{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 13.166556945358789, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0065832784726793945, "grad_norm": 4.667604446411133, "learning_rate": 1.8e-07, "loss": 0.3117, "step": 10 }, { "epoch": 0.013166556945358789, "grad_norm": 4.76257848739624, "learning_rate": 3.8e-07, "loss": 0.3056, "step": 20 }, { "epoch": 0.019749835418038184, "grad_norm": 4.510583400726318, "learning_rate": 5.8e-07, "loss": 0.3215, "step": 30 }, { "epoch": 0.026333113890717578, "grad_norm": 4.296549320220947, "learning_rate": 7.8e-07, "loss": 0.2798, "step": 40 }, { "epoch": 0.032916392363396975, "grad_norm": 3.3769516944885254, "learning_rate": 9.8e-07, "loss": 0.2167, "step": 50 }, { "epoch": 0.03949967083607637, "grad_norm": 1.7368696928024292, "learning_rate": 1.18e-06, "loss": 0.1576, "step": 60 }, { "epoch": 0.04608294930875576, "grad_norm": 1.2467026710510254, "learning_rate": 1.3800000000000001e-06, "loss": 0.0753, "step": 70 }, { "epoch": 0.052666227781435156, "grad_norm": 0.6001964807510376, "learning_rate": 1.5800000000000003e-06, "loss": 0.0596, "step": 80 }, { "epoch": 0.05924950625411455, "grad_norm": 0.700756847858429, "learning_rate": 1.7800000000000001e-06, "loss": 0.0377, "step": 90 }, { "epoch": 0.06583278472679395, "grad_norm": 0.520061731338501, "learning_rate": 1.98e-06, "loss": 0.0338, "step": 100 }, { "epoch": 0.07241606319947334, "grad_norm": 0.42451295256614685, "learning_rate": 2.1800000000000003e-06, "loss": 0.0238, "step": 110 }, { "epoch": 0.07899934167215274, "grad_norm": 0.4142055809497833, "learning_rate": 2.38e-06, "loss": 0.0294, "step": 120 }, { "epoch": 0.08558262014483213, "grad_norm": 0.250784307718277, "learning_rate": 2.5800000000000003e-06, "loss": 0.0172, "step": 130 }, { "epoch": 0.09216589861751152, "grad_norm": 0.3852202594280243, "learning_rate": 2.78e-06, "loss": 0.017, "step": 140 }, { "epoch": 0.09874917709019092, "grad_norm": 0.39633437991142273, "learning_rate": 2.9800000000000003e-06, "loss": 0.0208, "step": 150 }, { "epoch": 0.10533245556287031, "grad_norm": 0.20055508613586426, "learning_rate": 3.1800000000000005e-06, "loss": 0.0177, "step": 160 }, { "epoch": 0.1119157340355497, "grad_norm": 0.16524334251880646, "learning_rate": 3.38e-06, "loss": 0.0163, "step": 170 }, { "epoch": 0.1184990125082291, "grad_norm": 0.22953571379184723, "learning_rate": 3.58e-06, "loss": 0.0153, "step": 180 }, { "epoch": 0.1250822909809085, "grad_norm": 0.1979047656059265, "learning_rate": 3.7800000000000002e-06, "loss": 0.0194, "step": 190 }, { "epoch": 0.1316655694535879, "grad_norm": 0.3481867015361786, "learning_rate": 3.98e-06, "loss": 0.0186, "step": 200 }, { "epoch": 0.1382488479262673, "grad_norm": 0.47407621145248413, "learning_rate": 4.18e-06, "loss": 0.0169, "step": 210 }, { "epoch": 0.1448321263989467, "grad_norm": 0.37531784176826477, "learning_rate": 4.38e-06, "loss": 0.0179, "step": 220 }, { "epoch": 0.15141540487162608, "grad_norm": 0.27084881067276, "learning_rate": 4.58e-06, "loss": 0.0134, "step": 230 }, { "epoch": 0.15799868334430547, "grad_norm": 0.366066038608551, "learning_rate": 4.780000000000001e-06, "loss": 0.0164, "step": 240 }, { "epoch": 0.16458196181698487, "grad_norm": 0.2672060430049896, "learning_rate": 4.98e-06, "loss": 0.0162, "step": 250 }, { "epoch": 0.17116524028966426, "grad_norm": 0.3275546431541443, "learning_rate": 5.18e-06, "loss": 0.0134, "step": 260 }, { "epoch": 0.17774851876234365, "grad_norm": 0.11839713156223297, "learning_rate": 5.38e-06, "loss": 0.0132, "step": 270 }, { "epoch": 0.18433179723502305, "grad_norm": 0.41320323944091797, "learning_rate": 5.580000000000001e-06, "loss": 0.016, "step": 280 }, { "epoch": 0.19091507570770244, "grad_norm": 0.30657222867012024, "learning_rate": 5.78e-06, "loss": 0.0115, "step": 290 }, { "epoch": 0.19749835418038184, "grad_norm": 0.4881065785884857, "learning_rate": 5.98e-06, "loss": 0.0159, "step": 300 }, { "epoch": 0.20408163265306123, "grad_norm": 0.46792495250701904, "learning_rate": 6.18e-06, "loss": 0.014, "step": 310 }, { "epoch": 0.21066491112574062, "grad_norm": 0.42141374945640564, "learning_rate": 6.38e-06, "loss": 0.0119, "step": 320 }, { "epoch": 0.21724818959842002, "grad_norm": 0.3819800019264221, "learning_rate": 6.58e-06, "loss": 0.0148, "step": 330 }, { "epoch": 0.2238314680710994, "grad_norm": 0.4395577907562256, "learning_rate": 6.78e-06, "loss": 0.0115, "step": 340 }, { "epoch": 0.2304147465437788, "grad_norm": 0.36575037240982056, "learning_rate": 6.98e-06, "loss": 0.0119, "step": 350 }, { "epoch": 0.2369980250164582, "grad_norm": 0.2176530808210373, "learning_rate": 7.180000000000001e-06, "loss": 0.0118, "step": 360 }, { "epoch": 0.2435813034891376, "grad_norm": 0.15246424078941345, "learning_rate": 7.3800000000000005e-06, "loss": 0.0085, "step": 370 }, { "epoch": 0.250164581961817, "grad_norm": 0.25757896900177, "learning_rate": 7.580000000000001e-06, "loss": 0.011, "step": 380 }, { "epoch": 0.2567478604344964, "grad_norm": 0.40807631611824036, "learning_rate": 7.78e-06, "loss": 0.0132, "step": 390 }, { "epoch": 0.2633311389071758, "grad_norm": 0.22928398847579956, "learning_rate": 7.98e-06, "loss": 0.0082, "step": 400 }, { "epoch": 0.26991441737985516, "grad_norm": 0.32516878843307495, "learning_rate": 8.18e-06, "loss": 0.0102, "step": 410 }, { "epoch": 0.2764976958525346, "grad_norm": 0.38103151321411133, "learning_rate": 8.380000000000001e-06, "loss": 0.0156, "step": 420 }, { "epoch": 0.28308097432521395, "grad_norm": 0.2733783721923828, "learning_rate": 8.580000000000001e-06, "loss": 0.0109, "step": 430 }, { "epoch": 0.2896642527978934, "grad_norm": 0.25594833493232727, "learning_rate": 8.78e-06, "loss": 0.0101, "step": 440 }, { "epoch": 0.29624753127057274, "grad_norm": 0.36928215622901917, "learning_rate": 8.98e-06, "loss": 0.0103, "step": 450 }, { "epoch": 0.30283080974325216, "grad_norm": 0.2253832072019577, "learning_rate": 9.180000000000002e-06, "loss": 0.0081, "step": 460 }, { "epoch": 0.3094140882159315, "grad_norm": 0.43624746799468994, "learning_rate": 9.38e-06, "loss": 0.0114, "step": 470 }, { "epoch": 0.31599736668861095, "grad_norm": 0.23713339865207672, "learning_rate": 9.58e-06, "loss": 0.0084, "step": 480 }, { "epoch": 0.3225806451612903, "grad_norm": 0.32124054431915283, "learning_rate": 9.78e-06, "loss": 0.0097, "step": 490 }, { "epoch": 0.32916392363396973, "grad_norm": 0.26847419142723083, "learning_rate": 9.980000000000001e-06, "loss": 0.0069, "step": 500 }, { "epoch": 0.3357472021066491, "grad_norm": 0.33049169182777405, "learning_rate": 1.018e-05, "loss": 0.0096, "step": 510 }, { "epoch": 0.3423304805793285, "grad_norm": 0.1753293126821518, "learning_rate": 1.038e-05, "loss": 0.0064, "step": 520 }, { "epoch": 0.3489137590520079, "grad_norm": 0.29450568556785583, "learning_rate": 1.058e-05, "loss": 0.0101, "step": 530 }, { "epoch": 0.3554970375246873, "grad_norm": 0.4020543098449707, "learning_rate": 1.0780000000000002e-05, "loss": 0.0117, "step": 540 }, { "epoch": 0.3620803159973667, "grad_norm": 0.191111221909523, "learning_rate": 1.098e-05, "loss": 0.0065, "step": 550 }, { "epoch": 0.3686635944700461, "grad_norm": 0.28593379259109497, "learning_rate": 1.118e-05, "loss": 0.0092, "step": 560 }, { "epoch": 0.37524687294272546, "grad_norm": 0.5440319776535034, "learning_rate": 1.1380000000000001e-05, "loss": 0.0132, "step": 570 }, { "epoch": 0.3818301514154049, "grad_norm": 0.3086682856082916, "learning_rate": 1.1580000000000001e-05, "loss": 0.0109, "step": 580 }, { "epoch": 0.38841342988808425, "grad_norm": 0.31464439630508423, "learning_rate": 1.178e-05, "loss": 0.0082, "step": 590 }, { "epoch": 0.39499670836076367, "grad_norm": 0.534359872341156, "learning_rate": 1.198e-05, "loss": 0.0105, "step": 600 }, { "epoch": 0.40157998683344304, "grad_norm": 0.3846067786216736, "learning_rate": 1.2180000000000002e-05, "loss": 0.012, "step": 610 }, { "epoch": 0.40816326530612246, "grad_norm": 0.3279959559440613, "learning_rate": 1.238e-05, "loss": 0.0104, "step": 620 }, { "epoch": 0.4147465437788018, "grad_norm": 0.3157939016819, "learning_rate": 1.258e-05, "loss": 0.0082, "step": 630 }, { "epoch": 0.42132982225148125, "grad_norm": 0.19066298007965088, "learning_rate": 1.278e-05, "loss": 0.0065, "step": 640 }, { "epoch": 0.4279131007241606, "grad_norm": 0.2096276432275772, "learning_rate": 1.2980000000000001e-05, "loss": 0.0077, "step": 650 }, { "epoch": 0.43449637919684003, "grad_norm": 0.3505825698375702, "learning_rate": 1.3180000000000001e-05, "loss": 0.0079, "step": 660 }, { "epoch": 0.4410796576695194, "grad_norm": 0.28009364008903503, "learning_rate": 1.338e-05, "loss": 0.0049, "step": 670 }, { "epoch": 0.4476629361421988, "grad_norm": 0.2566313147544861, "learning_rate": 1.358e-05, "loss": 0.0099, "step": 680 }, { "epoch": 0.4542462146148782, "grad_norm": 0.18947407603263855, "learning_rate": 1.3780000000000002e-05, "loss": 0.0042, "step": 690 }, { "epoch": 0.4608294930875576, "grad_norm": 0.18175585567951202, "learning_rate": 1.3980000000000002e-05, "loss": 0.0054, "step": 700 }, { "epoch": 0.467412771560237, "grad_norm": 0.24939915537834167, "learning_rate": 1.4180000000000001e-05, "loss": 0.0077, "step": 710 }, { "epoch": 0.4739960500329164, "grad_norm": 0.26245540380477905, "learning_rate": 1.4380000000000001e-05, "loss": 0.0056, "step": 720 }, { "epoch": 0.48057932850559576, "grad_norm": 0.13992945849895477, "learning_rate": 1.4580000000000003e-05, "loss": 0.0075, "step": 730 }, { "epoch": 0.4871626069782752, "grad_norm": 0.21726356446743011, "learning_rate": 1.4779999999999999e-05, "loss": 0.0062, "step": 740 }, { "epoch": 0.4937458854509546, "grad_norm": 0.23025143146514893, "learning_rate": 1.4979999999999999e-05, "loss": 0.0066, "step": 750 }, { "epoch": 0.500329163923634, "grad_norm": 0.38760191202163696, "learning_rate": 1.518e-05, "loss": 0.0064, "step": 760 }, { "epoch": 0.5069124423963134, "grad_norm": 0.40132707357406616, "learning_rate": 1.538e-05, "loss": 0.0068, "step": 770 }, { "epoch": 0.5134957208689928, "grad_norm": 0.18108682334423065, "learning_rate": 1.558e-05, "loss": 0.008, "step": 780 }, { "epoch": 0.5200789993416721, "grad_norm": 0.27599403262138367, "learning_rate": 1.578e-05, "loss": 0.0048, "step": 790 }, { "epoch": 0.5266622778143516, "grad_norm": 0.17497989535331726, "learning_rate": 1.598e-05, "loss": 0.0063, "step": 800 }, { "epoch": 0.533245556287031, "grad_norm": 0.19714988768100739, "learning_rate": 1.618e-05, "loss": 0.0055, "step": 810 }, { "epoch": 0.5398288347597103, "grad_norm": 0.3557318150997162, "learning_rate": 1.6380000000000002e-05, "loss": 0.0138, "step": 820 }, { "epoch": 0.5464121132323897, "grad_norm": 0.2810114324092865, "learning_rate": 1.658e-05, "loss": 0.0071, "step": 830 }, { "epoch": 0.5529953917050692, "grad_norm": 0.31869757175445557, "learning_rate": 1.6780000000000002e-05, "loss": 0.0051, "step": 840 }, { "epoch": 0.5595786701777485, "grad_norm": 0.34567221999168396, "learning_rate": 1.698e-05, "loss": 0.007, "step": 850 }, { "epoch": 0.5661619486504279, "grad_norm": 0.3867408037185669, "learning_rate": 1.718e-05, "loss": 0.0091, "step": 860 }, { "epoch": 0.5727452271231073, "grad_norm": 0.23460763692855835, "learning_rate": 1.7380000000000003e-05, "loss": 0.0065, "step": 870 }, { "epoch": 0.5793285055957867, "grad_norm": 0.41180169582366943, "learning_rate": 1.758e-05, "loss": 0.0078, "step": 880 }, { "epoch": 0.5859117840684661, "grad_norm": 0.33165597915649414, "learning_rate": 1.7780000000000003e-05, "loss": 0.0076, "step": 890 }, { "epoch": 0.5924950625411455, "grad_norm": 0.17362745106220245, "learning_rate": 1.798e-05, "loss": 0.0061, "step": 900 }, { "epoch": 0.5990783410138248, "grad_norm": 0.35882455110549927, "learning_rate": 1.818e-05, "loss": 0.0078, "step": 910 }, { "epoch": 0.6056616194865043, "grad_norm": 0.2407483607530594, "learning_rate": 1.838e-05, "loss": 0.0124, "step": 920 }, { "epoch": 0.6122448979591837, "grad_norm": 0.38892731070518494, "learning_rate": 1.858e-05, "loss": 0.0123, "step": 930 }, { "epoch": 0.618828176431863, "grad_norm": 0.2332616001367569, "learning_rate": 1.878e-05, "loss": 0.0055, "step": 940 }, { "epoch": 0.6254114549045424, "grad_norm": 0.2040179818868637, "learning_rate": 1.898e-05, "loss": 0.0062, "step": 950 }, { "epoch": 0.6319947333772219, "grad_norm": 0.1975512057542801, "learning_rate": 1.918e-05, "loss": 0.0095, "step": 960 }, { "epoch": 0.6385780118499013, "grad_norm": 0.3402547836303711, "learning_rate": 1.938e-05, "loss": 0.0057, "step": 970 }, { "epoch": 0.6451612903225806, "grad_norm": 0.2078298032283783, "learning_rate": 1.9580000000000002e-05, "loss": 0.0073, "step": 980 }, { "epoch": 0.65174456879526, "grad_norm": 0.2692283093929291, "learning_rate": 1.978e-05, "loss": 0.0055, "step": 990 }, { "epoch": 0.6583278472679395, "grad_norm": 0.25150570273399353, "learning_rate": 1.9980000000000002e-05, "loss": 0.0053, "step": 1000 }, { "epoch": 0.6649111257406188, "grad_norm": 0.3064151704311371, "learning_rate": 2.0180000000000003e-05, "loss": 0.0087, "step": 1010 }, { "epoch": 0.6714944042132982, "grad_norm": 0.17607106268405914, "learning_rate": 2.038e-05, "loss": 0.0064, "step": 1020 }, { "epoch": 0.6780776826859776, "grad_norm": 0.3017083704471588, "learning_rate": 2.0580000000000003e-05, "loss": 0.0068, "step": 1030 }, { "epoch": 0.684660961158657, "grad_norm": 0.3750017583370209, "learning_rate": 2.078e-05, "loss": 0.006, "step": 1040 }, { "epoch": 0.6912442396313364, "grad_norm": 0.2683260440826416, "learning_rate": 2.098e-05, "loss": 0.0066, "step": 1050 }, { "epoch": 0.6978275181040158, "grad_norm": 0.2980571389198303, "learning_rate": 2.118e-05, "loss": 0.0093, "step": 1060 }, { "epoch": 0.7044107965766951, "grad_norm": 0.18826761841773987, "learning_rate": 2.138e-05, "loss": 0.0054, "step": 1070 }, { "epoch": 0.7109940750493746, "grad_norm": 0.16944971680641174, "learning_rate": 2.158e-05, "loss": 0.007, "step": 1080 }, { "epoch": 0.717577353522054, "grad_norm": 0.18174539506435394, "learning_rate": 2.178e-05, "loss": 0.0055, "step": 1090 }, { "epoch": 0.7241606319947334, "grad_norm": 0.09434781968593597, "learning_rate": 2.198e-05, "loss": 0.0042, "step": 1100 }, { "epoch": 0.7307439104674127, "grad_norm": 0.22035308182239532, "learning_rate": 2.218e-05, "loss": 0.0048, "step": 1110 }, { "epoch": 0.7373271889400922, "grad_norm": 0.15658259391784668, "learning_rate": 2.2380000000000003e-05, "loss": 0.0051, "step": 1120 }, { "epoch": 0.7439104674127716, "grad_norm": 0.14957945048809052, "learning_rate": 2.258e-05, "loss": 0.0049, "step": 1130 }, { "epoch": 0.7504937458854509, "grad_norm": 0.2875000834465027, "learning_rate": 2.2780000000000002e-05, "loss": 0.0038, "step": 1140 }, { "epoch": 0.7570770243581304, "grad_norm": 0.21444852650165558, "learning_rate": 2.298e-05, "loss": 0.0061, "step": 1150 }, { "epoch": 0.7636603028308098, "grad_norm": 0.22133076190948486, "learning_rate": 2.318e-05, "loss": 0.0047, "step": 1160 }, { "epoch": 0.7702435813034891, "grad_norm": 0.17518217861652374, "learning_rate": 2.3380000000000003e-05, "loss": 0.0054, "step": 1170 }, { "epoch": 0.7768268597761685, "grad_norm": 0.22834819555282593, "learning_rate": 2.358e-05, "loss": 0.0054, "step": 1180 }, { "epoch": 0.783410138248848, "grad_norm": 0.1731962412595749, "learning_rate": 2.3780000000000003e-05, "loss": 0.0057, "step": 1190 }, { "epoch": 0.7899934167215273, "grad_norm": 0.1858062893152237, "learning_rate": 2.398e-05, "loss": 0.0069, "step": 1200 }, { "epoch": 0.7965766951942067, "grad_norm": 0.16888290643692017, "learning_rate": 2.418e-05, "loss": 0.0082, "step": 1210 }, { "epoch": 0.8031599736668861, "grad_norm": 0.3752075135707855, "learning_rate": 2.438e-05, "loss": 0.007, "step": 1220 }, { "epoch": 0.8097432521395656, "grad_norm": 0.21810230612754822, "learning_rate": 2.4580000000000002e-05, "loss": 0.0046, "step": 1230 }, { "epoch": 0.8163265306122449, "grad_norm": 0.3144676983356476, "learning_rate": 2.478e-05, "loss": 0.007, "step": 1240 }, { "epoch": 0.8229098090849243, "grad_norm": 0.13102562725543976, "learning_rate": 2.498e-05, "loss": 0.0041, "step": 1250 }, { "epoch": 0.8294930875576036, "grad_norm": 0.15110915899276733, "learning_rate": 2.5180000000000003e-05, "loss": 0.0041, "step": 1260 }, { "epoch": 0.8360763660302831, "grad_norm": 0.150835782289505, "learning_rate": 2.5380000000000004e-05, "loss": 0.0044, "step": 1270 }, { "epoch": 0.8426596445029625, "grad_norm": 0.19511735439300537, "learning_rate": 2.5580000000000002e-05, "loss": 0.0059, "step": 1280 }, { "epoch": 0.8492429229756419, "grad_norm": 0.15513381361961365, "learning_rate": 2.5779999999999997e-05, "loss": 0.0041, "step": 1290 }, { "epoch": 0.8558262014483212, "grad_norm": 0.1765744835138321, "learning_rate": 2.598e-05, "loss": 0.0061, "step": 1300 }, { "epoch": 0.8624094799210007, "grad_norm": 0.1655944436788559, "learning_rate": 2.618e-05, "loss": 0.0045, "step": 1310 }, { "epoch": 0.8689927583936801, "grad_norm": 0.17169371247291565, "learning_rate": 2.6379999999999998e-05, "loss": 0.0055, "step": 1320 }, { "epoch": 0.8755760368663594, "grad_norm": 0.3155057728290558, "learning_rate": 2.658e-05, "loss": 0.0071, "step": 1330 }, { "epoch": 0.8821593153390388, "grad_norm": 0.31879231333732605, "learning_rate": 2.678e-05, "loss": 0.0093, "step": 1340 }, { "epoch": 0.8887425938117183, "grad_norm": 0.26626473665237427, "learning_rate": 2.698e-05, "loss": 0.0041, "step": 1350 }, { "epoch": 0.8953258722843976, "grad_norm": 0.31761062145233154, "learning_rate": 2.718e-05, "loss": 0.0098, "step": 1360 }, { "epoch": 0.901909150757077, "grad_norm": 0.22863753139972687, "learning_rate": 2.738e-05, "loss": 0.0048, "step": 1370 }, { "epoch": 0.9084924292297564, "grad_norm": 0.34291163086891174, "learning_rate": 2.758e-05, "loss": 0.005, "step": 1380 }, { "epoch": 0.9150757077024358, "grad_norm": 0.2496219128370285, "learning_rate": 2.778e-05, "loss": 0.006, "step": 1390 }, { "epoch": 0.9216589861751152, "grad_norm": 0.262248158454895, "learning_rate": 2.798e-05, "loss": 0.0052, "step": 1400 }, { "epoch": 0.9282422646477946, "grad_norm": 0.2531645894050598, "learning_rate": 2.818e-05, "loss": 0.0105, "step": 1410 }, { "epoch": 0.934825543120474, "grad_norm": 0.25929999351501465, "learning_rate": 2.8380000000000003e-05, "loss": 0.0058, "step": 1420 }, { "epoch": 0.9414088215931534, "grad_norm": 0.2439543604850769, "learning_rate": 2.858e-05, "loss": 0.0058, "step": 1430 }, { "epoch": 0.9479921000658328, "grad_norm": 0.3311539888381958, "learning_rate": 2.8780000000000002e-05, "loss": 0.0038, "step": 1440 }, { "epoch": 0.9545753785385122, "grad_norm": 0.15718941390514374, "learning_rate": 2.898e-05, "loss": 0.0103, "step": 1450 }, { "epoch": 0.9611586570111915, "grad_norm": 0.2077299803495407, "learning_rate": 2.9180000000000002e-05, "loss": 0.005, "step": 1460 }, { "epoch": 0.967741935483871, "grad_norm": 0.2793049216270447, "learning_rate": 2.9380000000000003e-05, "loss": 0.0036, "step": 1470 }, { "epoch": 0.9743252139565504, "grad_norm": 0.16314570605754852, "learning_rate": 2.958e-05, "loss": 0.0074, "step": 1480 }, { "epoch": 0.9809084924292297, "grad_norm": 0.28970712423324585, "learning_rate": 2.9780000000000003e-05, "loss": 0.0046, "step": 1490 }, { "epoch": 0.9874917709019092, "grad_norm": 0.163652703166008, "learning_rate": 2.998e-05, "loss": 0.0062, "step": 1500 }, { "epoch": 0.9940750493745886, "grad_norm": 0.08917911350727081, "learning_rate": 3.0180000000000002e-05, "loss": 0.0039, "step": 1510 }, { "epoch": 1.000658327847268, "grad_norm": 0.177604079246521, "learning_rate": 3.0380000000000004e-05, "loss": 0.0049, "step": 1520 }, { "epoch": 1.0072416063199474, "grad_norm": 0.21151091158390045, "learning_rate": 3.058e-05, "loss": 0.0045, "step": 1530 }, { "epoch": 1.0138248847926268, "grad_norm": 0.2265806794166565, "learning_rate": 3.078e-05, "loss": 0.005, "step": 1540 }, { "epoch": 1.0204081632653061, "grad_norm": 0.2583235204219818, "learning_rate": 3.0980000000000005e-05, "loss": 0.0065, "step": 1550 }, { "epoch": 1.0269914417379855, "grad_norm": 0.12779070436954498, "learning_rate": 3.118e-05, "loss": 0.0031, "step": 1560 }, { "epoch": 1.0335747202106649, "grad_norm": 0.186552032828331, "learning_rate": 3.138e-05, "loss": 0.004, "step": 1570 }, { "epoch": 1.0401579986833442, "grad_norm": 0.21109667420387268, "learning_rate": 3.1580000000000006e-05, "loss": 0.0042, "step": 1580 }, { "epoch": 1.0467412771560236, "grad_norm": 0.22061017155647278, "learning_rate": 3.1780000000000004e-05, "loss": 0.0056, "step": 1590 }, { "epoch": 1.0533245556287032, "grad_norm": 0.15494543313980103, "learning_rate": 3.198e-05, "loss": 0.0047, "step": 1600 }, { "epoch": 1.0599078341013826, "grad_norm": 0.27512314915657043, "learning_rate": 3.218e-05, "loss": 0.0071, "step": 1610 }, { "epoch": 1.066491112574062, "grad_norm": 0.14017550647258759, "learning_rate": 3.238e-05, "loss": 0.0051, "step": 1620 }, { "epoch": 1.0730743910467413, "grad_norm": 0.27555176615715027, "learning_rate": 3.2579999999999996e-05, "loss": 0.0093, "step": 1630 }, { "epoch": 1.0796576695194207, "grad_norm": 0.19666121900081635, "learning_rate": 3.278e-05, "loss": 0.0041, "step": 1640 }, { "epoch": 1.0862409479921, "grad_norm": 0.24898265302181244, "learning_rate": 3.298e-05, "loss": 0.0038, "step": 1650 }, { "epoch": 1.0928242264647794, "grad_norm": 0.23241955041885376, "learning_rate": 3.318e-05, "loss": 0.0083, "step": 1660 }, { "epoch": 1.0994075049374588, "grad_norm": 0.27159592509269714, "learning_rate": 3.338e-05, "loss": 0.004, "step": 1670 }, { "epoch": 1.1059907834101383, "grad_norm": 0.21334366500377655, "learning_rate": 3.358e-05, "loss": 0.0053, "step": 1680 }, { "epoch": 1.1125740618828177, "grad_norm": 0.15541066229343414, "learning_rate": 3.378e-05, "loss": 0.0098, "step": 1690 }, { "epoch": 1.119157340355497, "grad_norm": 0.21997252106666565, "learning_rate": 3.398e-05, "loss": 0.0042, "step": 1700 }, { "epoch": 1.1257406188281764, "grad_norm": 0.30237892270088196, "learning_rate": 3.418e-05, "loss": 0.0043, "step": 1710 }, { "epoch": 1.1323238973008558, "grad_norm": 0.24162374436855316, "learning_rate": 3.438e-05, "loss": 0.0056, "step": 1720 }, { "epoch": 1.1389071757735352, "grad_norm": 0.2436247169971466, "learning_rate": 3.4580000000000004e-05, "loss": 0.0045, "step": 1730 }, { "epoch": 1.1454904542462145, "grad_norm": 0.18317638337612152, "learning_rate": 3.478e-05, "loss": 0.0071, "step": 1740 }, { "epoch": 1.1520737327188941, "grad_norm": 0.23434752225875854, "learning_rate": 3.498e-05, "loss": 0.0042, "step": 1750 }, { "epoch": 1.1586570111915735, "grad_norm": 0.2120632380247116, "learning_rate": 3.518e-05, "loss": 0.0043, "step": 1760 }, { "epoch": 1.1652402896642529, "grad_norm": 0.2276468575000763, "learning_rate": 3.5380000000000003e-05, "loss": 0.0043, "step": 1770 }, { "epoch": 1.1718235681369322, "grad_norm": 0.17055390775203705, "learning_rate": 3.558e-05, "loss": 0.003, "step": 1780 }, { "epoch": 1.1784068466096116, "grad_norm": 0.27193719148635864, "learning_rate": 3.578e-05, "loss": 0.0059, "step": 1790 }, { "epoch": 1.184990125082291, "grad_norm": 0.21833619475364685, "learning_rate": 3.5980000000000004e-05, "loss": 0.0072, "step": 1800 }, { "epoch": 1.1915734035549703, "grad_norm": 0.2511771321296692, "learning_rate": 3.618e-05, "loss": 0.0037, "step": 1810 }, { "epoch": 1.1981566820276497, "grad_norm": 0.11682605743408203, "learning_rate": 3.638e-05, "loss": 0.0029, "step": 1820 }, { "epoch": 1.204739960500329, "grad_norm": 0.1463199406862259, "learning_rate": 3.6580000000000006e-05, "loss": 0.0048, "step": 1830 }, { "epoch": 1.2113232389730086, "grad_norm": 0.27067893743515015, "learning_rate": 3.6780000000000004e-05, "loss": 0.0049, "step": 1840 }, { "epoch": 1.217906517445688, "grad_norm": 0.11737799644470215, "learning_rate": 3.698e-05, "loss": 0.0037, "step": 1850 }, { "epoch": 1.2244897959183674, "grad_norm": 0.1815817803144455, "learning_rate": 3.7180000000000007e-05, "loss": 0.0038, "step": 1860 }, { "epoch": 1.2310730743910467, "grad_norm": 0.1771698296070099, "learning_rate": 3.7380000000000005e-05, "loss": 0.0034, "step": 1870 }, { "epoch": 1.237656352863726, "grad_norm": 0.17332947254180908, "learning_rate": 3.758e-05, "loss": 0.004, "step": 1880 }, { "epoch": 1.2442396313364055, "grad_norm": 0.2182808816432953, "learning_rate": 3.778000000000001e-05, "loss": 0.0061, "step": 1890 }, { "epoch": 1.2508229098090848, "grad_norm": 0.18819785118103027, "learning_rate": 3.7980000000000006e-05, "loss": 0.0081, "step": 1900 }, { "epoch": 1.2574061882817644, "grad_norm": 0.32689404487609863, "learning_rate": 3.818e-05, "loss": 0.0048, "step": 1910 }, { "epoch": 1.2639894667544438, "grad_norm": 