184417, "learning_rate": 3.7015945710905387e-08, "loss": 0.0457, "step": 20737 }, { "epoch": 0.9636617100371747, "grad_norm": 0.5471297689911496, "learning_rate": 3.692460462603909e-08, "loss": 0.0623, "step": 20738 }, { "epoch": 0.9637081784386617, "grad_norm": 0.5604862687150903, "learning_rate": 3.683337596005365e-08, "loss": 0.0392, "step": 20739 }, { "epoch": 0.9637546468401487, "grad_norm": 0.4810809076899891, "learning_rate": 3.674225971501632e-08, "loss": 0.0607, "step": 20740 }, { "epoch": 0.9638011152416357, "grad_norm": 0.7789691501125292, "learning_rate": 3.665125589299101e-08, "loss": 0.0713, "step": 20741 }, { "epoch": 0.9638475836431227, "grad_norm": 0.5886887273954874, "learning_rate": 3.656036449603773e-08, "loss": 0.0475, "step": 20742 }, { "epoch": 0.9638940520446097, "grad_norm": 0.41208758289056263, "learning_rate": 3.646958552621649e-08, "loss": 0.071, "step": 20743 }, { "epoch": 0.9639405204460967, "grad_norm": 0.36462359080877593, "learning_rate": 3.6378918985583435e-08, "loss": 0.0563, "step": 20744 }, { "epoch": 0.9639869888475836, "grad_norm": 0.3675830975852784, "learning_rate": 3.628836487619136e-08, "loss": 0.0545, "step": 20745 }, { "epoch": 0.9640334572490706, "grad_norm": 1.58313311803609, "learning_rate": 3.6197923200092525e-08, "loss": 0.0618, "step": 20746 }, { "epoch": 0.9640799256505577, "grad_norm": 0.4389880096020678, "learning_rate": 3.6107593959335276e-08, "loss": 0.0314, "step": 20747 }, { "epoch": 0.9641263940520446, "grad_norm": 0.37790555247747154, "learning_rate": 3.601737715596465e-08, "loss": 0.0523, "step": 20748 }, { "epoch": 0.9641728624535316, "grad_norm": 0.6504921972448438, "learning_rate": 3.592727279202513e-08, "loss": 0.061, "step": 20749 }, { "epoch": 0.9642193308550185, "grad_norm": 0.5123836471714549, "learning_rate": 3.5837280869557845e-08, "loss": 0.0467, "step": 20750 }, { "epoch": 0.9642657992565056, "grad_norm": 0.3199729752682597, "learning_rate": 3.574740139060062e-08, "loss": 0.0251, "step": 20751 }, { "epoch": 0.9643122676579926, "grad_norm": 0.8320244587939541, "learning_rate": 3.565763435718961e-08, "loss": 0.0574, "step": 20752 }, { "epoch": 0.9643587360594795, "grad_norm": 0.4511375187208447, "learning_rate": 3.556797977135817e-08, "loss": 0.0585, "step": 20753 }, { "epoch": 0.9644052044609666, "grad_norm": 0.40603852034853055, "learning_rate": 3.547843763513692e-08, "loss": 0.0442, "step": 20754 }, { "epoch": 0.9644516728624535, "grad_norm": 0.560378797620453, "learning_rate": 3.538900795055367e-08, "loss": 0.0597, "step": 20755 }, { "epoch": 0.9644981412639405, "grad_norm": 0.4785571249019565, "learning_rate": 3.529969071963457e-08, "loss": 0.0485, "step": 20756 }, { "epoch": 0.9645446096654275, "grad_norm": 0.4073387495687148, "learning_rate": 3.521048594440302e-08, "loss": 0.0493, "step": 20757 }, { "epoch": 0.9645910780669145, "grad_norm": 0.31968666956600816, "learning_rate": 3.5121393626879605e-08, "loss": 0.041, "step": 20758 }, { "epoch": 0.9646375464684015, "grad_norm": 0.5464393368780258, "learning_rate": 3.503241376908162e-08, "loss": 0.0464, "step": 20759 }, { "epoch": 0.9646840148698885, "grad_norm": 0.33301282217873307, "learning_rate": 3.4943546373025216e-08, "loss": 0.0554, "step": 20760 }, { "epoch": 0.9647304832713754, "grad_norm": 0.6377494494739584, "learning_rate": 3.485479144072323e-08, "loss": 0.0497, "step": 20761 }, { "epoch": 0.9647769516728625, "grad_norm": 0.4072821767769115, "learning_rate": 3.476614897418573e-08, "loss": 0.0473, "step": 20762 }, { "epoch": 0.9648234200743494, "grad_norm": 0.5638380298413167, "learning_rate": 3.467761897542055e-08, "loss": 0.0802, "step": 20763 }, { "epoch": 0.9648698884758364, "grad_norm": 0.43305778803081063, "learning_rate": 3.4589201446433295e-08, "loss": 0.0464, "step": 20764 }, { "epoch": 0.9649163568773235, "grad_norm": 0.552210588639056, "learning_rate": 3.450089638922738e-08, "loss": 0.0334, "step": 20765 }, { "epoch": 0.9649628252788104, "grad_norm": 0.8109840103433676, "learning_rate": 3.4412703805801196e-08, "loss": 0.0543, "step": 20766 }, { "epoch": 0.9650092936802974, "grad_norm": 0.4717592445743741, "learning_rate": 3.432462369815426e-08, "loss": 0.0649, "step": 20767 }, { "epoch": 0.9650557620817843, "grad_norm": 0.5163209314862829, "learning_rate": 3.4236656068280524e-08, "loss": 0.053, "step": 20768 }, { "epoch": 0.9651022304832714, "grad_norm": 0.4412476417470095, "learning_rate": 3.4148800918172855e-08, "loss": 0.0539, "step": 20769 }, { "epoch": 0.9651486988847584, "grad_norm": 0.4102225530432298, "learning_rate": 3.406105824982131e-08, "loss": 0.055, "step": 20770 }, { "epoch": 0.9651951672862453, "grad_norm": 0.30623445030749086, "learning_rate": 3.3973428065213755e-08, "loss": 0.0408, "step": 20771 }, { "epoch": 0.9652416356877324, "grad_norm": 0.42175768180861967, "learning_rate": 3.3885910366334154e-08, "loss": 0.0571, "step": 20772 }, { "epoch": 0.9652881040892193, "grad_norm": 0.4491089095635009, "learning_rate": 3.379850515516536e-08, "loss": 0.0492, "step": 20773 }, { "epoch": 0.9653345724907063, "grad_norm": 0.3863662590446217, "learning_rate": 3.371121243368747e-08, "loss": 0.0318, "step": 20774 }, { "epoch": 0.9653810408921933, "grad_norm": 0.56821663602286, "learning_rate": 3.3624032203877774e-08, "loss": 0.0376, "step": 20775 }, { "epoch": 0.9654275092936803, "grad_norm": 0.38254998828986936, "learning_rate": 3.353696446771082e-08, "loss": 0.0402, "step": 20776 }, { "epoch": 0.9654739776951673, "grad_norm": 0.3381370248601248, "learning_rate": 3.345000922715835e-08, "loss": 0.0283, "step": 20777 }, { "epoch": 0.9655204460966543, "grad_norm": 0.4325604651911912, "learning_rate": 3.336316648419047e-08, "loss": 0.0385, "step": 20778 }, { "epoch": 0.9655669144981412, "grad_norm": 0.5293208823772378, "learning_rate": 3.327643624077448e-08, "loss": 0.038, "step": 20779 }, { "epoch": 0.9656133828996283, "grad_norm": 0.3407243382973085, "learning_rate": 3.3189818498873815e-08, "loss": 0.0505, "step": 20780 }, { "epoch": 0.9656598513011152, "grad_norm": 0.4235013100508717, "learning_rate": 3.310331326045191e-08, "loss": 0.0473, "step": 20781 }, { "epoch": 0.9657063197026022, "grad_norm": 0.42334996337967085, "learning_rate": 3.301692052746774e-08, "loss": 0.0374, "step": 20782 }, { "epoch": 0.9657527881040893, "grad_norm": 0.5857097768286036, "learning_rate": 3.293064030187698e-08, "loss": 0.0727, "step": 20783 }, { "epoch": 0.9657992565055762, "grad_norm": 0.41666184072483675, "learning_rate": 3.284447258563583e-08, "loss": 0.0618, "step": 20784 }, { "epoch": 0.9658457249070632, "grad_norm": 0.4983437198522431, "learning_rate": 3.275841738069496e-08, "loss": 0.061, "step": 20785 }, { "epoch": 0.9658921933085501, "grad_norm": 0.5250639486230575, "learning_rate": 3.267247468900392e-08, "loss": 0.0576, "step": 20786 }, { "epoch": 0.9659386617100372, "grad_norm": 0.4713608000389867, "learning_rate": 3.258664451250892e-08, "loss": 0.0485, "step": 20787 }, { "epoch": 0.9659851301115242, "grad_norm": 0.6003347294303891, "learning_rate": 3.2500926853155094e-08, "loss": 0.0546, "step": 20788 }, { "epoch": 0.9660315985130111, "grad_norm": 0.5482433234967207, "learning_rate": 3.2415321712883643e-08, "loss": 0.0592, "step": 20789 }, { "epoch": 0.9660780669144982, "grad_norm": 0.4207767539830187, "learning_rate": 3.232982909363247e-08, "loss": 0.0433, "step": 20790 }, { "epoch": 0.9661245353159851, "grad_norm": 0.5809345157230114, "learning_rate": 3.224444899734003e-08, "loss": 0.0464, "step": 20791 }, { "epoch": 0.9661710037174721, "grad_norm": 0.9249867762371168, "learning_rate": 3.2159181425938655e-08, "loss": 0.0725, "step": 20792 }, { "epoch": 0.9662174721189591, "grad_norm": 0.5010323880204399, "learning_rate": 3.2074026381360704e-08, "loss": 0.055, "step": 20793 }, { "epoch": 0.9662639405204461, "grad_norm": 0.7492873204088191, "learning_rate": 3.1988983865534616e-08, "loss": 0.0515, "step": 20794 }, { "epoch": 0.9663104089219331, "grad_norm": 0.34197579329878586, "learning_rate": 3.19040538803872e-08, "loss": 0.0334, "step": 20795 }, { "epoch": 0.9663568773234201, "grad_norm": 0.5527204891551493, "learning_rate": 3.181923642784135e-08, "loss": 0.0393, "step": 20796 }, { "epoch": 0.966403345724907, "grad_norm": 0.42743341408337016, "learning_rate": 3.173453150981831e-08, "loss": 0.0275, "step": 20797 }, { "epoch": 0.9664498141263941, "grad_norm": 0.5797951433501864, "learning_rate": 3.1649939128237664e-08, "loss": 0.057, "step": 20798 }, { "epoch": 0.966496282527881, "grad_norm": 0.5233623929684483, "learning_rate": 3.156545928501509e-08, "loss": 0.0442, "step": 20799 }, { "epoch": 0.966542750929368, "grad_norm": 0.35871534588227416, "learning_rate": 3.148109198206351e-08, "loss": 0.0581, "step": 20800 }, { "epoch": 0.9665892193308551, "grad_norm": 0.30724318065832895, "learning_rate": 3.139683722129472e-08, "loss": 0.0417, "step": 20801 }, { "epoch": 0.966635687732342, "grad_norm": 0.3472341744633658, "learning_rate": 3.13126950046172e-08, "loss": 0.0359, "step": 20802 }, { "epoch": 0.966682156133829, "grad_norm": 0.5590058801401258, "learning_rate": 3.122866533393665e-08, "loss": 0.061, "step": 20803 }, { "epoch": 0.9667286245353159, "grad_norm": 0.6486549268533404, "learning_rate": 3.114474821115598e-08, "loss": 0.0504, "step": 20804 }, { "epoch": 0.966775092936803, "grad_norm": 0.5805423425767671, "learning_rate": 3.106094363817647e-08, "loss": 0.0606, "step": 20805 }, { "epoch": 0.96682156133829, "grad_norm": 0.729297323316162, "learning_rate": 3.0977251616896574e-08, "loss": 0.054, "step": 20806 }, { "epoch": 0.9668680297397769, "grad_norm": 0.5110775521782065, "learning_rate": 3.0893672149211464e-08, "loss": 0.0426, "step": 20807 }, { "epoch": 0.966914498141264, "grad_norm": 0.36034188355777874, "learning_rate": 3.0810205237014614e-08, "loss": 0.0413, "step": 20808 }, { "epoch": 0.9669609665427509, "grad_norm": 0.47944944199500844, "learning_rate": 3.072685088219618e-08, "loss": 0.0437, "step": 20809 }, { "epoch": 0.9670074349442379, "grad_norm": 0.5356143711762278, "learning_rate": 3.064360908664521e-08, "loss": 0.0642, "step": 20810 }, { "epoch": 0.9670539033457249, "grad_norm": 0.3891460052695473, "learning_rate": 3.0560479852246304e-08, "loss": 0.0464, "step": 20811 }, { "epoch": 0.9671003717472119, "grad_norm": 0.29387023601966555, "learning_rate": 3.0477463180883516e-08, "loss": 0.0442, "step": 20812 }, { "epoch": 0.9671468401486989, "grad_norm": 0.49997228447102066, "learning_rate": 3.039455907443589e-08, "loss": 0.031, "step": 20813 }, { "epoch": 0.9671933085501859, "grad_norm": 0.5381315768042888, "learning_rate": 3.031176753478138e-08, "loss": 0.0422, "step": 20814 }, { "epoch": 0.9672397769516728, "grad_norm": 0.7405955706191951, "learning_rate": 3.0229088563796803e-08, "loss": 0.0536, "step": 20815 }, { "epoch": 0.9672862453531599, "grad_norm": 0.3267407184600954, "learning_rate": 3.014652216335345e-08, "loss": 0.0445, "step": 20816 }, { "epoch": 0.9673327137546468, "grad_norm": 0.43312311639605483, "learning_rate": 3.006406833532205e-08, "loss": 0.0668, "step": 20817 }, { "epoch": 0.9673791821561338, "grad_norm": 0.5129139879284837, "learning_rate": 2.998172708157054e-08, "loss": 0.0496, "step": 20818 }, { "epoch": 0.9674256505576209, "grad_norm": 0.4842799881842804, "learning_rate": 2.9899498403963554e-08, "loss": 0.0425, "step": 20819 }, { "epoch": 0.9674721189591078, "grad_norm": 0.5259017975649146, "learning_rate": 2.981738230436404e-08, "loss": 0.056, "step": 20820 }, { "epoch": 0.9675185873605948, "grad_norm": 0.8759309666264034, "learning_rate": 2.973537878463162e-08, "loss": 0.0646, "step": 20821 }, { "epoch": 0.9675650557620817, "grad_norm": 0.7676878292679286, "learning_rate": 2.965348784662425e-08, "loss": 0.0684, "step": 20822 }, { "epoch": 0.9676115241635688, "grad_norm": 0.38339363284206074, "learning_rate": 2.9571709492196567e-08, "loss": 0.0448, "step": 20823 }, { "epoch": 0.9676579925650558, "grad_norm": 0.3927729599977107, "learning_rate": 2.949004372320041e-08, "loss": 0.0343, "step": 20824 }, { "epoch": 0.9677044609665427, "grad_norm": 0.3542141987648213, "learning_rate": 2.9408490541486532e-08, "loss": 0.0327, "step": 20825 }, { "epoch": 0.9677509293680298, "grad_norm": 0.6343021132478514, "learning_rate": 2.9327049948901786e-08, "loss": 0.0614, "step": 20826 }, { "epoch": 0.9677973977695167, "grad_norm": 0.4340902128734075, "learning_rate": 2.924572194729136e-08, "loss": 0.0568, "step": 20827 }, { "epoch": 0.9678438661710037, "grad_norm": 0.6273919317876535, "learning_rate": 2.9164506538496006e-08, "loss": 0.0247, "step": 20828 }, { "epoch": 0.9678903345724907, "grad_norm": 0.24442198189844452, "learning_rate": 2.908340372435703e-08, "loss": 0.0221, "step": 20829 }, { "epoch": 0.9679368029739777, "grad_norm": 0.4695389222965348, "learning_rate": 2.9002413506710736e-08, "loss": 0.0497, "step": 20830 }, { "epoch": 0.9679832713754647, "grad_norm": 0.4675137868346376, "learning_rate": 2.8921535887391218e-08, "loss": 0.0593, "step": 20831 }, { "epoch": 0.9680297397769517, "grad_norm": 0.5787304249997423, "learning_rate": 2.8840770868230893e-08, "loss": 0.0469, "step": 20832 }, { "epoch": 0.9680762081784386, "grad_norm": 0.6027708376765707, "learning_rate": 2.8760118451059416e-08, "loss": 0.0383, "step": 20833 }, { "epoch": 0.9681226765799257, "grad_norm": 0.37943873253323374, "learning_rate": 2.8679578637703098e-08, "loss": 0.0349, "step": 20834 }, { "epoch": 0.9681691449814126, "grad_norm": 0.3688013756286929, "learning_rate": 