0.17404583096504211, "learning_rate": 3.838e-05, "loss": 0.0029, "step": 1920 }, { "epoch": 1.2705727452271232, "grad_norm": 0.28862088918685913, "learning_rate": 3.858e-05, "loss": 0.0058, "step": 1930 }, { "epoch": 1.2771560236998025, "grad_norm": 0.09777224063873291, "learning_rate": 3.878e-05, "loss": 0.0032, "step": 1940 }, { "epoch": 1.2837393021724819, "grad_norm": 0.12917156517505646, "learning_rate": 3.898e-05, "loss": 0.0051, "step": 1950 }, { "epoch": 1.2903225806451613, "grad_norm": 0.24634969234466553, "learning_rate": 3.918e-05, "loss": 0.0047, "step": 1960 }, { "epoch": 1.2969058591178406, "grad_norm": 0.14665587246418, "learning_rate": 3.938e-05, "loss": 0.005, "step": 1970 }, { "epoch": 1.3034891375905202, "grad_norm": 0.24393214285373688, "learning_rate": 3.958e-05, "loss": 0.0097, "step": 1980 }, { "epoch": 1.3100724160631994, "grad_norm": 0.26606863737106323, "learning_rate": 3.978e-05, "loss": 0.0044, "step": 1990 }, { "epoch": 1.316655694535879, "grad_norm": 0.24726368486881256, "learning_rate": 3.998e-05, "loss": 0.0059, "step": 2000 }, { "epoch": 1.3232389730085583, "grad_norm": 0.26576346158981323, "learning_rate": 4.018e-05, "loss": 0.0049, "step": 2010 }, { "epoch": 1.3298222514812377, "grad_norm": 0.22501562535762787, "learning_rate": 4.038e-05, "loss": 0.0043, "step": 2020 }, { "epoch": 1.336405529953917, "grad_norm": 0.2221817970275879, "learning_rate": 4.058e-05, "loss": 0.0043, "step": 2030 }, { "epoch": 1.3429888084265964, "grad_norm": 0.2234380841255188, "learning_rate": 4.078e-05, "loss": 0.0039, "step": 2040 }, { "epoch": 1.3495720868992758, "grad_norm": 0.19398634135723114, "learning_rate": 4.0980000000000004e-05, "loss": 0.0038, "step": 2050 }, { "epoch": 1.3561553653719551, "grad_norm": 0.22415925562381744, "learning_rate": 4.118e-05, "loss": 0.0056, "step": 2060 }, { "epoch": 1.3627386438446347, "grad_norm": 0.1761099100112915, "learning_rate": 4.138e-05, "loss": 0.0044, "step": 2070 }, { "epoch": 1.369321922317314, "grad_norm": 0.35121166706085205, "learning_rate": 4.1580000000000005e-05, "loss": 0.0055, "step": 2080 }, { "epoch": 1.3759052007899935, "grad_norm": 0.22213824093341827, "learning_rate": 4.178e-05, "loss": 0.0039, "step": 2090 }, { "epoch": 1.3824884792626728, "grad_norm": 0.18536202609539032, "learning_rate": 4.198e-05, "loss": 0.0041, "step": 2100 }, { "epoch": 1.3890717577353522, "grad_norm": 0.18726810812950134, "learning_rate": 4.2180000000000006e-05, "loss": 0.006, "step": 2110 }, { "epoch": 1.3956550362080316, "grad_norm": 0.29389673471450806, "learning_rate": 4.2380000000000004e-05, "loss": 0.0091, "step": 2120 }, { "epoch": 1.402238314680711, "grad_norm": 0.26737287640571594, "learning_rate": 4.258e-05, "loss": 0.0071, "step": 2130 }, { "epoch": 1.4088215931533905, "grad_norm": 0.22132481634616852, "learning_rate": 4.278e-05, "loss": 0.0045, "step": 2140 }, { "epoch": 1.4154048716260696, "grad_norm": 0.1836686134338379, "learning_rate": 4.2980000000000005e-05, "loss": 0.0034, "step": 2150 }, { "epoch": 1.4219881500987492, "grad_norm": 0.17092525959014893, "learning_rate": 4.318e-05, "loss": 0.008, "step": 2160 }, { "epoch": 1.4285714285714286, "grad_norm": 0.11671039462089539, "learning_rate": 4.338e-05, "loss": 0.0028, "step": 2170 }, { "epoch": 1.435154707044108, "grad_norm": 0.1280469298362732, "learning_rate": 4.3580000000000006e-05, "loss": 0.0025, "step": 2180 }, { "epoch": 1.4417379855167873, "grad_norm": 0.18081294000148773, "learning_rate": 4.3780000000000004e-05, "loss": 0.0036, "step": 2190 }, { "epoch": 1.4483212639894667, "grad_norm": 0.17493580281734467, "learning_rate": 4.398e-05, "loss": 0.0056, "step": 2200 }, { "epoch": 1.454904542462146, "grad_norm": 0.20176613330841064, "learning_rate": 4.418000000000001e-05, "loss": 0.0064, "step": 2210 }, { "epoch": 1.4614878209348254, "grad_norm": 0.3107999861240387, "learning_rate": 4.438e-05, "loss": 0.0058, "step": 2220 }, { "epoch": 1.468071099407505, "grad_norm": 0.16614125669002533, "learning_rate": 4.458e-05, "loss": 0.0067, "step": 2230 }, { "epoch": 1.4746543778801844, "grad_norm": 0.13912639021873474, "learning_rate": 4.478e-05, "loss": 0.0058, "step": 2240 }, { "epoch": 1.4812376563528638, "grad_norm": 0.16735373437404633, "learning_rate": 4.498e-05, "loss": 0.0047, "step": 2250 }, { "epoch": 1.4878209348255431, "grad_norm": 0.19468317925930023, "learning_rate": 4.518e-05, "loss": 0.0112, "step": 2260 }, { "epoch": 1.4944042132982225, "grad_norm": 0.2687819302082062, "learning_rate": 4.538e-05, "loss": 0.0063, "step": 2270 }, { "epoch": 1.500987491770902, "grad_norm": 0.1778976023197174, "learning_rate": 4.558e-05, "loss": 0.0056, "step": 2280 }, { "epoch": 1.5075707702435812, "grad_norm": 0.23329800367355347, "learning_rate": 4.578e-05, "loss": 0.0062, "step": 2290 }, { "epoch": 1.5141540487162608, "grad_norm": 0.17943571507930756, "learning_rate": 4.5980000000000004e-05, "loss": 0.0049, "step": 2300 }, { "epoch": 1.52073732718894, "grad_norm": 0.2602374255657196, "learning_rate": 4.618e-05, "loss": 0.0039, "step": 2310 }, { "epoch": 1.5273206056616195, "grad_norm": 0.17935137450695038, "learning_rate": 4.638e-05, "loss": 0.005, "step": 2320 }, { "epoch": 1.533903884134299, "grad_norm": 0.32601723074913025, "learning_rate": 4.6580000000000005e-05, "loss": 0.004, "step": 2330 }, { "epoch": 1.5404871626069783, "grad_norm": 0.159331813454628, "learning_rate": 4.678e-05, "loss": 0.005, "step": 2340 }, { "epoch": 1.5470704410796576, "grad_norm": 0.3808997571468353, "learning_rate": 4.698e-05, "loss": 0.0058, "step": 2350 }, { "epoch": 1.553653719552337, "grad_norm": 0.25918102264404297, "learning_rate": 4.718e-05, "loss": 0.0046, "step": 2360 }, { "epoch": 1.5602369980250166, "grad_norm": 0.20101375877857208, "learning_rate": 4.7380000000000004e-05, "loss": 0.007, "step": 2370 }, { "epoch": 1.5668202764976957, "grad_norm": 0.2731170356273651, "learning_rate": 4.758e-05, "loss": 0.0054, "step": 2380 }, { "epoch": 1.5734035549703753, "grad_norm": 0.32701238989830017, "learning_rate": 4.778e-05, "loss": 0.0078, "step": 2390 }, { "epoch": 1.5799868334430547, "grad_norm": 0.2112145572900772, "learning_rate": 4.7980000000000005e-05, "loss": 0.0043, "step": 2400 }, { "epoch": 1.586570111915734, "grad_norm": 0.22174301743507385, "learning_rate": 4.818e-05, "loss": 0.0063, "step": 2410 }, { "epoch": 1.5931533903884134, "grad_norm": 0.21522150933742523, "learning_rate": 4.838e-05, "loss": 0.0064, "step": 2420 }, { "epoch": 1.5997366688610928, "grad_norm": 0.16287942230701447, "learning_rate": 4.8580000000000006e-05, "loss": 0.0044, "step": 2430 }, { "epoch": 1.6063199473337724, "grad_norm": 0.25932884216308594, "learning_rate": 4.8780000000000004e-05, "loss": 0.0042, "step": 2440 }, { "epoch": 1.6129032258064515, "grad_norm": 0.17687109112739563, "learning_rate": 4.898e-05, "loss": 0.0063, "step": 2450 }, { "epoch": 1.619486504279131, "grad_norm": 0.16744548082351685, "learning_rate": 4.918000000000001e-05, "loss": 0.0043, "step": 2460 }, { "epoch": 1.6260697827518102, "grad_norm": 0.22672729194164276, "learning_rate": 4.9380000000000005e-05, "loss": 0.004, "step": 2470 }, { "epoch": 1.6326530612244898, "grad_norm": 0.1761532574892044, "learning_rate": 4.958e-05, "loss": 0.0049, "step": 2480 }, { "epoch": 1.6392363396971692, "grad_norm": 0.290068119764328, "learning_rate": 4.978e-05, "loss": 0.0074, "step": 2490 }, { "epoch": 1.6458196181698486, "grad_norm": 0.191254660487175, "learning_rate": 4.9980000000000006e-05, "loss": 0.0082, "step": 2500 }, { "epoch": 1.652402896642528, "grad_norm": 0.24972333014011383, "learning_rate": 5.0180000000000004e-05, "loss": 0.005, "step": 2510 }, { "epoch": 1.6589861751152073, "grad_norm": 0.34614086151123047, "learning_rate": 5.038e-05, "loss": 0.005, "step": 2520 }, { "epoch": 1.6655694535878869, "grad_norm": 0.11589321494102478, "learning_rate": 5.058000000000001e-05, "loss": 0.0044, "step": 2530 }, { "epoch": 1.672152732060566, "grad_norm": 0.19550642371177673, "learning_rate": 5.0780000000000005e-05, "loss": 0.0076, "step": 2540 }, { "epoch": 1.6787360105332456, "grad_norm": 0.22189150750637054, "learning_rate": 5.098e-05, "loss": 0.0041, "step": 2550 }, { "epoch": 1.685319289005925, "grad_norm": 0.22148391604423523, "learning_rate": 5.118000000000001e-05, "loss": 0.0081, "step": 2560 }, { "epoch": 1.6919025674786043, "grad_norm": 0.21397729218006134, "learning_rate": 5.1380000000000006e-05, "loss": 0.0065, "step": 2570 }, { "epoch": 1.6984858459512837, "grad_norm": 0.17487972974777222, "learning_rate": 5.1580000000000004e-05, "loss": 0.0065, "step": 2580 }, { "epoch": 1.705069124423963, "grad_norm": 0.229610875248909, "learning_rate": 5.178000000000001e-05, "loss": 0.0042, "step": 2590 }, { "epoch": 1.7116524028966427, "grad_norm": 0.1317211389541626, "learning_rate": 5.198000000000001e-05, "loss": 0.0074, "step": 2600 }, { "epoch": 1.7182356813693218, "grad_norm": 0.2002493143081665, "learning_rate": 5.2180000000000005e-05, "loss": 0.0053, "step": 2610 }, { "epoch": 1.7248189598420014, "grad_norm": 0.19597384333610535, "learning_rate": 5.238000000000001e-05, "loss": 0.004, "step": 2620 }, { "epoch": 1.7314022383146808, "grad_norm": 0.26964783668518066, "learning_rate": 5.258000000000001e-05, "loss": 0.007, "step": 2630 }, { "epoch": 1.7379855167873601, "grad_norm": 0.2871486246585846, "learning_rate": 5.2780000000000006e-05, "loss": 0.0069, "step": 2640 }, { "epoch": 1.7445687952600395, "grad_norm": 0.23206603527069092, "learning_rate": 5.2980000000000004e-05, "loss": 0.0061, "step": 2650 }, { "epoch": 1.7511520737327189, "grad_norm": 0.190815731883049, "learning_rate": 5.318000000000001e-05, "loss": 0.0048, "step": 2660 }, { "epoch": 1.7577353522053984, "grad_norm": 0.22637800872325897, "learning_rate": 5.338000000000001e-05, "loss": 0.0067, "step": 2670 }, { "epoch": 1.7643186306780776, "grad_norm": 0.12689904868602753, "learning_rate": 5.3580000000000005e-05, "loss": 0.0039, "step": 2680 }, { "epoch": 1.7709019091507572, "grad_norm": 0.19234439730644226, "learning_rate": 5.378e-05, "loss": 0.0029, "step": 2690 }, { "epoch": 1.7774851876234363, "grad_norm": 0.16643893718719482, "learning_rate": 5.3979999999999995e-05, "loss": 0.0047, "step": 2700 }, { "epoch": 1.784068466096116, "grad_norm": 0.2089071273803711, "learning_rate": 5.418e-05, "loss": 0.0039, "step": 2710 }, { "epoch": 1.7906517445687953, "grad_norm": 0.14685404300689697, "learning_rate": 5.438e-05, "loss": 0.0053, "step": 2720 }, { "epoch": 1.7972350230414746, "grad_norm": 0.1672286093235016, "learning_rate": 5.4579999999999996e-05, "loss": 0.0025, "step": 2730 }, { "epoch": 1.803818301514154, "grad_norm": 0.17591333389282227, "learning_rate": 5.478e-05, "loss": 0.0034, "step": 2740 }, { "epoch": 1.8104015799868334, "grad_norm": 0.1953185796737671, "learning_rate": 5.498e-05, "loss": 0.0038, "step": 2750 }, { "epoch": 1.816984858459513, "grad_norm": 0.14522026479244232, "learning_rate": 5.518e-05, "loss": 0.0039, "step": 2760 }, { "epoch": 1.823568136932192, "grad_norm": 0.19128453731536865, "learning_rate": 5.538e-05, "loss": 0.0034, "step": 2770 }, { "epoch": 1.8301514154048717, "grad_norm": 0.16398826241493225, "learning_rate": 5.558e-05, "loss": 0.0035, "step": 2780 }, { "epoch": 1.836734693877551, "grad_norm": 0.29712462425231934, "learning_rate": 5.578e-05, "loss": 0.0055, "step": 2790 }, { "epoch": 1.8433179723502304, "grad_norm": 0.19872310757637024, "learning_rate": 5.5979999999999996e-05, "loss": 0.0045, "step": 2800 }, { "epoch": 1.8499012508229098, "grad_norm": 0.21916601061820984, "learning_rate": 5.618e-05, "loss": 0.004, "step": 2810 }, { "epoch": 1.8564845292955892, "grad_norm": 0.18947064876556396, "learning_rate": 5.638e-05, "loss": 0.0031, "step": 2820 }, { "epoch": 1.8630678077682687, "grad_norm": 0.17851956188678741, "learning_rate": 5.658e-05, "loss": 0.0097, "step": 2830 }, { "epoch": 1.869651086240948, "grad_norm": 0.18899938464164734, "learning_rate": 5.678e-05, "loss": 0.0047, "step": 2840 }, { "epoch": 1.8762343647136275, "grad_norm": 0.12696890532970428, "learning_rate": 5.698e-05, "loss": 0.0038, "step": 2850 }, { "epoch": 1.8828176431863066, "grad_norm": 0.2057206779718399, "learning_rate": 5.718e-05, "loss": 0.0047, "step": 2860 }, { "epoch": 1.8894009216589862, "grad_norm": 0.1901782900094986, "learning_rate": 5.738e-05, "loss": 0.0051, "step": 2870 }, { "epoch": 1.8959842001316656, "grad_norm": 0.10428407788276672, "learning_rate": 5.758e-05, "loss": 0.0039, "step": 2880 }, { "epoch": 1.902567478604345, "grad_norm": 0.29647666215896606, "learning_rate": 5.778e-05, "loss": 0.0061, "step": 2890 }, { "epoch": 1.9091507570770243, "grad_norm": 0.2557622790336609, "learning_rate": 5.7980000000000004e-05, "loss": 0.0053, "step": 2900 }, { "epoch": 1.9157340355497037, "grad_norm": 0.28392958641052246, "learning_rate": 5.818e-05, "loss": 0.0051, "step": 2910 }, { "epoch": 1.9223173140223833, "grad_norm": 0.19261664152145386, "learning_rate": 5.838e-05, "loss": 0.005, "step": 2920 }, { "epoch": 1.9289005924950624, "grad_norm": 0.17597192525863647, "learning_rate": 5.858e-05, "loss": 0.0045, "step": 2930 }, { "epoch": 1.935483870967742, "grad_norm": 0.17325453460216522, "learning_rate": 5.878e-05, "loss": 0.0056, "step": 2940 }, { "epoch": 1.9420671494404214, "grad_norm": 0.2147207111120224, "learning_rate": 5.898e-05, "loss": 0.0103, "step": 2950 }, { "epoch": 1.9486504279131007, "grad_norm": 0.2435470074415207, "learning_rate": 5.918e-05, "loss": 0.0044, "step": 2960 }, { "epoch": 1.95523370638578, "grad_norm": 0.1994706243276596, "learning_rate": 5.9380000000000004e-05, "loss": 0.0047, "step": 2970 }, { "epoch": 1.9618169848584595, "grad_norm": 0.19941651821136475, "learning_rate": 5.958e-05, "loss": 0.0038, "step": 2980 }, { "epoch": 1.968400263331139, "grad_norm": 0.21594572067260742, "learning_rate": 5.978e-05, "loss": 0.0053, "step": 2990 }, { "epoch": 1.9749835418038182, "grad_norm": 0.1423705667257309, "learning_rate": 5.9980000000000005e-05, "loss": 0.0096, "step": 3000 }, { "epoch": 1.9815668202764978, "grad_norm": 0.232273668050766, "learning_rate": 6.018e-05, "loss": 0.0047, "step": 3010 }, { "epoch": 1.9881500987491771, "grad_norm": 0.24501846730709076, "learning_rate": 6.038e-05, "loss": 0.0047, "step": 3020 }, { "epoch": 1.9947333772218565, "grad_norm": 0.2150702178478241, "learning_rate": 6.0580000000000006e-05, "loss": 0.0048, "step": 3030 }, { "epoch": 2.001316655694536, "grad_norm": 0.16417193412780762, "learning_rate": 6.0780000000000004e-05, "loss": 0.0054, "step": 3040 }, { "epoch": 2.0078999341672152, "grad_norm": 0.20016121864318848, "learning_rate": 6.098e-05, "loss": 0.0064, "step": 3050 }, { "epoch": 2.014483212639895, "grad_norm": 0.1666368693113327, "learning_rate": 6.118000000000001e-05, "loss": 0.0038, "step": 3060 }, { "epoch": 2.021066491112574, "grad_norm": 0.20008838176727295, "learning_rate": 6.138e-05, "loss": 0.004, "step": 3070 }, { "epoch": 2.0276497695852536, "grad_norm": 0.2996877133846283, "learning_rate": 6.158e-05, "loss": 0.0095, "step": 3080 }, { "epoch": 2.0342330480579327, "grad_norm": 0.18651555478572845, "learning_rate": 6.178000000000001e-05, "loss": 0.0065, "step": 3090 }, { "epoch": 2.0408163265306123, "grad_norm": 0.13422612845897675, "learning_rate": 6.198e-05, "loss": 0.0059, "step": 3100 }, { "epoch": 2.0473996050032914, "grad_norm": 0.1793995499610901, "learning_rate": 6.218e-05, "loss": 0.0042, "step": 3110 }, { "epoch": 2.053982883475971, "grad_norm": 0.27362483739852905, "learning_rate": 6.238000000000001e-05, "loss": 0.0059, "step": 3120 }, { "epoch": 2.0605661619486506, "grad_norm": 0.1651102751493454, "learning_rate": 6.258e-05, "loss": 0.004, "step": 3130 }, { "epoch": 2.0671494404213298, "grad_norm": 0.1674690842628479, "learning_rate": 6.278e-05, "loss": 0.0046, "step": 3140 }, { "epoch": 2.0737327188940093, "grad_norm": 0.2099774330854416, "learning_rate": 6.298000000000001e-05, "loss": 0.0072, "step": 3150 }, { "epoch": 2.0803159973666885, "grad_norm": 0.19672340154647827, "learning_rate": 6.318e-05, "loss": 0.0031, "step": 3160 }, { "epoch": 2.086899275839368, "grad_norm": 0.1912270188331604, "learning_rate": 6.338e-05, "loss": 0.006, "step": 3170 }, { "epoch": 2.093482554312047, "grad_norm": 0.21891021728515625, "learning_rate": 6.358000000000001e-05, "loss": 0.0044, "step": 3180 }, { "epoch": 2.100065832784727, "grad_norm": 0.17274227738380432, "learning_rate": 6.378e-05, "loss": 0.0034, "step": 3190 }, { "epoch": 2.1066491112574064, "grad_norm": 0.1768108606338501, "learning_rate": 6.398000000000001e-05, "loss": 0.0051, "step": 3200 }, { "epoch": 2.1132323897300855, "grad_norm": 0.20513848960399628, "learning_rate": 6.418000000000001e-05, "loss": 0.0044, "step": 3210 }, { "epoch": 2.119815668202765, "grad_norm": 0.1431579291820526, "learning_rate": 6.438e-05, "loss": 0.004, "step": 3220 }, { "epoch": 2.1263989466754443, "grad_norm": 0.13859818875789642, "learning_rate": 6.458000000000001e-05, "loss": 0.0034, "step": 3230 }, { "epoch": 2.132982225148124, "grad_norm": 0.2133711278438568, "learning_rate": 6.478000000000001e-05, "loss": 0.0036, "step": 3240 }, { "epoch": 2.139565503620803, "grad_norm": 0.18273620307445526, "learning_rate": 6.498e-05, "loss": 0.0051, "step": 3250 }, { "epoch": 2.1461487820934826, "grad_norm": 0.23271100223064423, "learning_rate": 6.518000000000001e-05, "loss": 0.0041, "step": 3260 }, { "epoch": 2.152732060566162, "grad_norm": 0.17983755469322205, "learning_rate": 6.538000000000001e-05, "loss": 0.0051, "step": 3270 }, { "epoch": 2.1593153390388413, "grad_norm": 0.225615993142128, "learning_rate": 6.558e-05, "loss": 0.0041, "step": 3280 }, { "epoch": 2.165898617511521, "grad_norm": 0.16896314918994904, "learning_rate": 6.578000000000001e-05, "loss": 0.0046, "step": 3290 }, { "epoch": 2.1724818959842, "grad_norm": 0.18113262951374054, "learning_rate": 6.598e-05, "loss": 0.0061, "step": 3300 }, { "epoch": 2.1790651744568796, "grad_norm": 0.2417910099029541, "learning_rate": 6.618e-05, "loss": 0.0042, "step": 3310 }, { "epoch": 2.185648452929559, "grad_norm": 0.1686246246099472, "learning_rate": 6.638e-05, "loss": 0.0037, "step": 3320 }, { "epoch": 2.1922317314022384, "grad_norm": 0.2682899832725525, "learning_rate": 6.658e-05, "loss": 0.0069, "step": 3330 }, { "epoch": 2.1988150098749175, "grad_norm": 0.2569248378276825, "learning_rate": 6.678e-05, "loss": 0.0052, "step": 3340 }, { "epoch": 2.205398288347597, "grad_norm": 0.2747572064399719, "learning_rate": 6.698e-05, "loss": 0.0063, "step": 3350 }, { "epoch": 2.2119815668202767, "grad_norm": 0.12319736182689667, "learning_rate": 6.718e-05, "loss": 0.0059, "step": 3360 }, { "epoch": 2.218564845292956, "grad_norm": 0.2230660319328308, "learning_rate": 6.738e-05, "loss": 0.0075, "step": 3370 }, { "epoch": 2.2251481237656354, "grad_norm": 0.222366064786911, "learning_rate": 6.758e-05, "loss": 0.0045, "step": 3380 }, { "epoch": 2.2317314022383146, "grad_norm": 0.2719450294971466, "learning_rate": 6.778e-05, "loss": 0.0056, "step": 3390 }, { "epoch": 2.238314680710994, "grad_norm": 0.24775299429893494, "learning_rate": 6.798e-05, "loss": 0.005, "step": 3400 }, { "epoch": 2.2448979591836733, "grad_norm": 0.2985354959964752, "learning_rate": 6.818e-05, "loss": 0.0055, "step": 3410 }, { "epoch": 2.251481237656353, "grad_norm": 0.1562320590019226, "learning_rate": 6.838e-05, "loss": 0.0044, "step": 3420 }, { "epoch": 2.258064516129032, "grad_norm": 0.22354333102703094, "learning_rate": 6.858e-05, "loss": 0.0044, "step": 3430 }, { "epoch": 2.2646477946017116, "grad_norm": 0.176690012216568, "learning_rate": 6.878e-05, "loss": 0.0078, "step": 3440 }, { "epoch": 2.271231073074391, "grad_norm": 0.18620379269123077, "learning_rate": 6.898e-05, "loss": 0.0054, "step": 3450 }, { "epoch": 2.2778143515470703, "grad_norm": 0.2027626931667328, "learning_rate": 6.918e-05, "loss": 0.0051, "step": 3460 }, { "epoch": 2.28439763001975, "grad_norm": 0.19690103828907013, "learning_rate": 6.938e-05, "loss": 0.004, "step": 3470 }, { "epoch": 2.290980908492429, "grad_norm": 0.16586609184741974, "learning_rate": 6.958e-05, "loss": 0.0033, "step": 3480 }, { "epoch": 2.2975641869651087, "grad_norm": 0.14822760224342346, "learning_rate": 6.978e-05, "loss": 0.0079, "step": 3490 }, { "epoch": 2.3041474654377883, "grad_norm": 0.24322155117988586, "learning_rate": 6.998e-05, "loss": 0.0044, "step": 3500 }, { "epoch": 2.3107307439104674, "grad_norm": 0.21890296041965485, "learning_rate": 7.018e-05, "loss": 0.0072, "step": 3510 }, { "epoch": 2.317314022383147, "grad_norm": 0.16870874166488647, "learning_rate": 7.038e-05, "loss": 0.0063, "step": 3520 }, { "epoch": 2.323897300855826, "grad_norm": 0.2116502821445465, "learning_rate": 7.058e-05, "loss": 0.0033, "step": 3530 }, { "epoch": 2.3304805793285057, "grad_norm": 0.3590920865535736, "learning_rate": 7.078e-05, "loss": 0.0085, "step": 3540 }, { "epoch": 2.337063857801185, "grad_norm": 0.28117936849594116, "learning_rate": 7.098e-05, "loss": 0.0048, "step": 3550 }, { "epoch": 2.3436471362738645, "grad_norm": 0.15749283134937286, "learning_rate": 7.118e-05, "loss": 0.0051, "step": 3560 }, { "epoch": 2.3502304147465436, "grad_norm": 0.17668424546718597, "learning_rate": 7.138e-05, "loss": 0.0122, "step": 3570 }, { "epoch": 2.356813693219223, "grad_norm": 0.2331148087978363, "learning_rate": 7.158e-05, "loss": 0.0074, "step": 3580 }, { "epoch": 2.3633969716919028, "grad_norm": 0.2188708633184433, "learning_rate": 7.178000000000001e-05, "loss": 0.0049, "step": 3590 }, { "epoch": 2.369980250164582, "grad_norm": 0.24049866199493408, "learning_rate": 7.198e-05, "loss": 0.0054, "step": 3600 }, { "epoch": 2.3765635286372615, "grad_norm": 0.22496403753757477, "learning_rate": 7.218e-05, "loss": 0.0048, "step": 3610 }, { "epoch": 2.3831468071099406, "grad_norm": 0.18597084283828735, "learning_rate": 7.238000000000001e-05, "loss": 0.0037, "step": 3620 }, { "epoch": 2.3897300855826202, "grad_norm": 0.16211439669132233, "learning_rate": 7.258e-05, "loss": 0.0041, "step": 3630 }, { "epoch": 2.3963133640552994, "grad_norm": 0.21170586347579956, "learning_rate": 7.278e-05, "loss": 0.0033, "step": 3640 }, { "epoch": 2.402896642527979, "grad_norm": 0.14216628670692444, "learning_rate": 7.298000000000001e-05, "loss": 0.0041, "step": 3650 }, { "epoch": 2.409479921000658, "grad_norm": 0.11140528321266174, "learning_rate": 7.318e-05, "loss": 0.0032, "step": 3660 }, { "epoch": 2.4160631994733377, "grad_norm": 0.19434642791748047, "learning_rate": 7.338e-05, "loss": 0.0048, "step": 3670 }, { "epoch": 2.4226464779460173, "grad_norm": 0.17758332192897797, "learning_rate": 7.358000000000001e-05, "loss": 0.0034, "step": 3680 }, { "epoch": 2.4292297564186964, "grad_norm": 0.13901595771312714, "learning_rate": 7.378e-05, "loss": 0.0034, "step": 3690 }, { "epoch": 2.435813034891376, "grad_norm": 0.2104514092206955, "learning_rate": 7.398e-05, "loss": 0.004, "step": 3700 }, { "epoch": 2.442396313364055, "grad_norm": 0.16535179316997528, "learning_rate": 7.418000000000001e-05, "loss": 0.0053, "step": 3710 }, { "epoch": 2.4489795918367347, "grad_norm": 0.21833057701587677, "learning_rate": 7.438e-05, "loss": 0.0056, "step": 3720 }, { "epoch": 2.4555628703094143, "grad_norm": 0.18802616000175476, "learning_rate": 7.458000000000001e-05, "loss": 0.0042, "step": 3730 }, { "epoch": 2.4621461487820935, "grad_norm": 0.38769403100013733, "learning_rate": 7.478e-05, "loss": 0.008, "step": 3740 }, { "epoch": 2.468729427254773, "grad_norm": 0.1622026115655899, "learning_rate": 7.498e-05, "loss": 0.0045, "step": 3750 }, { "epoch": 2.475312705727452, "grad_norm": 0.14058834314346313, "learning_rate": 7.518000000000001e-05, "loss": 0.0029, "step": 3760 }, { "epoch": 2.481895984200132, "grad_norm": 0.2033497840166092, "learning_rate": 7.538e-05, "loss": 0.0038, "step": 3770 }, { "epoch": 2.488479262672811, "grad_norm": 0.142676442861557, "learning_rate": 7.558e-05, "loss": 0.0035, "step": 3780 }, { "epoch": 2.4950625411454905, "grad_norm": 0.21690382063388824, "learning_rate": 7.578000000000001e-05, "loss": 0.0057, "step": 3790 }, { "epoch": 2.5016458196181697, "grad_norm": 0.16091883182525635, "learning_rate": 7.598e-05, "loss": 0.0073, "step": 3800 }, { "epoch": 2.5082290980908493, "grad_norm": 0.17600563168525696, "learning_rate": 7.618e-05, "loss": 0.0056, "step": 3810 }, { "epoch": 2.514812376563529, "grad_norm": 0.24199329316616058, "learning_rate": 7.638000000000001e-05, "loss": 0.005, "step": 3820 }, { "epoch": 2.521395655036208, "grad_norm": 0.1430000215768814, "learning_rate": 7.658e-05, "loss": 0.0033, "step": 3830 }, { "epoch": 2.5279789335088876, "grad_norm": 0.16358165442943573, "learning_rate": 7.678000000000001e-05, "loss": 0.0047, "step": 3840 }, { "epoch": 2.5345622119815667, "grad_norm": 0.2060249149799347, "learning_rate": 7.698000000000001e-05, "loss": 0.0033, "step": 3850 }, { "epoch": 2.5411454904542463, "grad_norm": 0.1935313493013382, "learning_rate": 7.718e-05, "loss": 0.0062, "step": 3860 }, { "epoch": 2.5477287689269255, "grad_norm": 0.12891928851604462, "learning_rate": 7.738000000000001e-05, "loss": 0.0056, "step": 3870 }, { "epoch": 2.554312047399605, "grad_norm": 0.2203224003314972, "learning_rate": 7.758000000000001e-05, "loss": 0.0048, "step": 3880 }, { "epoch": 2.560895325872284, "grad_norm": 0.32686325907707214, "learning_rate": 7.778e-05, "loss": 0.0072, "step": 3890 }, { "epoch": 2.5674786043449638, "grad_norm": 0.2307060956954956, "learning_rate": 7.798000000000001e-05, "loss": 0.0066, "step": 3900 }, { "epoch": 2.5740618828176434, "grad_norm": 0.2252025306224823, "learning_rate": 7.818000000000001e-05, "loss": 0.0039, "step": 3910 }, { "epoch": 2.5806451612903225, "grad_norm": 0.3624524474143982, "learning_rate": 7.838e-05, "loss": 0.0052, "step": 3920 }, { "epoch": 2.587228439763002, "grad_norm": 0.32195788621902466, "learning_rate": 7.858000000000001e-05, "loss": 0.005, "step": 3930 }, { "epoch": 2.5938117182356812, "grad_norm": 0.2710541784763336, "learning_rate": 7.878e-05, "loss": 0.0057, "step": 3940 }, { "epoch": 2.600394996708361, "grad_norm": 0.20396988093852997, "learning_rate": 7.897999999999999e-05, "loss": 0.0067, "step": 3950 }, { "epoch": 2.6069782751810404, "grad_norm": 0.2155308723449707, "learning_rate": 7.918e-05, "loss": 0.0087, "step": 3960 }, { "epoch": 2.6135615536537196, "grad_norm": 0.2822481691837311, "learning_rate": 7.938e-05, "loss": 0.0059, "step": 3970 }, { "epoch": 2.6201448321263987, "grad_norm": 0.1881878525018692, "learning_rate": 7.958e-05, "loss": 0.0048, "step": 3980 }, { "epoch": 2.6267281105990783, "grad_norm": 0.17815779149532318, "learning_rate": 7.978e-05, "loss": 0.004, "step": 3990 }, { "epoch": 2.633311389071758, "grad_norm": 0.1446467787027359, "learning_rate": 7.998e-05, "loss": 0.003, "step": 4000 }, { "epoch": 2.639894667544437, "grad_norm": 0.13284584879875183, "learning_rate": 8.018e-05, "loss": 0.0035, "step": 4010 }, { "epoch": 2.6464779460171166, "grad_norm": 0.16675715148448944, "learning_rate": 8.038e-05, "loss": 0.0027, "step": 4020 }, { "epoch": 2.6530612244897958, "grad_norm": 0.2089269757270813, "learning_rate": 8.058e-05, "loss": 0.0037, "step": 4030 }, { "epoch": 2.6596445029624753, "grad_norm": 0.2351311892271042, "learning_rate": 8.078e-05, "loss": 0.0071, "step": 4040 }, { "epoch": 2.666227781435155, "grad_norm": 0.2488633394241333, "learning_rate": 8.098e-05, "loss": 0.006, "step": 4050 }, { "epoch": 2.672811059907834, "grad_norm": 0.128318190574646, "learning_rate": 8.118e-05, "loss": 0.0072, "step": 4060 }, { "epoch": 2.679394338380513, "grad_norm": 0.288984090089798, "learning_rate": 8.138e-05, "loss": 0.0107, "step": 4070 }, { "epoch": 2.685977616853193, "grad_norm": 0.23555997014045715, "learning_rate": 8.158e-05, "loss": 0.0066, "step": 4080 }, { "epoch": 2.6925608953258724, "grad_norm": 0.1649252325296402, "learning_rate": 8.178e-05, "loss": 0.0049, "step": 4090 }, { "epoch": 2.6991441737985515, "grad_norm": 0.14338445663452148, "learning_rate": 8.198e-05, "loss": 0.005, "step": 4100 }, { "epoch": 2.705727452271231, "grad_norm": 0.13963057100772858, "learning_rate": 8.218e-05, "loss": 0.004, "step": 4110 }, { "epoch": 2.7123107307439103, "grad_norm": 0.2351156324148178, "learning_rate": 8.238000000000001e-05, "loss": 0.0045, "step": 4120 }, { "epoch": 2.71889400921659, "grad_norm": 0.1248876079916954, "learning_rate": 8.258e-05, "loss": 0.0033, "step": 4130 }, { "epoch": 2.7254772876892694, "grad_norm": 0.29248160123825073, "learning_rate": 8.278e-05, "loss": 0.0065, "step": 4140 }, { "epoch": 2.7320605661619486, "grad_norm": 0.12926040589809418, "learning_rate": 8.298000000000001e-05, "loss": 0.0043, "step": 4150 }, { "epoch": 2.738643844634628, "grad_norm": 0.14367802441120148, "learning_rate": 8.318e-05, "loss": 0.0027, "step": 4160 }, { "epoch": 2.7452271231073073, "grad_norm": 0.11659074574708939, "learning_rate": 8.338e-05, "loss": 0.0038, "step": 4170 }, { "epoch": 2.751810401579987, "grad_norm": 0.18045160174369812, "learning_rate": 8.358e-05, "loss": 0.0038, "step": 4180 }, { "epoch": 2.7583936800526665, "grad_norm": 0.18401756882667542, "learning_rate": 8.378e-05, "loss": 0.004, "step": 4190 }, { "epoch": 2.7649769585253456, "grad_norm": 0.16460269689559937, "learning_rate": 8.398e-05, "loss": 0.0031, "step": 4200 }, { "epoch": 2.771560236998025, "grad_norm": 0.11592137813568115, "learning_rate": 8.418e-05, "loss": 0.0032, "step": 4210 }, { "epoch": 2.7781435154707044, "grad_norm": 0.16036009788513184, "learning_rate": 8.438e-05, "loss": 0.0032, "step": 4220 }, { "epoch": 2.784726793943384, "grad_norm": 0.15967459976673126, "learning_rate": 8.458e-05, "loss": 0.0033, "step": 4230 }, { "epoch": 2.791310072416063, "grad_norm": 0.21757715940475464, "learning_rate": 8.478e-05, "loss": 0.0034, "step": 4240 }, { "epoch": 2.7978933508887427, "grad_norm": 0.20688943564891815, "learning_rate": 8.498e-05, "loss": 0.0056, "step": 4250 }, { "epoch": 2.804476629361422, "grad_norm": 0.1384371519088745, "learning_rate": 8.518000000000001e-05, "loss": 0.0027, "step": 4260 }, { "epoch": 2.8110599078341014, "grad_norm": 0.23236291110515594, "learning_rate": 8.538e-05, "loss": 0.0047, "step": 4270 }, { "epoch": 2.817643186306781, "grad_norm": 0.15110521018505096, "learning_rate": 8.558e-05, "loss": 0.005, "step": 4280 }, { "epoch": 2.82422646477946, "grad_norm": 0.1877370923757553, "learning_rate": 8.578000000000001e-05, "loss": 0.0099, "step": 4290 }, { "epoch": 2.8308097432521393, "grad_norm": 0.19951312243938446, "learning_rate": 8.598e-05, "loss": 0.0045, "step": 4300 }, { "epoch": 2.837393021724819, "grad_norm": 0.30475932359695435, "learning_rate": 8.618e-05, "loss": 0.0057, "step": 4310 }, { "epoch": 2.8439763001974985, "grad_norm": 0.17791634798049927, "learning_rate": 8.638000000000001e-05, "loss": 0.0039, "step": 4320 }, { "epoch": 2.8505595786701776, "grad_norm": 0.1921178251504898, "learning_rate": 8.658e-05, "loss": 0.005, "step": 4330 }, { "epoch": 2.857142857142857, "grad_norm": 0.14874717593193054, "learning_rate": 8.678e-05, "loss": 0.0029, "step": 4340 }, { "epoch": 2.8637261356155364, "grad_norm": 0.1675635427236557, "learning_rate": 8.698000000000001e-05, "loss": 0.0028, "step": 4350 }, { "epoch": 2.870309414088216, "grad_norm": 0.32213497161865234, "learning_rate": 8.718e-05, "loss": 0.008, "step": 4360 }, { "epoch": 2.8768926925608955, "grad_norm": 0.18042099475860596, "learning_rate": 8.738000000000001e-05, "loss": 0.004, "step": 4370 }, { "epoch": 2.8834759710335747, "grad_norm": 0.1826370656490326, "learning_rate": 8.758000000000001e-05, "loss": 0.0053, "step": 4380 }, { "epoch": 2.8900592495062543, "grad_norm": 0.17529167234897614, "learning_rate": 8.778e-05, "loss": 0.0087, "step": 4390 }, { "epoch": 2.8966425279789334, "grad_norm": 0.14037705957889557, "learning_rate": 8.798000000000001e-05, "loss": 0.0031, "step": 4400 }, { "epoch": 2.903225806451613, "grad_norm": 0.1587187796831131, "learning_rate": 8.818000000000001e-05, "loss": 0.004, "step": 4410 }, { "epoch": 2.909809084924292, "grad_norm": 0.21903245151042938, "learning_rate": 8.838e-05, "loss": 0.0042, "step": 4420 }, { "epoch": 2.9163923633969717, "grad_norm": 0.18052007257938385, "learning_rate": 8.858000000000001e-05, "loss": 0.0033, "step": 4430 }, { "epoch": 2.922975641869651, "grad_norm": 0.19172020256519318, "learning_rate": 8.878000000000001e-05, "loss": 0.0045, "step": 4440 }, { "epoch": 2.9295589203423305, "grad_norm": 0.16095320880413055, "learning_rate": 8.898e-05, "loss": 0.0059, "step": 4450 }, { "epoch": 2.93614219881501, "grad_norm": 0.28082239627838135, "learning_rate": 8.918000000000001e-05, "loss": 0.0056, "step": 4460 }, { "epoch": 2.942725477287689, "grad_norm": 0.18420974910259247, "learning_rate": 8.938e-05, "loss": 0.0045, "step": 4470 }, { "epoch": 2.9493087557603688, "grad_norm": 0.16270877420902252, "learning_rate": 8.958e-05, "loss": 0.0037, "step": 4480 }, { "epoch": 2.955892034233048, "grad_norm": 0.20542408525943756, "learning_rate": 8.978000000000001e-05, "loss": 0.004, "step": 4490 }, { "epoch": 2.9624753127057275, "grad_norm": 0.20075422525405884, "learning_rate": 8.998e-05, "loss": 0.0093, "step": 4500 }, { "epoch": 2.969058591178407, "grad_norm": 0.19178062677383423, "learning_rate": 9.018000000000001e-05, "loss": 0.0044, "step": 4510 }, { "epoch": 2.9756418696510862, "grad_norm": 0.16733354330062866, "learning_rate": 9.038000000000001e-05, "loss": 0.0069, "step": 4520 }, { "epoch": 2.9822251481237654, "grad_norm": 0.23701581358909607, "learning_rate": 9.058e-05, "loss": 0.0062, "step": 4530 }, { "epoch": 2.988808426596445, "grad_norm": 0.2159680277109146, "learning_rate": 9.078000000000001e-05, "loss": 0.0028, "step": 4540 }, { "epoch": 2.9953917050691246, "grad_norm": 0.205041766166687, "learning_rate": 9.098000000000001e-05, "loss": 0.0084, "step": 4550 }, { "epoch": 3.0019749835418037, "grad_norm": 0.26218873262405396, "learning_rate": 9.118e-05, "loss": 0.0046, "step": 4560 }, { "epoch": 3.0085582620144833, "grad_norm": 0.23402182757854462, "learning_rate": 9.138e-05, "loss": 0.0059, "step": 4570 }, { "epoch": 3.0151415404871624, "grad_norm": 0.16684837639331818, "learning_rate": 9.158e-05, "loss": 0.0054, "step": 4580 }, { "epoch": 3.021724818959842, "grad_norm": 0.1614120602607727, "learning_rate": 9.178e-05, "loss": 0.0043, "step": 4590 }, { "epoch": 3.0283080974325216, "grad_norm": 0.18508388102054596, "learning_rate": 9.198e-05, "loss": 0.0045, "step": 4600 }, { "epoch": 3.0348913759052007, "grad_norm": 0.2683901786804199, "learning_rate": 9.218e-05, "loss": 0.0083, "step": 4610 }, { "epoch": 3.0414746543778803, "grad_norm": 0.15951071679592133, "learning_rate": 9.238e-05, "loss": 0.0085, "step": 4620 }, { "epoch": 3.0480579328505595, "grad_norm": 0.18083246052265167, "learning_rate": 9.258e-05, "loss": 0.0044, "step": 4630 }, { "epoch": 3.054641211323239, "grad_norm": 0.16274961829185486, "learning_rate": 9.278e-05, "loss": 0.0053, "step": 4640 }, { "epoch": 3.061224489795918, "grad_norm": 0.3050588369369507, "learning_rate": 9.298e-05, "loss": 0.0146, "step": 4650 }, { "epoch": 3.067807768268598, "grad_norm": 0.3435303866863251, "learning_rate": 9.318e-05, "loss": 0.0122, "step": 4660 }, { "epoch": 3.074391046741277, "grad_norm": 0.24383644759655, "learning_rate": 9.338e-05, "loss": 0.0068, "step": 4670 }, { "epoch": 3.0809743252139565, "grad_norm": 0.2120666801929474, "learning_rate": 9.358e-05, "loss": 0.0047, "step": 4680 }, { "epoch": 3.087557603686636, "grad_norm": 0.255738765001297, "learning_rate": 9.378e-05, "loss": 0.011, "step": 4690 }, { "epoch": 3.0941408821593153, "grad_norm": 0.3285175561904907, "learning_rate": 9.398e-05, "loss": 0.0095, "step": 4700 }, { "epoch": 3.100724160631995, "grad_norm": 0.23807333409786224, "learning_rate": 9.418e-05, "loss": 0.0071, "step": 4710 }, { "epoch": 3.107307439104674, "grad_norm": 0.2007445991039276, "learning_rate": 9.438e-05, "loss": 0.0042, "step": 4720 }, { "epoch": 3.1138907175773536, "grad_norm": 0.216136172413826, "learning_rate": 9.458e-05, "loss": 0.0042, "step": 4730 }, { "epoch": 3.1204739960500327, "grad_norm": 0.25494450330734253, "learning_rate": 9.478e-05, "loss": 0.0086, "step": 4740 }, { "epoch": 3.1270572745227123, "grad_norm": 0.23773279786109924, "learning_rate": 9.498e-05, "loss": 0.0053, "step": 4750 }, { "epoch": 3.133640552995392, "grad_norm": 0.3192708492279053, "learning_rate": 9.518000000000001e-05, "loss": 0.0066, "step": 4760 }, { "epoch": 3.140223831468071, "grad_norm": 0.36344388127326965, "learning_rate": 9.538e-05, "loss": 0.0084, "step": 4770 }, { "epoch": 3.1468071099407506, "grad_norm": 0.33147770166397095, "learning_rate": 9.558e-05, "loss": 0.0076, "step": 4780 }, { "epoch": 3.15339038841343, "grad_norm": 0.2828924059867859, "learning_rate": 9.578000000000001e-05, "loss": 0.01, "step": 4790 }, { "epoch": 3.1599736668861094, "grad_norm": 0.3122549057006836, "learning_rate": 9.598e-05, "loss": 0.0055, "step": 4800 }, { "epoch": 3.1665569453587885, "grad_norm": 0.15197250247001648, "learning_rate": 9.618e-05, "loss": 0.0055, "step": 4810 }, { "epoch": 3.173140223831468, "grad_norm": 0.2090052217245102, "learning_rate": 9.638000000000001e-05, "loss": 0.0056, "step": 4820 }, { "epoch": 3.1797235023041477, "grad_norm": 0.25107434391975403, "learning_rate": 9.658e-05, "loss": 0.0056, "step": 4830 }, { "epoch": 3.186306780776827, "grad_norm": 0.20580177009105682, "learning_rate": 9.678e-05, "loss": 0.0044, "step": 4840 }, { "epoch": 3.1928900592495064, "grad_norm": 0.1427846997976303, "learning_rate": 9.698000000000001e-05, "loss": 0.0035, "step": 4850 }, { "epoch": 3.1994733377221856, "grad_norm": 0.25326648354530334, "learning_rate": 9.718e-05, "loss": 0.004, "step": 4860 }, { "epoch": 3.206056616194865, "grad_norm": 0.19157849252223969, "learning_rate": 9.738e-05, "loss": 0.0049, "step": 4870 }, { "epoch": 3.2126398946675443, "grad_norm": 0.17050795257091522, "learning_rate": 9.758000000000001e-05, "loss": 0.004, "step": 4880 }, { "epoch": 3.219223173140224, "grad_norm": 0.21727782487869263, "learning_rate": 9.778e-05, "loss": 0.0041, "step": 4890 }, { "epoch": 3.225806451612903, "grad_norm": 0.168072909116745, "learning_rate": 9.798000000000001e-05, "loss": 0.0033, "step": 4900 }, { "epoch": 3.2323897300855826, "grad_norm": 0.22176411747932434, "learning_rate": 9.818000000000001e-05, "loss": 0.0053, "step": 4910 }, { "epoch": 3.238973008558262, "grad_norm": 0.20157967507839203, "learning_rate": 9.838e-05, "loss": 0.0064, "step": 4920 }, { "epoch": 3.2455562870309413, "grad_norm": 0.19378389418125153, "learning_rate": 9.858000000000001e-05, "loss": 0.0058, "step": 4930 }, { "epoch": 3.252139565503621, "grad_norm": 0.21381482481956482, "learning_rate": 9.878e-05, "loss": 0.005, "step": 4940 }, { "epoch": 3.2587228439763, "grad_norm": 0.21356940269470215, "learning_rate": 9.898e-05, "loss": 0.0038, "step": 4950 }, { "epoch": 3.2653061224489797, "grad_norm": 0.24294263124465942, "learning_rate": 9.918000000000001e-05, "loss": 0.006, "step": 4960 }, { "epoch": 3.271889400921659, "grad_norm": 0.19664910435676575, "learning_rate": 9.938e-05, "loss": 0.0042, "step": 4970 }, { "epoch": 3.2784726793943384, "grad_norm": 0.20534099638462067, "learning_rate": 9.958e-05, "loss": 0.003, "step": 4980 }, { "epoch": 3.2850559578670175, "grad_norm": 0.15425042808055878, "learning_rate": 9.978000000000001e-05, "loss": 0.0051, "step": 4990 }, { "epoch": 3.291639236339697, "grad_norm": 0.20609745383262634, "learning_rate": 9.998e-05, "loss": 0.0045, "step": 5000 }, { "epoch": 3.2982225148123767, "grad_norm": 0.25262415409088135, "learning_rate": 9.999999778549045e-05, "loss": 0.0059, "step": 5010 }, { "epoch": 3.304805793285056, "grad_norm": 0.18249893188476562, "learning_rate": 9.999999013039593e-05, "loss": 0.0046, "step": 5020 }, { "epoch": 3.3113890717577354, "grad_norm": 0.23900967836380005, "learning_rate": 9.999997700737766e-05, "loss": 0.0061, "step": 5030 }, { "epoch": 3.3179723502304146, "grad_norm": 0.1903785616159439, "learning_rate": 9.999995841643709e-05, "loss": 0.0038, "step": 5040 }, { "epoch": 3.324555628703094, "grad_norm": 0.2228005975484848, "learning_rate": 9.999993435757623e-05, "loss": 0.0046, "step": 5050 }, { "epoch": 3.3311389071757738, "grad_norm": 0.13866274058818817, "learning_rate": 9.999990483079773e-05, "loss": 0.0033, "step": 5060 }, { "epoch": 3.337722185648453, "grad_norm": 0.19528432190418243, "learning_rate": 9.999986983610481e-05, "loss": 0.0047, "step": 5070 }, { "epoch": 3.3443054641211325, "grad_norm": 0.18838606774806976, "learning_rate": 9.99998293735013e-05, "loss": 0.0055, "step": 5080 }, { "epoch": 3.3508887425938116, "grad_norm": 0.1896565854549408, "learning_rate": 9.999978344299161e-05, "loss": 0.0056, "step": 5090 }, { "epoch": 3.3574720210664912, "grad_norm": 0.2896604537963867, "learning_rate": 9.99997320445808e-05, "loss": 0.0067, "step": 5100 }, { "epoch": 3.3640552995391704, "grad_norm": 0.15219822525978088, "learning_rate": 9.999967517827444e-05, "loss": 0.0043, "step": 5110 }, { "epoch": 3.37063857801185, "grad_norm": 0.16526977717876434, "learning_rate": 9.999961284407879e-05, "loss": 0.0059, "step": 5120 }, { "epoch": 3.377221856484529, "grad_norm": 0.20810434222221375, "learning_rate": 9.999954504200067e-05, "loss": 0.0033, "step": 5130 }, { "epoch": 3.3838051349572087, "grad_norm": 0.13080564141273499, "learning_rate": 9.999947177204744e-05, "loss": 0.0045, "step": 5140 }, { "epoch": 3.3903884134298883, "grad_norm": 0.19555552303791046, "learning_rate": 9.999939303422718e-05, "loss": 0.0051, "step": 5150 }, { "epoch": 3.3969716919025674, "grad_norm": 0.195923313498497, "learning_rate": 9.999930882854847e-05, "loss": 0.0038, "step": 5160 }, { "epoch": 3.403554970375247, "grad_norm": 0.24417954683303833, "learning_rate": 9.999921915502051e-05, "loss": 0.004, "step": 5170 }, { "epoch": 3.410138248847926, "grad_norm": 0.10848059505224228, "learning_rate": 9.99991240136531e-05, "loss": 0.0044, "step": 5180 }, { "epoch": 3.4167215273206057, "grad_norm": 0.17478244006633759, "learning_rate": 9.999902340445668e-05, "loss": 0.0044, "step": 5190 }, { "epoch": 3.423304805793285, "grad_norm": 0.24914295971393585, "learning_rate": 9.999891732744224e-05, "loss": 0.0042, "step": 5200 }, { "epoch": 3.4298880842659645, "grad_norm": 0.15586888790130615, "learning_rate": 9.999880578262135e-05, "loss": 0.0052, "step": 5210 }, { "epoch": 3.4364713627386436, "grad_norm": 0.17657938599586487, "learning_rate": 9.999868877000624e-05, "loss": 0.0031, "step": 5220 }, { "epoch": 3.443054641211323, "grad_norm": 0.27502259612083435, "learning_rate": 9.99985662896097e-05, "loss": 0.0051, "step": 5230 }, { "epoch": 3.449637919684003, "grad_norm": 0.12207293510437012, "learning_rate": 9.999843834144513e-05, "loss": 0.0034, "step": 5240 }, { "epoch": 3.456221198156682, "grad_norm": 0.13693022727966309, "learning_rate": 9.99983049255265e-05, "loss": 0.0025, "step": 5250 }, { "epoch": 3.4628044766293615, "grad_norm": 0.16259245574474335, "learning_rate": 9.999816604186843e-05, "loss": 0.0046, "step": 5260 }, { "epoch": 3.4693877551020407, "grad_norm": 0.14142292737960815, "learning_rate": 9.999802169048609e-05, "loss": 0.0036, "step": 5270 }, { "epoch": 3.4759710335747203, "grad_norm": 0.13523933291435242, "learning_rate": 9.999787187139527e-05, "loss": 0.002, "step": 5280 }, { "epoch": 3.4825543120474, "grad_norm": 0.19108827412128448, "learning_rate": 9.999771658461234e-05, "loss": 0.0031, "step": 5290 }, { "epoch": 3.489137590520079, "grad_norm": 0.1561547964811325, "learning_rate": 9.999755583015431e-05, "loss": 0.0061, "step": 5300 }, { "epoch": 3.4957208689927586, "grad_norm": 0.2339964061975479, "learning_rate": 9.999738960803874e-05, "loss": 0.0119, "step": 5310 }, { "epoch": 3.5023041474654377, "grad_norm": 0.298492968082428, "learning_rate": 9.99972179182838e-05, "loss": 0.006, "step": 5320 }, { "epoch": 3.5088874259381173, "grad_norm": 0.27646687626838684, "learning_rate": 9.99970407609083e-05, "loss": 0.0078, "step": 5330 }, { "epoch": 3.5154707044107965, "grad_norm": 0.20261631906032562, "learning_rate": 9.999685813593159e-05, "loss": 0.0052, "step": 5340 }, { "epoch": 3.522053982883476, "grad_norm": 0.19067029654979706, "learning_rate": 9.999667004337362e-05, "loss": 0.0107, "step": 5350 }, { "epoch": 3.528637261356155, "grad_norm": 0.2031271755695343, "learning_rate": 9.9996476483255e-05, "loss": 0.0062, "step": 5360 }, { "epoch": 3.5352205398288348, "grad_norm": 0.18474696576595306, "learning_rate": 9.999627745559688e-05, "loss": 0.0031, "step": 5370 }, { "epoch": 3.5418038183015144, "grad_norm": 0.13454397022724152, "learning_rate": 9.999607296042101e-05, "loss": 0.0057, "step": 5380 }, { "epoch": 3.5483870967741935, "grad_norm": 0.21881625056266785, "learning_rate": 9.99958629977498e-05, "loss": 0.0076, "step": 5390 }, { "epoch": 3.554970375246873, "grad_norm": 0.14093957841396332, "learning_rate": 9.999564756760615e-05, "loss": 0.0054, "step": 5400 }, { "epoch": 3.5615536537195522, "grad_norm": 0.13959524035453796, "learning_rate": 9.999542667001366e-05, "loss": 0.0042, "step": 5410 }, { "epoch": 3.568136932192232, "grad_norm": 0.15219472348690033, "learning_rate": 9.999520030499647e-05, "loss": 0.0062, "step": 5420 }, { "epoch": 3.574720210664911, "grad_norm": 0.12343954294919968, "learning_rate": 9.999496847257936e-05, "loss": 0.0041, "step": 5430 }, { "epoch": 3.5813034891375906, "grad_norm": 0.21494294703006744, "learning_rate": 9.999473117278764e-05, "loss": 0.005, "step": 5440 }, { "epoch": 3.5878867676102697, "grad_norm": 0.15911900997161865, "learning_rate": 9.999448840564731e-05, "loss": 0.0054, "step": 5450 }, { "epoch": 3.5944700460829493, "grad_norm": 0.18115796148777008, "learning_rate": 9.999424017118488e-05, "loss": 0.0038, "step": 5460 }, { "epoch": 3.601053324555629, "grad_norm": 0.25956225395202637, "learning_rate": 9.999398646942751e-05, "loss": 0.0064, "step": 5470 }, { "epoch": 3.607636603028308, "grad_norm": 0.2188800424337387, "learning_rate": 9.999372730040296e-05, "loss": 0.0031, "step": 5480 }, { "epoch": 3.6142198815009876, "grad_norm": 0.13523079454898834, "learning_rate": 9.999346266413953e-05, "loss": 0.0026, "step": 5490 }, { "epoch": 3.6208031599736668, "grad_norm": 0.1297069489955902, "learning_rate": 9.99931925606662e-05, "loss": 0.0035, "step": 5500 }, { "epoch": 3.6273864384463463, "grad_norm": 0.24267569184303284, "learning_rate": 9.99929169900125e-05, "loss": 0.0042, "step": 5510 }, { "epoch": 3.633969716919026, "grad_norm": 0.19329476356506348, "learning_rate": 9.999263595220855e-05, "loss": 0.0035, "step": 5520 }, { "epoch": 3.640552995391705, "grad_norm": 0.15211130678653717, "learning_rate": 9.99923494472851e-05, "loss": 0.0039, "step": 5530 }, { "epoch": 3.647136273864384, "grad_norm": 0.14313456416130066, "learning_rate": 9.999205747527348e-05, "loss": 0.0036, "step": 5540 }, { "epoch": 3.653719552337064, "grad_norm": 0.2151932716369629, "learning_rate": 9.999176003620561e-05, "loss": 0.0045, "step": 5550 }, { "epoch": 3.6603028308097434, "grad_norm": 0.15544962882995605, "learning_rate": 9.999145713011405e-05, "loss": 0.0043, "step": 5560 }, { "epoch": 3.6668861092824225, "grad_norm": 0.13948547840118408, "learning_rate": 9.999114875703186e-05, "loss": 0.0043, "step": 5570 }, { "epoch": 3.673469387755102, "grad_norm": 0.19837793707847595, "learning_rate": 9.999083491699281e-05, "loss": 0.0027, "step": 5580 }, { "epoch": 3.6800526662277813, "grad_norm": 0.1513727754354477, "learning_rate": 9.999051561003123e-05, "loss": 0.0033, "step": 5590 }, { "epoch": 3.686635944700461, "grad_norm": 0.1273980289697647, "learning_rate": 9.999019083618202e-05, "loss": 0.0038, "step": 5600 }, { "epoch": 3.6932192231731404, "grad_norm": 0.19568218290805817, "learning_rate": 9.99898605954807e-05, "loss": 0.0045, "step": 5610 }, { "epoch": 3.6998025016458196, "grad_norm": 0.16663530468940735, "learning_rate": 9.998952488796338e-05, "loss": 0.0057, "step": 5620 }, { "epoch": 3.706385780118499, "grad_norm": 0.1375395506620407, "learning_rate": 9.998918371366676e-05, "loss": 0.0023, "step": 5630 }, { "epoch": 3.7129690585911783, "grad_norm": 0.14415206015110016, "learning_rate": 9.99888370726282e-05, "loss": 0.0053, "step": 5640 }, { "epoch": 3.719552337063858, "grad_norm": 0.17417585849761963, "learning_rate": 9.998848496488556e-05, "loss": 0.0036, "step": 5650 }, { "epoch": 3.726135615536537, "grad_norm": 0.15360434353351593, "learning_rate": 9.998812739047736e-05, "loss": 0.0038, "step": 5660 }, { "epoch": 3.7327188940092166, "grad_norm": 0.21531397104263306, "learning_rate": 9.99877643494427e-05, "loss": 0.0031, "step": 5670 }, { "epoch": 3.739302172481896, "grad_norm": 0.1459791362285614, "learning_rate": 9.998739584182128e-05, "loss": 0.0077, "step": 5680 }, { "epoch": 3.7458854509545754, "grad_norm": 0.17074887454509735, "learning_rate": 9.998702186765342e-05, "loss": 0.0033, "step": 5690 }, { "epoch": 3.752468729427255, "grad_norm": 0.25759774446487427, "learning_rate": 9.998664242698e-05, "loss": 0.0056, "step": 5700 }, { "epoch": 3.759052007899934, "grad_norm": 0.1209205836057663, "learning_rate": 9.998625751984251e-05, "loss": 0.0062, "step": 5710 }, { "epoch": 3.7656352863726137, "grad_norm": 0.16733500361442566, "learning_rate": 9.998586714628307e-05, "loss": 0.0032, "step": 5720 }, { "epoch": 3.772218564845293, "grad_norm": 