2.8599151429986593e-08, "loss": 0.0447, "step": 20835 }, { "epoch": 0.9682156133828996, "grad_norm": 0.476700994608592, "learning_rate": 2.8518836829732332e-08, "loss": 0.047, "step": 20836 }, { "epoch": 0.9682620817843867, "grad_norm": 0.3521537130530519, "learning_rate": 2.8438634838758305e-08, "loss": 0.0346, "step": 20837 }, { "epoch": 0.9683085501858736, "grad_norm": 0.6378287945699991, "learning_rate": 2.8358545458881393e-08, "loss": 0.0574, "step": 20838 }, { "epoch": 0.9683550185873606, "grad_norm": 0.48099219304640595, "learning_rate": 2.8278568691916253e-08, "loss": 0.0413, "step": 20839 }, { "epoch": 0.9684014869888475, "grad_norm": 0.6935380516839638, "learning_rate": 2.8198704539674216e-08, "loss": 0.0568, "step": 20840 }, { "epoch": 0.9684479553903346, "grad_norm": 0.5893449054838544, "learning_rate": 2.8118953003963833e-08, "loss": 0.0539, "step": 20841 }, { "epoch": 0.9684944237918216, "grad_norm": 0.43237778490236417, "learning_rate": 2.803931408659255e-08, "loss": 0.0578, "step": 20842 }, { "epoch": 0.9685408921933085, "grad_norm": 0.594660631326716, "learning_rate": 2.795978778936337e-08, "loss": 0.0515, "step": 20843 }, { "epoch": 0.9685873605947956, "grad_norm": 0.40160086503189024, "learning_rate": 2.7880374114078178e-08, "loss": 0.0474, "step": 20844 }, { "epoch": 0.9686338289962825, "grad_norm": 0.5488772236318679, "learning_rate": 2.780107306253499e-08, "loss": 0.0499, "step": 20845 }, { "epoch": 0.9686802973977695, "grad_norm": 0.44136345716604275, "learning_rate": 2.772188463653125e-08, "loss": 0.0303, "step": 20846 }, { "epoch": 0.9687267657992565, "grad_norm": 0.3553099646326662, "learning_rate": 2.764280883785997e-08, "loss": 0.0491, "step": 20847 }, { "epoch": 0.9687732342007435, "grad_norm": 0.5164387513033659, "learning_rate": 2.7563845668312494e-08, "loss": 0.0586, "step": 20848 }, { "epoch": 0.9688197026022305, "grad_norm": 0.2690854075556186, "learning_rate": 2.748499512967684e-08, "loss": 0.0475, "step": 20849 }, { "epoch": 0.9688661710037175, "grad_norm": 0.6846120628049648, "learning_rate": 2.7406257223739908e-08, "loss": 0.0407, "step": 20850 }, { "epoch": 0.9689126394052044, "grad_norm": 0.5847023889042197, "learning_rate": 2.7327631952284715e-08, "loss": 0.0445, "step": 20851 }, { "epoch": 0.9689591078066915, "grad_norm": 0.7585233059814289, "learning_rate": 2.7249119317092064e-08, "loss": 0.0556, "step": 20852 }, { "epoch": 0.9690055762081784, "grad_norm": 0.6078106626882184, "learning_rate": 2.7170719319941085e-08, "loss": 0.0487, "step": 20853 }, { "epoch": 0.9690520446096654, "grad_norm": 0.5550888362615158, "learning_rate": 2.7092431962607023e-08, "loss": 0.0665, "step": 20854 }, { "epoch": 0.9690985130111525, "grad_norm": 0.5526482213200047, "learning_rate": 2.7014257246862907e-08, "loss": 0.0441, "step": 20855 }, { "epoch": 0.9691449814126394, "grad_norm": 0.6129478769992653, "learning_rate": 2.6936195174480095e-08, "loss": 0.0665, "step": 20856 }, { "epoch": 0.9691914498141264, "grad_norm": 0.2666010980013572, "learning_rate": 2.685824574722662e-08, "loss": 0.0386, "step": 20857 }, { "epoch": 0.9692379182156133, "grad_norm": 0.5205916664020372, "learning_rate": 2.6780408966867734e-08, "loss": 0.0531, "step": 20858 }, { "epoch": 0.9692843866171004, "grad_norm": 0.4382935507772394, "learning_rate": 2.6702684835167024e-08, "loss": 0.0317, "step": 20859 }, { "epoch": 0.9693308550185874, "grad_norm": 0.3691961942856205, "learning_rate": 2.6625073353884756e-08, "loss": 0.0455, "step": 20860 }, { "epoch": 0.9693773234200743, "grad_norm": 0.5106467622905364, "learning_rate": 2.6547574524778962e-08, "loss": 0.0448, "step": 20861 }, { "epoch": 0.9694237918215614, "grad_norm": 0.3782947233964926, "learning_rate": 2.6470188349604354e-08, "loss": 0.0281, "step": 20862 }, { "epoch": 0.9694702602230483, "grad_norm": 0.5334467343725341, "learning_rate": 2.6392914830115634e-08, "loss": 0.0463, "step": 20863 }, { "epoch": 0.9695167286245353, "grad_norm": 0.5677745850802192, "learning_rate": 2.6315753968061408e-08, "loss": 0.0569, "step": 20864 }, { "epoch": 0.9695631970260223, "grad_norm": 0.5301922148119573, "learning_rate": 2.6238705765190274e-08, "loss": 0.0488, "step": 20865 }, { "epoch": 0.9696096654275093, "grad_norm": 0.4718923780258107, "learning_rate": 2.616177022324695e-08, "loss": 0.0475, "step": 20866 }, { "epoch": 0.9696561338289963, "grad_norm": 0.3082617556917812, "learning_rate": 2.6084947343975042e-08, "loss": 0.0454, "step": 20867 }, { "epoch": 0.9697026022304833, "grad_norm": 0.5143790285660897, "learning_rate": 2.6008237129113712e-08, "loss": 0.0526, "step": 20868 }, { "epoch": 0.9697490706319702, "grad_norm": 0.6260635119230008, "learning_rate": 2.593163958040046e-08, "loss": 0.0514, "step": 20869 }, { "epoch": 0.9697955390334573, "grad_norm": 0.5360075489649235, "learning_rate": 2.5855154699571117e-08, "loss": 0.055, "step": 20870 }, { "epoch": 0.9698420074349442, "grad_norm": 0.39239564614990097, "learning_rate": 2.5778782488357635e-08, "loss": 0.0485, "step": 20871 }, { "epoch": 0.9698884758364312, "grad_norm": 0.49219359505020654, "learning_rate": 2.5702522948489738e-08, "loss": 0.0387, "step": 20872 }, { "epoch": 0.9699349442379183, "grad_norm": 0.29539679526917767, "learning_rate": 2.5626376081694936e-08, "loss": 0.047, "step": 20873 }, { "epoch": 0.9699814126394052, "grad_norm": 0.42270207766978585, "learning_rate": 2.5550341889698516e-08, "loss": 0.0493, "step": 20874 }, { "epoch": 0.9700278810408922, "grad_norm": 0.3032557660789108, "learning_rate": 2.547442037422243e-08, "loss": 0.0201, "step": 20875 }, { "epoch": 0.9700743494423791, "grad_norm": 0.43687583522315837, "learning_rate": 2.5398611536986416e-08, "loss": 0.0377, "step": 20876 }, { "epoch": 0.9701208178438662, "grad_norm": 0.3156319488581136, "learning_rate": 2.5322915379707437e-08, "loss": 0.0487, "step": 20877 }, { "epoch": 0.9701672862453532, "grad_norm": 0.5610974429966606, "learning_rate": 2.524733190409967e-08, "loss": 0.0428, "step": 20878 }, { "epoch": 0.9702137546468401, "grad_norm": 0.5944340144549677, "learning_rate": 2.5171861111876195e-08, "loss": 0.0499, "step": 20879 }, { "epoch": 0.9702602230483272, "grad_norm": 0.4522082135487036, "learning_rate": 2.509650300474564e-08, "loss": 0.0549, "step": 20880 }, { "epoch": 0.9703066914498141, "grad_norm": 0.5929591012046176, "learning_rate": 2.5021257584415536e-08, "loss": 0.0586, "step": 20881 }, { "epoch": 0.9703531598513011, "grad_norm": 0.40246920440795186, "learning_rate": 2.494612485259007e-08, "loss": 0.0391, "step": 20882 }, { "epoch": 0.9703996282527881, "grad_norm": 0.5388412662010886, "learning_rate": 2.487110481097066e-08, "loss": 0.0674, "step": 20883 }, { "epoch": 0.9704460966542751, "grad_norm": 0.4446909834172252, "learning_rate": 2.479619746125761e-08, "loss": 0.0403, "step": 20884 }, { "epoch": 0.9704925650557621, "grad_norm": 0.6123705091850222, "learning_rate": 2.4721402805146234e-08, "loss": 0.0597, "step": 20885 }, { "epoch": 0.9705390334572491, "grad_norm": 0.513590979902013, "learning_rate": 2.464672084433184e-08, "loss": 0.0483, "step": 20886 }, { "epoch": 0.970585501858736, "grad_norm": 0.49535233610254703, "learning_rate": 2.4572151580505853e-08, "loss": 0.0558, "step": 20887 }, { "epoch": 0.9706319702602231, "grad_norm": 0.7197736529542995, "learning_rate": 2.4497695015356926e-08, "loss": 0.0569, "step": 20888 }, { "epoch": 0.97067843866171, "grad_norm": 0.5448411495501967, "learning_rate": 2.442335115057204e-08, "loss": 0.0495, "step": 20889 }, { "epoch": 0.970724907063197, "grad_norm": 0.3020039447558317, "learning_rate": 2.4349119987834847e-08, "loss": 0.0371, "step": 20890 }, { "epoch": 0.9707713754646841, "grad_norm": 0.4983540671646315, "learning_rate": 2.427500152882678e-08, "loss": 0.0662, "step": 20891 }, { "epoch": 0.970817843866171, "grad_norm": 0.45577854715016475, "learning_rate": 2.4200995775226498e-08, "loss": 0.0594, "step": 20892 }, { "epoch": 0.970864312267658, "grad_norm": 0.39739671792504416, "learning_rate": 2.4127102728710995e-08, "loss": 0.0461, "step": 20893 }, { "epoch": 0.9709107806691449, "grad_norm": 0.46808487315720093, "learning_rate": 2.405332239095337e-08, "loss": 0.0543, "step": 20894 }, { "epoch": 0.970957249070632, "grad_norm": 0.40168212122531055, "learning_rate": 2.3979654763625625e-08, "loss": 0.0614, "step": 20895 }, { "epoch": 0.971003717472119, "grad_norm": 0.4302920914978188, "learning_rate": 2.3906099848395315e-08, "loss": 0.0424, "step": 20896 }, { "epoch": 0.9710501858736059, "grad_norm": 0.31952968728871606, "learning_rate": 2.3832657646929436e-08, "loss": 0.0497, "step": 20897 }, { "epoch": 0.971096654275093, "grad_norm": 0.5175622877332036, "learning_rate": 2.3759328160891106e-08, "loss": 0.0421, "step": 20898 }, { "epoch": 0.9711431226765799, "grad_norm": 0.5082953887516373, "learning_rate": 2.3686111391941213e-08, "loss": 0.0532, "step": 20899 }, { "epoch": 0.9711895910780669, "grad_norm": 0.5473633398882214, "learning_rate": 2.3613007341738437e-08, "loss": 0.0524, "step": 20900 }, { "epoch": 0.9712360594795539, "grad_norm": 0.4562750877570278, "learning_rate": 2.354001601193867e-08, "loss": 0.05, "step": 20901 }, { "epoch": 0.9712825278810409, "grad_norm": 0.5060424888605946, "learning_rate": 2.3467137404195038e-08, "loss": 0.0613, "step": 20902 }, { "epoch": 0.9713289962825279, "grad_norm": 0.43453194261835365, "learning_rate": 2.3394371520158443e-08, "loss": 0.0378, "step": 20903 }, { "epoch": 0.9713754646840149, "grad_norm": 0.7835494934687458, "learning_rate": 2.332171836147701e-08, "loss": 0.0494, "step": 20904 }, { "epoch": 0.9714219330855018, "grad_norm": 0.47274053023929197, "learning_rate": 2.324917792979664e-08, "loss": 0.0429, "step": 20905 }, { "epoch": 0.9714684014869889, "grad_norm": 0.46430586159395554, "learning_rate": 2.317675022676047e-08, "loss": 0.0512, "step": 20906 }, { "epoch": 0.9715148698884758, "grad_norm": 0.6178876667387737, "learning_rate": 2.3104435254008852e-08, "loss": 0.0436, "step": 20907 }, { "epoch": 0.9715613382899628, "grad_norm": 0.5594877679217696, "learning_rate": 2.3032233013179918e-08, "loss": 0.0608, "step": 20908 }, { "epoch": 0.9716078066914499, "grad_norm": 0.5358525754222969, "learning_rate": 2.2960143505909027e-08, "loss": 0.0481, "step": 20909 }, { "epoch": 0.9716542750929368, "grad_norm": 0.5654972849372059, "learning_rate": 2.288816673382932e-08, "loss": 0.0606, "step": 20910 }, { "epoch": 0.9717007434944238, "grad_norm": 0.5130698944313978, "learning_rate": 2.28163026985706e-08, "loss": 0.0693, "step": 20911 }, { "epoch": 0.9717472118959107, "grad_norm": 0.5079486629695246, "learning_rate": 2.2744551401761573e-08, "loss": 0.0669, "step": 20912 }, { "epoch": 0.9717936802973978, "grad_norm": 0.49142985408522794, "learning_rate": 2.2672912845025928e-08, "loss": 0.0514, "step": 20913 }, { "epoch": 0.9718401486988848, "grad_norm": 0.40200536709443735, "learning_rate": 2.260138702998793e-08, "loss": 0.0312, "step": 20914 }, { "epoch": 0.9718866171003717, "grad_norm": 0.6663342220436057, "learning_rate": 2.2529973958267394e-08, "loss": 0.0632, "step": 20915 }, { "epoch": 0.9719330855018588, "grad_norm": 0.3302707868467833, "learning_rate": 2.2458673631481354e-08, "loss": 0.0384, "step": 20916 }, { "epoch": 0.9719795539033457, "grad_norm": 0.2362792379796421, "learning_rate": 2.238748605124519e-08, "loss": 0.0236, "step": 20917 }, { "epoch": 0.9720260223048327, "grad_norm": 0.4806584099436155, "learning_rate": 2.231641121917094e-08, "loss": 0.0435, "step": 20918 }, { "epoch": 0.9720724907063197, "grad_norm": 0.41364578242431854, "learning_rate": 2.2245449136869545e-08, "loss": 0.0357, "step": 20919 }, { "epoch": 0.9721189591078067, "grad_norm": 0.5992669808586353, "learning_rate": 2.217459980594694e-08, "loss": 0.0466, "step": 20920 }, { "epoch": 0.9721654275092937, "grad_norm": 0.4555994292942271, "learning_rate": 2.2103863228009616e-08, "loss": 0.0613, "step": 20921 }, { "epoch": 0.9722118959107807, "grad_norm": 0.4485444618504643, "learning_rate": 2.2033239404658512e-08, "loss": 0.0521, "step": 20922 }, { "epoch": 0.9722583643122676, "grad_norm": 0.4745626552224529, "learning_rate": 2.1962728337493467e-08, "loss": 0.0465, "step": 20923 }, { "epoch": 0.9723048327137547, "grad_norm": 0.49200406529987467, "learning_rate": 2.1892330028112085e-08, "loss": 0.0632, "step": 20924 }, { "epoch": 0.9723513011152416, "grad_norm": 0.3819158321296803, "learning_rate": 2.1822044478109205e-08, "loss": 0.04, "step": 20925 }, { "epoch": 0.9723977695167286, "grad_norm": 0.6091166341288535, "learning_rate": 2.1751871689075775e-08, "loss": 0.0727, "step": 20926 }, { "epoch": 0.9724442379182157, "grad_norm": 0.5134991177142976, "learning_rate": 2.168181166260219e-08, "loss": 0.0654, "step": 20927 }, { "epoch": 0.9724907063197026, "grad_norm": 0.4834859065568656, "learning_rate": 2.161186440027552e-08, "loss": 0.0563, "step": 20928 }, { "epoch": 0.9725371747211896, "grad_norm": 0.4729622501036228, "learning_rate": 2.1542029903679485e-08, "loss": 0.0446, "step": 20929 }, { "epoch": 0.9725836431226765, "grad_norm": 0.7490175422894122, "learning_rate": 2.147230817439616e-08, "loss": 0.0522, "step": 20930 }, { "epoch": 0.9726301115241636, "grad_norm": 0.4070811933245576, "learning_rate": 2.140269921400484e-08, "loss": 0.0522, "step": 20931 }, { "epoch": 0.9726765799256506, "grad_norm": 0.9237901029380309, "learning_rate": 2.1333203024082594e-08, "loss": 0.0582, "step": 20932 }, { "epoch": 0.9727230483271375, "grad_norm": 0.5063135318619528, "learning_rate": 