0.18291418254375458, "learning_rate": 9.998547130634432e-05, "loss": 0.0058, "step": 5730 }, { "epoch": 3.7788018433179724, "grad_norm": 0.12644588947296143, "learning_rate": 9.99850700000696e-05, "loss": 0.003, "step": 5740 }, { "epoch": 3.785385121790652, "grad_norm": 0.1370575726032257, "learning_rate": 9.998466322750278e-05, "loss": 0.0041, "step": 5750 }, { "epoch": 3.791968400263331, "grad_norm": 0.1411421000957489, "learning_rate": 9.998425098868834e-05, "loss": 0.0033, "step": 5760 }, { "epoch": 3.7985516787360103, "grad_norm": 0.15249137580394745, "learning_rate": 9.998383328367136e-05, "loss": 0.0031, "step": 5770 }, { "epoch": 3.80513495720869, "grad_norm": 0.1673334687948227, "learning_rate": 9.99834101124975e-05, "loss": 0.0046, "step": 5780 }, { "epoch": 3.8117182356813695, "grad_norm": 0.16022978723049164, "learning_rate": 9.998298147521309e-05, "loss": 0.004, "step": 5790 }, { "epoch": 3.8183015141540486, "grad_norm": 0.17119403183460236, "learning_rate": 9.998254737186496e-05, "loss": 0.0038, "step": 5800 }, { "epoch": 3.824884792626728, "grad_norm": 0.2721160650253296, "learning_rate": 9.99821078025006e-05, "loss": 0.0072, "step": 5810 }, { "epoch": 3.8314680710994073, "grad_norm": 0.18358109891414642, "learning_rate": 9.998166276716807e-05, "loss": 0.0072, "step": 5820 }, { "epoch": 3.838051349572087, "grad_norm": 0.21073326468467712, "learning_rate": 9.998121226591606e-05, "loss": 0.0051, "step": 5830 }, { "epoch": 3.8446346280447665, "grad_norm": 0.22030693292617798, "learning_rate": 9.998075629879382e-05, "loss": 0.0044, "step": 5840 }, { "epoch": 3.8512179065174457, "grad_norm": 0.21124228835105896, "learning_rate": 9.99802948658512e-05, "loss": 0.0049, "step": 5850 }, { "epoch": 3.857801184990125, "grad_norm": 0.2661246359348297, "learning_rate": 9.99798279671387e-05, "loss": 0.0043, "step": 5860 }, { "epoch": 3.8643844634628044, "grad_norm": 0.24360588192939758, "learning_rate": 9.997935560270734e-05, "loss": 0.0034, "step": 5870 }, { "epoch": 3.870967741935484, "grad_norm": 0.13250988721847534, "learning_rate": 9.997887777260879e-05, "loss": 0.0066, "step": 5880 }, { "epoch": 3.877551020408163, "grad_norm": 0.13257840275764465, "learning_rate": 9.997839447689532e-05, "loss": 0.0033, "step": 5890 }, { "epoch": 3.8841342988808427, "grad_norm": 0.19003058969974518, "learning_rate": 9.997790571561978e-05, "loss": 0.0037, "step": 5900 }, { "epoch": 3.890717577353522, "grad_norm": 0.16157355904579163, "learning_rate": 9.99774114888356e-05, "loss": 0.0053, "step": 5910 }, { "epoch": 3.8973008558262014, "grad_norm": 0.19749705493450165, "learning_rate": 9.997691179659684e-05, "loss": 0.0055, "step": 5920 }, { "epoch": 3.903884134298881, "grad_norm": 0.12281320244073868, "learning_rate": 9.997640663895815e-05, "loss": 0.0042, "step": 5930 }, { "epoch": 3.91046741277156, "grad_norm": 0.1737818568944931, "learning_rate": 9.997589601597477e-05, "loss": 0.0062, "step": 5940 }, { "epoch": 3.9170506912442398, "grad_norm": 0.14455874264240265, "learning_rate": 9.997537992770252e-05, "loss": 0.003, "step": 5950 }, { "epoch": 3.923633969716919, "grad_norm": 0.15226136147975922, "learning_rate": 9.997485837419788e-05, "loss": 0.0041, "step": 5960 }, { "epoch": 3.9302172481895985, "grad_norm": 0.13709214329719543, "learning_rate": 9.997433135551786e-05, "loss": 0.0033, "step": 5970 }, { "epoch": 3.936800526662278, "grad_norm": 0.15622080862522125, "learning_rate": 9.997379887172009e-05, "loss": 0.0021, "step": 5980 }, { "epoch": 3.9433838051349572, "grad_norm": 0.27198338508605957, "learning_rate": 9.997326092286281e-05, "loss": 0.0048, "step": 5990 }, { "epoch": 3.9499670836076364, "grad_norm": 0.3133968114852905, "learning_rate": 9.997271750900486e-05, "loss": 0.0064, "step": 6000 }, { "epoch": 3.956550362080316, "grad_norm": 0.14745965600013733, "learning_rate": 9.997216863020565e-05, "loss": 0.0045, "step": 6010 }, { "epoch": 3.9631336405529956, "grad_norm": 0.2773269712924957, "learning_rate": 9.99716142865252e-05, "loss": 0.0044, "step": 6020 }, { "epoch": 3.9697169190256747, "grad_norm": 0.11389656364917755, "learning_rate": 9.997105447802415e-05, "loss": 0.0046, "step": 6030 }, { "epoch": 3.9763001974983543, "grad_norm": 0.24610108137130737, "learning_rate": 9.997048920476373e-05, "loss": 0.006, "step": 6040 }, { "epoch": 3.9828834759710334, "grad_norm": 0.18511781096458435, "learning_rate": 9.996991846680572e-05, "loss": 0.0044, "step": 6050 }, { "epoch": 3.989466754443713, "grad_norm": 0.18410849571228027, "learning_rate": 9.996934226421257e-05, "loss": 0.0043, "step": 6060 }, { "epoch": 3.9960500329163926, "grad_norm": 0.16596458852291107, "learning_rate": 9.996876059704726e-05, "loss": 0.0035, "step": 6070 }, { "epoch": 4.002633311389072, "grad_norm": 0.09848038852214813, "learning_rate": 9.996817346537343e-05, "loss": 0.0028, "step": 6080 }, { "epoch": 4.009216589861751, "grad_norm": 0.18131598830223083, "learning_rate": 9.996758086925526e-05, "loss": 0.0056, "step": 6090 }, { "epoch": 4.0157998683344305, "grad_norm": 0.20494620501995087, "learning_rate": 9.996698280875759e-05, "loss": 0.0056, "step": 6100 }, { "epoch": 4.02238314680711, "grad_norm": 0.16492784023284912, "learning_rate": 9.99663792839458e-05, "loss": 0.0063, "step": 6110 }, { "epoch": 4.02896642527979, "grad_norm": 0.1459585577249527, "learning_rate": 9.99657702948859e-05, "loss": 0.0037, "step": 6120 }, { "epoch": 4.035549703752468, "grad_norm": 0.20403306186199188, "learning_rate": 9.996515584164448e-05, "loss": 0.0063, "step": 6130 }, { "epoch": 4.042132982225148, "grad_norm": 0.22532984614372253, "learning_rate": 9.996453592428873e-05, "loss": 0.0048, "step": 6140 }, { "epoch": 4.0487162606978275, "grad_norm": 0.20675784349441528, "learning_rate": 9.996391054288646e-05, "loss": 0.0046, "step": 6150 }, { "epoch": 4.055299539170507, "grad_norm": 0.249908909201622, "learning_rate": 9.996327969750605e-05, "loss": 0.0048, "step": 6160 }, { "epoch": 4.061882817643187, "grad_norm": 0.15669399499893188, "learning_rate": 9.996264338821649e-05, "loss": 0.005, "step": 6170 }, { "epoch": 4.068466096115865, "grad_norm": 0.2215421348810196, "learning_rate": 9.996200161508735e-05, "loss": 0.0046, "step": 6180 }, { "epoch": 4.075049374588545, "grad_norm": 0.1439313292503357, "learning_rate": 9.996135437818885e-05, "loss": 0.0034, "step": 6190 }, { "epoch": 4.081632653061225, "grad_norm": 0.22410021722316742, "learning_rate": 9.996070167759175e-05, "loss": 0.0051, "step": 6200 }, { "epoch": 4.088215931533904, "grad_norm": 0.1737864464521408, "learning_rate": 9.996004351336743e-05, "loss": 0.0039, "step": 6210 }, { "epoch": 4.094799210006583, "grad_norm": 0.20563966035842896, "learning_rate": 9.995937988558785e-05, "loss": 0.0073, "step": 6220 }, { "epoch": 4.1013824884792625, "grad_norm": 0.2625710964202881, "learning_rate": 9.995871079432561e-05, "loss": 0.006, "step": 6230 }, { "epoch": 4.107965766951942, "grad_norm": 0.17079569399356842, "learning_rate": 9.995803623965389e-05, "loss": 0.0032, "step": 6240 }, { "epoch": 4.114549045424622, "grad_norm": 0.14084303379058838, "learning_rate": 9.995735622164641e-05, "loss": 0.0046, "step": 6250 }, { "epoch": 4.121132323897301, "grad_norm": 0.22147543728351593, "learning_rate": 9.995667074037758e-05, "loss": 0.0067, "step": 6260 }, { "epoch": 4.12771560236998, "grad_norm": 0.15355825424194336, "learning_rate": 9.995597979592232e-05, "loss": 0.0053, "step": 6270 }, { "epoch": 4.1342988808426595, "grad_norm": 0.19798825681209564, "learning_rate": 9.995528338835625e-05, "loss": 0.0042, "step": 6280 }, { "epoch": 4.140882159315339, "grad_norm": 0.16669896245002747, "learning_rate": 9.995458151775547e-05, "loss": 0.0036, "step": 6290 }, { "epoch": 4.147465437788019, "grad_norm": 0.18412400782108307, "learning_rate": 9.995387418419677e-05, "loss": 0.006, "step": 6300 }, { "epoch": 4.154048716260698, "grad_norm": 0.21070927381515503, "learning_rate": 9.99531613877575e-05, "loss": 0.0037, "step": 6310 }, { "epoch": 4.160631994733377, "grad_norm": 0.25839361548423767, "learning_rate": 9.995244312851559e-05, "loss": 0.0051, "step": 6320 }, { "epoch": 4.167215273206057, "grad_norm": 0.21170639991760254, "learning_rate": 9.995171940654961e-05, "loss": 0.0057, "step": 6330 }, { "epoch": 4.173798551678736, "grad_norm": 0.20744511485099792, "learning_rate": 9.995099022193871e-05, "loss": 0.0063, "step": 6340 }, { "epoch": 4.180381830151416, "grad_norm": 0.2542797029018402, "learning_rate": 9.995025557476261e-05, "loss": 0.005, "step": 6350 }, { "epoch": 4.186965108624094, "grad_norm": 0.17993493378162384, "learning_rate": 9.994951546510165e-05, "loss": 0.0039, "step": 6360 }, { "epoch": 4.193548387096774, "grad_norm": 0.17484597861766815, "learning_rate": 9.994876989303679e-05, "loss": 0.004, "step": 6370 }, { "epoch": 4.200131665569454, "grad_norm": 0.23395459353923798, "learning_rate": 9.994801885864955e-05, "loss": 0.0053, "step": 6380 }, { "epoch": 4.206714944042133, "grad_norm": 0.15392647683620453, "learning_rate": 9.994726236202205e-05, "loss": 0.0065, "step": 6390 }, { "epoch": 4.213298222514813, "grad_norm": 0.15251809358596802, "learning_rate": 9.994650040323704e-05, "loss": 0.005, "step": 6400 }, { "epoch": 4.2198815009874915, "grad_norm": 0.16726380586624146, "learning_rate": 9.994573298237784e-05, "loss": 0.005, "step": 6410 }, { "epoch": 4.226464779460171, "grad_norm": 0.12131811678409576, "learning_rate": 9.994496009952837e-05, "loss": 0.0043, "step": 6420 }, { "epoch": 4.233048057932851, "grad_norm": 0.17872901260852814, "learning_rate": 9.994418175477316e-05, "loss": 0.0043, "step": 6430 }, { "epoch": 4.23963133640553, "grad_norm": 0.14288176596164703, "learning_rate": 9.994339794819733e-05, "loss": 0.003, "step": 6440 }, { "epoch": 4.246214614878209, "grad_norm": 0.16294293105602264, "learning_rate": 9.994260867988658e-05, "loss": 0.0052, "step": 6450 }, { "epoch": 4.2527978933508885, "grad_norm": 0.1564975529909134, "learning_rate": 9.994181394992723e-05, "loss": 0.003, "step": 6460 }, { "epoch": 4.259381171823568, "grad_norm": 0.170480877161026, "learning_rate": 9.994101375840618e-05, "loss": 0.0043, "step": 6470 }, { "epoch": 4.265964450296248, "grad_norm": 0.14930586516857147, "learning_rate": 9.994020810541098e-05, "loss": 0.0028, "step": 6480 }, { "epoch": 4.272547728768927, "grad_norm": 0.1282031238079071, "learning_rate": 9.99393969910297e-05, "loss": 0.0027, "step": 6490 }, { "epoch": 4.279131007241606, "grad_norm": 0.1819976419210434, "learning_rate": 9.993858041535104e-05, "loss": 0.0035, "step": 6500 }, { "epoch": 4.285714285714286, "grad_norm": 0.20018666982650757, "learning_rate": 9.99377583784643e-05, "loss": 0.0082, "step": 6510 }, { "epoch": 4.292297564186965, "grad_norm": 0.16977563500404358, "learning_rate": 9.993693088045939e-05, "loss": 0.0051, "step": 6520 }, { "epoch": 4.298880842659645, "grad_norm": 0.16830118000507355, "learning_rate": 9.99360979214268e-05, "loss": 0.0028, "step": 6530 }, { "epoch": 4.305464121132324, "grad_norm": 0.21772077679634094, "learning_rate": 9.99352595014576e-05, "loss": 0.0061, "step": 6540 }, { "epoch": 4.312047399605003, "grad_norm": 0.13900883495807648, "learning_rate": 9.993441562064354e-05, "loss": 0.0043, "step": 6550 }, { "epoch": 4.318630678077683, "grad_norm": 0.12345195561647415, "learning_rate": 9.993356627907685e-05, "loss": 0.0024, "step": 6560 }, { "epoch": 4.325213956550362, "grad_norm": 0.15612512826919556, "learning_rate": 9.99327114768504e-05, "loss": 0.0051, "step": 6570 }, { "epoch": 4.331797235023042, "grad_norm": 0.11138295382261276, "learning_rate": 9.99318512140577e-05, "loss": 0.0029, "step": 6580 }, { "epoch": 4.3383805134957205, "grad_norm": 0.10942983627319336, "learning_rate": 9.993098549079284e-05, "loss": 0.0034, "step": 6590 }, { "epoch": 4.3449637919684, "grad_norm": 0.2337152361869812, "learning_rate": 9.993011430715047e-05, "loss": 0.0032, "step": 6600 }, { "epoch": 4.35154707044108, "grad_norm": 0.12469164282083511, "learning_rate": 9.992923766322586e-05, "loss": 0.0016, "step": 6610 }, { "epoch": 4.358130348913759, "grad_norm": 0.14959700405597687, "learning_rate": 9.99283555591149e-05, "loss": 0.0058, "step": 6620 }, { "epoch": 4.364713627386438, "grad_norm": 0.18529990315437317, "learning_rate": 9.992746799491404e-05, "loss": 0.0028, "step": 6630 }, { "epoch": 4.371296905859118, "grad_norm": 0.12093222886323929, "learning_rate": 9.992657497072033e-05, "loss": 0.0026, "step": 6640 }, { "epoch": 4.377880184331797, "grad_norm": 0.0720699355006218, "learning_rate": 9.992567648663147e-05, "loss": 0.0027, "step": 6650 }, { "epoch": 4.384463462804477, "grad_norm": 0.3394845724105835, "learning_rate": 9.992477254274568e-05, "loss": 0.0057, "step": 6660 }, { "epoch": 4.391046741277156, "grad_norm": 0.2064623236656189, "learning_rate": 9.992386313916183e-05, "loss": 0.0042, "step": 6670 }, { "epoch": 4.397630019749835, "grad_norm": 0.15707750618457794, "learning_rate": 9.992294827597934e-05, "loss": 0.0038, "step": 6680 }, { "epoch": 4.404213298222515, "grad_norm": 0.13891290128231049, "learning_rate": 9.992202795329831e-05, "loss": 0.0057, "step": 6690 }, { "epoch": 4.410796576695194, "grad_norm": 0.15383563935756683, "learning_rate": 9.992110217121936e-05, "loss": 0.0053, "step": 6700 }, { "epoch": 4.417379855167874, "grad_norm": 0.13387784361839294, "learning_rate": 9.992017092984372e-05, "loss": 0.0037, "step": 6710 }, { "epoch": 4.423963133640553, "grad_norm": 0.2049536406993866, "learning_rate": 9.991923422927326e-05, "loss": 0.0036, "step": 6720 }, { "epoch": 4.430546412113232, "grad_norm": 0.11599751561880112, "learning_rate": 9.991829206961037e-05, "loss": 0.007, "step": 6730 }, { "epoch": 4.437129690585912, "grad_norm": 0.2536364495754242, "learning_rate": 9.991734445095813e-05, "loss": 0.0084, "step": 6740 }, { "epoch": 4.443712969058591, "grad_norm": 0.16394062340259552, "learning_rate": 9.991639137342015e-05, "loss": 0.0051, "step": 6750 }, { "epoch": 4.450296247531271, "grad_norm": 0.15859438478946686, "learning_rate": 9.991543283710064e-05, "loss": 0.0075, "step": 6760 }, { "epoch": 4.45687952600395, "grad_norm": 0.16036082804203033, "learning_rate": 9.991446884210445e-05, "loss": 0.0062, "step": 6770 }, { "epoch": 4.463462804476629, "grad_norm": 0.2094184309244156, "learning_rate": 9.9913499388537e-05, "loss": 0.005, "step": 6780 }, { "epoch": 4.470046082949309, "grad_norm": 0.1476239711046219, "learning_rate": 9.99125244765043e-05, "loss": 0.004, "step": 6790 }, { "epoch": 4.476629361421988, "grad_norm": 0.20211604237556458, "learning_rate": 9.991154410611296e-05, "loss": 0.008, "step": 6800 }, { "epoch": 4.483212639894668, "grad_norm": 0.17130358517169952, "learning_rate": 9.99105582774702e-05, "loss": 0.0049, "step": 6810 }, { "epoch": 4.489795918367347, "grad_norm": 0.2183646261692047, "learning_rate": 9.990956699068384e-05, "loss": 0.0039, "step": 6820 }, { "epoch": 4.496379196840026, "grad_norm": 0.18767009675502777, "learning_rate": 9.990857024586224e-05, "loss": 0.0052, "step": 6830 }, { "epoch": 4.502962475312706, "grad_norm": 0.15425646305084229, "learning_rate": 9.990756804311446e-05, "loss": 0.0034, "step": 6840 }, { "epoch": 4.509545753785385, "grad_norm": 0.21681298315525055, "learning_rate": 9.990656038255006e-05, "loss": 0.0048, "step": 6850 }, { "epoch": 4.516129032258064, "grad_norm": 0.21347884833812714, "learning_rate": 9.990554726427926e-05, "loss": 0.0035, "step": 6860 }, { "epoch": 4.522712310730744, "grad_norm": 0.15742377936840057, "learning_rate": 9.990452868841284e-05, "loss": 0.003, "step": 6870 }, { "epoch": 4.529295589203423, "grad_norm": 0.16803038120269775, "learning_rate": 9.99035046550622e-05, "loss": 0.0074, "step": 6880 }, { "epoch": 4.535878867676103, "grad_norm": 0.15852800011634827, "learning_rate": 9.99024751643393e-05, "loss": 0.0029, "step": 6890 }, { "epoch": 4.542462146148782, "grad_norm": 0.20096071064472198, "learning_rate": 9.990144021635677e-05, "loss": 0.0043, "step": 6900 }, { "epoch": 4.549045424621461, "grad_norm": 0.18715864419937134, "learning_rate": 9.990039981122775e-05, "loss": 0.0074, "step": 6910 }, { "epoch": 4.555628703094141, "grad_norm": 0.17612719535827637, "learning_rate": 9.989935394906602e-05, "loss": 0.005, "step": 6920 }, { "epoch": 4.56221198156682, "grad_norm": 0.1360718011856079, "learning_rate": 9.989830262998598e-05, "loss": 0.0036, "step": 6930 }, { "epoch": 4.5687952600395, "grad_norm": 0.15484477579593658, "learning_rate": 9.989724585410259e-05, "loss": 0.0041, "step": 6940 }, { "epoch": 4.5753785385121795, "grad_norm": 0.12353093177080154, "learning_rate": 9.989618362153139e-05, "loss": 0.0052, "step": 6950 }, { "epoch": 4.581961816984858, "grad_norm": 0.1712733507156372, "learning_rate": 9.989511593238859e-05, "loss": 0.0049, "step": 6960 }, { "epoch": 4.588545095457538, "grad_norm": 0.1357087343931198, "learning_rate": 9.98940427867909e-05, "loss": 0.0044, "step": 6970 }, { "epoch": 4.595128373930217, "grad_norm": 0.1642552614212036, "learning_rate": 9.989296418485573e-05, "loss": 0.0043, "step": 6980 }, { "epoch": 4.601711652402897, "grad_norm": 0.283719539642334, "learning_rate": 9.989188012670101e-05, "loss": 0.0088, "step": 6990 }, { "epoch": 4.6082949308755765, "grad_norm": 0.17903223633766174, "learning_rate": 9.989079061244528e-05, "loss": 0.0052, "step": 7000 }, { "epoch": 4.614878209348255, "grad_norm": 0.247084841132164, "learning_rate": 9.988969564220769e-05, "loss": 0.0059, "step": 7010 }, { "epoch": 4.621461487820935, "grad_norm": 0.23811039328575134, "learning_rate": 9.988859521610801e-05, "loss": 0.0053, "step": 7020 }, { "epoch": 4.628044766293614, "grad_norm": 0.14293409883975983, "learning_rate": 9.988748933426656e-05, "loss": 0.0037, "step": 7030 }, { "epoch": 4.634628044766294, "grad_norm": 0.19719170033931732, "learning_rate": 9.988637799680428e-05, "loss": 0.006, "step": 7040 }, { "epoch": 4.641211323238973, "grad_norm": 0.19844700396060944, "learning_rate": 9.98852612038427e-05, "loss": 0.0044, "step": 7050 }, { "epoch": 4.647794601711652, "grad_norm": 0.2322036772966385, "learning_rate": 9.988413895550397e-05, "loss": 0.0042, "step": 7060 }, { "epoch": 4.654377880184332, "grad_norm": 0.15980470180511475, "learning_rate": 9.98830112519108e-05, "loss": 0.006, "step": 7070 }, { "epoch": 4.660961158657011, "grad_norm": 0.2137414813041687, "learning_rate": 9.98818780931865e-05, "loss": 0.0069, "step": 7080 }, { "epoch": 4.66754443712969, "grad_norm": 0.163013756275177, "learning_rate": 9.988073947945502e-05, "loss": 0.0043, "step": 7090 }, { "epoch": 4.67412771560237, "grad_norm": 0.20377755165100098, "learning_rate": 9.987959541084087e-05, "loss": 0.0051, "step": 7100 }, { "epoch": 4.680710994075049, "grad_norm": 0.15359339118003845, "learning_rate": 9.987844588746915e-05, "loss": 0.0053, "step": 7110 }, { "epoch": 4.687294272547729, "grad_norm": 0.2149839550256729, "learning_rate": 9.987729090946558e-05, "loss": 0.0035, "step": 7120 }, { "epoch": 4.6938775510204085, "grad_norm": 0.19012287259101868, "learning_rate": 9.987613047695647e-05, "loss": 0.0052, "step": 7130 }, { "epoch": 4.700460829493087, "grad_norm": 0.19843228161334991, "learning_rate": 9.987496459006871e-05, "loss": 0.0044, "step": 7140 }, { "epoch": 4.707044107965767, "grad_norm": 0.15949580073356628, "learning_rate": 9.987379324892982e-05, "loss": 0.0037, "step": 7150 }, { "epoch": 4.713627386438446, "grad_norm": 0.12308900058269501, "learning_rate": 9.987261645366788e-05, "loss": 0.0093, "step": 7160 }, { "epoch": 4.720210664911126, "grad_norm": 0.09453386068344116, "learning_rate": 9.987143420441158e-05, "loss": 0.0031, "step": 7170 }, { "epoch": 4.7267939433838055, "grad_norm": 0.11401432007551193, "learning_rate": 9.987024650129022e-05, "loss": 0.0035, "step": 7180 }, { "epoch": 4.733377221856484, "grad_norm": 0.15365423262119293, "learning_rate": 9.986905334443368e-05, "loss": 0.0033, "step": 7190 }, { "epoch": 4.739960500329164, "grad_norm": 0.27689075469970703, "learning_rate": 9.986785473397245e-05, "loss": 0.0063, "step": 7200 }, { "epoch": 4.746543778801843, "grad_norm": 0.17181693017482758, "learning_rate": 9.98666506700376e-05, "loss": 0.0046, "step": 7210 }, { "epoch": 4.753127057274523, "grad_norm": 0.12398321181535721, "learning_rate": 9.986544115276081e-05, "loss": 0.0031, "step": 7220 }, { "epoch": 4.759710335747203, "grad_norm": 0.1839093714952469, "learning_rate": 9.986422618227433e-05, "loss": 0.0035, "step": 7230 }, { "epoch": 4.766293614219881, "grad_norm": 0.1597033590078354, "learning_rate": 9.986300575871106e-05, "loss": 0.0039, "step": 7240 }, { "epoch": 4.772876892692561, "grad_norm": 0.2728452980518341, "learning_rate": 9.986177988220444e-05, "loss": 0.0066, "step": 7250 }, { "epoch": 4.7794601711652405, "grad_norm": 0.12375461310148239, "learning_rate": 9.986054855288856e-05, "loss": 0.0045, "step": 7260 }, { "epoch": 4.78604344963792, "grad_norm": 0.18078511953353882, "learning_rate": 9.985931177089802e-05, "loss": 0.0071, "step": 7270 }, { "epoch": 4.792626728110599, "grad_norm": 0.24507707357406616, "learning_rate": 9.985806953636814e-05, "loss": 0.0034, "step": 7280 }, { "epoch": 4.799210006583278, "grad_norm": 0.1518697291612625, "learning_rate": 9.985682184943471e-05, "loss": 0.0042, "step": 7290 }, { "epoch": 4.805793285055958, "grad_norm": 0.11414860188961029, "learning_rate": 9.98555687102342e-05, "loss": 0.007, "step": 7300 }, { "epoch": 4.8123765635286375, "grad_norm": 0.14914824068546295, "learning_rate": 9.985431011890367e-05, "loss": 0.0031, "step": 7310 }, { "epoch": 4.818959842001316, "grad_norm": 0.21440480649471283, "learning_rate": 9.985304607558075e-05, "loss": 0.0037, "step": 7320 }, { "epoch": 4.825543120473996, "grad_norm": 0.12512162327766418, "learning_rate": 9.985177658040364e-05, "loss": 0.004, "step": 7330 }, { "epoch": 4.832126398946675, "grad_norm": 0.22395510971546173, "learning_rate": 9.985050163351119e-05, "loss": 0.0051, "step": 7340 }, { "epoch": 4.838709677419355, "grad_norm": 0.17169657349586487, "learning_rate": 9.984922123504286e-05, "loss": 0.0042, "step": 7350 }, { "epoch": 4.845292955892035, "grad_norm": 0.17394252121448517, "learning_rate": 9.984793538513862e-05, "loss": 0.0053, "step": 7360 }, { "epoch": 4.851876234364713, "grad_norm": 0.14337463676929474, "learning_rate": 9.984664408393912e-05, "loss": 0.0049, "step": 7370 }, { "epoch": 4.858459512837393, "grad_norm": 0.1761924922466278, "learning_rate": 9.984534733158556e-05, "loss": 0.0052, "step": 7380 }, { "epoch": 4.8650427913100724, "grad_norm": 0.16465042531490326, "learning_rate": 9.984404512821977e-05, "loss": 0.0043, "step": 7390 }, { "epoch": 4.871626069782752, "grad_norm": 0.1757267862558365, "learning_rate": 9.984273747398411e-05, "loss": 0.0035, "step": 7400 }, { "epoch": 4.878209348255432, "grad_norm": 0.12685853242874146, "learning_rate": 9.984142436902165e-05, "loss": 0.0033, "step": 7410 }, { "epoch": 4.88479262672811, "grad_norm": 0.17150230705738068, "learning_rate": 9.984010581347596e-05, "loss": 0.0109, "step": 7420 }, { "epoch": 4.89137590520079, "grad_norm": 0.16668912768363953, "learning_rate": 9.983878180749121e-05, "loss": 0.0067, "step": 7430 }, { "epoch": 4.8979591836734695, "grad_norm": 0.13767383992671967, "learning_rate": 9.983745235121222e-05, "loss": 0.0053, "step": 7440 }, { "epoch": 4.904542462146149, "grad_norm": 0.13094231486320496, "learning_rate": 9.983611744478438e-05, "loss": 0.0035, "step": 7450 }, { "epoch": 4.911125740618829, "grad_norm": 0.12874433398246765, "learning_rate": 9.983477708835365e-05, "loss": 0.0037, "step": 7460 }, { "epoch": 4.917709019091507, "grad_norm": 0.1483183056116104, "learning_rate": 9.983343128206664e-05, "loss": 0.0049, "step": 7470 }, { "epoch": 4.924292297564187, "grad_norm": 0.20683136582374573, "learning_rate": 9.983208002607049e-05, "loss": 0.0103, "step": 7480 }, { "epoch": 4.9308755760368665, "grad_norm": 0.14720121026039124, "learning_rate": 9.9830723320513e-05, "loss": 0.0051, "step": 7490 }, { "epoch": 4.937458854509546, "grad_norm": 0.19889312982559204, "learning_rate": 9.982936116554254e-05, "loss": 0.0054, "step": 7500 }, { "epoch": 4.944042132982225, "grad_norm": 0.18466399610042572, "learning_rate": 9.982799356130803e-05, "loss": 0.0031, "step": 7510 }, { "epoch": 4.950625411454904, "grad_norm": 0.1842978149652481, "learning_rate": 9.982662050795908e-05, "loss": 0.0071, "step": 7520 }, { "epoch": 4.957208689927584, "grad_norm": 0.23563820123672485, "learning_rate": 9.982524200564583e-05, "loss": 0.0099, "step": 7530 }, { "epoch": 4.963791968400264, "grad_norm": 0.22789046168327332, "learning_rate": 9.982385805451901e-05, "loss": 0.006, "step": 7540 }, { "epoch": 4.970375246872942, "grad_norm": 