2.1263819606203163e-08, "loss": 0.0556, "step": 20933 }, { "epoch": 0.9727695167286246, "grad_norm": 0.374182953024212, "learning_rate": 2.1194548961938066e-08, "loss": 0.0433, "step": 20934 }, { "epoch": 0.9728159851301115, "grad_norm": 0.5067885074205487, "learning_rate": 2.1125391092856607e-08, "loss": 0.0375, "step": 20935 }, { "epoch": 0.9728624535315985, "grad_norm": 0.4118965054587774, "learning_rate": 2.1056346000525307e-08, "loss": 0.0644, "step": 20936 }, { "epoch": 0.9729089219330855, "grad_norm": 0.5611911960242948, "learning_rate": 2.0987413686507363e-08, "loss": 0.0391, "step": 20937 }, { "epoch": 0.9729553903345725, "grad_norm": 0.6231744654967727, "learning_rate": 2.091859415236541e-08, "loss": 0.0528, "step": 20938 }, { "epoch": 0.9730018587360595, "grad_norm": 0.3370584837872074, "learning_rate": 2.0849887399657655e-08, "loss": 0.0374, "step": 20939 }, { "epoch": 0.9730483271375465, "grad_norm": 0.4847019271777615, "learning_rate": 2.078129342994062e-08, "loss": 0.0396, "step": 20940 }, { "epoch": 0.9730947955390334, "grad_norm": 0.44991241710599955, "learning_rate": 2.071281224476751e-08, "loss": 0.0409, "step": 20941 }, { "epoch": 0.9731412639405205, "grad_norm": 0.6325630076753952, "learning_rate": 2.0644443845689866e-08, "loss": 0.0469, "step": 20942 }, { "epoch": 0.9731877323420074, "grad_norm": 0.5061479553285044, "learning_rate": 2.0576188234256443e-08, "loss": 0.0733, "step": 20943 }, { "epoch": 0.9732342007434944, "grad_norm": 0.4933163141107407, "learning_rate": 2.050804541201268e-08, "loss": 0.0475, "step": 20944 }, { "epoch": 0.9732806691449815, "grad_norm": 0.358011882822917, "learning_rate": 2.0440015380502888e-08, "loss": 0.051, "step": 20945 }, { "epoch": 0.9733271375464684, "grad_norm": 0.4285230968973945, "learning_rate": 2.0372098141267503e-08, "loss": 0.0509, "step": 20946 }, { "epoch": 0.9733736059479554, "grad_norm": 0.6082327643302002, "learning_rate": 2.0304293695844745e-08, "loss": 0.045, "step": 20947 }, { "epoch": 0.9734200743494423, "grad_norm": 0.5886586050613941, "learning_rate": 2.023660204577116e-08, "loss": 0.063, "step": 20948 }, { "epoch": 0.9734665427509294, "grad_norm": 0.44611697966393377, "learning_rate": 2.0169023192579407e-08, "loss": 0.059, "step": 20949 }, { "epoch": 0.9735130111524164, "grad_norm": 0.4673055362343897, "learning_rate": 2.0101557137801042e-08, "loss": 0.0523, "step": 20950 }, { "epoch": 0.9735594795539033, "grad_norm": 0.47221850701179885, "learning_rate": 2.003420388296262e-08, "loss": 0.0381, "step": 20951 }, { "epoch": 0.9736059479553903, "grad_norm": 0.3466975459224005, "learning_rate": 1.9966963429591812e-08, "loss": 0.0455, "step": 20952 }, { "epoch": 0.9736524163568773, "grad_norm": 0.6159871535084818, "learning_rate": 1.9899835779210175e-08, "loss": 0.0618, "step": 20953 }, { "epoch": 0.9736988847583643, "grad_norm": 0.44266885340717793, "learning_rate": 1.9832820933338714e-08, "loss": 0.0497, "step": 20954 }, { "epoch": 0.9737453531598513, "grad_norm": 0.5372767705466654, "learning_rate": 1.9765918893495663e-08, "loss": 0.051, "step": 20955 }, { "epoch": 0.9737918215613383, "grad_norm": 0.41556970758813366, "learning_rate": 1.9699129661195915e-08, "loss": 0.0457, "step": 20956 }, { "epoch": 0.9738382899628253, "grad_norm": 0.4556100046388512, "learning_rate": 1.963245323795271e-08, "loss": 0.0392, "step": 20957 }, { "epoch": 0.9738847583643123, "grad_norm": 0.6445561279750462, "learning_rate": 1.9565889625275945e-08, "loss": 0.0469, "step": 20958 }, { "epoch": 0.9739312267657992, "grad_norm": 0.5197465412411332, "learning_rate": 1.9499438824673867e-08, "loss": 0.0474, "step": 20959 }, { "epoch": 0.9739776951672863, "grad_norm": 0.6996509461505228, "learning_rate": 1.9433100837651374e-08, "loss": 0.0596, "step": 20960 }, { "epoch": 0.9740241635687732, "grad_norm": 0.49802168223719406, "learning_rate": 1.9366875665710607e-08, "loss": 0.0435, "step": 20961 }, { "epoch": 0.9740706319702602, "grad_norm": 0.6493280451100981, "learning_rate": 1.9300763310352577e-08, "loss": 0.0593, "step": 20962 }, { "epoch": 0.9741171003717473, "grad_norm": 0.7112538622213078, "learning_rate": 1.9234763773074428e-08, "loss": 0.0533, "step": 20963 }, { "epoch": 0.9741635687732342, "grad_norm": 0.3572365928102734, "learning_rate": 1.9168877055370516e-08, "loss": 0.0515, "step": 20964 }, { "epoch": 0.9742100371747212, "grad_norm": 0.4939931231070474, "learning_rate": 1.9103103158734092e-08, "loss": 0.0549, "step": 20965 }, { "epoch": 0.9742565055762081, "grad_norm": 0.45917788500103673, "learning_rate": 1.903744208465508e-08, "loss": 0.046, "step": 20966 }, { "epoch": 0.9743029739776952, "grad_norm": 0.33912242318546143, "learning_rate": 1.8971893834620058e-08, "loss": 0.038, "step": 20967 }, { "epoch": 0.9743494423791822, "grad_norm": 0.6383430783240971, "learning_rate": 1.8906458410113403e-08, "loss": 0.0582, "step": 20968 }, { "epoch": 0.9743959107806691, "grad_norm": 0.7852767838564692, "learning_rate": 1.8841135812618928e-08, "loss": 0.0587, "step": 20969 }, { "epoch": 0.9744423791821561, "grad_norm": 0.5182590455028029, "learning_rate": 1.877592604361489e-08, "loss": 0.0605, "step": 20970 }, { "epoch": 0.9744888475836431, "grad_norm": 0.4592310339232495, "learning_rate": 1.8710829104579e-08, "loss": 0.044, "step": 20971 }, { "epoch": 0.9745353159851301, "grad_norm": 0.439500948618399, "learning_rate": 1.8645844996985073e-08, "loss": 0.0387, "step": 20972 }, { "epoch": 0.9745817843866171, "grad_norm": 0.495069141998658, "learning_rate": 1.8580973722305828e-08, "loss": 0.0464, "step": 20973 }, { "epoch": 0.974628252788104, "grad_norm": 0.4476898628912753, "learning_rate": 1.851621528201064e-08, "loss": 0.0355, "step": 20974 }, { "epoch": 0.9746747211895911, "grad_norm": 0.3835345399800252, "learning_rate": 1.845156967756556e-08, "loss": 0.0301, "step": 20975 }, { "epoch": 0.9747211895910781, "grad_norm": 0.4451293231024047, "learning_rate": 1.838703691043553e-08, "loss": 0.0532, "step": 20976 }, { "epoch": 0.974767657992565, "grad_norm": 0.32966341142308664, "learning_rate": 1.8322616982082707e-08, "loss": 0.0341, "step": 20977 }, { "epoch": 0.9748141263940521, "grad_norm": 0.3885421120649628, "learning_rate": 1.8258309893965375e-08, "loss": 0.0547, "step": 20978 }, { "epoch": 0.974860594795539, "grad_norm": 0.5512207518320321, "learning_rate": 1.8194115647540145e-08, "loss": 0.0474, "step": 20979 }, { "epoch": 0.974907063197026, "grad_norm": 0.8568048575047674, "learning_rate": 1.8130034244261963e-08, "loss": 0.0563, "step": 20980 }, { "epoch": 0.9749535315985131, "grad_norm": 0.40242586311728795, "learning_rate": 1.8066065685581892e-08, "loss": 0.0275, "step": 20981 }, { "epoch": 0.975, "grad_norm": 0.44897457348985437, "learning_rate": 1.8002209972948215e-08, "loss": 0.0338, "step": 20982 }, { "epoch": 0.975046468401487, "grad_norm": 0.507647860502095, "learning_rate": 1.7938467107808664e-08, "loss": 0.0571, "step": 20983 }, { "epoch": 0.9750929368029739, "grad_norm": 0.645849883297124, "learning_rate": 1.787483709160598e-08, "loss": 0.0514, "step": 20984 }, { "epoch": 0.975139405204461, "grad_norm": 0.538282255226924, "learning_rate": 1.7811319925781778e-08, "loss": 0.0507, "step": 20985 }, { "epoch": 0.975185873605948, "grad_norm": 0.5104016884547597, "learning_rate": 1.7747915611774913e-08, "loss": 0.0517, "step": 20986 }, { "epoch": 0.9752323420074349, "grad_norm": 0.587995020341447, "learning_rate": 1.7684624151021457e-08, "loss": 0.0672, "step": 20987 }, { "epoch": 0.975278810408922, "grad_norm": 0.7409609131881169, "learning_rate": 1.7621445544955262e-08, "loss": 0.0541, "step": 20988 }, { "epoch": 0.9753252788104089, "grad_norm": 0.5033982849651072, "learning_rate": 1.7558379795006853e-08, "loss": 0.051, "step": 20989 }, { "epoch": 0.9753717472118959, "grad_norm": 0.523728073236341, "learning_rate": 1.7495426902605084e-08, "loss": 0.0448, "step": 20990 }, { "epoch": 0.9754182156133829, "grad_norm": 0.5363668775291445, "learning_rate": 1.7432586869176038e-08, "loss": 0.0329, "step": 20991 }, { "epoch": 0.9754646840148699, "grad_norm": 0.5952193587770317, "learning_rate": 1.736985969614302e-08, "loss": 0.0632, "step": 20992 }, { "epoch": 0.9755111524163569, "grad_norm": 0.45816086327329913, "learning_rate": 1.7307245384926562e-08, "loss": 0.0452, "step": 20993 }, { "epoch": 0.9755576208178439, "grad_norm": 0.5653199925404925, "learning_rate": 1.724474393694553e-08, "loss": 0.0577, "step": 20994 }, { "epoch": 0.9756040892193308, "grad_norm": 0.527270703370926, "learning_rate": 1.71823553536149e-08, "loss": 0.0364, "step": 20995 }, { "epoch": 0.9756505576208179, "grad_norm": 0.6211832030376042, "learning_rate": 1.7120079636348542e-08, "loss": 0.0585, "step": 20996 }, { "epoch": 0.9756970260223048, "grad_norm": 0.4994626859666145, "learning_rate": 1.7057916786556994e-08, "loss": 0.0599, "step": 20997 }, { "epoch": 0.9757434944237918, "grad_norm": 0.2694267486705168, "learning_rate": 1.6995866805647466e-08, "loss": 0.0425, "step": 20998 }, { "epoch": 0.9757899628252789, "grad_norm": 0.44812627869083826, "learning_rate": 1.6933929695026607e-08, "loss": 0.0435, "step": 20999 }, { "epoch": 0.9758364312267658, "grad_norm": 0.44812627869083826, "learning_rate": 1.6933929695026607e-08, "loss": 0.067, "step": 21000 }, { "epoch": 0.9758828996282528, "grad_norm": 0.4062792573229617, "learning_rate": 1.6872105456096634e-08, "loss": 0.0571, "step": 21001 }, { "epoch": 0.9759293680297397, "grad_norm": 0.2762718120239235, "learning_rate": 1.6810394090258087e-08, "loss": 0.0345, "step": 21002 }, { "epoch": 0.9759758364312268, "grad_norm": 0.3489579299079282, "learning_rate": 1.674879559890874e-08, "loss": 0.035, "step": 21003 }, { "epoch": 0.9760223048327138, "grad_norm": 0.4390289365273609, "learning_rate": 1.6687309983444144e-08, "loss": 0.0554, "step": 21004 }, { "epoch": 0.9760687732342007, "grad_norm": 0.4287489510297815, "learning_rate": 1.662593724525652e-08, "loss": 0.0355, "step": 21005 }, { "epoch": 0.9761152416356877, "grad_norm": 0.7146564469111866, "learning_rate": 1.6564677385736415e-08, "loss": 0.0522, "step": 21006 }, { "epoch": 0.9761617100371747, "grad_norm": 0.4399145137162616, "learning_rate": 1.6503530406271617e-08, "loss": 0.0538, "step": 21007 }, { "epoch": 0.9762081784386617, "grad_norm": 0.9060347548723697, "learning_rate": 1.644249630824657e-08, "loss": 0.0803, "step": 21008 }, { "epoch": 0.9762546468401487, "grad_norm": 0.4930661879580747, "learning_rate": 1.638157509304461e-08, "loss": 0.0482, "step": 21009 }, { "epoch": 0.9763011152416357, "grad_norm": 0.32619061456073756, "learning_rate": 1.632076676204464e-08, "loss": 0.0453, "step": 21010 }, { "epoch": 0.9763475836431227, "grad_norm": 0.32336336279812106, "learning_rate": 1.6260071316624437e-08, "loss": 0.0353, "step": 21011 }, { "epoch": 0.9763940520446097, "grad_norm": 0.8435580484311901, "learning_rate": 1.619948875815902e-08, "loss": 0.0678, "step": 21012 }, { "epoch": 0.9764405204460966, "grad_norm": 0.35071625897905884, "learning_rate": 1.6139019088020626e-08, "loss": 0.0514, "step": 21013 }, { "epoch": 0.9764869888475837, "grad_norm": 0.5641142874483928, "learning_rate": 1.6078662307579262e-08, "loss": 0.0692, "step": 21014 }, { "epoch": 0.9765334572490706, "grad_norm": 0.4347735230019975, "learning_rate": 1.6018418418201066e-08, "loss": 0.0623, "step": 21015 }, { "epoch": 0.9765799256505576, "grad_norm": 0.3365697094549434, "learning_rate": 1.5958287421251605e-08, "loss": 0.0269, "step": 21016 }, { "epoch": 0.9766263940520447, "grad_norm": 0.5954032903693686, "learning_rate": 1.5898269318092018e-08, "loss": 0.0672, "step": 21017 }, { "epoch": 0.9766728624535316, "grad_norm": 0.6113591789087964, "learning_rate": 1.5838364110082883e-08, "loss": 0.0551, "step": 21018 }, { "epoch": 0.9767193308550186, "grad_norm": 0.5201263907797911, "learning_rate": 1.5778571798580334e-08, "loss": 0.0293, "step": 21019 }, { "epoch": 0.9767657992565055, "grad_norm": 0.6406926639944861, "learning_rate": 1.5718892384938843e-08, "loss": 0.0438, "step": 21020 }, { "epoch": 0.9768122676579926, "grad_norm": 0.500540456285436, "learning_rate": 1.5659325870510668e-08, "loss": 0.0572, "step": 21021 }, { "epoch": 0.9768587360594796, "grad_norm": 0.5050854051726361, "learning_rate": 1.5599872256644167e-08, "loss": 0.0487, "step": 21022 }, { "epoch": 0.9769052044609665, "grad_norm": 0.4386292969075879, "learning_rate": 1.5540531544686597e-08, "loss": 0.0547, "step": 21023 }, { "epoch": 0.9769516728624535, "grad_norm": 0.7296955559938356, "learning_rate": 1.5481303735982443e-08, "loss": 0.056, "step": 21024 }, { "epoch": 0.9769981412639405, "grad_norm": 0.46871303706500544, "learning_rate": 1.5422188831872852e-08, "loss": 0.0567, "step": 21025 }, { "epoch": 0.9770446096654275, "grad_norm": 0.6483941184558268, "learning_rate": 1.5363186833696198e-08, "loss": 0.0506, "step": 21026 }, { "epoch": 0.9770910780669145, "grad_norm": 0.3151067210394621, "learning_rate": 1.5304297742790298e-08, "loss": 0.0471, "step": 21027 }, { "epoch": 0.9771375464684015, "grad_norm": 0.7351544723586388, "learning_rate": 1.5245521560487974e-08, "loss": 0.0552, "step": 21028 }, { "epoch": 0.9771840148698885, "grad_norm": 0.6085420023213005, "learning_rate": 1.5186858288120944e-08, "loss": 0.0585, "step": 21029 }, { "epoch": 0.9772304832713755, "grad_norm": 0.5641854985710733, "learning_rate": 1.5128307927018136e-08, "loss": 0.0608, "step": 21030 }, { "epoch": 0.9772769516728624, "grad_norm": 0.4478220939458067, "learning_rate": 1.506987047850572e-08, "loss": 