0.1759812831878662, "learning_rate": 9.982246865472998e-05, "loss": 0.0052, "step": 7550 }, { "epoch": 4.976958525345622, "grad_norm": 0.18048153817653656, "learning_rate": 9.982107380643069e-05, "loss": 0.004, "step": 7560 }, { "epoch": 4.9835418038183015, "grad_norm": 0.11625604331493378, "learning_rate": 9.981967350977368e-05, "loss": 0.0028, "step": 7570 }, { "epoch": 4.990125082290981, "grad_norm": 0.1672743260860443, "learning_rate": 9.981826776491208e-05, "loss": 0.0046, "step": 7580 }, { "epoch": 4.996708360763661, "grad_norm": 0.14576631784439087, "learning_rate": 9.98168565719996e-05, "loss": 0.0065, "step": 7590 }, { "epoch": 5.003291639236339, "grad_norm": 0.21726883947849274, "learning_rate": 9.98154399311906e-05, "loss": 0.005, "step": 7600 }, { "epoch": 5.009874917709019, "grad_norm": 0.11858858913183212, "learning_rate": 9.981401784263997e-05, "loss": 0.0034, "step": 7610 }, { "epoch": 5.0164581961816985, "grad_norm": 0.19020964205265045, "learning_rate": 9.981259030650326e-05, "loss": 0.0026, "step": 7620 }, { "epoch": 5.023041474654378, "grad_norm": 0.1685321182012558, "learning_rate": 9.981115732293655e-05, "loss": 0.0099, "step": 7630 }, { "epoch": 5.029624753127058, "grad_norm": 0.15001019835472107, "learning_rate": 9.980971889209659e-05, "loss": 0.0038, "step": 7640 }, { "epoch": 5.036208031599736, "grad_norm": 0.13302698731422424, "learning_rate": 9.980827501414064e-05, "loss": 0.0051, "step": 7650 }, { "epoch": 5.042791310072416, "grad_norm": 0.15681104362010956, "learning_rate": 9.980682568922663e-05, "loss": 0.0066, "step": 7660 }, { "epoch": 5.049374588545096, "grad_norm": 0.13218951225280762, "learning_rate": 9.980537091751304e-05, "loss": 0.0042, "step": 7670 }, { "epoch": 5.055957867017775, "grad_norm": 0.1735256165266037, "learning_rate": 9.980391069915897e-05, "loss": 0.0047, "step": 7680 }, { "epoch": 5.062541145490454, "grad_norm": 0.1385265439748764, "learning_rate": 9.98024450343241e-05, "loss": 0.0047, "step": 7690 }, { "epoch": 5.0691244239631335, "grad_norm": 0.18062913417816162, "learning_rate": 9.980097392316872e-05, "loss": 0.0047, "step": 7700 }, { "epoch": 5.075707702435813, "grad_norm": 0.12201916426420212, "learning_rate": 9.97994973658537e-05, "loss": 0.0034, "step": 7710 }, { "epoch": 5.082290980908493, "grad_norm": 0.12323544919490814, "learning_rate": 9.979801536254054e-05, "loss": 0.0063, "step": 7720 }, { "epoch": 5.088874259381172, "grad_norm": 0.203764870762825, "learning_rate": 9.979652791339127e-05, "loss": 0.0055, "step": 7730 }, { "epoch": 5.095457537853851, "grad_norm": 0.17137792706489563, "learning_rate": 9.97950350185686e-05, "loss": 0.0039, "step": 7740 }, { "epoch": 5.1020408163265305, "grad_norm": 0.23326125741004944, "learning_rate": 9.979353667823574e-05, "loss": 0.0059, "step": 7750 }, { "epoch": 5.10862409479921, "grad_norm": 0.20196546614170074, "learning_rate": 9.979203289255658e-05, "loss": 0.005, "step": 7760 }, { "epoch": 5.11520737327189, "grad_norm": 0.17507046461105347, "learning_rate": 9.979052366169557e-05, "loss": 0.0039, "step": 7770 }, { "epoch": 5.121790651744568, "grad_norm": 0.16280853748321533, "learning_rate": 9.978900898581775e-05, "loss": 0.0039, "step": 7780 }, { "epoch": 5.128373930217248, "grad_norm": 0.15298143029212952, "learning_rate": 9.978748886508875e-05, "loss": 0.0055, "step": 7790 }, { "epoch": 5.1349572086899276, "grad_norm": 0.17174504697322845, "learning_rate": 9.978596329967484e-05, "loss": 0.0034, "step": 7800 }, { "epoch": 5.141540487162607, "grad_norm": 0.12979206442832947, "learning_rate": 9.978443228974284e-05, "loss": 0.0041, "step": 7810 }, { "epoch": 5.148123765635287, "grad_norm": 0.16857776045799255, "learning_rate": 9.978289583546015e-05, "loss": 0.004, "step": 7820 }, { "epoch": 5.154707044107965, "grad_norm": 0.22906222939491272, "learning_rate": 9.978135393699484e-05, "loss": 0.004, "step": 7830 }, { "epoch": 5.161290322580645, "grad_norm": 0.12299106270074844, "learning_rate": 9.977980659451548e-05, "loss": 0.0026, "step": 7840 }, { "epoch": 5.167873601053325, "grad_norm": 0.15928016602993011, "learning_rate": 9.977825380819135e-05, "loss": 0.0034, "step": 7850 }, { "epoch": 5.174456879526004, "grad_norm": 0.16443081200122833, "learning_rate": 9.97766955781922e-05, "loss": 0.0029, "step": 7860 }, { "epoch": 5.181040157998684, "grad_norm": 0.092191681265831, "learning_rate": 9.977513190468848e-05, "loss": 0.0038, "step": 7870 }, { "epoch": 5.1876234364713625, "grad_norm": 0.17411775887012482, "learning_rate": 9.977356278785116e-05, "loss": 0.0034, "step": 7880 }, { "epoch": 5.194206714944042, "grad_norm": 0.15928712487220764, "learning_rate": 9.977198822785184e-05, "loss": 0.0034, "step": 7890 }, { "epoch": 5.200789993416722, "grad_norm": 0.18201106786727905, "learning_rate": 9.977040822486273e-05, "loss": 0.0056, "step": 7900 }, { "epoch": 5.207373271889401, "grad_norm": 0.14058929681777954, "learning_rate": 9.97688227790566e-05, "loss": 0.0037, "step": 7910 }, { "epoch": 5.21395655036208, "grad_norm": 0.1664869487285614, "learning_rate": 9.976723189060684e-05, "loss": 0.0033, "step": 7920 }, { "epoch": 5.2205398288347595, "grad_norm": 0.1868186593055725, "learning_rate": 9.976563555968742e-05, "loss": 0.0043, "step": 7930 }, { "epoch": 5.227123107307439, "grad_norm": 0.16662606596946716, "learning_rate": 9.976403378647292e-05, "loss": 0.0053, "step": 7940 }, { "epoch": 5.233706385780119, "grad_norm": 0.11939506232738495, "learning_rate": 9.97624265711385e-05, "loss": 0.0037, "step": 7950 }, { "epoch": 5.240289664252798, "grad_norm": 0.1555798500776291, "learning_rate": 9.976081391385993e-05, "loss": 0.0036, "step": 7960 }, { "epoch": 5.246872942725477, "grad_norm": 0.16300635039806366, "learning_rate": 9.975919581481356e-05, "loss": 0.0045, "step": 7970 }, { "epoch": 5.253456221198157, "grad_norm": 0.162973091006279, "learning_rate": 9.975757227417634e-05, "loss": 0.0029, "step": 7980 }, { "epoch": 5.260039499670836, "grad_norm": 0.15504707396030426, "learning_rate": 9.975594329212586e-05, "loss": 0.0045, "step": 7990 }, { "epoch": 5.266622778143516, "grad_norm": 0.10465308278799057, "learning_rate": 9.97543088688402e-05, "loss": 0.0049, "step": 8000 }, { "epoch": 5.2732060566161945, "grad_norm": 0.12009970843791962, "learning_rate": 9.975266900449814e-05, "loss": 0.004, "step": 8010 }, { "epoch": 5.279789335088874, "grad_norm": 0.17839759588241577, "learning_rate": 9.975102369927898e-05, "loss": 0.0038, "step": 8020 }, { "epoch": 5.286372613561554, "grad_norm": 0.18020717799663544, "learning_rate": 9.974937295336269e-05, "loss": 0.0028, "step": 8030 }, { "epoch": 5.292955892034233, "grad_norm": 0.16900689899921417, "learning_rate": 9.974771676692975e-05, "loss": 0.0029, "step": 8040 }, { "epoch": 5.299539170506913, "grad_norm": 0.13639606535434723, "learning_rate": 9.974605514016131e-05, "loss": 0.0041, "step": 8050 }, { "epoch": 5.3061224489795915, "grad_norm": 0.2025255411863327, "learning_rate": 9.974438807323907e-05, "loss": 0.005, "step": 8060 }, { "epoch": 5.312705727452271, "grad_norm": 0.13803504407405853, "learning_rate": 9.974271556634535e-05, "loss": 0.0039, "step": 8070 }, { "epoch": 5.319289005924951, "grad_norm": 0.13224107027053833, "learning_rate": 9.974103761966302e-05, "loss": 0.0055, "step": 8080 }, { "epoch": 5.32587228439763, "grad_norm": 0.1025618463754654, "learning_rate": 9.973935423337563e-05, "loss": 0.003, "step": 8090 }, { "epoch": 5.33245556287031, "grad_norm": 0.19223064184188843, "learning_rate": 9.973766540766722e-05, "loss": 0.0041, "step": 8100 }, { "epoch": 5.339038841342989, "grad_norm": 0.11431878060102463, "learning_rate": 9.97359711427225e-05, "loss": 0.0041, "step": 8110 }, { "epoch": 5.345622119815668, "grad_norm": 0.14973682165145874, "learning_rate": 9.973427143872677e-05, "loss": 0.006, "step": 8120 }, { "epoch": 5.352205398288348, "grad_norm": 0.16735172271728516, "learning_rate": 9.973256629586589e-05, "loss": 0.004, "step": 8130 }, { "epoch": 5.358788676761027, "grad_norm": 0.10929079353809357, "learning_rate": 9.973085571432632e-05, "loss": 0.0035, "step": 8140 }, { "epoch": 5.365371955233706, "grad_norm": 0.0864635780453682, "learning_rate": 9.972913969429513e-05, "loss": 0.0025, "step": 8150 }, { "epoch": 5.371955233706386, "grad_norm": 0.27760031819343567, "learning_rate": 9.972741823596e-05, "loss": 0.0046, "step": 8160 }, { "epoch": 5.378538512179065, "grad_norm": 0.18094976246356964, "learning_rate": 9.972569133950917e-05, "loss": 0.005, "step": 8170 }, { "epoch": 5.385121790651745, "grad_norm": 0.19732160866260529, "learning_rate": 9.972395900513151e-05, "loss": 0.0035, "step": 8180 }, { "epoch": 5.391705069124424, "grad_norm": 0.1464812159538269, "learning_rate": 9.972222123301645e-05, "loss": 0.0045, "step": 8190 }, { "epoch": 5.398288347597103, "grad_norm": 0.18641170859336853, "learning_rate": 9.972047802335403e-05, "loss": 0.0044, "step": 8200 }, { "epoch": 5.404871626069783, "grad_norm": 0.13720044493675232, "learning_rate": 9.971872937633488e-05, "loss": 0.0045, "step": 8210 }, { "epoch": 5.411454904542462, "grad_norm": 0.16702471673488617, "learning_rate": 9.971697529215024e-05, "loss": 0.0044, "step": 8220 }, { "epoch": 5.418038183015142, "grad_norm": 0.1949428915977478, "learning_rate": 9.971521577099192e-05, "loss": 0.004, "step": 8230 }, { "epoch": 5.4246214614878205, "grad_norm": 0.1615414023399353, "learning_rate": 9.971345081305236e-05, "loss": 0.0046, "step": 8240 }, { "epoch": 5.4312047399605, "grad_norm": 0.1464558094739914, "learning_rate": 9.971168041852456e-05, "loss": 0.0035, "step": 8250 }, { "epoch": 5.43778801843318, "grad_norm": 0.13962870836257935, "learning_rate": 9.970990458760215e-05, "loss": 0.0038, "step": 8260 }, { "epoch": 5.444371296905859, "grad_norm": 0.1537601798772812, "learning_rate": 9.970812332047929e-05, "loss": 0.0039, "step": 8270 }, { "epoch": 5.450954575378539, "grad_norm": 0.141169935464859, "learning_rate": 9.97063366173508e-05, "loss": 0.004, "step": 8280 }, { "epoch": 5.457537853851218, "grad_norm": 0.12763714790344238, "learning_rate": 9.970454447841207e-05, "loss": 0.004, "step": 8290 }, { "epoch": 5.464121132323897, "grad_norm": 0.13144730031490326, "learning_rate": 9.970274690385909e-05, "loss": 0.002, "step": 8300 }, { "epoch": 5.470704410796577, "grad_norm": 0.13234877586364746, "learning_rate": 9.970094389388844e-05, "loss": 0.0017, "step": 8310 }, { "epoch": 5.477287689269256, "grad_norm": 0.1626807600259781, "learning_rate": 9.969913544869728e-05, "loss": 0.0036, "step": 8320 }, { "epoch": 5.483870967741936, "grad_norm": 0.06375545263290405, "learning_rate": 9.96973215684834e-05, "loss": 0.0031, "step": 8330 }, { "epoch": 5.490454246214615, "grad_norm": 0.1641947478055954, "learning_rate": 9.969550225344513e-05, "loss": 0.0065, "step": 8340 }, { "epoch": 5.497037524687294, "grad_norm": 0.1599036008119583, "learning_rate": 9.969367750378147e-05, "loss": 0.0026, "step": 8350 }, { "epoch": 5.503620803159974, "grad_norm": 0.17979788780212402, "learning_rate": 9.969184731969194e-05, "loss": 0.0073, "step": 8360 }, { "epoch": 5.510204081632653, "grad_norm": 0.18441416323184967, "learning_rate": 9.96900117013767e-05, "loss": 0.0086, "step": 8370 }, { "epoch": 5.516787360105332, "grad_norm": 0.1496792435646057, "learning_rate": 9.96881706490365e-05, "loss": 0.0047, "step": 8380 }, { "epoch": 5.523370638578012, "grad_norm": 0.18887898325920105, "learning_rate": 9.968632416287265e-05, "loss": 0.0042, "step": 8390 }, { "epoch": 5.529953917050691, "grad_norm": 0.13192118704319, "learning_rate": 9.96844722430871e-05, "loss": 0.0031, "step": 8400 }, { "epoch": 5.536537195523371, "grad_norm": 0.1630241423845291, "learning_rate": 9.968261488988235e-05, "loss": 0.0033, "step": 8410 }, { "epoch": 5.54312047399605, "grad_norm": 0.17109939455986023, "learning_rate": 9.968075210346155e-05, "loss": 0.0028, "step": 8420 }, { "epoch": 5.549703752468729, "grad_norm": 0.18551169335842133, "learning_rate": 9.967888388402839e-05, "loss": 0.0042, "step": 8430 }, { "epoch": 5.556287030941409, "grad_norm": 0.13661743700504303, "learning_rate": 9.967701023178717e-05, "loss": 0.0025, "step": 8440 }, { "epoch": 5.562870309414088, "grad_norm": 0.19721899926662445, "learning_rate": 9.967513114694282e-05, "loss": 0.0035, "step": 8450 }, { "epoch": 5.569453587886768, "grad_norm": 0.1455097496509552, "learning_rate": 9.967324662970079e-05, "loss": 0.0026, "step": 8460 }, { "epoch": 5.576036866359447, "grad_norm": 0.10771012306213379, "learning_rate": 9.96713566802672e-05, "loss": 0.0047, "step": 8470 }, { "epoch": 5.582620144832126, "grad_norm": 0.2040475457906723, "learning_rate": 9.966946129884873e-05, "loss": 0.0048, "step": 8480 }, { "epoch": 5.589203423304806, "grad_norm": 0.1334817111492157, "learning_rate": 9.966756048565265e-05, "loss": 0.0037, "step": 8490 }, { "epoch": 5.595786701777485, "grad_norm": 0.22243735194206238, "learning_rate": 9.966565424088681e-05, "loss": 0.006, "step": 8500 }, { "epoch": 5.602369980250165, "grad_norm": 0.18287323415279388, "learning_rate": 9.96637425647597e-05, "loss": 0.0113, "step": 8510 }, { "epoch": 5.608953258722844, "grad_norm": 0.11420335620641708, "learning_rate": 9.966182545748038e-05, "loss": 0.0036, "step": 8520 }, { "epoch": 5.615536537195523, "grad_norm": 0.24647900462150574, "learning_rate": 9.96599029192585e-05, "loss": 0.006, "step": 8530 }, { "epoch": 5.622119815668203, "grad_norm": 0.19122307002544403, "learning_rate": 9.965797495030428e-05, "loss": 0.0036, "step": 8540 }, { "epoch": 5.628703094140882, "grad_norm": 0.12318669259548187, "learning_rate": 9.96560415508286e-05, "loss": 0.0036, "step": 8550 }, { "epoch": 5.635286372613562, "grad_norm": 0.11982472985982895, "learning_rate": 9.965410272104286e-05, "loss": 0.0037, "step": 8560 }, { "epoch": 5.641869651086241, "grad_norm": 0.11362205445766449, "learning_rate": 9.96521584611591e-05, "loss": 0.0025, "step": 8570 }, { "epoch": 5.64845292955892, "grad_norm": 0.1960466057062149, "learning_rate": 9.965020877138994e-05, "loss": 0.0051, "step": 8580 }, { "epoch": 5.6550362080316, "grad_norm": 0.18973922729492188, "learning_rate": 9.964825365194861e-05, "loss": 0.0048, "step": 8590 }, { "epoch": 5.6616194865042795, "grad_norm": 0.20897580683231354, "learning_rate": 9.96462931030489e-05, "loss": 0.0044, "step": 8600 }, { "epoch": 5.668202764976958, "grad_norm": 0.18020887672901154, "learning_rate": 9.96443271249052e-05, "loss": 0.0034, "step": 8610 }, { "epoch": 5.674786043449638, "grad_norm": 0.1803027242422104, "learning_rate": 9.964235571773255e-05, "loss": 0.0027, "step": 8620 }, { "epoch": 5.681369321922317, "grad_norm": 0.168981671333313, "learning_rate": 9.96403788817465e-05, "loss": 0.0065, "step": 8630 }, { "epoch": 5.687952600394997, "grad_norm": 0.13090486824512482, "learning_rate": 9.963839661716325e-05, "loss": 0.0025, "step": 8640 }, { "epoch": 5.694535878867676, "grad_norm": 0.10865361988544464, "learning_rate": 9.963640892419958e-05, "loss": 0.0042, "step": 8650 }, { "epoch": 5.701119157340355, "grad_norm": 0.16280893981456757, "learning_rate": 9.963441580307286e-05, "loss": 0.0028, "step": 8660 }, { "epoch": 5.707702435813035, "grad_norm": 0.12608975172042847, "learning_rate": 9.963241725400104e-05, "loss": 0.004, "step": 8670 }, { "epoch": 5.714285714285714, "grad_norm": 0.20025120675563812, "learning_rate": 9.963041327720271e-05, "loss": 0.0036, "step": 8680 }, { "epoch": 5.720868992758394, "grad_norm": 0.2411954700946808, "learning_rate": 9.962840387289697e-05, "loss": 0.0036, "step": 8690 }, { "epoch": 5.727452271231073, "grad_norm": 0.11536764353513718, "learning_rate": 9.962638904130363e-05, "loss": 0.0035, "step": 8700 }, { "epoch": 5.734035549703752, "grad_norm": 0.19403482973575592, "learning_rate": 9.962436878264298e-05, "loss": 0.0029, "step": 8710 }, { "epoch": 5.740618828176432, "grad_norm": 0.16452810168266296, "learning_rate": 9.962234309713598e-05, "loss": 0.0035, "step": 8720 }, { "epoch": 5.7472021066491115, "grad_norm": 0.11054060608148575, "learning_rate": 9.962031198500414e-05, "loss": 0.0033, "step": 8730 }, { "epoch": 5.753785385121791, "grad_norm": 0.14504188299179077, "learning_rate": 9.961827544646958e-05, "loss": 0.0028, "step": 8740 }, { "epoch": 5.76036866359447, "grad_norm": 0.11901656538248062, "learning_rate": 9.961623348175501e-05, "loss": 0.0025, "step": 8750 }, { "epoch": 5.766951942067149, "grad_norm": 0.20375290513038635, "learning_rate": 9.961418609108377e-05, "loss": 0.0037, "step": 8760 }, { "epoch": 5.773535220539829, "grad_norm": 0.13122229278087616, "learning_rate": 9.961213327467971e-05, "loss": 0.0043, "step": 8770 }, { "epoch": 5.7801184990125085, "grad_norm": 0.190016970038414, "learning_rate": 9.961007503276736e-05, "loss": 0.0062, "step": 8780 }, { "epoch": 5.786701777485188, "grad_norm": 0.13874346017837524, "learning_rate": 9.960801136557179e-05, "loss": 0.0044, "step": 8790 }, { "epoch": 5.793285055957867, "grad_norm": 0.22725573182106018, "learning_rate": 9.960594227331866e-05, "loss": 0.0072, "step": 8800 }, { "epoch": 5.799868334430546, "grad_norm": 0.22128959000110626, "learning_rate": 9.960386775623429e-05, "loss": 0.0051, "step": 8810 }, { "epoch": 5.806451612903226, "grad_norm": 0.14470231533050537, "learning_rate": 9.96017878145455e-05, "loss": 0.0035, "step": 8820 }, { "epoch": 5.813034891375906, "grad_norm": 0.2545865476131439, "learning_rate": 9.959970244847977e-05, "loss": 0.0064, "step": 8830 }, { "epoch": 5.819618169848584, "grad_norm": 0.22424954175949097, "learning_rate": 9.959761165826518e-05, "loss": 0.0045, "step": 8840 }, { "epoch": 5.826201448321264, "grad_norm": 0.13974206149578094, "learning_rate": 9.959551544413033e-05, "loss": 0.0053, "step": 8850 }, { "epoch": 5.832784726793943, "grad_norm": 0.13326512277126312, "learning_rate": 9.959341380630448e-05, "loss": 0.0059, "step": 8860 }, { "epoch": 5.839368005266623, "grad_norm": 0.25552642345428467, "learning_rate": 9.959130674501746e-05, "loss": 0.0057, "step": 8870 }, { "epoch": 5.845951283739302, "grad_norm": 0.21832486987113953, "learning_rate": 9.958919426049968e-05, "loss": 0.0082, "step": 8880 }, { "epoch": 5.852534562211981, "grad_norm": 0.20759667456150055, "learning_rate": 9.958707635298219e-05, "loss": 0.0055, "step": 8890 }, { "epoch": 5.859117840684661, "grad_norm": 0.1811683475971222, "learning_rate": 9.958495302269657e-05, "loss": 0.0027, "step": 8900 }, { "epoch": 5.8657011191573405, "grad_norm": 0.1952507644891739, "learning_rate": 9.958282426987503e-05, "loss": 0.0051, "step": 8910 }, { "epoch": 5.87228439763002, "grad_norm": 0.19793321192264557, "learning_rate": 9.95806900947504e-05, "loss": 0.0037, "step": 8920 }, { "epoch": 5.878867676102699, "grad_norm": 0.1996155083179474, "learning_rate": 9.957855049755604e-05, "loss": 0.0035, "step": 8930 }, { "epoch": 5.885450954575378, "grad_norm": 0.13972783088684082, "learning_rate": 9.957640547852593e-05, "loss": 0.0038, "step": 8940 }, { "epoch": 5.892034233048058, "grad_norm": 0.090321384370327, "learning_rate": 9.957425503789466e-05, "loss": 0.0035, "step": 8950 }, { "epoch": 5.8986175115207375, "grad_norm": 0.19212886691093445, "learning_rate": 9.957209917589738e-05, "loss": 0.0043, "step": 8960 }, { "epoch": 5.905200789993417, "grad_norm": 0.13126474618911743, "learning_rate": 9.956993789276987e-05, "loss": 0.0031, "step": 8970 }, { "epoch": 5.911784068466096, "grad_norm": 0.2920091152191162, "learning_rate": 9.956777118874847e-05, "loss": 0.0068, "step": 8980 }, { "epoch": 5.918367346938775, "grad_norm": 0.16958096623420715, "learning_rate": 9.956559906407016e-05, "loss": 0.0067, "step": 8990 }, { "epoch": 5.924950625411455, "grad_norm": 0.20367182791233063, "learning_rate": 9.956342151897245e-05, "loss": 0.005, "step": 9000 }, { "epoch": 5.931533903884135, "grad_norm": 0.17489448189735413, "learning_rate": 9.956123855369346e-05, "loss": 0.0067, "step": 9010 }, { "epoch": 5.938117182356814, "grad_norm": 0.21678303182125092, "learning_rate": 9.955905016847196e-05, "loss": 0.0068, "step": 9020 }, { "epoch": 5.944700460829493, "grad_norm": 0.19352510571479797, "learning_rate": 9.955685636354723e-05, "loss": 0.0046, "step": 9030 }, { "epoch": 5.9512837393021725, "grad_norm": 0.12152896821498871, "learning_rate": 9.95546571391592e-05, "loss": 0.0039, "step": 9040 }, { "epoch": 5.957867017774852, "grad_norm": 0.1558520793914795, "learning_rate": 9.955245249554837e-05, "loss": 0.0048, "step": 9050 }, { "epoch": 5.964450296247532, "grad_norm": 0.20076215267181396, "learning_rate": 9.955024243295582e-05, "loss": 0.0031, "step": 9060 }, { "epoch": 5.97103357472021, "grad_norm": 0.13397449254989624, "learning_rate": 9.954802695162328e-05, "loss": 0.004, "step": 9070 }, { "epoch": 5.97761685319289, "grad_norm": 0.12151654064655304, "learning_rate": 9.954580605179302e-05, "loss": 0.0032, "step": 9080 }, { "epoch": 5.9842001316655695, "grad_norm": 0.2205461710691452, "learning_rate": 9.954357973370788e-05, "loss": 0.0039, "step": 9090 }, { "epoch": 5.990783410138249, "grad_norm": 0.12803231179714203, "learning_rate": 9.954134799761135e-05, "loss": 0.0038, "step": 9100 }, { "epoch": 5.997366688610928, "grad_norm": 0.15411901473999023, "learning_rate": 9.953911084374748e-05, "loss": 0.0035, "step": 9110 }, { "epoch": 6.003949967083607, "grad_norm": 0.2388690859079361, "learning_rate": 9.953686827236093e-05, "loss": 0.0033, "step": 9120 }, { "epoch": 6.010533245556287, "grad_norm": 0.12717680633068085, "learning_rate": 9.953462028369695e-05, "loss": 0.0035, "step": 9130 }, { "epoch": 6.017116524028967, "grad_norm": 0.11664164811372757, "learning_rate": 9.953236687800136e-05, "loss": 0.003, "step": 9140 }, { "epoch": 6.023699802501646, "grad_norm": 0.207468181848526, "learning_rate": 9.95301080555206e-05, "loss": 0.0031, "step": 9150 }, { "epoch": 6.030283080974325, "grad_norm": 0.11771786957979202, "learning_rate": 9.952784381650171e-05, "loss": 0.0038, "step": 9160 }, { "epoch": 6.0368663594470044, "grad_norm": 0.10141989588737488, "learning_rate": 9.952557416119226e-05, "loss": 0.0033, "step": 9170 }, { "epoch": 6.043449637919684, "grad_norm": 0.14697608351707458, "learning_rate": 9.95232990898405e-05, "loss": 0.0024, "step": 9180 }, { "epoch": 6.050032916392364, "grad_norm": 0.19397738575935364, "learning_rate": 9.95210186026952e-05, "loss": 0.0033, "step": 9190 }, { "epoch": 6.056616194865043, "grad_norm": 0.09810974448919296, "learning_rate": 9.951873270000576e-05, "loss": 0.0035, "step": 9200 }, { "epoch": 6.063199473337722, "grad_norm": 0.18403321504592896, "learning_rate": 9.951644138202216e-05, "loss": 0.004, "step": 9210 }, { "epoch": 6.0697827518104015, "grad_norm": 0.1438199132680893, "learning_rate": 9.951414464899498e-05, "loss": 0.0027, "step": 9220 }, { "epoch": 6.076366030283081, "grad_norm": 0.2365601509809494, "learning_rate": 9.951184250117538e-05, "loss": 0.0051, "step": 9230 }, { "epoch": 6.082949308755761, "grad_norm": 0.1394122838973999, "learning_rate": 9.950953493881513e-05, "loss": 0.0044, "step": 9240 }, { "epoch": 6.089532587228439, "grad_norm": 0.12571769952774048, "learning_rate": 9.950722196216658e-05, "loss": 0.0043, "step": 9250 }, { "epoch": 6.096115865701119, "grad_norm": 0.16777974367141724, "learning_rate": 9.950490357148265e-05, "loss": 0.0032, "step": 9260 }, { "epoch": 6.1026991441737986, "grad_norm": 0.1852402538061142, "learning_rate": 9.950257976701692e-05, "loss": 0.0034, "step": 9270 }, { "epoch": 6.109282422646478, "grad_norm": 0.12286099046468735, "learning_rate": 9.950025054902348e-05, "loss": 0.0061, "step": 9280 }, { "epoch": 6.115865701119158, "grad_norm": 0.18889589607715607, "learning_rate": 9.949791591775706e-05, "loss": 0.0034, "step": 9290 }, { "epoch": 6.122448979591836, "grad_norm": 0.21235306560993195, "learning_rate": 9.949557587347298e-05, "loss": 0.0061, "step": 9300 }, { "epoch": 6.129032258064516, "grad_norm": 0.21616734564304352, "learning_rate": 9.949323041642713e-05, "loss": 0.0044, "step": 9310 }, { "epoch": 6.135615536537196, "grad_norm": 0.20806770026683807, "learning_rate": 9.949087954687602e-05, "loss": 0.0128, "step": 9320 }, { "epoch": 6.142198815009875, "grad_norm": 0.21776287257671356, "learning_rate": 9.948852326507672e-05, "loss": 0.0048, "step": 9330 }, { "epoch": 6.148782093482554, "grad_norm": 0.21416909992694855, "learning_rate": 9.948616157128694e-05, "loss": 0.0053, "step": 9340 }, { "epoch": 6.1553653719552335, "grad_norm": 0.1293257623910904, "learning_rate": 9.948379446576493e-05, "loss": 0.0032, "step": 9350 }, { "epoch": 6.161948650427913, "grad_norm": 0.14805038273334503, "learning_rate": 9.948142194876952e-05, "loss": 0.0038, "step": 9360 }, { "epoch": 6.168531928900593, "grad_norm": 0.14277274906635284, "learning_rate": 9.947904402056024e-05, "loss": 0.0083, "step": 9370 }, { "epoch": 6.175115207373272, "grad_norm": 0.1809692531824112, "learning_rate": 9.947666068139708e-05, "loss": 0.0034, "step": 9380 }, { "epoch": 6.181698485845951, "grad_norm": 0.17821776866912842, "learning_rate": 9.947427193154071e-05, "loss": 0.0116, "step": 9390 }, { "epoch": 6.1882817643186305, "grad_norm": 0.21515420079231262, "learning_rate": 9.947187777125233e-05, "loss": 0.0039, "step": 9400 }, { "epoch": 6.19486504279131, "grad_norm": 0.2290203720331192, "learning_rate": 9.946947820079377e-05, "loss": 0.0043, "step": 9410 }, { "epoch": 6.20144832126399, "grad_norm": 0.18981987237930298, "learning_rate": 9.946707322042747e-05, "loss": 0.0035, "step": 9420 }, { "epoch": 6.208031599736669, "grad_norm": 0.16829422116279602, "learning_rate": 9.94646628304164e-05, "loss": 0.0043, "step": 9430 }, { "epoch": 6.214614878209348, "grad_norm": 0.1351151466369629, "learning_rate": 9.946224703102418e-05, "loss": 0.0035, "step": 9440 }, { "epoch": 6.221198156682028, "grad_norm": 0.16584992408752441, "learning_rate": 9.945982582251498e-05, "loss": 0.0067, "step": 9450 }, { "epoch": 6.227781435154707, "grad_norm": 0.13488180935382843, "learning_rate": 9.94573992051536e-05, "loss": 0.0029, "step": 9460 }, { "epoch": 6.234364713627387, "grad_norm": 0.1701415330171585, "learning_rate": 9.94549671792054e-05, "loss": 0.0045, "step": 9470 }, { "epoch": 6.2409479921000655, "grad_norm": 0.12126437574625015, "learning_rate": 9.945252974493635e-05, "loss": 0.0041, "step": 9480 }, { "epoch": 6.247531270572745, "grad_norm": 0.249045729637146, "learning_rate": 9.9450086902613e-05, "loss": 0.0042, "step": 9490 }, { "epoch": 6.254114549045425, "grad_norm": 0.19939686357975006, "learning_rate": 9.944763865250248e-05, "loss": 0.0036, "step": 9500 }, { "epoch": 6.260697827518104, "grad_norm": 0.14088821411132812, "learning_rate": 9.944518499487254e-05, "loss": 0.0045, "step": 9510 }, { "epoch": 6.267281105990784, "grad_norm": 0.133262500166893, "learning_rate": 9.944272592999151e-05, "loss": 0.0039, "step": 9520 }, { "epoch": 6.2738643844634625, "grad_norm": 0.1991509199142456, "learning_rate": 9.94402614581283e-05, "loss": 0.005, "step": 9530 }, { "epoch": 6.280447662936142, "grad_norm": 0.17973390221595764, "learning_rate": 9.943779157955244e-05, "loss": 0.0038, "step": 9540 }, { "epoch": 6.287030941408822, "grad_norm": 0.17198804020881653, "learning_rate": 9.943531629453403e-05, "loss": 0.0063, "step": 9550 }, { "epoch": 6.293614219881501, "grad_norm": 0.10564780235290527, "learning_rate": 9.943283560334375e-05, "loss": 0.0034, "step": 9560 }, { "epoch": 6.30019749835418, "grad_norm": 0.10555972158908844, "learning_rate": 9.943034950625288e-05, "loss": 0.0049, "step": 9570 }, { "epoch": 6.30678077682686, "grad_norm": 0.13958945870399475, "learning_rate": 9.942785800353332e-05, "loss": 0.0025, "step": 9580 }, { "epoch": 6.313364055299539, "grad_norm": 0.13087177276611328, "learning_rate": 9.942536109545751e-05, "loss": 0.003, "step": 9590 }, { "epoch": 6.319947333772219, "grad_norm": 0.12887345254421234, "learning_rate": 9.942285878229853e-05, "loss": 0.0097, "step": 9600 }, { "epoch": 6.326530612244898, "grad_norm": 0.1350538581609726, "learning_rate": 9.942035106433001e-05, "loss": 0.0026, "step": 9610 }, { "epoch": 6.333113890717577, "grad_norm": 0.14022459089756012, "learning_rate": 9.94178379418262e-05, "loss": 0.0033, "step": 9620 }, { "epoch": 6.339697169190257, "grad_norm": 0.08853518217802048, "learning_rate": 9.941531941506194e-05, "loss": 0.0025, "step": 9630 }, { "epoch": 6.346280447662936, "grad_norm": 0.16832095384597778, "learning_rate": 9.941279548431263e-05, "loss": 0.0025, "step": 9640 }, { "epoch": 6.352863726135616, "grad_norm": 0.20263363420963287, "learning_rate": 9.941026614985431e-05, "loss": 0.0041, "step": 9650 }, { "epoch": 6.359447004608295, "grad_norm": 0.21911534667015076, "learning_rate": 9.940773141196357e-05, "loss": 0.0045, "step": 9660 }, { "epoch": 6.366030283080974, "grad_norm": 0.2004307210445404, "learning_rate": 9.94051912709176e-05, "loss": 0.0176, "step": 9670 }, { "epoch": 6.372613561553654, "grad_norm": 0.2188200056552887, "learning_rate": 9.940264572699421e-05, "loss": 0.005, "step": 9680 }, { "epoch": 6.379196840026333, "grad_norm": 0.14378902316093445, "learning_rate": 9.940009478047174e-05, "loss": 0.0085, "step": 9690 }, { "epoch": 6.385780118499013, "grad_norm": 0.11846856772899628, "learning_rate": 9.939753843162918e-05, "loss": 0.0044, "step": 9700 }, { "epoch": 6.3923633969716915, "grad_norm": 0.15709710121154785, "learning_rate": 9.939497668074609e-05, "loss": 0.0032, "step": 9710 }, { "epoch": 6.398946675444371, "grad_norm": 0.1545567810535431, "learning_rate": 9.93924095281026e-05, "loss": 0.007, "step": 9720 }, { "epoch": 6.405529953917051, "grad_norm": 0.07838129997253418, "learning_rate": 9.938983697397948e-05, "loss": 0.0033, "step": 9730 }, { "epoch": 6.41211323238973, "grad_norm": 0.1433374583721161, "learning_rate": 9.938725901865805e-05, "loss": 0.0046, "step": 9740 }, { "epoch": 6.41869651086241, "grad_norm": 0.0984044224023819, "learning_rate": 9.93846756624202e-05, "loss": 0.007, "step": 9750 }, { "epoch": 6.425279789335089, "grad_norm": 0.16751566529273987, "learning_rate": 9.938208690554849e-05, "loss": 0.0052, "step": 9760 }, { "epoch": 6.431863067807768, "grad_norm": 0.20086205005645752, "learning_rate": 9.9379492748326e-05, "loss": 0.0029, "step": 9770 }, { "epoch": 6.438446346280448, "grad_norm": 0.24653929471969604, "learning_rate": 9.937689319103641e-05, "loss": 0.005, "step": 9780 }, { "epoch": 6.445029624753127, "grad_norm": 0.19016973674297333, "learning_rate": 9.937428823396404e-05, "loss": 0.0042, "step": 9790 }, { "epoch": 6.451612903225806, "grad_norm": 0.2207612544298172, "learning_rate": 9.937167787739372e-05, "loss": 0.0051, "step": 9800 }, { "epoch": 6.458196181698486, "grad_norm": 0.19792930781841278, "learning_rate": 9.936906212161095e-05, "loss": 0.0045, "step": 9810 }, { "epoch": 6.464779460171165, "grad_norm": 0.11168951541185379, "learning_rate": 9.936644096690176e-05, "loss": 0.0045, "step": 9820 }, { "epoch": 6.471362738643845, "grad_norm": 0.15161779522895813, "learning_rate": 9.936381441355282e-05, "loss": 0.0035, "step": 9830 }, { "epoch": 6.477946017116524, "grad_norm": 0.1703379601240158, "learning_rate": 9.936118246185136e-05, "loss": 0.0031, "step": 9840 }, { "epoch": 6.484529295589203, "grad_norm": 0.17777413129806519, "learning_rate": 9.935854511208518e-05, "loss": 0.0043, "step": 9850 }, { "epoch": 6.491112574061883, "grad_norm": 0.2139226347208023, "learning_rate": 9.935590236454272e-05, "loss": 0.0043, "step": 9860 }, { "epoch": 6.497695852534562, "grad_norm": 0.14054365456104279, "learning_rate": 9.935325421951298e-05, "loss": 0.0029, "step": 9870 }, { "epoch": 6.504279131007242, "grad_norm": 0.1933344304561615, "learning_rate": 9.935060067728557e-05, "loss": 0.0043, "step": 9880 }, { "epoch": 6.5108624094799215, "grad_norm": 0.19287428259849548, "learning_rate": 9.934794173815067e-05, "loss": 0.0045, "step": 9890 }, { "epoch": 6.5174456879526, "grad_norm": 0.1585034430027008, "learning_rate": 9.934527740239906e-05, "loss": 0.0054, "step": 9900 }, { "epoch": 6.52402896642528, "grad_norm": 0.1511351615190506, "learning_rate": 9.934260767032209e-05, "loss": 0.0042, "step": 9910 }, { "epoch": 6.530612244897959, "grad_norm": 0.1429518312215805, "learning_rate": 9.933993254221172e-05, "loss": 0.0023, "step": 9920 }, { "epoch": 6.537195523370639, "grad_norm": 0.27772828936576843, "learning_rate": 9.933725201836053e-05, "loss": 0.0092, "step": 9930 }, { "epoch": 6.543778801843318, "grad_norm": 0.1809336096048355, "learning_rate": 9.933456609906162e-05, "loss": 0.0044, "step": 9940 }, { "epoch": 6.550362080315997, "grad_norm": 0.20517009496688843, "learning_rate": 9.933187478460875e-05, "loss": 0.0097, "step": 9950 }, { "epoch": 6.556945358788677, "grad_norm": 0.14628200232982635, "learning_rate": 9.93291780752962e-05, "loss": 0.0037, "step": 9960 }, { "epoch": 6.563528637261356, "grad_norm": 0.17714634537696838, "learning_rate": 9.932647597141893e-05, "loss": 0.0037, "step": 9970 }, { "epoch": 6.570111915734035, "grad_norm": 0.16576094925403595, "learning_rate": 9.932376847327239e-05, "loss": 0.0037, "step": 9980 }, { "epoch": 6.576695194206715, "grad_norm": 0.52414870262146, "learning_rate": 9.932105558115268e-05, "loss": 0.0037, "step": 9990 }, { "epoch": 6.583278472679394, "grad_norm": 0.21010442078113556, "learning_rate": 9.931833729535651e-05, "loss": 0.004, "step": 10000 }, { "epoch": 6.589861751152074, "grad_norm": 0.15539118647575378, "learning_rate": 9.931561361618111e-05, "loss": 0.0036, "step": 10010 }, { "epoch": 6.596445029624753, "grad_norm": 0.12251342087984085, "learning_rate": 9.931288454392435e-05, "loss": 0.0082, "step": 10020 }, { "epoch": 6.603028308097432, "grad_norm": 0.1267755776643753, "learning_rate": 9.931015007888467e-05, "loss": 0.0033, "step": 10030 }, { "epoch": 6.609611586570112, "grad_norm": 0.1436208188533783, "learning_rate": 9.930741022136112e-05, "loss": 0.0045, "step": 10040 }, { "epoch": 6.616194865042791, "grad_norm": 0.14464491605758667, "learning_rate": 9.930466497165333e-05, "loss": 0.0053, "step": 10050 }, { "epoch": 6.622778143515471, "grad_norm": 0.13660353422164917, "learning_rate": 9.93019143300615e-05, "loss": 0.0038, "step": 10060 }, { "epoch": 6.6293614219881505, "grad_norm": 0.10352449864149094, "learning_rate": 9.929915829688644e-05, "loss": 0.0039, "step": 10070 }, { "epoch": 6.635944700460829, "grad_norm": 0.18688145279884338, "learning_rate": 9.929639687242955e-05, "loss": 0.0043, "step": 10080 }, { "epoch": 6.642527978933509, "grad_norm": 0.24405129253864288, "learning_rate": 9.929363005699281e-05, "loss": 0.0077, "step": 10090 }, { "epoch": 6.649111257406188, "grad_norm": 0.1363968402147293, "learning_rate": 9.92908578508788e-05, "loss": 0.0035, "step": 10100 }, { "epoch": 6.655694535878868, "grad_norm": 0.11649111658334732, "learning_rate": 9.928808025439069e-05, "loss": 0.0035, "step": 10110 }, { "epoch": 6.6622778143515475, "grad_norm": 0.1950749158859253, "learning_rate": 9.928529726783223e-05, "loss": 0.006, "step": 10120 }, { "epoch": 6.668861092824226, "grad_norm": 0.14854936301708221, "learning_rate": 9.928250889150774e-05, "loss": 0.0027, "step": 10130 }, { "epoch": 6.675444371296906, "grad_norm": 0.13930678367614746, "learning_rate": 9.92797151257222e-05, "loss": 0.0026, "step": 10140 }, { "epoch": 6.682027649769585, "grad_norm": 0.09725431352853775, "learning_rate": 9.927691597078108e-05, "loss": 0.0046, "step": 10150 }, { "epoch": 6.688610928242265, "grad_norm": 0.14973102509975433, "learning_rate": 9.927411142699053e-05, "loss": 0.0049, "step": 10160 }, { "epoch": 6.695194206714944, "grad_norm": 0.23969830572605133, "learning_rate": 9.927130149465725e-05, "loss": 0.0046, "step": 10170 }, { "epoch": 6.701777485187623, "grad_norm": 0.17006459832191467, "learning_rate": 9.92684861740885e-05, "loss": 0.0041, "step": 10180 }, { "epoch": 6.708360763660303, "grad_norm": 0.24886395037174225, "learning_rate": 9.926566546559217e-05, "loss": 0.0096, "step": 10190 }, { "epoch": 6.7149440421329825, "grad_norm": 0.16121667623519897, "learning_rate": 9.926283936947673e-05, "loss": 0.0039, "step": 10200 }, { "epoch": 6.721527320605661, "grad_norm": 0.182964488863945, "learning_rate": 9.926000788605126e-05, "loss": 0.0039, "step": 10210 }, { "epoch": 6.728110599078341, "grad_norm": 0.1523403823375702, "learning_rate": 9.92571710156254e-05, "loss": 0.0038, "step": 10220 }, { "epoch": 6.73469387755102, "grad_norm": 0.23936374485492706, "learning_rate": 9.925432875850936e-05, "loss": 0.0084, "step": 10230 }, { "epoch": 6.7412771560237, "grad_norm": 0.23110678791999817, "learning_rate": 9.925148111501396e-05, "loss": 0.0059, "step": 10240 }, { "epoch": 6.7478604344963795, "grad_norm": 0.16122660040855408, "learning_rate": 9.924862808545066e-05, "loss": 0.0043, "step": 10250 }, { "epoch": 6.754443712969058, "grad_norm": 0.18625423312187195, "learning_rate": 9.924576967013141e-05, "loss": 0.003, "step": 10260 }, { "epoch": 6.761026991441738, "grad_norm": 0.21271024644374847, "learning_rate": 9.924290586936887e-05, "loss": 0.0048, "step": 10270 }, { "epoch": 6.767610269914417, "grad_norm": 0.13807539641857147, "learning_rate": 9.924003668347614e-05, "loss": 0.0037, "step": 10280 }, { "epoch": 6.774193548387097, "grad_norm": 0.10380616039037704, "learning_rate": 9.923716211276704e-05, "loss": 0.0028, "step": 10290 }, { "epoch": 6.780776826859777, "grad_norm": 0.14279644191265106, "learning_rate": 9.923428215755594e-05, "loss": 0.0044, "step": 10300 }, { "epoch": 6.787360105332455, "grad_norm": 0.12905529141426086, "learning_rate": 9.923139681815775e-05, "loss": 0.0055, "step": 10310 }, { "epoch": 6.793943383805135, "grad_norm": 0.17455433309078217, "learning_rate": 9.922850609488801e-05, "loss": 0.0034, "step": 10320 }, { "epoch": 6.800526662277814, "grad_norm": 0.1157543882727623, "learning_rate": 9.922560998806287e-05, "loss": 0.0025, "step": 10330 }, { "epoch": 6.807109940750494, "grad_norm": 0.13407455384731293, "learning_rate": 9.922270849799905e-05, "loss": 0.0024, "step": 10340 }, { "epoch": 6.813693219223174, "grad_norm": 0.13540343940258026, "learning_rate": 9.92198016250138e-05, "loss": 0.0037, "step": 10350 }, { "epoch": 6.820276497695852, "grad_norm": 0.13606594502925873, "learning_rate": 9.921688936942506e-05, "loss": 0.0034, "step": 10360 }, { "epoch": 6.826859776168532, "grad_norm": 0.09382519125938416, "learning_rate": 9.921397173155129e-05, "loss": 0.0076, "step": 10370 }, { "epoch": 6.8334430546412115, "grad_norm": 0.14648884534835815, "learning_rate": 9.921104871171157e-05, "loss": 0.0046, "step": 10380 }, { "epoch": 6.840026333113891, "grad_norm": 0.15996293723583221, "learning_rate": 9.920812031022554e-05, "loss": 0.0062, "step": 10390 }, { "epoch": 6.84660961158657, "grad_norm": 0.21178320050239563, "learning_rate": 9.920518652741348e-05, "loss": 0.0047, "step": 10400 }, { "epoch": 6.853192890059249, "grad_norm": 0.20493119955062866, "learning_rate": 9.920224736359618e-05, "loss": 0.0041, "step": 10410 }, { "epoch": 6.859776168531929, "grad_norm": 0.1787489503622055, "learning_rate": 9.91993028190951e-05, "loss": 0.0035, "step": 10420 }, { "epoch": 6.8663594470046085, "grad_norm": 0.17640015482902527, "learning_rate": 9.919635289423222e-05, "loss": 0.0044, "step": 10430 }, { "epoch": 6.872942725477287, "grad_norm": 0.1444956511259079, "learning_rate": 9.919339758933015e-05, "loss": 0.0039, "step": 10440 }, { "epoch": 6.879526003949967, "grad_norm": 0.13196717202663422, "learning_rate": 9.919043690471209e-05, "loss": 0.0044, "step": 10450 }, { "epoch": 6.886109282422646, "grad_norm": 0.14714795351028442, "learning_rate": 9.91874708407018e-05, "loss": 0.0045, "step": 10460 }, { "epoch": 6.892692560895326, "grad_norm": 0.10407540947198868, "learning_rate": 9.918449939762367e-05, "loss": 0.003, "step": 10470 }, { "epoch": 6.899275839368006, "grad_norm": 0.1378217190504074, "learning_rate": 9.91815225758026e-05, "loss": 0.0044, "step": 10480 }, { "epoch": 6.905859117840684, "grad_norm": 0.11981435865163803, "learning_rate": 9.917854037556419e-05, "loss": 0.0022, "step": 10490 }, { "epoch": 6.912442396313364, "grad_norm": 0.15841242671012878, "learning_rate": 9.917555279723454e-05, "loss": 0.003, "step": 10500 }, { "epoch": 6.9190256747860435, "grad_norm": 0.126947283744812, "learning_rate": 9.917255984114036e-05, "loss": 0.0036, "step": 10510 }, { "epoch": 6.925608953258723, "grad_norm": 0.12283198535442352, "learning_rate": 9.916956150760896e-05, "loss": 0.0049, "step": 10520 }, { "epoch": 6.932192231731403, "grad_norm": 0.12930618226528168, "learning_rate": 9.916655779696826e-05, "loss": 0.0043, "step": 10530 }, { "epoch": 6.938775510204081, "grad_norm": 0.12479476630687714, "learning_rate": 9.916354870954671e-05, "loss": 0.0038, "step": 10540 }, { "epoch": 6.945358788676761, "grad_norm": 0.19381871819496155, "learning_rate": 9.91605342456734e-05, "loss": 0.0047, "step": 10550 }, { "epoch": 6.9519420671494405, "grad_norm": 0.21987885236740112, "learning_rate": 9.915751440567795e-05, "loss": 0.003, "step": 10560 }, { "epoch": 6.95852534562212, "grad_norm": 0.16727258265018463, "learning_rate": 9.915448918989066e-05, "loss": 0.0043, "step": 10570 }, { "epoch": 6.9651086240948, "grad_norm": 0.12566395103931427, "learning_rate": 9.915145859864232e-05, "loss": 0.0031, "step": 10580 }, { "epoch": 6.971691902567478, "grad_norm": 0.13294143974781036, "learning_rate": 9.914842263226437e-05, "loss": 0.0031, "step": 10590 }, { "epoch": 6.978275181040158, "grad_norm": 0.10941747575998306, "learning_rate": 9.914538129108882e-05, "loss": 0.0041, "step": 10600 }, { "epoch": 6.984858459512838, "grad_norm": 0.1465856432914734, "learning_rate": 9.914233457544825e-05, "loss": 0.003, "step": 10610 }, { "epoch": 6.991441737985517, "grad_norm": 0.11623964458703995, "learning_rate": 9.913928248567586e-05, "loss": 0.0028, "step": 10620 }, { "epoch": 6.998025016458196, "grad_norm": 0.11098870635032654, "learning_rate": 9.913622502210542e-05, "loss": 0.0035, "step": 10630 }, { "epoch": 7.0046082949308754, "grad_norm": 0.1629747748374939, "learning_rate": 9.913316218507128e-05, "loss": 0.0039, "step": 10640 }, { "epoch": 7.011191573403555, "grad_norm": 0.11901634186506271, "learning_rate": 9.91300939749084e-05, "loss": 0.003, "step": 10650 }, { "epoch": 7.017774851876235, "grad_norm": 0.11372654885053635, "learning_rate": 9.91270203919523e-05, "loss": 0.0029, "step": 10660 }, { "epoch": 7.024358130348913, "grad_norm": 0.09347504377365112, "learning_rate": 9.912394143653912e-05, "loss": 0.0042, "step": 10670 }, { "epoch": 7.030941408821593, "grad_norm": 0.18111474812030792, "learning_rate": 9.912085710900555e-05, "loss": 0.0047, "step": 10680 }, { "epoch": 7.0375246872942725, "grad_norm": 0.10008914768695831, "learning_rate": 9.911776740968892e-05, "loss": 0.003, "step": 10690 }, { "epoch": 7.044107965766952, "grad_norm": 0.11528833955526352, "learning_rate": 9.911467233892709e-05, "loss": 0.0034, "step": 10700 }, { "epoch": 7.050691244239632, "grad_norm": 0.10422374308109283, "learning_rate": 9.911157189705853e-05, "loss": 0.0054, "step": 10710 }, { "epoch": 7.05727452271231, "grad_norm": 0.10589602589607239, "learning_rate": 9.910846608442229e-05, "loss": 0.0027, "step": 10720 }, { "epoch": 7.06385780118499, "grad_norm": 0.12106330692768097, "learning_rate": 9.910535490135805e-05, "loss": 0.0051, "step": 10730 }, { "epoch": 7.0704410796576695, "grad_norm": 0.1140916496515274, "learning_rate": 9.910223834820603e-05, "loss": 0.0048, "step": 10740 }, { "epoch": 7.077024358130349, "grad_norm": 0.19260945916175842, "learning_rate": 9.909911642530703e-05, "loss": 0.0034, "step": 10750 }, { "epoch": 7.083607636603029, "grad_norm": 0.13183332979679108, "learning_rate": 9.909598913300249e-05, "loss": 0.0038, "step": 10760 }, { "epoch": 7.090190915075707, "grad_norm": 0.09123890846967697, "learning_rate": 9.909285647163438e-05, "loss": 0.0029, "step": 10770 }, { "epoch": 7.096774193548387, "grad_norm": 0.11492792516946793, "learning_rate": 9.908971844154531e-05, "loss": 0.003, "step": 10780 }, { "epoch": 7.103357472021067, "grad_norm": 0.13432548940181732, "learning_rate": 9.908657504307843e-05, "loss": 0.0024, "step": 10790 }, { "epoch": 7.109940750493746, "grad_norm": 0.09666921943426132, "learning_rate": 9.908342627657751e-05, "loss": 0.003, "step": 10800 }, { "epoch": 7.116524028966425, "grad_norm": 0.10835039615631104, "learning_rate": 9.908027214238689e-05, "loss": 0.0037, "step": 10810 }, { "epoch": 7.1231073074391045, "grad_norm": 0.07529739290475845, "learning_rate": 9.90771126408515e-05, "loss": 0.0026, "step": 10820 }, { "epoch": 7.129690585911784, "grad_norm": 0.1297670304775238, "learning_rate": 9.907394777231685e-05, "loss": 0.0024, "step": 10830 }, { "epoch": 7.136273864384464, "grad_norm": 0.07843507826328278, "learning_rate": 9.907077753712905e-05, "loss": 0.0019, "step": 10840 }, { "epoch": 7.142857142857143, "grad_norm": 0.1917518973350525, "learning_rate": 9.906760193563482e-05, "loss": 0.003, "step": 10850 }, { "epoch": 7.149440421329822, "grad_norm": 0.18574674427509308, "learning_rate": 9.906442096818139e-05, "loss": 0.003, "step": 10860 }, { "epoch": 7.1560236998025015, "grad_norm": 0.10715876519680023, "learning_rate": 9.906123463511665e-05, "loss": 0.0022, "step": 10870 }, { "epoch": 7.162606978275181, "grad_norm": 0.19454669952392578, "learning_rate": 9.905804293678907e-05, "loss": 0.0045, "step": 10880 }, { "epoch": 7.169190256747861, "grad_norm": 0.19382624328136444, "learning_rate": 9.905484587354766e-05, "loss": 0.0068, "step": 10890 }, { "epoch": 7.175773535220539, "grad_norm": 0.1616256982088089, "learning_rate": 9.905164344574205e-05, "loss": 0.0036, "step": 10900 }, { "epoch": 7.182356813693219, "grad_norm": 0.11448747664690018, "learning_rate": 9.904843565372248e-05, "loss": 0.0022, "step": 10910 }, { "epoch": 7.188940092165899, "grad_norm": 0.21234630048274994, "learning_rate": 9.904522249783972e-05, "loss": 0.0037, "step": 10920 }, { "epoch": 7.195523370638578, "grad_norm": 0.14499087631702423, "learning_rate": 9.904200397844517e-05, "loss": 0.0044, "step": 10930 }, { "epoch": 7.202106649111258, "grad_norm": 0.09941845387220383, "learning_rate": 9.903878009589078e-05, "loss": 0.0026, "step": 10940 }, { "epoch": 7.2086899275839365, "grad_norm": 0.10217706859111786, "learning_rate": 9.903555085052915e-05, "loss": 0.0023, "step": 10950 }, { "epoch": 7.215273206056616, "grad_norm": 0.17881952226161957, "learning_rate": 9.903231624271338e-05, "loss": 0.0039, "step": 10960 }, { "epoch": 7.221856484529296, "grad_norm": 0.22151139378547668, "learning_rate": 9.902907627279724e-05, "loss": 0.0027, "step": 10970 }, { "epoch": 7.228439763001975, "grad_norm": 0.18364161252975464, "learning_rate": 9.902583094113504e-05, "loss": 0.0027, "step": 10980 }, { "epoch": 7.235023041474655, "grad_norm": 0.13275617361068726, "learning_rate": 9.902258024808168e-05, "loss": 0.0032, "step": 10990 }, { "epoch": 7.2416063199473335, "grad_norm": 0.16427932679653168, "learning_rate": 9.901932419399264e-05, "loss": 0.0031, "step": 11000 }, { "epoch": 7.248189598420013, "grad_norm": 0.11626508086919785, "learning_rate": 9.9016062779224e-05, "loss": 0.0034, "step": 11010 }, { "epoch": 7.254772876892693, "grad_norm": 0.16740843653678894, "learning_rate": 9.901279600413242e-05, "loss": 0.0046, "step": 11020 }, { "epoch": 7.261356155365372, "grad_norm": 0.1254817247390747, "learning_rate": 9.900952386907518e-05, "loss": 0.0027, "step": 11030 }, { "epoch": 7.267939433838051, "grad_norm": 0.1403561383485794, "learning_rate": 9.90062463744101e-05, "loss": 0.0034, "step": 11040 }, { "epoch": 7.2745227123107306, "grad_norm": 0.11541598290205002, "learning_rate": 9.900296352049558e-05, "loss": 0.0041, "step": 11050 }, { "epoch": 7.28110599078341, "grad_norm": 0.1053864136338234, "learning_rate": 9.899967530769065e-05, "loss": 0.0104, "step": 11060 }, { "epoch": 7.28768926925609, "grad_norm": 0.17408975958824158, "learning_rate": 9.899638173635489e-05, "loss": 0.0033, "step": 11070 }, { "epoch": 7.294272547728769, "grad_norm": 0.08440624177455902, "learning_rate": 9.899308280684849e-05, "loss": 0.0028, "step": 11080 }, { "epoch": 7.300855826201448, "grad_norm": 0.13369396328926086, "learning_rate": 9.898977851953222e-05, "loss": 0.0019, "step": 11090 }, { "epoch": 7.307439104674128, "grad_norm": 0.07885158807039261, "learning_rate": 9.898646887476741e-05, "loss": 0.0025, "step": 11100 }, { "epoch": 7.314022383146807, "grad_norm": 0.10353007912635803, "learning_rate": 9.898315387291603e-05, "loss": 0.0025, "step": 11110 }, { "epoch": 7.320605661619487, "grad_norm": 0.18878936767578125, "learning_rate": 9.89798335143406e-05, "loss": 0.0034, "step": 11120 }, { "epoch": 7.3271889400921655, "grad_norm": 0.21067515015602112, "learning_rate": 9.897650779940419e-05, "loss": 0.0034, "step": 11130 }, { "epoch": 7.333772218564845, "grad_norm": 0.1723841428756714, "learning_rate": 9.897317672847054e-05, "loss": 0.0031, "step": 11140 }, { "epoch": 7.340355497037525, "grad_norm": 0.16209867596626282, "learning_rate": 9.89698403019039e-05, "loss": 0.0028, "step": 11150 }, { "epoch": 7.346938775510204, "grad_norm": 0.1939551830291748, "learning_rate": 9.896649852006917e-05, "loss": 0.0049, "step": 11160 }, { "epoch": 7.353522053982884, "grad_norm": 0.1688978374004364, "learning_rate": 9.896315138333177e-05, "loss": 0.0046, "step": 11170 }, { "epoch": 7.3601053324555625, "grad_norm": 0.10818961262702942, "learning_rate": 