0.0558, "step": 21031 }, { "epoch": 0.9773234200743495, "grad_norm": 0.3807843697079736, "learning_rate": 1.501154594390708e-08, "loss": 0.0403, "step": 21032 }, { "epoch": 0.9773698884758364, "grad_norm": 0.40466752425692165, "learning_rate": 1.4953334324543932e-08, "loss": 0.0434, "step": 21033 }, { "epoch": 0.9774163568773234, "grad_norm": 0.6711395831485013, "learning_rate": 1.4895235621734117e-08, "loss": 0.051, "step": 21034 }, { "epoch": 0.9774628252788105, "grad_norm": 0.5316215572342027, "learning_rate": 1.4837249836793799e-08, "loss": 0.0583, "step": 21035 }, { "epoch": 0.9775092936802974, "grad_norm": 0.5226468415804654, "learning_rate": 1.477937697103693e-08, "loss": 0.0516, "step": 21036 }, { "epoch": 0.9775557620817844, "grad_norm": 0.4481419770968727, "learning_rate": 1.4721617025773571e-08, "loss": 0.0481, "step": 21037 }, { "epoch": 0.9776022304832713, "grad_norm": 0.5225142885375125, "learning_rate": 1.4663970002312678e-08, "loss": 0.0493, "step": 21038 }, { "epoch": 0.9776486988847584, "grad_norm": 0.46033570008631747, "learning_rate": 1.460643590195987e-08, "loss": 0.0456, "step": 21039 }, { "epoch": 0.9776951672862454, "grad_norm": 0.471468305786756, "learning_rate": 1.454901472601855e-08, "loss": 0.0489, "step": 21040 }, { "epoch": 0.9777416356877323, "grad_norm": 0.3162405690984564, "learning_rate": 1.4491706475788791e-08, "loss": 0.0357, "step": 21041 }, { "epoch": 0.9777881040892193, "grad_norm": 0.3802290975730854, "learning_rate": 1.4434511152568997e-08, "loss": 0.0546, "step": 21042 }, { "epoch": 0.9778345724907063, "grad_norm": 0.47962002713344465, "learning_rate": 1.4377428757655354e-08, "loss": 0.0473, "step": 21043 }, { "epoch": 0.9778810408921933, "grad_norm": 0.5816535538684646, "learning_rate": 1.4320459292339605e-08, "loss": 0.0564, "step": 21044 }, { "epoch": 0.9779275092936803, "grad_norm": 0.43275021498694355, "learning_rate": 1.4263602757912942e-08, "loss": 0.0665, "step": 21045 }, { "epoch": 0.9779739776951673, "grad_norm": 0.7618260480516621, "learning_rate": 1.4206859155663223e-08, "loss": 0.0422, "step": 21046 }, { "epoch": 0.9780204460966543, "grad_norm": 0.5321489316491246, "learning_rate": 1.4150228486875528e-08, "loss": 0.0505, "step": 21047 }, { "epoch": 0.9780669144981413, "grad_norm": 0.6089357336627158, "learning_rate": 1.4093710752832169e-08, "loss": 0.0453, "step": 21048 }, { "epoch": 0.9781133828996282, "grad_norm": 0.6675730338021746, "learning_rate": 1.4037305954814895e-08, "loss": 0.0548, "step": 21049 }, { "epoch": 0.9781598513011153, "grad_norm": 0.5935274392796154, "learning_rate": 1.3981014094099354e-08, "loss": 0.0372, "step": 21050 }, { "epoch": 0.9782063197026022, "grad_norm": 0.46129875816272287, "learning_rate": 1.3924835171961748e-08, "loss": 0.053, "step": 21051 }, { "epoch": 0.9782527881040892, "grad_norm": 0.4014572941109965, "learning_rate": 1.3868769189674391e-08, "loss": 0.0187, "step": 21052 }, { "epoch": 0.9782992565055763, "grad_norm": 0.3051198305971669, "learning_rate": 1.381281614850738e-08, "loss": 0.0309, "step": 21053 }, { "epoch": 0.9783457249070632, "grad_norm": 0.6319980183023746, "learning_rate": 1.3756976049727477e-08, "loss": 0.0612, "step": 21054 }, { "epoch": 0.9783921933085502, "grad_norm": 0.5218302458368795, "learning_rate": 1.3701248894600894e-08, "loss": 0.0471, "step": 21055 }, { "epoch": 0.9784386617100371, "grad_norm": 0.5991493330579646, "learning_rate": 1.3645634684388287e-08, "loss": 0.0518, "step": 21056 }, { "epoch": 0.9784851301115242, "grad_norm": 0.5395067105770911, "learning_rate": 1.3590133420350315e-08, "loss": 0.0409, "step": 21057 }, { "epoch": 0.9785315985130112, "grad_norm": 0.47547547966534515, "learning_rate": 1.3534745103743752e-08, "loss": 0.045, "step": 21058 }, { "epoch": 0.9785780669144981, "grad_norm": 0.5825095577866265, "learning_rate": 1.3479469735823702e-08, "loss": 0.0626, "step": 21059 }, { "epoch": 0.9786245353159851, "grad_norm": 0.5246711530518197, "learning_rate": 1.342430731784139e-08, "loss": 0.0399, "step": 21060 }, { "epoch": 0.9786710037174721, "grad_norm": 0.5893398254908717, "learning_rate": 1.336925785104748e-08, "loss": 0.0252, "step": 21061 }, { "epoch": 0.9787174721189591, "grad_norm": 0.37294919967069556, "learning_rate": 1.3314321336687641e-08, "loss": 0.0431, "step": 21062 }, { "epoch": 0.9787639405204461, "grad_norm": 0.6539574826602703, "learning_rate": 1.325949777600699e-08, "loss": 0.0554, "step": 21063 }, { "epoch": 0.978810408921933, "grad_norm": 0.5463704577746786, "learning_rate": 1.3204787170247312e-08, "loss": 0.0519, "step": 21064 }, { "epoch": 0.9788568773234201, "grad_norm": 0.7575800817381907, "learning_rate": 1.3150189520647615e-08, "loss": 0.0518, "step": 21065 }, { "epoch": 0.9789033457249071, "grad_norm": 0.5272868674675909, "learning_rate": 1.3095704828444689e-08, "loss": 0.0532, "step": 21066 }, { "epoch": 0.978949814126394, "grad_norm": 0.3273779642424293, "learning_rate": 1.3041333094872543e-08, "loss": 0.0295, "step": 21067 }, { "epoch": 0.9789962825278811, "grad_norm": 0.3738810900194511, "learning_rate": 1.2987074321162974e-08, "loss": 0.0511, "step": 21068 }, { "epoch": 0.979042750929368, "grad_norm": 0.40624876163831997, "learning_rate": 1.2932928508544995e-08, "loss": 0.0385, "step": 21069 }, { "epoch": 0.979089219330855, "grad_norm": 0.5521895206275373, "learning_rate": 1.287889565824485e-08, "loss": 0.0484, "step": 21070 }, { "epoch": 0.979135687732342, "grad_norm": 0.31432264806168375, "learning_rate": 1.2824975771486558e-08, "loss": 0.0367, "step": 21071 }, { "epoch": 0.979182156133829, "grad_norm": 0.3660451832564744, "learning_rate": 1.2771168849491367e-08, "loss": 0.0479, "step": 21072 }, { "epoch": 0.979228624535316, "grad_norm": 0.5000933711505653, "learning_rate": 1.27174748934783e-08, "loss": 0.0549, "step": 21073 }, { "epoch": 0.9792750929368029, "grad_norm": 0.4512282987750791, "learning_rate": 1.2663893904663604e-08, "loss": 0.0512, "step": 21074 }, { "epoch": 0.97932156133829, "grad_norm": 0.6059196728907051, "learning_rate": 1.2610425884260758e-08, "loss": 0.0584, "step": 21075 }, { "epoch": 0.979368029739777, "grad_norm": 0.3742600891588266, "learning_rate": 1.2557070833480455e-08, "loss": 0.0293, "step": 21076 }, { "epoch": 0.9794144981412639, "grad_norm": 0.46509902861765495, "learning_rate": 1.2503828753532288e-08, "loss": 0.0417, "step": 21077 }, { "epoch": 0.9794609665427509, "grad_norm": 0.38709737624330604, "learning_rate": 1.24506996456214e-08, "loss": 0.0342, "step": 21078 }, { "epoch": 0.9795074349442379, "grad_norm": 0.5892430131111365, "learning_rate": 1.2397683510951275e-08, "loss": 0.0516, "step": 21079 }, { "epoch": 0.9795539033457249, "grad_norm": 0.3817615275455642, "learning_rate": 1.2344780350723174e-08, "loss": 0.0401, "step": 21080 }, { "epoch": 0.9796003717472119, "grad_norm": 0.42441783080980916, "learning_rate": 1.2291990166135026e-08, "loss": 0.0385, "step": 21081 }, { "epoch": 0.9796468401486988, "grad_norm": 0.3862166813088833, "learning_rate": 1.2239312958383098e-08, "loss": 0.0598, "step": 21082 }, { "epoch": 0.9796933085501859, "grad_norm": 0.41079146087847457, "learning_rate": 1.2186748728660325e-08, "loss": 0.0496, "step": 21083 }, { "epoch": 0.9797397769516729, "grad_norm": 0.42907297388849974, "learning_rate": 1.2134297478156865e-08, "loss": 0.054, "step": 21084 }, { "epoch": 0.9797862453531598, "grad_norm": 0.3986343491251733, "learning_rate": 1.2081959208061211e-08, "loss": 0.0431, "step": 21085 }, { "epoch": 0.9798327137546469, "grad_norm": 0.358179827199868, "learning_rate": 1.2029733919559083e-08, "loss": 0.0532, "step": 21086 }, { "epoch": 0.9798791821561338, "grad_norm": 0.6036438216798876, "learning_rate": 1.1977621613833423e-08, "loss": 0.0505, "step": 21087 }, { "epoch": 0.9799256505576208, "grad_norm": 0.6275492538009219, "learning_rate": 1.1925622292063843e-08, "loss": 0.0568, "step": 21088 }, { "epoch": 0.9799721189591079, "grad_norm": 0.46891475036118807, "learning_rate": 1.1873735955428845e-08, "loss": 0.0411, "step": 21089 }, { "epoch": 0.9800185873605948, "grad_norm": 0.6730742182143418, "learning_rate": 1.1821962605104154e-08, "loss": 0.0593, "step": 21090 }, { "epoch": 0.9800650557620818, "grad_norm": 0.3717321064854196, "learning_rate": 1.1770302242261611e-08, "loss": 0.0439, "step": 21091 }, { "epoch": 0.9801115241635687, "grad_norm": 0.5821391219567055, "learning_rate": 1.1718754868071947e-08, "loss": 0.0496, "step": 21092 }, { "epoch": 0.9801579925650558, "grad_norm": 0.771958581200628, "learning_rate": 1.166732048370256e-08, "loss": 0.0585, "step": 21093 }, { "epoch": 0.9802044609665428, "grad_norm": 0.5930947088520473, "learning_rate": 1.1615999090318075e-08, "loss": 0.0476, "step": 21094 }, { "epoch": 0.9802509293680297, "grad_norm": 0.46100045043917454, "learning_rate": 1.1564790689082006e-08, "loss": 0.0511, "step": 21095 }, { "epoch": 0.9802973977695167, "grad_norm": 0.5235295801779514, "learning_rate": 1.1513695281153424e-08, "loss": 0.0662, "step": 21096 }, { "epoch": 0.9803438661710037, "grad_norm": 0.4080940442964432, "learning_rate": 1.146271286768974e-08, "loss": 0.0562, "step": 21097 }, { "epoch": 0.9803903345724907, "grad_norm": 0.37832999840333326, "learning_rate": 1.1411843449846138e-08, "loss": 0.0442, "step": 21098 }, { "epoch": 0.9804368029739777, "grad_norm": 0.6241077844293926, "learning_rate": 1.1361087028774476e-08, "loss": 0.045, "step": 21099 }, { "epoch": 0.9804832713754646, "grad_norm": 0.4866319414084008, "learning_rate": 1.1310443605624943e-08, "loss": 0.0508, "step": 21100 }, { "epoch": 0.9805297397769517, "grad_norm": 0.3766304934090862, "learning_rate": 1.1259913181543847e-08, "loss": 0.0409, "step": 21101 }, { "epoch": 0.9805762081784387, "grad_norm": 0.47660024904115506, "learning_rate": 1.1209495757676936e-08, "loss": 0.0631, "step": 21102 }, { "epoch": 0.9806226765799256, "grad_norm": 0.736891074733972, "learning_rate": 1.1159191335164966e-08, "loss": 0.0443, "step": 21103 }, { "epoch": 0.9806691449814127, "grad_norm": 0.47780565808878683, "learning_rate": 1.1108999915148132e-08, "loss": 0.0521, "step": 21104 }, { "epoch": 0.9807156133828996, "grad_norm": 0.6756804133135162, "learning_rate": 1.1058921498763308e-08, "loss": 0.0557, "step": 21105 }, { "epoch": 0.9807620817843866, "grad_norm": 0.32441002966532795, "learning_rate": 1.1008956087144585e-08, "loss": 0.035, "step": 21106 }, { "epoch": 0.9808085501858737, "grad_norm": 0.427310067193165, "learning_rate": 1.0959103681423832e-08, "loss": 0.0421, "step": 21107 }, { "epoch": 0.9808550185873606, "grad_norm": 0.4134255283365426, "learning_rate": 1.090936428273015e-08, "loss": 0.0466, "step": 21108 }, { "epoch": 0.9809014869888476, "grad_norm": 0.40572089112319404, "learning_rate": 1.0859737892190414e-08, "loss": 0.044, "step": 21109 }, { "epoch": 0.9809479553903345, "grad_norm": 0.6102562668015967, "learning_rate": 1.0810224510928724e-08, "loss": 0.0619, "step": 21110 }, { "epoch": 0.9809944237918216, "grad_norm": 0.7019895168439139, "learning_rate": 1.0760824140066407e-08, "loss": 0.0379, "step": 21111 }, { "epoch": 0.9810408921933086, "grad_norm": 0.6038469471571443, "learning_rate": 1.0711536780722564e-08, "loss": 0.0474, "step": 21112 }, { "epoch": 0.9810873605947955, "grad_norm": 0.3301673125389479, "learning_rate": 1.0662362434013529e-08, "loss": 0.0389, "step": 21113 }, { "epoch": 0.9811338289962825, "grad_norm": 0.720992397812073, "learning_rate": 1.0613301101052853e-08, "loss": 0.065, "step": 21114 }, { "epoch": 0.9811802973977695, "grad_norm": 0.6279333084774013, "learning_rate": 1.0564352782952426e-08, "loss": 0.0468, "step": 21115 }, { "epoch": 0.9812267657992565, "grad_norm": 0.4578427393597497, "learning_rate": 1.0515517480820803e-08, "loss": 0.0371, "step": 21116 }, { "epoch": 0.9812732342007435, "grad_norm": 0.5123618853156775, "learning_rate": 1.0466795195763768e-08, "loss": 0.0563, "step": 21117 }, { "epoch": 0.9813197026022304, "grad_norm": 0.3328640999848618, "learning_rate": 1.0418185928884883e-08, "loss": 0.0344, "step": 21118 }, { "epoch": 0.9813661710037175, "grad_norm": 0.34193035326680693, "learning_rate": 1.0369689681286044e-08, "loss": 0.0391, "step": 21119 }, { "epoch": 0.9814126394052045, "grad_norm": 0.5917306078458265, "learning_rate": 1.0321306454065261e-08, "loss": 0.0504, "step": 21120 }, { "epoch": 0.9814591078066914, "grad_norm": 0.7310773752832329, "learning_rate": 1.0273036248318325e-08, "loss": 0.082, "step": 21121 }, { "epoch": 0.9815055762081785, "grad_norm": 0.48458805120661236, "learning_rate": 1.0224879065138804e-08, "loss": 0.0366, "step": 21122 }, { "epoch": 0.9815520446096654, "grad_norm": 0.626275216149129, "learning_rate": 1.017683490561694e-08, "loss": 0.0691, "step": 21123 }, { "epoch": 0.9815985130111524, "grad_norm": 0.3317866192037875, "learning_rate": 1.0128903770841859e-08, "loss": 0.0392, "step": 21124 }, { "epoch": 0.9816449814126395, "grad_norm": 0.5630911647725335, "learning_rate": 1.0081085661898804e-08, "loss": 0.0539, "step": 21125 }, { "epoch": 0.9816914498141264, "grad_norm": 0.6295179863175934, "learning_rate": 1.0033380579871354e-08, "loss": 0.0645, "step": 21126 }, { "epoch": 0.9817379182156134, "grad_norm": 0.7409864401964615, "learning_rate": 9.985788525839201e-09, "loss": 0.0455, "step": 21127 }, { "epoch": 0.9817843866171003, "grad_norm": 0.46417004851136956, "learning_rate": 9.938309500880928e-09, "loss": 0.0561, "step": 21128 }, { "epoch": 0.9818308550185874, "grad_norm": 0.522176154499943, "learning_rate": 9.890943506071781e-09, "loss": 0.0476, "step": 21129 }, { "epoch": 0.9818773234200744, "grad_norm": 0.7145167591229411, "learning_rate": 9.843690542484796e-09, "loss": 0.0468, "step": 21130 }, { "epoch": 0.9819237918215613, "grad_norm": 0.49411716089313734, "learning_rate": 9.796550611190226e-09, "loss": 0.044, "step": 21131 }, { "epoch": 0.9819702602230483, "grad_norm": 0.3923711412749968, "learning_rate": 9.749523713256104e-09, "loss": 0.0494, "step": 21132 }, { "epoch": 0.9820167286245353, "grad_norm": 0.32476154666469775, "learning_rate": 9.702609849747136e-09, "loss": 0.0475, "step": 21133 }, { "epoch": 0.9820631970260223, "grad_norm": 0.5435621809623252, "learning_rate": 9.65580902172636e-09, "loss": 0.0518, "step": 21134 }, { "epoch": 0.9821096654275093, "grad_norm": 0.3999616736066063, "learning_rate": 9.609121230254037e-09, "loss": 0.0565, "step": 21135 }, { "epoch": 0.9821561338289962, "grad_norm": 0.3625962423222723, "learning_rate": 9.562546476387102e-09, "loss": 0.0375, "step": 21136 }, { "epoch": 0.9822026022304833, "grad_norm": 0.6095107842475249, "learning_rate": 9.516084761180822e-09, "loss": 0.0571, "step": 21137 }, { "epoch": 0.9822490706319703, "grad_norm": 0.3448551935787631, "learning_rate": 9.469736085687686e-09, "loss": 0.0356, "step": 21138 }, { "epoch": 0.9822955390334572, "grad_norm": 0.5555454505066815, "learning_rate": 9.423500450957413e-09, "loss": 0.0471, "step": 21139 }, { "epoch": 0.9823420074349443, "grad_norm": 0.6555463115867594, "learning_rate": 9.377377858037495e-09, "loss": 0.0391, "step": 21140 }, { "epoch": 0.9823884758364312, "grad_norm": 0.4385019196218419, "learning_rate": 9.3313683079721e-09, "loss": 0.0368, "step": 21141 }, { "epoch": 0.9824349442379182, "grad_norm": 0.40560943202657507, "learning_rate": 9.28547180180428e-09, "loss": 0.0372, "step": 21142 }, { "epoch": 0.9824814126394052, "grad_norm": 0.4555517991596808, "learning_rate": 9.23968834057265e-09, "loss": 0.068, "step": 21143 }, { "epoch": 0.9825278810408922, "grad_norm": 0.48439926965274804, "learning_rate": 9.194017925315269e-09, "loss": 0.044, "step": 21144 }, { "epoch": 0.9825743494423792, "grad_norm": 0.4300147893432051, "learning_rate": 9.148460557065753e-09, "loss": 0.0628, "step": 21145 }, { "epoch": 0.9826208178438661, "grad_norm": 0.43969621647649365, "learning_rate": 9.10301623685661e-09, "loss": 0.0238, "step": 21146 }, { "epoch": 0.9826672862453532, "grad_norm": 0.42677854603760484, "learning_rate": 9.057684965717018e-09, "loss": 0.0331, "step": 21147 }, { "epoch": 0.9827137546468402, "grad_norm": 0.409880449045529, "learning_rate": 9.012466744673376e-09, "loss": 0.0519, "step": 21148 }, { "epoch": 0.9827602230483271, "grad_norm": 0.6378170918160057, "learning_rate": 8.967361574750976e-09, "loss": 0.0466, "step": 21149 }, { "epoch": 0.9828066914498141, "grad_norm": 0.5495365229361219, "learning_rate": 8.922369456970115e-09, "loss": 0.0617, "step": 21150 }, { "epoch": 0.9828531598513011, "grad_norm": 0.5883345291451234, "learning_rate": 8.877490392351084e-09, "loss": 0.0569, "step": 21151 }, { "epoch": 0.9828996282527881, "grad_norm": 0.3976956028662417, "learning_rate": 8.832724381909186e-09, "loss": 0.0441, "step": 21152 }, { "epoch": 0.9829460966542751, "grad_norm": 0.46820317161594605, "learning_rate": 8.788071426659717e-09, "loss": 0.0382, "step": 21153 }, { "epoch": 0.982992565055762, "grad_norm": 0.5022192481967883, "learning_rate": 8.743531527613535e-09, "loss": 0.0325, "step": 21154 }, { "epoch": 0.9830390334572491, "grad_norm": 0.5995155075392239, "learning_rate": 8.699104685779835e-09, "loss": 0.0533, "step": 21155 }, { "epoch": 0.9830855018587361, "grad_norm": 0.535825578148785, "learning_rate": 8.654790902164478e-09, "loss": 0.0487, "step": 21156 }, { "epoch": 0.983131970260223, "grad_norm": 0.6188063243907345, "learning_rate": 8.610590177771106e-09, "loss": 0.0424, "step": 21157 }, { "epoch": 0.9831784386617101, "grad_norm": 0.40175694301762255, "learning_rate": 8.566502513601693e-09, "loss": 0.038, "step": 21158 }, { "epoch": 0.983224907063197, "grad_norm": 0.4027669179904187, "learning_rate": 8.522527910654332e-09, "loss": 0.0437, "step": 21159 }, { "epoch": 0.983271375464684, "grad_norm": 0.37245707781450355, "learning_rate": 8.478666369924893e-09, "loss": 0.0457, "step": 21160 }, { "epoch": 0.983317843866171, "grad_norm": 0.8502683476146291, "learning_rate": 8.434917892407024e-09, "loss": 0.0541, "step": 21161 }, { "epoch": 0.983364312267658, "grad_norm": 0.3419487865784326, "learning_rate": 8.391282479092156e-09, "loss": 0.059, "step": 21162 }, { "epoch": 0.983410780669145, "grad_norm": 0.6531628593546527, "learning_rate": 8.34776013096783e-09, "loss": 0.0464, "step": 21163 }, { "epoch": 0.9834572490706319, "grad_norm": 0.6331825618736874, "learning_rate": 8.304350849021036e-09, "loss": 0.0431, "step": 21164 }, { "epoch": 0.983503717472119, "grad_norm": 0.46336602421355333, "learning_rate": 8.261054634233767e-09, "loss": 0.0487, "step": 21165 }, { "epoch": 0.983550185873606, "grad_norm": 0.7119598830360224, "learning_rate": 8.217871487587458e-09, "loss": 0.07, "step": 21166 }, { "epoch": 0.9835966542750929, "grad_norm": 0.4514164763990355, "learning_rate": 8.174801410060219e-09, "loss": 0.0524, "step": 21167 }, { "epoch": 0.9836431226765799, "grad_norm": 0.3598377032584458, "learning_rate": 8.131844402627931e-09, "loss": 0.0467, "step": 21168 }, { "epoch": 0.9836895910780669, "grad_norm": 0.5755945872419289, "learning_rate": 8.089000466262597e-09, "loss": 0.0613, "step": 21169 }, { "epoch": 0.9837360594795539, "grad_norm": 0.521123268020295, "learning_rate": 8.046269601935108e-09, "loss": 0.0449, "step": 21170 }, { "epoch": 0.9837825278810409, "grad_norm": 0.6463276028862667, "learning_rate": 8.003651810614132e-09, "loss": 0.0447, "step": 21171 }, { "epoch": 0.9838289962825278, "grad_norm": 0.42024737647588173, "learning_rate": 7.961147093264453e-09, "loss": 0.0589, "step": 21172 }, { "epoch": 0.9838754646840149, "grad_norm": 2.3072966712351515, "learning_rate": 7.918755450848636e-09, "loss": 0.0709, "step": 21173 }, { "epoch": 0.9839219330855019, "grad_norm": 0.4403797478951066, "learning_rate": 7.876476884327022e-09, "loss": 0.0456, "step": 21174 }, { "epoch": 0.9839684014869888, "grad_norm": 0.36499317587901986, "learning_rate": 7.834311394657178e-09, "loss": 0.0473, "step": 21175 }, { "epoch": 0.9840148698884759, "grad_norm": 0.40668254263086945, "learning_rate": 7.792258982794454e-09, "loss": 0.0448, "step": 21176 }, { "epoch": 0.9840613382899628, "grad_norm": 0.3960695869869476, "learning_rate": 7.750319649691419e-09, "loss": 0.052, "step": 21177 }, { "epoch": 0.9841078066914498, "grad_norm": 0.49218173899675327, "learning_rate": 7.708493396297868e-09, "loss": 0.052, "step": 21178 }, { "epoch": 0.9841542750929368, "grad_norm": 0.44804136904899894, "learning_rate": 7.666780223560821e-09, "loss": 0.0417, "step": 21179 }, { "epoch": 0.9842007434944238, "grad_norm": 0.5665484013494141, "learning_rate": 7.625180132426191e-09, "loss": 0.0496, "step": 21180 }, { "epoch": 0.9842472118959108, "grad_norm": 0.5622611458270489, "learning_rate": 7.58369312383489e-09, "loss": 0.0632, "step": 21181 }, { "epoch": 0.9842936802973977, "grad_norm": 0.5767563292957282, "learning_rate": 7.542319198727832e-09, "loss": 0.0517, "step": 21182 }, { "epoch": 0.9843401486988848, "grad_norm": 0.7163434104663284, "learning_rate": 7.501058358041491e-09, "loss": 0.0715, "step": 21183 }, { "epoch": 0.9843866171003718, "grad_norm": 0.5590533425816481, "learning_rate": 7.459910602711228e-09, "loss": 0.0644, "step": 21184 }, { "epoch": 0.9844330855018587, "grad_norm": 0.4711451454682705, "learning_rate": 7.4188759336679684e-09, "loss": 0.0486, "step": 21185 }, { "epoch": 0.9844795539033457, "grad_norm": 0.37041633362896503, "learning_rate": 7.377954351842076e-09, "loss": 0.0444, "step": 21186 }, { "epoch": 0.9845260223048327, "grad_norm": 0.44697017458868865, "learning_rate": 7.337145858160033e-09, "loss": 0.0292, "step": 21187 }, { "epoch": 0.9845724907063197, "grad_norm": 0.5676056440582841, "learning_rate": 7.2964504535461e-09, "loss": 0.0467, "step": 21188 }, { "epoch": 0.9846189591078067, "grad_norm": 0.4719960666121306, "learning_rate": 7.255868138922318e-09, "loss": 0.0466, "step": 21189 }, { "epoch": 0.9846654275092936, "grad_norm": 0.5321062030104636, "learning_rate": 7.215398915208505e-09, "loss": 0.0499, "step": 21190 }, { "epoch": 0.9847118959107807, "grad_norm": 0.4651055146250468, "learning_rate": 7.175042783320041e-09, "loss": 0.0356, "step": 21191 }, { "epoch": 0.9847583643122677, "grad_norm": 0.501695437778729, "learning_rate": 7.13479974417175e-09, "loss": 0.0499, "step": 21192 }, { "epoch": 0.9848048327137546, "grad_norm": 0.6309068317915976, "learning_rate": 7.0946697986756795e-09, "loss": 0.0731, "step": 21193 }, { "epoch": 0.9848513011152417, "grad_norm": 0.7378919880369244, "learning_rate": 7.054652947739992e-09, "loss": 0.0574, "step": 21194 }, { "epoch": 0.9848977695167286, "grad_norm": 0.6177257609331819, "learning_rate": 7.014749192271186e-09, "loss": 0.0579, "step": 21195 }, { "epoch": 0.9849442379182156, "grad_norm": 0.6664316268289275, "learning_rate": 6.974958533173537e-09, "loss": 0.0404, "step": 21196 }, { "epoch": 0.9849907063197026, "grad_norm": 0.2739281105829411, "learning_rate": 6.9352809713479905e-09, "loss": 0.0485, "step": 21197 }, { "epoch": 0.9850371747211896, "grad_norm": 0.6150172188409193, "learning_rate": 6.8957165076938285e-09, "loss": 0.064, "step": 21198 }, { "epoch": 0.9850836431226766, "grad_norm": 0.6514599766139564, "learning_rate": 6.856265143106444e-09, "loss": 0.0454, "step": 21199 }, { "epoch": 0.9851301115241635, "grad_norm": 0.3713332828472559, "learning_rate": 6.8169268784795685e-09, "loss": 0.0368, "step": 21200 }, { "epoch": 0.9851765799256506, "grad_norm": 0.5030608001168609, "learning_rate": 6.777701714705265e-09, "loss": 0.0478, "step": 21201 }, { "epoch": 0.9852230483271376, "grad_norm": 0.5988094457043267, "learning_rate": 6.738589652670602e-09, "loss": 0.0471, "step": 21202 }, { "epoch": 0.9852695167286245, "grad_norm": 0.5660245135024379, "learning_rate": 6.699590693262648e-09, "loss": 0.0471, "step": 21203 }, { "epoch": 0.9853159851301115, "grad_norm": 0.4106256946367721, "learning_rate": 6.660704837364029e-09, "loss": 0.0475, "step": 21204 }, { "epoch": 0.9853624535315985, "grad_norm": 0.7834040145113329, "learning_rate": 6.621932085856264e-09, "loss": 0.0613, "step": 21205 }, { "epoch": 0.9854089219330855, "grad_norm": 0.6262065792299423, "learning_rate": 6.583272439616983e-09, "loss": 0.0556, "step": 21206 }, { "epoch": 0.9854553903345725, "grad_norm": 0.5386789674793798, "learning_rate": 6.544725899521598e-09, "loss": 0.0497, "step": 21207 }, { "epoch": 0.9855018587360594, "grad_norm": 0.46044600868077445, "learning_rate": 6.506292466444408e-09, "loss": 0.0379, "step": 21208 }, { "epoch": 0.9855483271375465, "grad_norm": 0.45539478625120544, "learning_rate": 6.467972141254719e-09, "loss": 0.0568, "step": 21209 }, { "epoch": 0.9855947955390335, "grad_norm": 0.3254896613903166, "learning_rate": 6.4297649248212804e-09, "loss": 0.0498, "step": 21210 }, { "epoch": 0.9856412639405204, "grad_norm": 0.5347006781469721, "learning_rate": 6.391670818008955e-09, "loss": 0.0602, "step": 21211 }, { "epoch": 0.9856877323420075, "grad_norm": 0.5229969184560852, "learning_rate": 6.353689821681497e-09, "loss": 0.0517, "step": 21212 }, { "epoch": 0.9857342007434944, "grad_norm": 0.6444199662273027, "learning_rate": 6.315821936698774e-09, "loss": 0.0564, "step": 21213 }, { "epoch": 0.9857806691449814, "grad_norm": 0.4521853671789507, "learning_rate": 6.278067163917878e-09, "loss": 0.0529, "step": 21214 }, { "epoch": 0.9858271375464684, "grad_norm": 0.6119909157942276, "learning_rate": 6.240425504194792e-09, "loss": 0.0561, "step": 21215 }, { "epoch": 0.9858736059479554, "grad_norm": 0.5573733500855507, "learning_rate": 6.202896958382165e-09, "loss": 0.053, "step": 21216 }, { "epoch": 0.9859200743494424, "grad_norm": 0.5392216978414717, "learning_rate": 6.16548152732932e-09, "loss": 0.0563, "step": 21217 }, { "epoch": 0.9859665427509293, "grad_norm": 0.5349411771686844, "learning_rate": 6.1281792118844666e-09, "loss": 0.0478, "step": 21218 }, { "epoch": 0.9860130111524164, "grad_norm": 0.24965792131853085, "learning_rate": 6.090990012892484e-09, "loss": 0.0204, "step": 21219 }, { "epoch": 0.9860594795539034, "grad_norm": 0.579742011965338, "learning_rate": 6.053913931195476e-09, "loss": 0.0442, "step": 21220 }, { "epoch": 0.9861059479553903, "grad_norm": 0.5695036612690652, "learning_rate": 6.016950967633328e-09, "loss": 0.052, "step": 21221 }, { "epoch": 0.9861524163568773, "grad_norm": 0.5165365872953739, "learning_rate": 5.980101123043147e-09, "loss": 0.0539, "step": 21222 }, { "epoch": 0.9861988847583643, "grad_norm": 0.4614612318748222, "learning_rate": 5.943364398259821e-09, "loss": 0.0686, "step": 21223 }, { "epoch": 0.9862453531598513, "grad_norm": 0.5915520532634284, "learning_rate": 5.906740794115462e-09, "loss": 0.0485, "step": 21224 }, { "epoch": 0.9862918215613383, "grad_norm": 0.5617520845951702, "learning_rate": 5.870230311439406e-09, "loss": 0.043, "step": 21225 }, { "epoch": 0.9863382899628252, "grad_norm": 0.5438357891485298, "learning_rate": 5.833832951059326e-09, "loss": 0.0595, "step": 21226 }, { "epoch": 0.9863847583643123, "grad_norm": 0.7559821732710353, "learning_rate": 5.797548713799006e-09, "loss": 0.056, "step": 21227 }, { "epoch": 0.9864312267657993, "grad_norm": 