9.895979889205774e-05, "loss": 0.004, "step": 11180 }, { "epoch": 7.366688610928242, "grad_norm": 0.1496257334947586, "learning_rate": 9.895644104661372e-05, "loss": 0.0055, "step": 11190 }, { "epoch": 7.373271889400922, "grad_norm": 0.12881360948085785, "learning_rate": 9.895307784736691e-05, "loss": 0.0039, "step": 11200 }, { "epoch": 7.379855167873601, "grad_norm": 0.15952281653881073, "learning_rate": 9.894970929468512e-05, "loss": 0.0069, "step": 11210 }, { "epoch": 7.386438446346281, "grad_norm": 0.17643862962722778, "learning_rate": 9.89463353889367e-05, "loss": 0.0035, "step": 11220 }, { "epoch": 7.39302172481896, "grad_norm": 0.2160326987504959, "learning_rate": 9.894295613049065e-05, "loss": 0.0049, "step": 11230 }, { "epoch": 7.399605003291639, "grad_norm": 0.17517884075641632, "learning_rate": 9.893957151971649e-05, "loss": 0.0053, "step": 11240 }, { "epoch": 7.406188281764319, "grad_norm": 0.12693043053150177, "learning_rate": 9.893618155698436e-05, "loss": 0.0035, "step": 11250 }, { "epoch": 7.412771560236998, "grad_norm": 0.11771489679813385, "learning_rate": 9.8932786242665e-05, "loss": 0.0044, "step": 11260 }, { "epoch": 7.419354838709677, "grad_norm": 0.14587187767028809, "learning_rate": 9.89293855771297e-05, "loss": 0.0045, "step": 11270 }, { "epoch": 7.425938117182357, "grad_norm": 0.21254634857177734, "learning_rate": 9.892597956075036e-05, "loss": 0.0107, "step": 11280 }, { "epoch": 7.432521395655036, "grad_norm": 0.1268506646156311, "learning_rate": 9.892256819389947e-05, "loss": 0.0027, "step": 11290 }, { "epoch": 7.439104674127716, "grad_norm": 0.07757087051868439, "learning_rate": 9.891915147695006e-05, "loss": 0.0022, "step": 11300 }, { "epoch": 7.445687952600395, "grad_norm": 0.17634238302707672, "learning_rate": 9.891572941027577e-05, "loss": 0.0028, "step": 11310 }, { "epoch": 7.452271231073074, "grad_norm": 0.11021340638399124, "learning_rate": 9.89123019942509e-05, "loss": 0.0049, "step": 11320 }, { "epoch": 7.458854509545754, "grad_norm": 0.2406524270772934, "learning_rate": 9.89088692292502e-05, "loss": 0.0038, "step": 11330 }, { "epoch": 7.465437788018433, "grad_norm": 0.22155822813510895, "learning_rate": 9.89054311156491e-05, "loss": 0.0028, "step": 11340 }, { "epoch": 7.472021066491113, "grad_norm": 0.1595885306596756, "learning_rate": 9.890198765382357e-05, "loss": 0.0026, "step": 11350 }, { "epoch": 7.478604344963792, "grad_norm": 0.16504448652267456, "learning_rate": 9.889853884415021e-05, "loss": 0.0052, "step": 11360 }, { "epoch": 7.485187623436471, "grad_norm": 0.09254878759384155, "learning_rate": 9.889508468700614e-05, "loss": 0.0028, "step": 11370 }, { "epoch": 7.491770901909151, "grad_norm": 0.20768098533153534, "learning_rate": 9.889162518276915e-05, "loss": 0.006, "step": 11380 }, { "epoch": 7.49835418038183, "grad_norm": 0.15622158348560333, "learning_rate": 9.888816033181752e-05, "loss": 0.0029, "step": 11390 }, { "epoch": 7.50493745885451, "grad_norm": 0.09749166667461395, "learning_rate": 9.888469013453018e-05, "loss": 0.0021, "step": 11400 }, { "epoch": 7.511520737327189, "grad_norm": 0.17173795402050018, "learning_rate": 9.888121459128663e-05, "loss": 0.0047, "step": 11410 }, { "epoch": 7.518104015799868, "grad_norm": 0.17633898556232452, "learning_rate": 9.887773370246693e-05, "loss": 0.0047, "step": 11420 }, { "epoch": 7.524687294272548, "grad_norm": 0.1491798758506775, "learning_rate": 9.887424746845177e-05, "loss": 0.0081, "step": 11430 }, { "epoch": 7.531270572745227, "grad_norm": 0.26718375086784363, "learning_rate": 9.887075588962239e-05, "loss": 0.005, "step": 11440 }, { "epoch": 7.537853851217907, "grad_norm": 0.160284623503685, "learning_rate": 9.88672589663606e-05, "loss": 0.0061, "step": 11450 }, { "epoch": 7.544437129690586, "grad_norm": 0.1604444682598114, "learning_rate": 9.886375669904886e-05, "loss": 0.005, "step": 11460 }, { "epoch": 7.551020408163265, "grad_norm": 0.1519317626953125, "learning_rate": 9.886024908807014e-05, "loss": 0.0027, "step": 11470 }, { "epoch": 7.557603686635945, "grad_norm": 0.12274865806102753, "learning_rate": 9.885673613380806e-05, "loss": 0.0055, "step": 11480 }, { "epoch": 7.564186965108624, "grad_norm": 0.1968243569135666, "learning_rate": 9.885321783664676e-05, "loss": 0.003, "step": 11490 }, { "epoch": 7.570770243581303, "grad_norm": 0.12619735300540924, "learning_rate": 9.884969419697101e-05, "loss": 0.0031, "step": 11500 }, { "epoch": 7.577353522053983, "grad_norm": 0.1267111450433731, "learning_rate": 9.884616521516614e-05, "loss": 0.0024, "step": 11510 }, { "epoch": 7.583936800526662, "grad_norm": 0.12901253998279572, "learning_rate": 9.88426308916181e-05, "loss": 0.0041, "step": 11520 }, { "epoch": 7.590520078999342, "grad_norm": 0.1887347251176834, "learning_rate": 9.883909122671335e-05, "loss": 0.0051, "step": 11530 }, { "epoch": 7.597103357472021, "grad_norm": 0.22598102688789368, "learning_rate": 9.883554622083904e-05, "loss": 0.0035, "step": 11540 }, { "epoch": 7.6036866359447, "grad_norm": 0.15657243132591248, "learning_rate": 9.88319958743828e-05, "loss": 0.0037, "step": 11550 }, { "epoch": 7.61026991441738, "grad_norm": 0.13063113391399384, "learning_rate": 9.882844018773291e-05, "loss": 0.0056, "step": 11560 }, { "epoch": 7.616853192890059, "grad_norm": 0.14112794399261475, "learning_rate": 9.882487916127823e-05, "loss": 0.0029, "step": 11570 }, { "epoch": 7.623436471362739, "grad_norm": 0.22803080081939697, "learning_rate": 9.882131279540815e-05, "loss": 0.0092, "step": 11580 }, { "epoch": 7.630019749835418, "grad_norm": 0.08649920672178268, "learning_rate": 9.881774109051271e-05, "loss": 0.0032, "step": 11590 }, { "epoch": 7.636603028308097, "grad_norm": 0.11383189260959625, "learning_rate": 9.881416404698252e-05, "loss": 0.0036, "step": 11600 }, { "epoch": 7.643186306780777, "grad_norm": 0.09686729311943054, "learning_rate": 9.881058166520873e-05, "loss": 0.0025, "step": 11610 }, { "epoch": 7.649769585253456, "grad_norm": 0.12086573988199234, "learning_rate": 9.880699394558311e-05, "loss": 0.0028, "step": 11620 }, { "epoch": 7.656352863726136, "grad_norm": 0.10092036426067352, "learning_rate": 9.880340088849801e-05, "loss": 0.0038, "step": 11630 }, { "epoch": 7.662936142198815, "grad_norm": 0.13803012669086456, "learning_rate": 9.879980249434637e-05, "loss": 0.0035, "step": 11640 }, { "epoch": 7.669519420671494, "grad_norm": 0.13205309212207794, "learning_rate": 9.879619876352168e-05, "loss": 0.0067, "step": 11650 }, { "epoch": 7.676102699144174, "grad_norm": 0.22816671431064606, "learning_rate": 9.879258969641809e-05, "loss": 0.0053, "step": 11660 }, { "epoch": 7.6826859776168535, "grad_norm": 0.17496293783187866, "learning_rate": 9.878897529343023e-05, "loss": 0.004, "step": 11670 }, { "epoch": 7.689269256089533, "grad_norm": 0.2030470371246338, "learning_rate": 9.878535555495338e-05, "loss": 0.004, "step": 11680 }, { "epoch": 7.695852534562212, "grad_norm": 0.12472641468048096, "learning_rate": 9.87817304813834e-05, "loss": 0.0066, "step": 11690 }, { "epoch": 7.702435813034891, "grad_norm": 0.2193479835987091, "learning_rate": 9.877810007311671e-05, "loss": 0.01, "step": 11700 }, { "epoch": 7.709019091507571, "grad_norm": 0.12892645597457886, "learning_rate": 9.877446433055035e-05, "loss": 0.0047, "step": 11710 }, { "epoch": 7.7156023699802505, "grad_norm": 0.1597674936056137, "learning_rate": 9.877082325408191e-05, "loss": 0.0046, "step": 11720 }, { "epoch": 7.722185648452929, "grad_norm": 0.11522781848907471, "learning_rate": 9.876717684410954e-05, "loss": 0.0039, "step": 11730 }, { "epoch": 7.728768926925609, "grad_norm": 0.08793069422245026, "learning_rate": 9.876352510103204e-05, "loss": 0.003, "step": 11740 }, { "epoch": 7.735352205398288, "grad_norm": 0.18073928356170654, "learning_rate": 9.875986802524875e-05, "loss": 0.0062, "step": 11750 }, { "epoch": 7.741935483870968, "grad_norm": 0.2535589635372162, "learning_rate": 9.87562056171596e-05, "loss": 0.0089, "step": 11760 }, { "epoch": 7.748518762343647, "grad_norm": 0.22579056024551392, "learning_rate": 9.875253787716511e-05, "loss": 0.0041, "step": 11770 }, { "epoch": 7.755102040816326, "grad_norm": 0.168572336435318, "learning_rate": 9.874886480566637e-05, "loss": 0.0031, "step": 11780 }, { "epoch": 7.761685319289006, "grad_norm": 0.2847021818161011, "learning_rate": 9.874518640306507e-05, "loss": 0.0036, "step": 11790 }, { "epoch": 7.768268597761685, "grad_norm": 0.16013234853744507, "learning_rate": 9.874150266976347e-05, "loss": 0.0073, "step": 11800 }, { "epoch": 7.774851876234365, "grad_norm": 0.1752183735370636, "learning_rate": 9.873781360616443e-05, "loss": 0.005, "step": 11810 }, { "epoch": 7.781435154707044, "grad_norm": 0.19422337412834167, "learning_rate": 9.873411921267137e-05, "loss": 0.0052, "step": 11820 }, { "epoch": 7.788018433179723, "grad_norm": 0.13301026821136475, "learning_rate": 9.873041948968829e-05, "loss": 0.0052, "step": 11830 }, { "epoch": 7.794601711652403, "grad_norm": 0.1488548070192337, "learning_rate": 9.872671443761981e-05, "loss": 0.0059, "step": 11840 }, { "epoch": 7.8011849901250825, "grad_norm": 0.1882699877023697, "learning_rate": 9.872300405687109e-05, "loss": 0.0038, "step": 11850 }, { "epoch": 7.807768268597762, "grad_norm": 0.13138292729854584, "learning_rate": 9.871928834784792e-05, "loss": 0.003, "step": 11860 }, { "epoch": 7.814351547070441, "grad_norm": 0.13954825699329376, "learning_rate": 9.871556731095661e-05, "loss": 0.0046, "step": 11870 }, { "epoch": 7.82093482554312, "grad_norm": 0.14339083433151245, "learning_rate": 9.871184094660411e-05, "loss": 0.0035, "step": 11880 }, { "epoch": 7.8275181040158, "grad_norm": 0.19191016256809235, "learning_rate": 9.870810925519791e-05, "loss": 0.003, "step": 11890 }, { "epoch": 7.8341013824884795, "grad_norm": 0.12699520587921143, "learning_rate": 9.870437223714612e-05, "loss": 0.0032, "step": 11900 }, { "epoch": 7.840684660961159, "grad_norm": 0.18587861955165863, "learning_rate": 9.87006298928574e-05, "loss": 0.0032, "step": 11910 }, { "epoch": 7.847267939433838, "grad_norm": 0.11701507121324539, "learning_rate": 9.869688222274103e-05, "loss": 0.0026, "step": 11920 }, { "epoch": 7.853851217906517, "grad_norm": 0.09564034640789032, "learning_rate": 9.869312922720681e-05, "loss": 0.0021, "step": 11930 }, { "epoch": 7.860434496379197, "grad_norm": 0.09878816455602646, "learning_rate": 9.868937090666521e-05, "loss": 0.003, "step": 11940 }, { "epoch": 7.867017774851877, "grad_norm": 0.12026241421699524, "learning_rate": 9.86856072615272e-05, "loss": 0.0027, "step": 11950 }, { "epoch": 7.873601053324555, "grad_norm": 0.10108808428049088, "learning_rate": 9.868183829220438e-05, "loss": 0.0036, "step": 11960 }, { "epoch": 7.880184331797235, "grad_norm": 0.127727210521698, "learning_rate": 9.867806399910893e-05, "loss": 0.0018, "step": 11970 }, { "epoch": 7.8867676102699145, "grad_norm": 0.16336168348789215, "learning_rate": 9.867428438265356e-05, "loss": 0.004, "step": 11980 }, { "epoch": 7.893350888742594, "grad_norm": 0.12333890795707703, "learning_rate": 9.867049944325165e-05, "loss": 0.0031, "step": 11990 }, { "epoch": 7.899934167215273, "grad_norm": 0.15159986913204193, "learning_rate": 9.86667091813171e-05, "loss": 0.0036, "step": 12000 }, { "epoch": 7.906517445687952, "grad_norm": 0.1300847828388214, "learning_rate": 9.866291359726438e-05, "loss": 0.0025, "step": 12010 }, { "epoch": 7.913100724160632, "grad_norm": 0.2058093398809433, "learning_rate": 9.865911269150861e-05, "loss": 0.0033, "step": 12020 }, { "epoch": 7.9196840026333115, "grad_norm": 0.1887059509754181, "learning_rate": 9.865530646446544e-05, "loss": 0.0039, "step": 12030 }, { "epoch": 7.926267281105991, "grad_norm": 0.10858837515115738, "learning_rate": 9.86514949165511e-05, "loss": 0.003, "step": 12040 }, { "epoch": 7.93285055957867, "grad_norm": 0.10430429130792618, "learning_rate": 9.864767804818243e-05, "loss": 0.0026, "step": 12050 }, { "epoch": 7.939433838051349, "grad_norm": 0.08872396498918533, "learning_rate": 9.86438558597768e-05, "loss": 0.0017, "step": 12060 }, { "epoch": 7.946017116524029, "grad_norm": 0.08273135125637054, "learning_rate": 9.864002835175225e-05, "loss": 0.0048, "step": 12070 }, { "epoch": 7.952600394996709, "grad_norm": 0.12898515164852142, "learning_rate": 9.863619552452734e-05, "loss": 0.0032, "step": 12080 }, { "epoch": 7.959183673469388, "grad_norm": 0.08116790652275085, "learning_rate": 9.863235737852119e-05, "loss": 0.0019, "step": 12090 }, { "epoch": 7.965766951942067, "grad_norm": 0.14680612087249756, "learning_rate": 9.862851391415356e-05, "loss": 0.0022, "step": 12100 }, { "epoch": 7.972350230414746, "grad_norm": 0.1673414707183838, "learning_rate": 9.862466513184477e-05, "loss": 0.0034, "step": 12110 }, { "epoch": 7.978933508887426, "grad_norm": 0.09442789852619171, "learning_rate": 9.86208110320157e-05, "loss": 0.0029, "step": 12120 }, { "epoch": 7.985516787360106, "grad_norm": 0.142278790473938, "learning_rate": 9.861695161508784e-05, "loss": 0.0022, "step": 12130 }, { "epoch": 7.992100065832785, "grad_norm": 0.1867281198501587, "learning_rate": 9.861308688148324e-05, "loss": 0.0026, "step": 12140 }, { "epoch": 7.998683344305464, "grad_norm": 0.09046067297458649, "learning_rate": 9.860921683162455e-05, "loss": 0.0048, "step": 12150 }, { "epoch": 8.005266622778144, "grad_norm": 0.11400073021650314, "learning_rate": 9.860534146593499e-05, "loss": 0.0037, "step": 12160 }, { "epoch": 8.011849901250823, "grad_norm": 0.11901703476905823, "learning_rate": 9.860146078483836e-05, "loss": 0.0023, "step": 12170 }, { "epoch": 8.018433179723502, "grad_norm": 0.11556803435087204, "learning_rate": 9.859757478875905e-05, "loss": 0.0037, "step": 12180 }, { "epoch": 8.025016458196182, "grad_norm": 0.10101506114006042, "learning_rate": 9.859368347812204e-05, "loss": 0.0031, "step": 12190 }, { "epoch": 8.031599736668861, "grad_norm": 0.18180590867996216, "learning_rate": 9.858978685335285e-05, "loss": 0.0025, "step": 12200 }, { "epoch": 8.03818301514154, "grad_norm": 0.13066363334655762, "learning_rate": 9.858588491487763e-05, "loss": 0.0023, "step": 12210 }, { "epoch": 8.04476629361422, "grad_norm": 0.08643079549074173, "learning_rate": 9.858197766312308e-05, "loss": 0.0022, "step": 12220 }, { "epoch": 8.051349572086899, "grad_norm": 0.13291996717453003, "learning_rate": 9.857806509851649e-05, "loss": 0.0035, "step": 12230 }, { "epoch": 8.05793285055958, "grad_norm": 0.12433474510908127, "learning_rate": 9.857414722148574e-05, "loss": 0.004, "step": 12240 }, { "epoch": 8.064516129032258, "grad_norm": 0.11932418495416641, "learning_rate": 9.857022403245928e-05, "loss": 0.0032, "step": 12250 }, { "epoch": 8.071099407504937, "grad_norm": 0.17608383297920227, "learning_rate": 9.856629553186615e-05, "loss": 0.0037, "step": 12260 }, { "epoch": 8.077682685977617, "grad_norm": 0.22695602476596832, "learning_rate": 9.856236172013595e-05, "loss": 0.0029, "step": 12270 }, { "epoch": 8.084265964450296, "grad_norm": 0.07442466914653778, "learning_rate": 9.85584225976989e-05, "loss": 0.003, "step": 12280 }, { "epoch": 8.090849242922976, "grad_norm": 0.1624513566493988, "learning_rate": 9.855447816498575e-05, "loss": 0.0022, "step": 12290 }, { "epoch": 8.097432521395655, "grad_norm": 0.1513073444366455, "learning_rate": 9.855052842242787e-05, "loss": 0.0024, "step": 12300 }, { "epoch": 8.104015799868334, "grad_norm": 0.10735592246055603, "learning_rate": 9.85465733704572e-05, "loss": 0.0034, "step": 12310 }, { "epoch": 8.110599078341014, "grad_norm": 0.10396716743707657, "learning_rate": 9.854261300950624e-05, "loss": 0.0044, "step": 12320 }, { "epoch": 8.117182356813693, "grad_norm": 0.12171872705221176, "learning_rate": 9.853864734000813e-05, "loss": 0.0036, "step": 12330 }, { "epoch": 8.123765635286373, "grad_norm": 0.1290580779314041, "learning_rate": 9.85346763623965e-05, "loss": 0.0041, "step": 12340 }, { "epoch": 8.130348913759052, "grad_norm": 0.16103459894657135, "learning_rate": 9.853070007710564e-05, "loss": 0.0032, "step": 12350 }, { "epoch": 8.13693219223173, "grad_norm": 0.13467156887054443, "learning_rate": 9.85267184845704e-05, "loss": 0.0036, "step": 12360 }, { "epoch": 8.143515470704411, "grad_norm": 0.14061225950717926, "learning_rate": 9.852273158522616e-05, "loss": 0.0034, "step": 12370 }, { "epoch": 8.15009874917709, "grad_norm": 0.15284621715545654, "learning_rate": 9.851873937950896e-05, "loss": 0.0045, "step": 12380 }, { "epoch": 8.156682027649769, "grad_norm": 0.22863613069057465, "learning_rate": 9.851474186785537e-05, "loss": 0.0038, "step": 12390 }, { "epoch": 8.16326530612245, "grad_norm": 0.1020992323756218, "learning_rate": 9.851073905070254e-05, "loss": 0.0027, "step": 12400 }, { "epoch": 8.169848584595128, "grad_norm": 0.14904101192951202, "learning_rate": 9.850673092848824e-05, "loss": 0.0029, "step": 12410 }, { "epoch": 8.176431863067808, "grad_norm": 0.13345691561698914, "learning_rate": 9.850271750165077e-05, "loss": 0.0039, "step": 12420 }, { "epoch": 8.183015141540487, "grad_norm": 0.1406468152999878, "learning_rate": 9.849869877062902e-05, "loss": 0.0026, "step": 12430 }, { "epoch": 8.189598420013166, "grad_norm": 0.1567605435848236, "learning_rate": 9.849467473586252e-05, "loss": 0.005, "step": 12440 }, { "epoch": 8.196181698485846, "grad_norm": 0.12626947462558746, "learning_rate": 9.849064539779127e-05, "loss": 0.0052, "step": 12450 }, { "epoch": 8.202764976958525, "grad_norm": 0.14794841408729553, "learning_rate": 9.848661075685594e-05, "loss": 0.0027, "step": 12460 }, { "epoch": 8.209348255431205, "grad_norm": 0.15645621716976166, "learning_rate": 9.848257081349778e-05, "loss": 0.0033, "step": 12470 }, { "epoch": 8.215931533903884, "grad_norm": 0.2500571310520172, "learning_rate": 9.847852556815856e-05, "loss": 0.0043, "step": 12480 }, { "epoch": 8.222514812376563, "grad_norm": 0.11785180866718292, "learning_rate": 9.847447502128067e-05, "loss": 0.0036, "step": 12490 }, { "epoch": 8.229098090849243, "grad_norm": 0.16784904897212982, "learning_rate": 9.847041917330708e-05, "loss": 0.0071, "step": 12500 }, { "epoch": 8.235681369321922, "grad_norm": 0.1360204964876175, "learning_rate": 9.846635802468132e-05, "loss": 0.0039, "step": 12510 }, { "epoch": 8.242264647794602, "grad_norm": 0.1316796839237213, "learning_rate": 9.84622915758475e-05, "loss": 0.0068, "step": 12520 }, { "epoch": 8.248847926267281, "grad_norm": 0.13549868762493134, "learning_rate": 9.845821982725034e-05, "loss": 0.0039, "step": 12530 }, { "epoch": 8.25543120473996, "grad_norm": 0.2503281235694885, "learning_rate": 9.845414277933514e-05, "loss": 0.0033, "step": 12540 }, { "epoch": 8.26201448321264, "grad_norm": 0.14901553094387054, "learning_rate": 9.845006043254771e-05, "loss": 0.0042, "step": 12550 }, { "epoch": 8.268597761685319, "grad_norm": 0.1619502305984497, "learning_rate": 9.844597278733451e-05, "loss": 0.0038, "step": 12560 }, { "epoch": 8.275181040158, "grad_norm": 0.23243741691112518, "learning_rate": 9.844187984414259e-05, "loss": 0.0094, "step": 12570 }, { "epoch": 8.281764318630678, "grad_norm": 0.11851873993873596, "learning_rate": 9.84377816034195e-05, "loss": 0.0032, "step": 12580 }, { "epoch": 8.288347597103357, "grad_norm": 0.12821359932422638, "learning_rate": 9.843367806561345e-05, "loss": 0.0028, "step": 12590 }, { "epoch": 8.294930875576037, "grad_norm": 0.1294981837272644, "learning_rate": 9.842956923117317e-05, "loss": 0.0033, "step": 12600 }, { "epoch": 8.301514154048716, "grad_norm": 0.16726182401180267, "learning_rate": 9.842545510054802e-05, "loss": 0.0037, "step": 12610 }, { "epoch": 8.308097432521397, "grad_norm": 0.16419383883476257, "learning_rate": 9.842133567418792e-05, "loss": 0.0035, "step": 12620 }, { "epoch": 8.314680710994075, "grad_norm": 0.06982065737247467, "learning_rate": 9.841721095254333e-05, "loss": 0.0016, "step": 12630 }, { "epoch": 8.321263989466754, "grad_norm": 0.10907136648893356, "learning_rate": 9.841308093606537e-05, "loss": 0.0032, "step": 12640 }, { "epoch": 8.327847267939434, "grad_norm": 0.15429718792438507, "learning_rate": 9.840894562520565e-05, "loss": 0.0032, "step": 12650 }, { "epoch": 8.334430546412113, "grad_norm": 0.12580859661102295, "learning_rate": 9.840480502041642e-05, "loss": 0.0026, "step": 12660 }, { "epoch": 8.341013824884792, "grad_norm": 0.1364290863275528, "learning_rate": 9.840065912215049e-05, "loss": 0.0041, "step": 12670 }, { "epoch": 8.347597103357472, "grad_norm": 0.16115640103816986, "learning_rate": 9.839650793086124e-05, "loss": 0.0031, "step": 12680 }, { "epoch": 8.354180381830151, "grad_norm": 0.09809307754039764, "learning_rate": 9.839235144700265e-05, "loss": 0.0033, "step": 12690 }, { "epoch": 8.360763660302831, "grad_norm": 0.15195265412330627, "learning_rate": 9.838818967102926e-05, "loss": 0.0021, "step": 12700 }, { "epoch": 8.36734693877551, "grad_norm": 0.12715238332748413, "learning_rate": 9.83840226033962e-05, "loss": 0.0023, "step": 12710 }, { "epoch": 8.373930217248189, "grad_norm": 0.11065050214529037, "learning_rate": 9.837985024455918e-05, "loss": 0.0036, "step": 12720 }, { "epoch": 8.38051349572087, "grad_norm": 0.14545685052871704, "learning_rate": 9.837567259497447e-05, "loss": 0.0044, "step": 12730 }, { "epoch": 8.387096774193548, "grad_norm": 0.17027585208415985, "learning_rate": 9.837148965509894e-05, "loss": 0.004, "step": 12740 }, { "epoch": 8.393680052666229, "grad_norm": 0.15783797204494476, "learning_rate": 9.836730142539001e-05, "loss": 0.0033, "step": 12750 }, { "epoch": 8.400263331138907, "grad_norm": 0.186026468873024, "learning_rate": 9.836310790630574e-05, "loss": 0.0033, "step": 12760 }, { "epoch": 8.406846609611586, "grad_norm": 0.142345130443573, "learning_rate": 9.83589090983047e-05, "loss": 0.0045, "step": 12770 }, { "epoch": 8.413429888084266, "grad_norm": 0.16055217385292053, "learning_rate": 9.835470500184605e-05, "loss": 0.0028, "step": 12780 }, { "epoch": 8.420013166556945, "grad_norm": 0.18976756930351257, "learning_rate": 9.835049561738957e-05, "loss": 0.0033, "step": 12790 }, { "epoch": 8.426596445029626, "grad_norm": 0.13760510087013245, "learning_rate": 9.834628094539558e-05, "loss": 0.0034, "step": 12800 }, { "epoch": 8.433179723502304, "grad_norm": 0.17087586224079132, "learning_rate": 9.834206098632499e-05, "loss": 0.0049, "step": 12810 }, { "epoch": 8.439763001974983, "grad_norm": 0.11391337215900421, "learning_rate": 9.833783574063931e-05, "loss": 0.0027, "step": 12820 }, { "epoch": 8.446346280447663, "grad_norm": 0.1796637773513794, "learning_rate": 9.833360520880058e-05, "loss": 0.0029, "step": 12830 }, { "epoch": 8.452929558920342, "grad_norm": 0.10328694432973862, "learning_rate": 9.832936939127144e-05, "loss": 0.0051, "step": 12840 }, { "epoch": 8.45951283739302, "grad_norm": 0.09079291671514511, "learning_rate": 9.832512828851515e-05, "loss": 0.0024, "step": 12850 }, { "epoch": 8.466096115865701, "grad_norm": 0.11398924142122269, "learning_rate": 9.832088190099546e-05, "loss": 0.0025, "step": 12860 }, { "epoch": 8.47267939433838, "grad_norm": 0.14047811925411224, "learning_rate": 9.831663022917679e-05, "loss": 0.0041, "step": 12870 }, { "epoch": 8.47926267281106, "grad_norm": 0.09377261996269226, "learning_rate": 9.831237327352407e-05, "loss": 0.0024, "step": 12880 }, { "epoch": 8.48584595128374, "grad_norm": 0.1399117112159729, "learning_rate": 9.830811103450286e-05, "loss": 0.0022, "step": 12890 }, { "epoch": 8.492429229756418, "grad_norm": 0.09656792879104614, "learning_rate": 9.830384351257924e-05, "loss": 0.0023, "step": 12900 }, { "epoch": 8.499012508229098, "grad_norm": 0.09565888345241547, "learning_rate": 9.829957070821993e-05, "loss": 0.002, "step": 12910 }, { "epoch": 8.505595786701777, "grad_norm": 0.11005685478448868, "learning_rate": 9.829529262189218e-05, "loss": 0.0031, "step": 12920 }, { "epoch": 8.512179065174458, "grad_norm": 0.09890520572662354, "learning_rate": 9.829100925406385e-05, "loss": 0.0029, "step": 12930 }, { "epoch": 8.518762343647136, "grad_norm": 0.19052959978580475, "learning_rate": 9.828672060520333e-05, "loss": 0.0062, "step": 12940 }, { "epoch": 8.525345622119815, "grad_norm": 0.16169655323028564, "learning_rate": 9.828242667577966e-05, "loss": 0.0027, "step": 12950 }, { "epoch": 8.531928900592495, "grad_norm": 0.1486174613237381, "learning_rate": 9.82781274662624e-05, "loss": 0.0037, "step": 12960 }, { "epoch": 8.538512179065174, "grad_norm": 0.12793396413326263, "learning_rate": 9.82738229771217e-05, "loss": 0.0019, "step": 12970 }, { "epoch": 8.545095457537855, "grad_norm": 0.12173007428646088, "learning_rate": 9.826951320882829e-05, "loss": 0.0035, "step": 12980 }, { "epoch": 8.551678736010533, "grad_norm": 0.15295392274856567, "learning_rate": 