0.6038534893989, "learning_rate": 5.761377600480567e-09, "loss": 0.0453, "step": 21228 }, { "epoch": 0.9864776951672862, "grad_norm": 0.48389404447929446, "learning_rate": 5.725319611923352e-09, "loss": 0.0552, "step": 21229 }, { "epoch": 0.9865241635687733, "grad_norm": 0.4480019092394804, "learning_rate": 5.689374748943932e-09, "loss": 0.0437, "step": 21230 }, { "epoch": 0.9865706319702602, "grad_norm": 0.44974067089885106, "learning_rate": 5.653543012356099e-09, "loss": 0.0492, "step": 21231 }, { "epoch": 0.9866171003717472, "grad_norm": 0.6853385493704237, "learning_rate": 5.6178244029725386e-09, "loss": 0.0811, "step": 21232 }, { "epoch": 0.9866635687732342, "grad_norm": 0.6820566795551337, "learning_rate": 5.58221892160149e-09, "loss": 0.073, "step": 21233 }, { "epoch": 0.9867100371747212, "grad_norm": 0.3959739558274262, "learning_rate": 5.5467265690500875e-09, "loss": 0.0415, "step": 21234 }, { "epoch": 0.9867565055762082, "grad_norm": 0.44727628119160995, "learning_rate": 5.511347346121576e-09, "loss": 0.0483, "step": 21235 }, { "epoch": 0.9868029739776951, "grad_norm": 0.43276178242748825, "learning_rate": 5.476081253617538e-09, "loss": 0.0432, "step": 21236 }, { "epoch": 0.9868494423791822, "grad_norm": 0.4302753809426774, "learning_rate": 5.440928292337333e-09, "loss": 0.0549, "step": 21237 }, { "epoch": 0.9868959107806692, "grad_norm": 0.6737469377952574, "learning_rate": 5.4058884630764365e-09, "loss": 0.0594, "step": 21238 }, { "epoch": 0.9869423791821561, "grad_norm": 0.6541368795393501, "learning_rate": 5.370961766629212e-09, "loss": 0.0705, "step": 21239 }, { "epoch": 0.9869888475836431, "grad_norm": 0.4835203314672047, "learning_rate": 5.33614820378614e-09, "loss": 0.0343, "step": 21240 }, { "epoch": 0.9870353159851301, "grad_norm": 0.4113080814047607, "learning_rate": 5.301447775336033e-09, "loss": 0.05, "step": 21241 }, { "epoch": 0.9870817843866171, "grad_norm": 0.6939525529110386, "learning_rate": 5.266860482064928e-09, "loss": 0.0464, "step": 21242 }, { "epoch": 0.9871282527881041, "grad_norm": 0.7089054025925758, "learning_rate": 5.232386324756644e-09, "loss": 0.0528, "step": 21243 }, { "epoch": 0.987174721189591, "grad_norm": 0.5502283455857174, "learning_rate": 5.198025304191668e-09, "loss": 0.0554, "step": 21244 }, { "epoch": 0.9872211895910781, "grad_norm": 0.43507514683959897, "learning_rate": 5.163777421147709e-09, "loss": 0.0484, "step": 21245 }, { "epoch": 0.9872676579925651, "grad_norm": 0.6193978273183378, "learning_rate": 5.129642676401925e-09, "loss": 0.0872, "step": 21246 }, { "epoch": 0.987314126394052, "grad_norm": 0.7254788485261973, "learning_rate": 5.095621070726475e-09, "loss": 0.0604, "step": 21247 }, { "epoch": 0.9873605947955391, "grad_norm": 0.5733612187285764, "learning_rate": 5.0617126048924105e-09, "loss": 0.0388, "step": 21248 }, { "epoch": 0.987407063197026, "grad_norm": 0.5444047875120757, "learning_rate": 5.0279172796674494e-09, "loss": 0.0437, "step": 21249 }, { "epoch": 0.987453531598513, "grad_norm": 0.5769258384940917, "learning_rate": 4.99423509581709e-09, "loss": 0.0468, "step": 21250 }, { "epoch": 0.9875, "grad_norm": 0.7034315195522783, "learning_rate": 4.9606660541046124e-09, "loss": 0.0541, "step": 21251 }, { "epoch": 0.987546468401487, "grad_norm": 0.5299588768841779, "learning_rate": 4.927210155290518e-09, "loss": 0.0479, "step": 21252 }, { "epoch": 0.987592936802974, "grad_norm": 0.8595780044756347, "learning_rate": 4.89386740013198e-09, "loss": 0.0683, "step": 21253 }, { "epoch": 0.9876394052044609, "grad_norm": 0.6147125206084316, "learning_rate": 4.860637789384503e-09, "loss": 0.0581, "step": 21254 }, { "epoch": 0.987685873605948, "grad_norm": 0.38069447329257305, "learning_rate": 4.827521323801376e-09, "loss": 0.0423, "step": 21255 }, { "epoch": 0.987732342007435, "grad_norm": 0.5967339323829706, "learning_rate": 4.794518004131998e-09, "loss": 0.0534, "step": 21256 }, { "epoch": 0.9877788104089219, "grad_norm": 0.5504637220307526, "learning_rate": 4.761627831124105e-09, "loss": 0.0659, "step": 21257 }, { "epoch": 0.9878252788104089, "grad_norm": 0.6399650983860544, "learning_rate": 4.728850805522656e-09, "loss": 0.0646, "step": 21258 }, { "epoch": 0.9878717472118959, "grad_norm": 0.7276981596638991, "learning_rate": 4.696186928069834e-09, "loss": 0.0676, "step": 21259 }, { "epoch": 0.9879182156133829, "grad_norm": 0.7618324808580456, "learning_rate": 4.663636199506161e-09, "loss": 0.0508, "step": 21260 }, { "epoch": 0.9879646840148699, "grad_norm": 0.5598768181364971, "learning_rate": 4.631198620568822e-09, "loss": 0.0412, "step": 21261 }, { "epoch": 0.9880111524163568, "grad_norm": 0.5471716935754284, "learning_rate": 4.598874191992231e-09, "loss": 0.0589, "step": 21262 }, { "epoch": 0.9880576208178439, "grad_norm": 0.9266901818577009, "learning_rate": 4.56666291450858e-09, "loss": 0.083, "step": 21263 }, { "epoch": 0.9881040892193309, "grad_norm": 0.3896390789018858, "learning_rate": 4.534564788847839e-09, "loss": 0.0412, "step": 21264 }, { "epoch": 0.9881505576208178, "grad_norm": 0.5043921217305855, "learning_rate": 4.502579815736652e-09, "loss": 0.0386, "step": 21265 }, { "epoch": 0.9881970260223049, "grad_norm": 0.555241970605041, "learning_rate": 4.470707995899992e-09, "loss": 0.0387, "step": 21266 }, { "epoch": 0.9882434944237918, "grad_norm": 0.3573267943588484, "learning_rate": 4.438949330059506e-09, "loss": 0.0369, "step": 21267 }, { "epoch": 0.9882899628252788, "grad_norm": 0.3213526004037683, "learning_rate": 4.407303818934061e-09, "loss": 0.0451, "step": 21268 }, { "epoch": 0.9883364312267658, "grad_norm": 0.47873519464596537, "learning_rate": 4.3757714632408634e-09, "loss": 0.0609, "step": 21269 }, { "epoch": 0.9883828996282528, "grad_norm": 0.619746952854045, "learning_rate": 4.344352263694895e-09, "loss": 0.0472, "step": 21270 }, { "epoch": 0.9884293680297398, "grad_norm": 0.4811095890908126, "learning_rate": 4.3130462210067e-09, "loss": 0.0485, "step": 21271 }, { "epoch": 0.9884758364312267, "grad_norm": 0.4760863164387637, "learning_rate": 4.281853335886266e-09, "loss": 0.0383, "step": 21272 }, { "epoch": 0.9885223048327137, "grad_norm": 0.7131174078013716, "learning_rate": 4.250773609039694e-09, "loss": 0.0485, "step": 21273 }, { "epoch": 0.9885687732342008, "grad_norm": 0.4515981792792344, "learning_rate": 4.219807041170865e-09, "loss": 0.0416, "step": 21274 }, { "epoch": 0.9886152416356877, "grad_norm": 0.8120829203173587, "learning_rate": 4.188953632980886e-09, "loss": 0.0605, "step": 21275 }, { "epoch": 0.9886617100371747, "grad_norm": 0.43812794209077616, "learning_rate": 4.158213385169752e-09, "loss": 0.0416, "step": 21276 }, { "epoch": 0.9887081784386617, "grad_norm": 0.46702607807341034, "learning_rate": 4.1275862984330175e-09, "loss": 0.0282, "step": 21277 }, { "epoch": 0.9887546468401487, "grad_norm": 0.5510741715995114, "learning_rate": 4.097072373464572e-09, "loss": 0.0375, "step": 21278 }, { "epoch": 0.9888011152416357, "grad_norm": 0.5457641352361239, "learning_rate": 4.066671610954975e-09, "loss": 0.0552, "step": 21279 }, { "epoch": 0.9888475836431226, "grad_norm": 0.40209558764594433, "learning_rate": 4.03638401159423e-09, "loss": 0.0413, "step": 21280 }, { "epoch": 0.9888940520446097, "grad_norm": 0.5023257648978406, "learning_rate": 4.006209576066788e-09, "loss": 0.0355, "step": 21281 }, { "epoch": 0.9889405204460967, "grad_norm": 0.4639357702307041, "learning_rate": 3.976148305057659e-09, "loss": 0.0678, "step": 21282 }, { "epoch": 0.9889869888475836, "grad_norm": 0.6659585216255958, "learning_rate": 3.946200199246297e-09, "loss": 0.0491, "step": 21283 }, { "epoch": 0.9890334572490707, "grad_norm": 0.38286543723965616, "learning_rate": 3.91636525931216e-09, "loss": 0.045, "step": 21284 }, { "epoch": 0.9890799256505576, "grad_norm": 0.3665052930317231, "learning_rate": 3.886643485930264e-09, "loss": 0.0416, "step": 21285 }, { "epoch": 0.9891263940520446, "grad_norm": 0.6035639839720587, "learning_rate": 3.857034879774513e-09, "loss": 0.0687, "step": 21286 }, { "epoch": 0.9891728624535316, "grad_norm": 0.4471766517310278, "learning_rate": 3.827539441514927e-09, "loss": 0.0562, "step": 21287 }, { "epoch": 0.9892193308550186, "grad_norm": 0.5719933639139145, "learning_rate": 3.798157171820416e-09, "loss": 0.0687, "step": 21288 }, { "epoch": 0.9892657992565056, "grad_norm": 0.5235427801697831, "learning_rate": 3.768888071356003e-09, "loss": 0.05, "step": 21289 }, { "epoch": 0.9893122676579925, "grad_norm": 0.650920177144801, "learning_rate": 3.739732140784491e-09, "loss": 0.0513, "step": 21290 }, { "epoch": 0.9893587360594795, "grad_norm": 0.42351218164895543, "learning_rate": 3.7106893807664633e-09, "loss": 0.0688, "step": 21291 }, { "epoch": 0.9894052044609666, "grad_norm": 0.4045988607533507, "learning_rate": 3.681759791960282e-09, "loss": 0.0514, "step": 21292 }, { "epoch": 0.9894516728624535, "grad_norm": 0.7020125930062017, "learning_rate": 3.6529433750198683e-09, "loss": 0.0524, "step": 21293 }, { "epoch": 0.9894981412639405, "grad_norm": 0.825192680622216, "learning_rate": 3.6242401305996988e-09, "loss": 0.0465, "step": 21294 }, { "epoch": 0.9895446096654275, "grad_norm": 0.5233227471851765, "learning_rate": 3.5956500593486986e-09, "loss": 0.0526, "step": 21295 }, { "epoch": 0.9895910780669145, "grad_norm": 0.4230807443560036, "learning_rate": 3.5671731619152385e-09, "loss": 0.051, "step": 21296 }, { "epoch": 0.9896375464684015, "grad_norm": 0.42767379947487266, "learning_rate": 3.538809438943247e-09, "loss": 0.0517, "step": 21297 }, { "epoch": 0.9896840148698884, "grad_norm": 0.6979719430942141, "learning_rate": 3.5105588910766543e-09, "loss": 0.0565, "step": 21298 }, { "epoch": 0.9897304832713755, "grad_norm": 0.37132345493907554, "learning_rate": 3.482421518953838e-09, "loss": 0.0467, "step": 21299 }, { "epoch": 0.9897769516728625, "grad_norm": 0.3950862568565735, "learning_rate": 3.454397323213732e-09, "loss": 0.0533, "step": 21300 }, { "epoch": 0.9898234200743494, "grad_norm": 0.8846250697173696, "learning_rate": 3.426486304489718e-09, "loss": 0.0415, "step": 21301 }, { "epoch": 0.9898698884758365, "grad_norm": 0.6164281810384568, "learning_rate": 3.3986884634146232e-09, "loss": 0.0513, "step": 21302 }, { "epoch": 0.9899163568773234, "grad_norm": 0.41065957063144964, "learning_rate": 3.3710038006185e-09, "loss": 0.0563, "step": 21303 }, { "epoch": 0.9899628252788104, "grad_norm": 0.5506469657052776, "learning_rate": 3.343432316727513e-09, "loss": 0.041, "step": 21304 }, { "epoch": 0.9900092936802974, "grad_norm": 0.32675967023105146, "learning_rate": 3.3159740123667184e-09, "loss": 0.0362, "step": 21305 }, { "epoch": 0.9900557620817844, "grad_norm": 0.27844198969506484, "learning_rate": 3.288628888158396e-09, "loss": 0.0433, "step": 21306 }, { "epoch": 0.9901022304832714, "grad_norm": 0.7278105285233385, "learning_rate": 3.2613969447209406e-09, "loss": 0.0594, "step": 21307 }, { "epoch": 0.9901486988847583, "grad_norm": 0.41582605308488146, "learning_rate": 3.2342781826727453e-09, "loss": 0.0597, "step": 21308 }, { "epoch": 0.9901951672862453, "grad_norm": 0.6362826456528418, "learning_rate": 3.2072726026266544e-09, "loss": 0.0701, "step": 21309 }, { "epoch": 0.9902416356877324, "grad_norm": 0.36228286246667185, "learning_rate": 3.1803802051944e-09, "loss": 0.0392, "step": 21310 }, { "epoch": 0.9902881040892193, "grad_norm": 0.4585694206463931, "learning_rate": 3.1536009909860498e-09, "loss": 0.0497, "step": 21311 }, { "epoch": 0.9903345724907063, "grad_norm": 0.3698651607874529, "learning_rate": 3.126934960607786e-09, "loss": 0.0526, "step": 21312 }, { "epoch": 0.9903810408921933, "grad_norm": 0.711800349034463, "learning_rate": 3.100382114663569e-09, "loss": 0.0537, "step": 21313 }, { "epoch": 0.9904275092936803, "grad_norm": 0.6450704173668242, "learning_rate": 3.073942453754586e-09, "loss": 0.0554, "step": 21314 }, { "epoch": 0.9904739776951673, "grad_norm": 0.642486720654186, "learning_rate": 3.047615978480356e-09, "loss": 0.0549, "step": 21315 }, { "epoch": 0.9905204460966542, "grad_norm": 0.42413680398062864, "learning_rate": 3.0214026894365143e-09, "loss": 0.0414, "step": 21316 }, { "epoch": 0.9905669144981413, "grad_norm": 0.2997976457968695, "learning_rate": 2.9953025872170306e-09, "loss": 0.0168, "step": 21317 }, { "epoch": 0.9906133828996283, "grad_norm": 0.4779987737554055, "learning_rate": 2.969315672413653e-09, "loss": 0.0487, "step": 21318 }, { "epoch": 0.9906598513011152, "grad_norm": 0.5101120029306291, "learning_rate": 2.943441945614245e-09, "loss": 0.0655, "step": 21319 }, { "epoch": 0.9907063197026023, "grad_norm": 0.6831952647138649, "learning_rate": 2.9176814074050042e-09, "loss": 0.0432, "step": 21320 }, { "epoch": 0.9907527881040892, "grad_norm": 0.5380577632092345, "learning_rate": 2.892034058369908e-09, "loss": 0.0798, "step": 21321 }, { "epoch": 0.9907992565055762, "grad_norm": 0.4269323351854658, "learning_rate": 2.8664998990896033e-09, "loss": 0.0517, "step": 21322 }, { "epoch": 0.9908457249070632, "grad_norm": 0.5848111100879312, "learning_rate": 2.8410789301425156e-09, "loss": 0.0501, "step": 21323 }, { "epoch": 0.9908921933085502, "grad_norm": 0.3630019306645977, "learning_rate": 2.8157711521042962e-09, "loss": 0.0288, "step": 21324 }, { "epoch": 0.9909386617100372, "grad_norm": 0.4792075038241589, "learning_rate": 2.7905765655478202e-09, "loss": 0.0412, "step": 21325 }, { "epoch": 0.9909851301115241, "grad_norm": 0.4216252732925, "learning_rate": 2.7654951710448517e-09, "loss": 0.0315, "step": 21326 }, { "epoch": 0.9910315985130111, "grad_norm": 0.5641600694366339, "learning_rate": 2.7405269691627156e-09, "loss": 0.0536, "step": 21327 }, { "epoch": 0.9910780669144982, "grad_norm": 0.5919109439080054, "learning_rate": 2.715671960467625e-09, "loss": 0.0527, "step": 21328 }, { "epoch": 0.9911245353159851, "grad_norm": 0.38065752598256136, "learning_rate": 2.6909301455213534e-09, "loss": 0.0515, "step": 21329 }, { "epoch": 0.9911710037174721, "grad_norm": 0.592929810517329, "learning_rate": 2.6663015248856727e-09, "loss": 0.0518, "step": 21330 }, { "epoch": 0.991217472118959, "grad_norm": 0.5009656823583287, "learning_rate": 2.6417860991173605e-09, "loss": 0.0492, "step": 21331 }, { "epoch": 0.9912639405204461, "grad_norm": 0.4971947874208012, "learning_rate": 2.617383868772638e-09, "loss": 0.0392, "step": 21332 }, { "epoch": 0.9913104089219331, "grad_norm": 0.4889612679131501, "learning_rate": 2.593094834403842e-09, "loss": 0.0654, "step": 21333 }, { "epoch": 0.99135687732342, "grad_norm": 0.36267644301212815, "learning_rate": 2.568918996560532e-09, "loss": 0.0475, "step": 21334 }, { "epoch": 0.9914033457249071, "grad_norm": 0.4356893408318882, "learning_rate": 2.544856355791714e-09, "loss": 0.0229, "step": 21335 }, { "epoch": 0.9914498141263941, "grad_norm": 0.42660087831302684, "learning_rate": 2.5209069126408413e-09, "loss": 0.0361, "step": 21336 }, { "epoch": 0.991496282527881, "grad_norm": 0.41842831568713923, "learning_rate": 2.497070667651924e-09, "loss": 0.0507, "step": 21337 }, { "epoch": 0.991542750929368, "grad_norm": 0.586427016268973, "learning_rate": 2.47334762136342e-09, "loss": 0.0616, "step": 21338 }, { "epoch": 0.991589219330855, "grad_norm": 0.3007424661159943, "learning_rate": 2.4497377743137875e-09, "loss": 0.0279, "step": 21339 }, { "epoch": 0.991635687732342, "grad_norm": 0.5830910482732587, "learning_rate": 2.4262411270375986e-09, "loss": 0.06, "step": 21340 }, { "epoch": 0.991682156133829, "grad_norm": 0.5459870342512807, "learning_rate": 2.40285768006665e-09, "loss": 0.0529, "step": 21341 }, { "epoch": 0.991728624535316, "grad_norm": 0.6347897932017045, "learning_rate": 2.379587433930519e-09, "loss": 0.0442, "step": 21342 }, { "epoch": 0.991775092936803, "grad_norm": 0.49944308065298754, "learning_rate": 2.356430389157116e-09, "loss": 0.0601, "step": 21343 }, { "epoch": 0.9918215613382899, "grad_norm": 0.6939185990707704, "learning_rate": 2.3333865462704665e-09, "loss": 0.06, "step": 21344 }, { "epoch": 0.991868029739777, "grad_norm": 0.6223449547774778, "learning_rate": 2.3104559057923747e-09, "loss": 0.0292, "step": 21345 }, { "epoch": 0.991914498141264, "grad_norm": 0.4790091139938601, "learning_rate": 2.287638468242426e-09, "loss": 0.048, "step": 21346 }, { "epoch": 0.9919609665427509, "grad_norm": 0.6605465825927129, "learning_rate": 2.264934234137983e-09, "loss": 0.0596, "step": 21347 }, { "epoch": 0.9920074349442379, "grad_norm": 0.29688145222502316, "learning_rate": 2.24234320399197e-09, "loss": 0.037, "step": 21348 }, { "epoch": 0.9920539033457249, "grad_norm": 0.44992180277339633, "learning_rate": 2.2198653783173097e-09, "loss": 0.0398, "step": 21349 }, { "epoch": 0.9921003717472119, "grad_norm": 0.43694723774767336, "learning_rate": 2.197500757623039e-09, "loss": 0.0335, "step": 21350 }, { "epoch": 0.9921468401486989, "grad_norm": 0.5861045094436292, "learning_rate": 2.1752493424148647e-09, "loss": 0.0548, "step": 21351 }, { "epoch": 0.9921933085501858, "grad_norm": 0.517440463907901, "learning_rate": 2.1531111331979382e-09, "loss": 0.0563, "step": 21352 }, { "epoch": 0.9922397769516729, "grad_norm": 0.604293911471738, "learning_rate": 2.13108613047297e-09, "loss": 0.0414, "step": 21353 }, { "epoch": 0.9922862453531599, "grad_norm": 0.6559776834391914, "learning_rate": 2.109174334739006e-09, "loss": 0.0747, "step": 21354 }, { "epoch": 0.9923327137546468, "grad_norm": 0.3654921201147781, "learning_rate": 2.08737574649287e-09, "loss": 0.0412, "step": 21355 }, { "epoch": 0.9923791821561339, "grad_norm": 0.29817817077715786, "learning_rate": 2.0656903662275017e-09, "loss": 0.0542, "step": 21356 }, { "epoch": 0.9924256505576208, "grad_norm": 0.39480974269555014, "learning_rate": 2.0441181944341746e-09, "loss": 0.0463, "step": 21357 }, { "epoch": 0.9924721189591078, "grad_norm": 0.6204184344029047, "learning_rate": 2.0226592316024972e-09, "loss": 0.0584, "step": 21358 }, { "epoch": 0.9925185873605948, "grad_norm": 0.41413737879488566, "learning_rate": 2.001313478217637e-09, "loss": 0.0437, "step": 21359 }, { "epoch": 0.9925650557620818, "grad_norm": 0.5671059633203933, "learning_rate": 1.9800809347630957e-09, "loss": 0.0521, "step": 21360 }, { "epoch": 0.9926115241635688, "grad_norm": 0.6716486037148744, "learning_rate": 1.9589616017201553e-09, "loss": 0.0601, "step": 21361 }, { "epoch": 0.9926579925650557, "grad_norm": 0.5782205517550106, "learning_rate": 1.9379554795667666e-09, "loss": 0.0479, "step": 21362 }, { "epoch": 0.9927044609665427, "grad_norm": 0.3708817118522432, "learning_rate": 1.9170625687797705e-09, "loss": 0.052, "step": 21363 }, { "epoch": 0.9927509293680298, "grad_norm": 0.5633849870528991, "learning_rate": 1.896282869831012e-09, "loss": 0.0461, "step": 21364 }, { "epoch": 0.9927973977695167, "grad_norm": 0.7630261227350116, "learning_rate": 1.875616383192336e-09, "loss": 0.0685, "step": 21365 }, { "epoch": 0.9928438661710037, "grad_norm": 0.4680799044817494, "learning_rate": 1.8550631093311455e-09, "loss": 0.046, "step": 21366 }, { "epoch": 0.9928903345724907, "grad_norm": 0.612159488926037, "learning_rate": 1.8346230487131799e-09, "loss": 0.0608, "step": 21367 }, { "epoch": 0.9929368029739777, "grad_norm": 0.39029033153203213, "learning_rate": 1.8142962018014022e-09, "loss": 0.0454, "step": 21368 }, { "epoch": 0.9929832713754647, "grad_norm": 0.4381609542636117, "learning_rate": 1.7940825690565543e-09, "loss": 0.0425, "step": 21369 }, { "epoch": 0.9930297397769516, "grad_norm": 0.5228273755214433, "learning_rate": 1.773982150936049e-09, "loss": 0.0416, "step": 21370 }, { "epoch": 0.9930762081784387, "grad_norm": 0.6283079550078563, "learning_rate": 1.753994947895632e-09, "loss": 0.0493, "step": 21371 }, { "epoch": 0.9931226765799257, "grad_norm": 0.5478276907127188, "learning_rate": 1.7341209603877196e-09, "loss": 0.0618, "step": 21372 }, { "epoch": 0.9931691449814126, "grad_norm": 0.9146054101677725, "learning_rate": 1.7143601888625072e-09, "loss": 0.0436, "step": 21373 }, { "epoch": 0.9932156133828997, "grad_norm": 0.5083346606607265, "learning_rate": 1.6947126337674147e-09, "loss": 0.0481, "step": 21374 }, { "epoch": 0.9932620817843866, "grad_norm": 0.6771018100461813, "learning_rate": 1.6751782955476415e-09, "loss": 0.0503, "step": 21375 }, { "epoch": 0.9933085501858736, "grad_norm": 0.36065129527085443, "learning_rate": 1.6557571746461665e-09, "loss": 0.0497, "step": 21376 }, { "epoch": 0.9933550185873606, "grad_norm": 0.30204296850400036, "learning_rate": 1.636449271502083e-09, "loss": 0.0372, "step": 21377 }, { "epoch": 0.9934014869888476, "grad_norm": 0.5421196524469658, "learning_rate": 1.6172545865533739e-09, "loss": 0.057, "step": 21378 }, { "epoch": 0.9934479553903346, "grad_norm": 0.5907416407312075, "learning_rate": 1.5981731202346917e-09, "loss": 0.0512, "step": 21379 }, { "epoch": 0.9934944237918215, "grad_norm": 0.33073620825835054, "learning_rate": 1.579204872977913e-09, "loss": 0.03, "step": 21380 }, { "epoch": 0.9935408921933085, "grad_norm": 0.466451778130556, "learning_rate": 1.5603498452126941e-09, "loss": 0.0447, "step": 21381 }, { "epoch": 0.9935873605947956, "grad_norm": 0.4157860624192296, "learning_rate": 1.541608037366471e-09, "loss": 0.0589, "step": 21382 }, { "epoch": 0.9936338289962825, "grad_norm": 0.363163840508445, "learning_rate": 1.5229794498633488e-09, "loss": 0.0319, "step": 21383 }, { "epoch": 0.9936802973977695, "grad_norm": 0.522398430830263, "learning_rate": 1.5044640831257672e-09, "loss": 0.0522, "step": 21384 }, { "epoch": 0.9937267657992565, "grad_norm": 0.4791341128671058, "learning_rate": 1.486061937573391e-09, "loss": 0.0682, "step": 21385 }, { "epoch": 0.9937732342007435, "grad_norm": 0.5051284753949044, "learning_rate": 1.467773013621998e-09, "loss": 0.0503, "step": 21386 }, { "epoch": 0.9938197026022305, "grad_norm": 0.3152792278707421, "learning_rate": 1.4495973116862572e-09, "loss": 0.0264, "step": 21387 }, { "epoch": 0.9938661710037174, "grad_norm": 0.33089207733504344, "learning_rate": 1.431534832178616e-09, "loss": 0.0418, "step": 21388 }, { "epoch": 0.9939126394052045, "grad_norm": 0.5031935485677386, "learning_rate": 1.4135855755070815e-09, "loss": 0.0407, "step": 21389 }, { "epoch": 0.9939591078066915, "grad_norm": 0.6399432373370183, "learning_rate": 1.3957495420785505e-09, "loss": 0.0618, "step": 21390 }, { "epoch": 0.9940055762081784, "grad_norm": 0.5159272083311556, "learning_rate": 1.3780267322976992e-09, "loss": 0.0517, "step": 21391 }, { "epoch": 0.9940520446096655, "grad_norm": 0.6797880759600208, "learning_rate": 1.3604171465653183e-09, "loss": 0.0539, "step": 21392 }, { "epoch": 0.9940985130111524, "grad_norm": 0.5237564859252095, "learning_rate": 1.3429207852805325e-09, "loss": 0.0593, "step": 21393 }, { "epoch": 0.9941449814126394, "grad_norm": 0.6261122374556014, "learning_rate": 1.325537648839692e-09, "loss": 0.0593, "step": 21394 }, { "epoch": 0.9941914498141264, "grad_norm": 0.5246408301475352, "learning_rate": 1.3082677376363706e-09, "loss": 0.0599, "step": 21395 }, { "epoch": 0.9942379182156134, "grad_norm": 0.5852717138126591, "learning_rate": 1.2911110520619218e-09, "loss": 0.0556, "step": 21396 }, { "epoch": 0.9942843866171004, "grad_norm": 0.4551616939813242, "learning_rate": 1.274067592504924e-09, "loss": 0.049, "step": 21397 }, { "epoch": 0.9943308550185873, "grad_norm": 0.26235684471241616, "learning_rate": 1.2571373593506242e-09, "loss": 0.0324, "step": 21398 }, { "epoch": 0.9943773234200743, "grad_norm": 0.703393879322758, "learning_rate": 1.240320352983715e-09, "loss": 0.0494, "step": 21399 }, { "epoch": 0.9944237918215614, "grad_norm": 0.678044843665512, "learning_rate": 1.2236165737850025e-09, "loss": 0.0535, "step": 21400 }, { "epoch": 0.9944702602230483, "grad_norm": 0.4117415136199933, "learning_rate": 1.2070260221319629e-09, "loss": 0.0397, "step": 21401 }, { "epoch": 0.9945167286245353, "grad_norm": 0.4287132916219151, "learning_rate": 1.1905486984015169e-09, "loss": 0.0636, "step": 21402 }, { "epoch": 0.9945631970260222, "grad_norm": 0.3619552998686585, "learning_rate": 1.1741846029655891e-09, "loss": 0.0327, "step": 21403 }, { "epoch": 0.9946096654275093, "grad_norm": 0.781710189172368, "learning_rate": 1.1579337361961041e-09, "loss": 0.0676, "step": 21404 }, { "epoch": 0.9946561338289963, "grad_norm": 0.4891200728004565, "learning_rate": 1.1417960984605459e-09, "loss": 0.0535, "step": 21405 }, { "epoch": 0.9947026022304832, "grad_norm": 0.5439718670736411, "learning_rate": 1.1257716901241778e-09, "loss": 0.0445, "step": 21406 }, { "epoch": 0.9947490706319703, "grad_norm": 0.5855249366447466, "learning_rate": 1.1098605115500427e-09, "loss": 0.0528, "step": 21407 }, { "epoch": 0.9947955390334573, "grad_norm": 0.6299421782223786, "learning_rate": 1.0940625630995182e-09, "loss": 0.0522, "step": 21408 }, { "epoch": 0.9948420074349442, "grad_norm": 0.47133100500032904, "learning_rate": 1.0783778451295412e-09, "loss": 0.0429, "step": 21409 }, { "epoch": 0.9948884758364313, "grad_norm": 0.4678471468590808, "learning_rate": 1.0628063579953829e-09, "loss": 0.0472, "step": 21410 }, { "epoch": 0.9949349442379182, "grad_norm": 0.4201327189561144, "learning_rate": 1.0473481020500943e-09, "loss": 0.0425, "step": 21411 }, { "epoch": 0.9949814126394052, "grad_norm": 0.3945006086005891, "learning_rate": 1.0320030776433953e-09, "loss": 0.0434, "step": 21412 }, { "epoch": 0.9950278810408922, "grad_norm": 0.3124758465624035, "learning_rate": 1.0167712851233413e-09, "loss": 0.0672, "step": 21413 }, { "epoch": 0.9950743494423792, "grad_norm": 0.3673138010764901, "learning_rate": 1.0016527248346563e-09, "loss": 0.037, "step": 21414 }, { "epoch": 0.9951208178438662, "grad_norm": 0.5871735647870985, "learning_rate": 9.866473971203993e-10, "loss": 0.0604, "step": 21415 }, { "epoch": 0.9951672862453531, "grad_norm": 0.35889111560864156, "learning_rate": 9.717553023191883e-10, "loss": 0.0328, "step": 21416 }, { "epoch": 0.9952137546468401, "grad_norm": 0.3571995100605344, "learning_rate": 9.569764407696413e-10, "loss": 0.0399, "step": 21417 }, { "epoch": 0.9952602230483272, "grad_norm": 0.44472738018506514, "learning_rate": 9.423108128059355e-10, "loss": 0.0561, "step": 21418 }, { "epoch": 0.9953066914498141, "grad_norm": 0.3991888618586933, "learning_rate": 9.277584187600275e-10, "loss": 0.0407, "step": 21419 }, { "epoch": 0.9953531598513011, "grad_norm": 0.44029569174297306, "learning_rate": 9.133192589616535e-10, "loss": 0.0446, "step": 21420 }, { "epoch": 0.995399628252788, "grad_norm": 0.42536861580224644, "learning_rate": 8.989933337388846e-10, "loss": 0.052, "step": 21421 }, { "epoch": 0.9954460966542751, "grad_norm": 0.45270796211430936, "learning_rate": 8.847806434147954e-10, "loss": 0.06, "step": 21422 }, { "epoch": 0.9954925650557621, "grad_norm": 0.6600873896794306, "learning_rate": 8.70681188311906e-10, "loss": 0.0539, "step": 21423 }, { "epoch": 0.995539033457249, "grad_norm": 0.49913836282300433, "learning_rate": 8.566949687494053e-10, "loss": 0.0467, "step": 21424 }, { "epoch": 