9.826519816185351e-05, "loss": 0.0028, "step": 12990 }, { "epoch": 8.558262014483212, "grad_norm": 0.12468356639146805, "learning_rate": 9.826087783666921e-05, "loss": 0.0031, "step": 13000 }, { "epoch": 8.564845292955892, "grad_norm": 0.1285577267408371, "learning_rate": 9.825655223374787e-05, "loss": 0.0031, "step": 13010 }, { "epoch": 8.571428571428571, "grad_norm": 0.13097038865089417, "learning_rate": 9.825222135356253e-05, "loss": 0.0038, "step": 13020 }, { "epoch": 8.578011849901252, "grad_norm": 0.10667772591114044, "learning_rate": 9.82478851965868e-05, "loss": 0.0029, "step": 13030 }, { "epoch": 8.58459512837393, "grad_norm": 0.15021389722824097, "learning_rate": 9.82435437632949e-05, "loss": 0.0064, "step": 13040 }, { "epoch": 8.591178406846609, "grad_norm": 0.13568520545959473, "learning_rate": 9.823919705416158e-05, "loss": 0.0046, "step": 13050 }, { "epoch": 8.59776168531929, "grad_norm": 0.15769396722316742, "learning_rate": 9.82348450696622e-05, "loss": 0.003, "step": 13060 }, { "epoch": 8.604344963791968, "grad_norm": 0.09940027445554733, "learning_rate": 9.823048781027268e-05, "loss": 0.0031, "step": 13070 }, { "epoch": 8.610928242264649, "grad_norm": 0.14519630372524261, "learning_rate": 9.822612527646953e-05, "loss": 0.0048, "step": 13080 }, { "epoch": 8.617511520737327, "grad_norm": 0.19395820796489716, "learning_rate": 9.822175746872984e-05, "loss": 0.0038, "step": 13090 }, { "epoch": 8.624094799210006, "grad_norm": 0.14362339675426483, "learning_rate": 9.821738438753123e-05, "loss": 0.0038, "step": 13100 }, { "epoch": 8.630678077682687, "grad_norm": 0.20641876757144928, "learning_rate": 9.821300603335196e-05, "loss": 0.0043, "step": 13110 }, { "epoch": 8.637261356155365, "grad_norm": 0.180748850107193, "learning_rate": 9.820862240667085e-05, "loss": 0.0034, "step": 13120 }, { "epoch": 8.643844634628044, "grad_norm": 0.13828176259994507, "learning_rate": 9.820423350796726e-05, "loss": 0.0053, "step": 13130 }, { "epoch": 8.650427913100724, "grad_norm": 0.15445657074451447, "learning_rate": 9.819983933772118e-05, "loss": 0.0078, "step": 13140 }, { "epoch": 8.657011191573403, "grad_norm": 0.1252973973751068, "learning_rate": 9.819543989641314e-05, "loss": 0.0066, "step": 13150 }, { "epoch": 8.663594470046084, "grad_norm": 0.1962350457906723, "learning_rate": 9.819103518452423e-05, "loss": 0.006, "step": 13160 }, { "epoch": 8.670177748518762, "grad_norm": 0.17580729722976685, "learning_rate": 9.818662520253618e-05, "loss": 0.0038, "step": 13170 }, { "epoch": 8.676761026991441, "grad_norm": 0.1166810542345047, "learning_rate": 9.818220995093126e-05, "loss": 0.0034, "step": 13180 }, { "epoch": 8.683344305464122, "grad_norm": 0.12761376798152924, "learning_rate": 9.817778943019228e-05, "loss": 0.0026, "step": 13190 }, { "epoch": 8.6899275839368, "grad_norm": 0.15387585759162903, "learning_rate": 9.81733636408027e-05, "loss": 0.0048, "step": 13200 }, { "epoch": 8.69651086240948, "grad_norm": 0.09185472875833511, "learning_rate": 9.816893258324649e-05, "loss": 0.0022, "step": 13210 }, { "epoch": 8.70309414088216, "grad_norm": 0.16528670489788055, "learning_rate": 9.816449625800823e-05, "loss": 0.0043, "step": 13220 }, { "epoch": 8.709677419354838, "grad_norm": 0.13746030628681183, "learning_rate": 9.816005466557308e-05, "loss": 0.0053, "step": 13230 }, { "epoch": 8.716260697827519, "grad_norm": 0.13486793637275696, "learning_rate": 9.815560780642674e-05, "loss": 0.0028, "step": 13240 }, { "epoch": 8.722843976300197, "grad_norm": 0.10134762525558472, "learning_rate": 9.815115568105555e-05, "loss": 0.004, "step": 13250 }, { "epoch": 8.729427254772876, "grad_norm": 0.14097189903259277, "learning_rate": 9.814669828994638e-05, "loss": 0.0058, "step": 13260 }, { "epoch": 8.736010533245556, "grad_norm": 0.1717834770679474, "learning_rate": 9.814223563358665e-05, "loss": 0.0036, "step": 13270 }, { "epoch": 8.742593811718235, "grad_norm": 0.10857930034399033, "learning_rate": 9.813776771246443e-05, "loss": 0.0032, "step": 13280 }, { "epoch": 8.749177090190916, "grad_norm": 0.1173538938164711, "learning_rate": 9.813329452706829e-05, "loss": 0.0023, "step": 13290 }, { "epoch": 8.755760368663594, "grad_norm": 0.18737493455410004, "learning_rate": 9.812881607788744e-05, "loss": 0.004, "step": 13300 }, { "epoch": 8.762343647136273, "grad_norm": 0.13430064916610718, "learning_rate": 9.812433236541163e-05, "loss": 0.0033, "step": 13310 }, { "epoch": 8.768926925608953, "grad_norm": 0.18540328741073608, "learning_rate": 9.811984339013116e-05, "loss": 0.0031, "step": 13320 }, { "epoch": 8.775510204081632, "grad_norm": 0.21997332572937012, "learning_rate": 9.811534915253698e-05, "loss": 0.0041, "step": 13330 }, { "epoch": 8.782093482554313, "grad_norm": 0.103426493704319, "learning_rate": 9.811084965312056e-05, "loss": 0.0032, "step": 13340 }, { "epoch": 8.788676761026991, "grad_norm": 0.2641832232475281, "learning_rate": 9.810634489237396e-05, "loss": 0.0055, "step": 13350 }, { "epoch": 8.79526003949967, "grad_norm": 0.14761470258235931, "learning_rate": 9.81018348707898e-05, "loss": 0.0036, "step": 13360 }, { "epoch": 8.80184331797235, "grad_norm": 0.1175559014081955, "learning_rate": 9.809731958886131e-05, "loss": 0.0024, "step": 13370 }, { "epoch": 8.80842659644503, "grad_norm": 0.14466352760791779, "learning_rate": 9.809279904708224e-05, "loss": 0.004, "step": 13380 }, { "epoch": 8.81500987491771, "grad_norm": 0.1096877008676529, "learning_rate": 9.808827324594699e-05, "loss": 0.0027, "step": 13390 }, { "epoch": 8.821593153390388, "grad_norm": 0.10725880414247513, "learning_rate": 9.808374218595046e-05, "loss": 0.003, "step": 13400 }, { "epoch": 8.828176431863067, "grad_norm": 0.10435739904642105, "learning_rate": 9.80792058675882e-05, "loss": 0.0021, "step": 13410 }, { "epoch": 8.834759710335748, "grad_norm": 0.18115316331386566, "learning_rate": 9.807466429135627e-05, "loss": 0.0031, "step": 13420 }, { "epoch": 8.841342988808426, "grad_norm": 0.09213712066411972, "learning_rate": 9.807011745775132e-05, "loss": 0.0028, "step": 13430 }, { "epoch": 8.847926267281107, "grad_norm": 0.19900372624397278, "learning_rate": 9.806556536727061e-05, "loss": 0.0028, "step": 13440 }, { "epoch": 8.854509545753785, "grad_norm": 0.15528027713298798, "learning_rate": 9.806100802041193e-05, "loss": 0.0032, "step": 13450 }, { "epoch": 8.861092824226464, "grad_norm": 0.20445214211940765, "learning_rate": 9.805644541767368e-05, "loss": 0.0042, "step": 13460 }, { "epoch": 8.867676102699145, "grad_norm": 0.15152792632579803, "learning_rate": 9.805187755955478e-05, "loss": 0.0047, "step": 13470 }, { "epoch": 8.874259381171823, "grad_norm": 0.12591654062271118, "learning_rate": 9.804730444655483e-05, "loss": 0.0035, "step": 13480 }, { "epoch": 8.880842659644504, "grad_norm": 0.08646434545516968, "learning_rate": 9.804272607917388e-05, "loss": 0.0033, "step": 13490 }, { "epoch": 8.887425938117183, "grad_norm": 0.14181238412857056, "learning_rate": 9.803814245791265e-05, "loss": 0.0027, "step": 13500 }, { "epoch": 8.894009216589861, "grad_norm": 0.20367440581321716, "learning_rate": 9.803355358327239e-05, "loss": 0.0026, "step": 13510 }, { "epoch": 8.900592495062542, "grad_norm": 0.15828967094421387, "learning_rate": 9.802895945575492e-05, "loss": 0.0035, "step": 13520 }, { "epoch": 8.90717577353522, "grad_norm": 0.15288794040679932, "learning_rate": 9.802436007586266e-05, "loss": 0.002, "step": 13530 }, { "epoch": 8.9137590520079, "grad_norm": 0.09474591910839081, "learning_rate": 9.801975544409858e-05, "loss": 0.0032, "step": 13540 }, { "epoch": 8.92034233048058, "grad_norm": 0.16030067205429077, "learning_rate": 9.801514556096625e-05, "loss": 0.0023, "step": 13550 }, { "epoch": 8.926925608953258, "grad_norm": 0.1420895904302597, "learning_rate": 9.801053042696977e-05, "loss": 0.0038, "step": 13560 }, { "epoch": 8.933508887425939, "grad_norm": 0.1566780060529709, "learning_rate": 9.800591004261388e-05, "loss": 0.0039, "step": 13570 }, { "epoch": 8.940092165898617, "grad_norm": 0.10788200050592422, "learning_rate": 9.800128440840385e-05, "loss": 0.0026, "step": 13580 }, { "epoch": 8.946675444371296, "grad_norm": 0.1695166826248169, "learning_rate": 9.799665352484552e-05, "loss": 0.0042, "step": 13590 }, { "epoch": 8.953258722843977, "grad_norm": 0.23227865993976593, "learning_rate": 9.799201739244532e-05, "loss": 0.0051, "step": 13600 }, { "epoch": 8.959842001316655, "grad_norm": 0.1746222972869873, "learning_rate": 9.798737601171025e-05, "loss": 0.0023, "step": 13610 }, { "epoch": 8.966425279789336, "grad_norm": 0.13414688408374786, "learning_rate": 9.79827293831479e-05, "loss": 0.0024, "step": 13620 }, { "epoch": 8.973008558262014, "grad_norm": 0.10067746043205261, "learning_rate": 9.797807750726638e-05, "loss": 0.0038, "step": 13630 }, { "epoch": 8.979591836734693, "grad_norm": 0.10164303332567215, "learning_rate": 9.797342038457446e-05, "loss": 0.0018, "step": 13640 }, { "epoch": 8.986175115207374, "grad_norm": 0.07559173554182053, "learning_rate": 9.796875801558141e-05, "loss": 0.0014, "step": 13650 }, { "epoch": 8.992758393680052, "grad_norm": 0.13463075459003448, "learning_rate": 9.79640904007971e-05, "loss": 0.0034, "step": 13660 }, { "epoch": 8.999341672152733, "grad_norm": 0.11943522840738297, "learning_rate": 9.795941754073199e-05, "loss": 0.0019, "step": 13670 }, { "epoch": 9.005924950625412, "grad_norm": 0.23368771374225616, "learning_rate": 9.795473943589705e-05, "loss": 0.005, "step": 13680 }, { "epoch": 9.01250822909809, "grad_norm": 0.14208076894283295, "learning_rate": 9.795005608680394e-05, "loss": 0.0028, "step": 13690 }, { "epoch": 9.01909150757077, "grad_norm": 0.11541754752397537, "learning_rate": 9.794536749396477e-05, "loss": 0.0023, "step": 13700 }, { "epoch": 9.02567478604345, "grad_norm": 0.1525447964668274, "learning_rate": 9.79406736578923e-05, "loss": 0.0029, "step": 13710 }, { "epoch": 9.03225806451613, "grad_norm": 0.16002915799617767, "learning_rate": 9.793597457909984e-05, "loss": 0.0045, "step": 13720 }, { "epoch": 9.038841342988809, "grad_norm": 0.10856320708990097, "learning_rate": 9.793127025810127e-05, "loss": 0.0032, "step": 13730 }, { "epoch": 9.045424621461487, "grad_norm": 0.17371220886707306, "learning_rate": 9.792656069541104e-05, "loss": 0.0057, "step": 13740 }, { "epoch": 9.052007899934168, "grad_norm": 0.1856151819229126, "learning_rate": 9.79218458915442e-05, "loss": 0.0048, "step": 13750 }, { "epoch": 9.058591178406846, "grad_norm": 0.07466883957386017, "learning_rate": 9.791712584701634e-05, "loss": 0.0043, "step": 13760 }, { "epoch": 9.065174456879525, "grad_norm": 0.08220570534467697, "learning_rate": 9.791240056234364e-05, "loss": 0.0034, "step": 13770 }, { "epoch": 9.071757735352206, "grad_norm": 0.08555582165718079, "learning_rate": 9.790767003804283e-05, "loss": 0.0033, "step": 13780 }, { "epoch": 9.078341013824884, "grad_norm": 0.08291485160589218, "learning_rate": 9.790293427463126e-05, "loss": 0.0034, "step": 13790 }, { "epoch": 9.084924292297565, "grad_norm": 0.19654907286167145, "learning_rate": 9.789819327262684e-05, "loss": 0.0028, "step": 13800 }, { "epoch": 9.091507570770244, "grad_norm": 0.09666986763477325, "learning_rate": 9.7893447032548e-05, "loss": 0.0032, "step": 13810 }, { "epoch": 9.098090849242922, "grad_norm": 0.09953656792640686, "learning_rate": 9.78886955549138e-05, "loss": 0.0025, "step": 13820 }, { "epoch": 9.104674127715603, "grad_norm": 0.18908856809139252, "learning_rate": 9.788393884024387e-05, "loss": 0.006, "step": 13830 }, { "epoch": 9.111257406188281, "grad_norm": 0.13021788001060486, "learning_rate": 9.787917688905836e-05, "loss": 0.0044, "step": 13840 }, { "epoch": 9.117840684660962, "grad_norm": 0.18915122747421265, "learning_rate": 9.787440970187807e-05, "loss": 0.0027, "step": 13850 }, { "epoch": 9.12442396313364, "grad_norm": 0.12839360535144806, "learning_rate": 9.786963727922429e-05, "loss": 0.0021, "step": 13860 }, { "epoch": 9.13100724160632, "grad_norm": 0.11671461910009384, "learning_rate": 9.786485962161897e-05, "loss": 0.0021, "step": 13870 }, { "epoch": 9.137590520079, "grad_norm": 0.08409273624420166, "learning_rate": 9.786007672958455e-05, "loss": 0.0019, "step": 13880 }, { "epoch": 9.144173798551678, "grad_norm": 0.1172027587890625, "learning_rate": 9.78552886036441e-05, "loss": 0.005, "step": 13890 }, { "epoch": 9.150757077024359, "grad_norm": 0.1534237116575241, "learning_rate": 9.785049524432124e-05, "loss": 0.0037, "step": 13900 }, { "epoch": 9.157340355497038, "grad_norm": 0.1558518409729004, "learning_rate": 9.784569665214016e-05, "loss": 0.0037, "step": 13910 }, { "epoch": 9.163923633969716, "grad_norm": 0.15426068007946014, "learning_rate": 9.784089282762563e-05, "loss": 0.0036, "step": 13920 }, { "epoch": 9.170506912442397, "grad_norm": 0.1244080439209938, "learning_rate": 9.7836083771303e-05, "loss": 0.0023, "step": 13930 }, { "epoch": 9.177090190915075, "grad_norm": 0.15417607128620148, "learning_rate": 9.783126948369817e-05, "loss": 0.0085, "step": 13940 }, { "epoch": 9.183673469387756, "grad_norm": 0.15456146001815796, "learning_rate": 9.78264499653376e-05, "loss": 0.0032, "step": 13950 }, { "epoch": 9.190256747860435, "grad_norm": 0.11314090341329575, "learning_rate": 9.782162521674838e-05, "loss": 0.003, "step": 13960 }, { "epoch": 9.196840026333113, "grad_norm": 0.07908967137336731, "learning_rate": 9.781679523845812e-05, "loss": 0.0032, "step": 13970 }, { "epoch": 9.203423304805794, "grad_norm": 0.1301450878381729, "learning_rate": 9.781196003099502e-05, "loss": 0.0067, "step": 13980 }, { "epoch": 9.210006583278473, "grad_norm": 0.09574795514345169, "learning_rate": 9.780711959488786e-05, "loss": 0.0019, "step": 13990 }, { "epoch": 9.216589861751151, "grad_norm": 0.21665827929973602, "learning_rate": 9.780227393066599e-05, "loss": 0.0047, "step": 14000 }, { "epoch": 9.223173140223832, "grad_norm": 0.17507702112197876, "learning_rate": 9.77974230388593e-05, "loss": 0.0036, "step": 14010 }, { "epoch": 9.22975641869651, "grad_norm": 0.15830545127391815, "learning_rate": 9.779256691999829e-05, "loss": 0.0035, "step": 14020 }, { "epoch": 9.236339697169191, "grad_norm": 0.12650786340236664, "learning_rate": 9.778770557461403e-05, "loss": 0.0022, "step": 14030 }, { "epoch": 9.24292297564187, "grad_norm": 0.17364948987960815, "learning_rate": 9.778283900323812e-05, "loss": 0.0032, "step": 14040 }, { "epoch": 9.249506254114548, "grad_norm": 0.11233724653720856, "learning_rate": 9.777796720640277e-05, "loss": 0.0035, "step": 14050 }, { "epoch": 9.256089532587229, "grad_norm": 0.11931813508272171, "learning_rate": 9.777309018464078e-05, "loss": 0.0039, "step": 14060 }, { "epoch": 9.262672811059907, "grad_norm": 0.20263682305812836, "learning_rate": 9.776820793848547e-05, "loss": 0.0025, "step": 14070 }, { "epoch": 9.269256089532588, "grad_norm": 0.11366266757249832, "learning_rate": 9.776332046847075e-05, "loss": 0.0054, "step": 14080 }, { "epoch": 9.275839368005267, "grad_norm": 0.12975242733955383, "learning_rate": 9.775842777513111e-05, "loss": 0.0049, "step": 14090 }, { "epoch": 9.282422646477945, "grad_norm": 0.16860324144363403, "learning_rate": 9.775352985900163e-05, "loss": 0.0078, "step": 14100 }, { "epoch": 9.289005924950626, "grad_norm": 0.1079597994685173, "learning_rate": 9.774862672061791e-05, "loss": 0.0033, "step": 14110 }, { "epoch": 9.295589203423305, "grad_norm": 0.22344449162483215, "learning_rate": 9.774371836051616e-05, "loss": 0.0032, "step": 14120 }, { "epoch": 9.302172481895985, "grad_norm": 0.15691903233528137, "learning_rate": 9.773880477923315e-05, "loss": 0.0054, "step": 14130 }, { "epoch": 9.308755760368664, "grad_norm": 0.14857572317123413, "learning_rate": 9.773388597730623e-05, "loss": 0.0028, "step": 14140 }, { "epoch": 9.315339038841342, "grad_norm": 0.0919775664806366, "learning_rate": 9.77289619552733e-05, "loss": 0.0023, "step": 14150 }, { "epoch": 9.321922317314023, "grad_norm": 0.11303535103797913, "learning_rate": 9.772403271367285e-05, "loss": 0.0021, "step": 14160 }, { "epoch": 9.328505595786702, "grad_norm": 0.1006031334400177, "learning_rate": 9.771909825304396e-05, "loss": 0.0032, "step": 14170 }, { "epoch": 9.33508887425938, "grad_norm": 0.12553997337818146, "learning_rate": 9.771415857392619e-05, "loss": 0.0019, "step": 14180 }, { "epoch": 9.34167215273206, "grad_norm": 0.16176943480968475, "learning_rate": 9.770921367685978e-05, "loss": 0.0032, "step": 14190 }, { "epoch": 9.34825543120474, "grad_norm": 0.08055946230888367, "learning_rate": 9.770426356238551e-05, "loss": 0.0029, "step": 14200 }, { "epoch": 9.35483870967742, "grad_norm": 0.172756165266037, "learning_rate": 9.769930823104469e-05, "loss": 0.0031, "step": 14210 }, { "epoch": 9.361421988150099, "grad_norm": 0.09307631850242615, "learning_rate": 9.769434768337926e-05, "loss": 0.0025, "step": 14220 }, { "epoch": 9.368005266622777, "grad_norm": 0.1605590432882309, "learning_rate": 9.768938191993164e-05, "loss": 0.0029, "step": 14230 }, { "epoch": 9.374588545095458, "grad_norm": 0.18443773686885834, "learning_rate": 9.768441094124494e-05, "loss": 0.0042, "step": 14240 }, { "epoch": 9.381171823568137, "grad_norm": 0.13564659655094147, "learning_rate": 9.767943474786275e-05, "loss": 0.0026, "step": 14250 }, { "epoch": 9.387755102040817, "grad_norm": 0.10971985012292862, "learning_rate": 9.767445334032923e-05, "loss": 0.0043, "step": 14260 }, { "epoch": 9.394338380513496, "grad_norm": 0.1185365617275238, "learning_rate": 9.766946671918919e-05, "loss": 0.0035, "step": 14270 }, { "epoch": 9.400921658986174, "grad_norm": 0.10340417921543121, "learning_rate": 9.766447488498796e-05, "loss": 0.0037, "step": 14280 }, { "epoch": 9.407504937458855, "grad_norm": 0.162923663854599, "learning_rate": 9.765947783827139e-05, "loss": 0.006, "step": 14290 }, { "epoch": 9.414088215931534, "grad_norm": 0.12016482651233673, "learning_rate": 9.765447557958599e-05, "loss": 0.0028, "step": 14300 }, { "epoch": 9.420671494404214, "grad_norm": 0.09548697620630264, "learning_rate": 9.764946810947879e-05, "loss": 0.007, "step": 14310 }, { "epoch": 9.427254772876893, "grad_norm": 0.0995086282491684, "learning_rate": 9.764445542849738e-05, "loss": 0.0027, "step": 14320 }, { "epoch": 9.433838051349571, "grad_norm": 0.15329615771770477, "learning_rate": 9.763943753718998e-05, "loss": 0.0031, "step": 14330 }, { "epoch": 9.440421329822252, "grad_norm": 0.15115195512771606, "learning_rate": 9.76344144361053e-05, "loss": 0.0035, "step": 14340 }, { "epoch": 9.44700460829493, "grad_norm": 0.18265020847320557, "learning_rate": 9.762938612579269e-05, "loss": 0.0046, "step": 14350 }, { "epoch": 9.453587886767611, "grad_norm": 0.12191231548786163, "learning_rate": 9.762435260680202e-05, "loss": 0.0036, "step": 14360 }, { "epoch": 9.46017116524029, "grad_norm": 0.13637147843837738, "learning_rate": 9.761931387968373e-05, "loss": 0.0079, "step": 14370 }, { "epoch": 9.466754443712968, "grad_norm": 0.09438101947307587, "learning_rate": 9.76142699449889e-05, "loss": 0.0047, "step": 14380 }, { "epoch": 9.473337722185649, "grad_norm": 0.19698798656463623, "learning_rate": 9.760922080326908e-05, "loss": 0.0034, "step": 14390 }, { "epoch": 9.479921000658328, "grad_norm": 0.1290300339460373, "learning_rate": 9.760416645507644e-05, "loss": 0.0035, "step": 14400 }, { "epoch": 9.486504279131008, "grad_norm": 0.17004182934761047, "learning_rate": 9.759910690096375e-05, "loss": 0.0029, "step": 14410 }, { "epoch": 9.493087557603687, "grad_norm": 0.12916478514671326, "learning_rate": 9.759404214148429e-05, "loss": 0.0021, "step": 14420 }, { "epoch": 9.499670836076366, "grad_norm": 0.10555191338062286, "learning_rate": 9.758897217719191e-05, "loss": 0.0037, "step": 14430 }, { "epoch": 9.506254114549046, "grad_norm": 0.08133132755756378, "learning_rate": 9.758389700864113e-05, "loss": 0.0055, "step": 14440 }, { "epoch": 9.512837393021725, "grad_norm": 0.16655434668064117, "learning_rate": 9.757881663638688e-05, "loss": 0.0028, "step": 14450 }, { "epoch": 9.519420671494403, "grad_norm": 0.125949427485466, "learning_rate": 9.757373106098478e-05, "loss": 0.0032, "step": 14460 }, { "epoch": 9.526003949967084, "grad_norm": 0.14204567670822144, "learning_rate": 9.756864028299097e-05, "loss": 0.0027, "step": 14470 }, { "epoch": 9.532587228439763, "grad_norm": 0.12339302897453308, "learning_rate": 9.75635443029622e-05, "loss": 0.0036, "step": 14480 }, { "epoch": 9.539170506912443, "grad_norm": 0.10802753269672394, "learning_rate": 9.755844312145572e-05, "loss": 0.0024, "step": 14490 }, { "epoch": 9.545753785385122, "grad_norm": 0.10697764903306961, "learning_rate": 9.755333673902941e-05, "loss": 0.0021, "step": 14500 }, { "epoch": 9.5523370638578, "grad_norm": 0.10028508305549622, "learning_rate": 9.75482251562417e-05, "loss": 0.0028, "step": 14510 }, { "epoch": 9.558920342330481, "grad_norm": 0.20441381633281708, "learning_rate": 9.754310837365155e-05, "loss": 0.0033, "step": 14520 }, { "epoch": 9.56550362080316, "grad_norm": 0.12866373360157013, "learning_rate": 9.753798639181856e-05, "loss": 0.0026, "step": 14530 }, { "epoch": 9.57208689927584, "grad_norm": 0.10690212994813919, "learning_rate": 9.753285921130286e-05, "loss": 0.003, "step": 14540 }, { "epoch": 9.578670177748519, "grad_norm": 0.12540453672409058, "learning_rate": 9.752772683266512e-05, "loss": 0.0031, "step": 14550 }, { "epoch": 9.585253456221198, "grad_norm": 0.09787749499082565, "learning_rate": 9.752258925646665e-05, "loss": 0.0022, "step": 14560 }, { "epoch": 9.591836734693878, "grad_norm": 0.13307100534439087, "learning_rate": 9.751744648326926e-05, "loss": 0.0019, "step": 14570 }, { "epoch": 9.598420013166557, "grad_norm": 0.07867175340652466, "learning_rate": 9.751229851363536e-05, "loss": 0.0038, "step": 14580 }, { "epoch": 9.605003291639237, "grad_norm": 0.15991991758346558, "learning_rate": 9.750714534812793e-05, "loss": 0.0041, "step": 14590 }, { "epoch": 9.611586570111916, "grad_norm": 0.10468481481075287, "learning_rate": 9.750198698731053e-05, "loss": 0.0028, "step": 14600 }, { "epoch": 9.618169848584595, "grad_norm": 0.20081333816051483, "learning_rate": 9.749682343174722e-05, "loss": 0.0056, "step": 14610 }, { "epoch": 9.624753127057275, "grad_norm": 0.13003017008304596, "learning_rate": 9.749165468200272e-05, "loss": 0.0068, "step": 14620 }, { "epoch": 9.631336405529954, "grad_norm": 0.14617370069026947, "learning_rate": 9.748648073864229e-05, "loss": 0.0031, "step": 14630 }, { "epoch": 9.637919684002632, "grad_norm": 0.1215650886297226, "learning_rate": 9.748130160223168e-05, "loss": 0.003, "step": 14640 }, { "epoch": 9.644502962475313, "grad_norm": 0.15294747054576874, "learning_rate": 9.747611727333734e-05, "loss": 0.0025, "step": 14650 }, { "epoch": 9.651086240947992, "grad_norm": 0.10920429974794388, "learning_rate": 9.74709277525262e-05, "loss": 0.005, "step": 14660 }, { "epoch": 9.657669519420672, "grad_norm": 0.11421952396631241, "learning_rate": 9.746573304036576e-05, "loss": 0.0025, "step": 14670 }, { "epoch": 9.66425279789335, "grad_norm": 0.15861693024635315, "learning_rate": 9.746053313742412e-05, "loss": 0.0037, "step": 14680 }, { "epoch": 9.67083607636603, "grad_norm": 0.1239711344242096, "learning_rate": 9.745532804426994e-05, "loss": 0.0034, "step": 14690 }, { "epoch": 9.67741935483871, "grad_norm": 0.1393122673034668, "learning_rate": 9.745011776147242e-05, "loss": 0.0027, "step": 14700 }, { "epoch": 9.684002633311389, "grad_norm": 0.13880792260169983, "learning_rate": 9.744490228960138e-05, "loss": 0.0031, "step": 14710 }, { "epoch": 9.69058591178407, "grad_norm": 0.08720485866069794, "learning_rate": 9.743968162922713e-05, "loss": 0.003, "step": 14720 }, { "epoch": 9.697169190256748, "grad_norm": 0.10602348297834396, "learning_rate": 9.743445578092064e-05, "loss": 0.0028, "step": 14730 }, { "epoch": 9.703752468729427, "grad_norm": 0.14484144747257233, "learning_rate": 9.742922474525338e-05, "loss": 0.0035, "step": 14740 }, { "epoch": 9.710335747202107, "grad_norm": 0.16590657830238342, "learning_rate": 9.742398852279741e-05, "loss": 0.004, "step": 14750 }, { "epoch": 9.716919025674786, "grad_norm": 0.1763269603252411, "learning_rate": 9.741874711412535e-05, "loss": 0.0021, "step": 14760 }, { "epoch": 9.723502304147466, "grad_norm": 0.140358567237854, "learning_rate": 9.741350051981042e-05, "loss": 0.0082, "step": 14770 }, { "epoch": 9.730085582620145, "grad_norm": 0.10798827558755875, "learning_rate": 9.740824874042633e-05, "loss": 0.0021, "step": 14780 }, { "epoch": 9.736668861092824, "grad_norm": 0.084516741335392, "learning_rate": 9.740299177654746e-05, "loss": 0.0041, "step": 14790 }, { "epoch": 9.743252139565504, "grad_norm": 0.11820462346076965, "