0.9955855018587361, "grad_norm": 0.5742512017205784, "learning_rate": 8.428219850442621e-10, "loss": 0.0569, "step": 21425 }, { "epoch": 0.9956319702602231, "grad_norm": 0.4311072032442494, "learning_rate": 8.290622375112246e-10, "loss": 0.0584, "step": 21426 }, { "epoch": 0.99567843866171, "grad_norm": 0.4517778764446377, "learning_rate": 8.15415726461155e-10, "loss": 0.0466, "step": 21427 }, { "epoch": 0.995724907063197, "grad_norm": 0.43922124452307054, "learning_rate": 8.018824522032509e-10, "loss": 0.0677, "step": 21428 }, { "epoch": 0.995771375464684, "grad_norm": 0.5022543891744221, "learning_rate": 7.884624150444887e-10, "loss": 0.0426, "step": 21429 }, { "epoch": 0.995817843866171, "grad_norm": 0.48698548004176545, "learning_rate": 7.751556152885142e-10, "loss": 0.0504, "step": 21430 }, { "epoch": 0.995864312267658, "grad_norm": 0.3871329877078751, "learning_rate": 7.619620532373084e-10, "loss": 0.0454, "step": 21431 }, { "epoch": 0.995910780669145, "grad_norm": 0.2753621980682036, "learning_rate": 7.488817291889661e-10, "loss": 0.0414, "step": 21432 }, { "epoch": 0.995957249070632, "grad_norm": 0.5274894046869661, "learning_rate": 7.359146434399167e-10, "loss": 0.0459, "step": 21433 }, { "epoch": 0.9960037174721189, "grad_norm": 0.9004304947938316, "learning_rate": 7.230607962838143e-10, "loss": 0.0798, "step": 21434 }, { "epoch": 0.9960501858736059, "grad_norm": 0.7833260175935599, "learning_rate": 7.103201880126476e-10, "loss": 0.0484, "step": 21435 }, { "epoch": 0.996096654275093, "grad_norm": 0.5877478617214331, "learning_rate": 6.976928189139642e-10, "loss": 0.0582, "step": 21436 }, { "epoch": 0.9961431226765799, "grad_norm": 0.9070819582880475, "learning_rate": 6.851786892742018e-10, "loss": 0.0477, "step": 21437 }, { "epoch": 0.9961895910780669, "grad_norm": 0.617047661412132, "learning_rate": 6.727777993770224e-10, "loss": 0.0606, "step": 21438 }, { "epoch": 0.9962360594795538, "grad_norm": 0.31740757361059563, "learning_rate": 6.604901495033123e-10, "loss": 0.0358, "step": 21439 }, { "epoch": 0.9962825278810409, "grad_norm": 0.6686647508935507, "learning_rate": 6.483157399306272e-10, "loss": 0.0626, "step": 21440 }, { "epoch": 0.9963289962825279, "grad_norm": 0.7714521692094132, "learning_rate": 6.36254570935968e-10, "loss": 0.0774, "step": 21441 }, { "epoch": 0.9963754646840148, "grad_norm": 0.6929039900168729, "learning_rate": 6.243066427913391e-10, "loss": 0.0588, "step": 21442 }, { "epoch": 0.9964219330855019, "grad_norm": 0.32845030106673845, "learning_rate": 6.124719557681902e-10, "loss": 0.0307, "step": 21443 }, { "epoch": 0.9964684014869889, "grad_norm": 0.39952059475803753, "learning_rate": 6.00750510134085e-10, "loss": 0.0543, "step": 21444 }, { "epoch": 0.9965148698884758, "grad_norm": 0.6848867500281024, "learning_rate": 5.891423061549217e-10, "loss": 0.0533, "step": 21445 }, { "epoch": 0.9965613382899629, "grad_norm": 0.5308624243715182, "learning_rate": 5.776473440938235e-10, "loss": 0.0447, "step": 21446 }, { "epoch": 0.9966078066914498, "grad_norm": 0.5066328945911951, "learning_rate": 5.662656242100273e-10, "loss": 0.0533, "step": 21447 }, { "epoch": 0.9966542750929368, "grad_norm": 0.42428813751965644, "learning_rate": 5.549971467627701e-10, "loss": 0.0648, "step": 21448 }, { "epoch": 0.9967007434944238, "grad_norm": 0.4041982412909981, "learning_rate": 5.438419120062933e-10, "loss": 0.0514, "step": 21449 }, { "epoch": 0.9967472118959108, "grad_norm": 0.39496719227631544, "learning_rate": 5.327999201937273e-10, "loss": 0.0598, "step": 21450 }, { "epoch": 0.9967936802973978, "grad_norm": 0.6079567608597324, "learning_rate": 5.218711715754276e-10, "loss": 0.048, "step": 21451 }, { "epoch": 0.9968401486988847, "grad_norm": 0.5737584267049087, "learning_rate": 5.110556663978639e-10, "loss": 0.0673, "step": 21452 }, { "epoch": 0.9968866171003717, "grad_norm": 0.566997988688321, "learning_rate": 5.003534049075054e-10, "loss": 0.0493, "step": 21453 }, { "epoch": 0.9969330855018588, "grad_norm": 0.5663238071531956, "learning_rate": 4.897643873452706e-10, "loss": 0.066, "step": 21454 }, { "epoch": 0.9969795539033457, "grad_norm": 0.29929307089467905, "learning_rate": 4.792886139526332e-10, "loss": 0.0409, "step": 21455 }, { "epoch": 0.9970260223048327, "grad_norm": 0.4458248304706019, "learning_rate": 4.689260849655152e-10, "loss": 0.0464, "step": 21456 }, { "epoch": 0.9970724907063196, "grad_norm": 0.2613276907836697, "learning_rate": 4.586768006192843e-10, "loss": 0.0432, "step": 21457 }, { "epoch": 0.9971189591078067, "grad_norm": 1.0362190287431674, "learning_rate": 4.4854076114542177e-10, "loss": 0.0492, "step": 21458 }, { "epoch": 0.9971654275092937, "grad_norm": 0.46553369150751894, "learning_rate": 4.38517966774854e-10, "loss": 0.0493, "step": 21459 }, { "epoch": 0.9972118959107806, "grad_norm": 0.44052654826956256, "learning_rate": 4.2860841773295636e-10, "loss": 0.0361, "step": 21460 }, { "epoch": 0.9972583643122677, "grad_norm": 0.4377289024674464, "learning_rate": 4.1881211424565914e-10, "loss": 0.0371, "step": 21461 }, { "epoch": 0.9973048327137547, "grad_norm": 0.6832341885791621, "learning_rate": 4.091290565338968e-10, "loss": 0.0442, "step": 21462 }, { "epoch": 0.9973513011152416, "grad_norm": 0.4984586225837554, "learning_rate": 3.995592448174934e-10, "loss": 0.0441, "step": 21463 }, { "epoch": 0.9973977695167286, "grad_norm": 0.535683681640588, "learning_rate": 3.9010267931294257e-10, "loss": 0.0312, "step": 21464 }, { "epoch": 0.9974442379182156, "grad_norm": 0.5290879248148223, "learning_rate": 3.807593602350723e-10, "loss": 0.0614, "step": 21465 }, { "epoch": 0.9974907063197026, "grad_norm": 0.4300318433494996, "learning_rate": 3.7152928779427e-10, "loss": 0.0356, "step": 21466 }, { "epoch": 0.9975371747211896, "grad_norm": 0.7069408130206188, "learning_rate": 3.6241246220092287e-10, "loss": 0.0569, "step": 21467 }, { "epoch": 0.9975836431226766, "grad_norm": 0.4546433325920944, "learning_rate": 3.534088836604221e-10, "loss": 0.0675, "step": 21468 }, { "epoch": 0.9976301115241636, "grad_norm": 0.6779141726914745, "learning_rate": 3.4451855237760403e-10, "loss": 0.0552, "step": 21469 }, { "epoch": 0.9976765799256505, "grad_norm": 0.5065648827063638, "learning_rate": 3.3574146855341885e-10, "loss": 0.0521, "step": 21470 }, { "epoch": 0.9977230483271375, "grad_norm": 0.9496368008763894, "learning_rate": 3.270776323871516e-10, "loss": 0.036, "step": 21471 }, { "epoch": 0.9977695167286246, "grad_norm": 0.3906928524736156, "learning_rate": 3.1852704407420163e-10, "loss": 0.0449, "step": 21472 }, { "epoch": 0.9978159851301115, "grad_norm": 0.4395825268090959, "learning_rate": 3.1008970380830285e-10, "loss": 0.0542, "step": 21473 }, { "epoch": 0.9978624535315985, "grad_norm": 0.4459103609169576, "learning_rate": 3.0176561178152377e-10, "loss": 0.0489, "step": 21474 }, { "epoch": 0.9979089219330854, "grad_norm": 0.3831689542971272, "learning_rate": 2.935547681814921e-10, "loss": 0.0384, "step": 21475 }, { "epoch": 0.9979553903345725, "grad_norm": 0.47990299139425696, "learning_rate": 2.854571731947253e-10, "loss": 0.0493, "step": 21476 }, { "epoch": 0.9980018587360595, "grad_norm": 0.5683010926025452, "learning_rate": 2.7747282700441025e-10, "loss": 0.0476, "step": 21477 }, { "epoch": 0.9980483271375464, "grad_norm": 0.5680247859255182, "learning_rate": 2.6960172979151324e-10, "loss": 0.0502, "step": 21478 }, { "epoch": 0.9980947955390335, "grad_norm": 0.411215269827534, "learning_rate": 2.61843881734225e-10, "loss": 0.0649, "step": 21479 }, { "epoch": 0.9981412639405205, "grad_norm": 0.5734247879408877, "learning_rate": 2.54199283008516e-10, "loss": 0.0419, "step": 21480 }, { "epoch": 0.9981877323420074, "grad_norm": 0.473152193275705, "learning_rate": 2.466679337870259e-10, "loss": 0.0457, "step": 21481 }, { "epoch": 0.9982342007434944, "grad_norm": 0.3665159725173964, "learning_rate": 2.39249834240729e-10, "loss": 0.0307, "step": 21482 }, { "epoch": 0.9982806691449814, "grad_norm": 0.6611136487816289, "learning_rate": 2.319449845378241e-10, "loss": 0.0569, "step": 21483 }, { "epoch": 0.9983271375464684, "grad_norm": 0.6769180255555843, "learning_rate": 2.2475338484317933e-10, "loss": 0.0508, "step": 21484 }, { "epoch": 0.9983736059479554, "grad_norm": 0.41613224231097873, "learning_rate": 2.1767503531999746e-10, "loss": 0.0355, "step": 21485 }, { "epoch": 0.9984200743494424, "grad_norm": 0.32385210825409616, "learning_rate": 2.1070993612870571e-10, "loss": 0.0445, "step": 21486 }, { "epoch": 0.9984665427509294, "grad_norm": 0.5235107038780454, "learning_rate": 2.038580874275109e-10, "loss": 0.0623, "step": 21487 }, { "epoch": 0.9985130111524163, "grad_norm": 0.5566512238382185, "learning_rate": 1.9711948937017888e-10, "loss": 0.0484, "step": 21488 }, { "epoch": 0.9985594795539033, "grad_norm": 0.5749937956151495, "learning_rate": 1.9049414211103068e-10, "loss": 0.0436, "step": 21489 }, { "epoch": 0.9986059479553904, "grad_norm": 0.2986303897424674, "learning_rate": 1.8398204579939128e-10, "loss": 0.0476, "step": 21490 }, { "epoch": 0.9986524163568773, "grad_norm": 0.2480874459771145, "learning_rate": 1.7758320058236522e-10, "loss": 0.026, "step": 21491 }, { "epoch": 0.9986988847583643, "grad_norm": 0.34783282805253607, "learning_rate": 1.7129760660539174e-10, "loss": 0.0464, "step": 21492 }, { "epoch": 0.9987453531598512, "grad_norm": 0.6749153007476693, "learning_rate": 1.6512526401113448e-10, "loss": 0.0537, "step": 21493 }, { "epoch": 0.9987918215613383, "grad_norm": 0.5220038454117126, "learning_rate": 1.5906617293892644e-10, "loss": 0.0442, "step": 21494 }, { "epoch": 0.9988382899628253, "grad_norm": 0.5356147121610284, "learning_rate": 1.531203335258802e-10, "loss": 0.0526, "step": 21495 }, { "epoch": 0.9988847583643122, "grad_norm": 0.7044250870156203, "learning_rate": 1.4728774590688776e-10, "loss": 0.0543, "step": 21496 }, { "epoch": 0.9989312267657993, "grad_norm": 0.5607426845500709, "learning_rate": 1.4156841021406574e-10, "loss": 0.0434, "step": 21497 }, { "epoch": 0.9989776951672863, "grad_norm": 0.255176702794927, "learning_rate": 1.3596232657675512e-10, "loss": 0.0388, "step": 21498 }, { "epoch": 0.9990241635687732, "grad_norm": 0.5183731497877468, "learning_rate": 1.304694951226315e-10, "loss": 0.0472, "step": 21499 }, { "epoch": 0.9990706319702602, "grad_norm": 0.3516080742966754, "learning_rate": 1.250899159754848e-10, "loss": 0.0373, "step": 21500 }, { "epoch": 0.9991171003717472, "grad_norm": 0.5182385464482984, "learning_rate": 1.1982358925743952e-10, "loss": 0.0312, "step": 21501 }, { "epoch": 0.9991635687732342, "grad_norm": 0.49580086552675373, "learning_rate": 1.1467051508728955e-10, "loss": 0.0485, "step": 21502 }, { "epoch": 0.9992100371747212, "grad_norm": 0.5416369787678053, "learning_rate": 1.0963069358271849e-10, "loss": 0.055, "step": 21503 }, { "epoch": 0.9992565055762082, "grad_norm": 0.4724464854814933, "learning_rate": 1.0470412485696912e-10, "loss": 0.0454, "step": 21504 }, { "epoch": 0.9993029739776952, "grad_norm": 0.4814115310034121, "learning_rate": 9.989080902161886e-11, "loss": 0.0451, "step": 21505 }, { "epoch": 0.9993494423791821, "grad_norm": 0.5588311117025482, "learning_rate": 9.519074618657975e-11, "loss": 0.0412, "step": 21506 }, { "epoch": 0.9993959107806691, "grad_norm": 0.4800879251788432, "learning_rate": 9.060393645732302e-11, "loss": 0.0402, "step": 21507 }, { "epoch": 0.9994423791821562, "grad_norm": 0.4017529884511829, "learning_rate": 8.613037993876471e-11, "loss": 0.0441, "step": 21508 }, { "epoch": 0.9994888475836431, "grad_norm": 0.5451574941594673, "learning_rate": 8.177007673138004e-11, "loss": 0.0517, "step": 21509 }, { "epoch": 0.9995353159851301, "grad_norm": 0.36920980674093806, "learning_rate": 7.752302693397884e-11, "loss": 0.0458, "step": 21510 }, { "epoch": 0.999581784386617, "grad_norm": 0.7538769716315065, "learning_rate": 7.338923064315052e-11, "loss": 0.0635, "step": 21511 }, { "epoch": 0.9996282527881041, "grad_norm": 0.7551318450524457, "learning_rate": 6.93686879527089e-11, "loss": 0.0405, "step": 21512 }, { "epoch": 0.9996747211895911, "grad_norm": 0.3643639920150027, "learning_rate": 6.546139895313719e-11, "loss": 0.0607, "step": 21513 }, { "epoch": 0.999721189591078, "grad_norm": 0.48196306046611576, "learning_rate": 6.166736373325322e-11, "loss": 0.0339, "step": 21514 }, { "epoch": 0.9997676579925651, "grad_norm": 0.6862428592420815, "learning_rate": 5.798658237854415e-11, "loss": 0.0335, "step": 21515 }, { "epoch": 0.9998141263940521, "grad_norm": 0.35984589659527005, "learning_rate": 5.441905497338695e-11, "loss": 0.0313, "step": 21516 }, { "epoch": 0.999860594795539, "grad_norm": 0.6971786579021827, "learning_rate": 5.096478159771767e-11, "loss": 0.0645, "step": 21517 }, { "epoch": 0.999907063197026, "grad_norm": 0.3005469731341458, "learning_rate": 4.762376232980703e-11, "loss": 0.0266, "step": 21518 }, { "epoch": 0.999953531598513, "grad_norm": 0.43353664264004166, "learning_rate": 4.4395997246260424e-11, "loss": 0.0484, "step": 21519 }, { "epoch": 1.0, "grad_norm": 0.6619521928086528, "learning_rate": 4.1281486418687235e-11, "loss": 0.0429, "step": 21520 }, { "epoch": 1.0, "step": 21520, "total_flos": 107554027631616.0, "train_loss": 0.07908512394811237, "train_runtime": 40769.7596, "train_samples_per_second": 2.111, "train_steps_per_second": 0.528 } ], "logging_steps": 1.0, "max_steps": 21520, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 107554027631616.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }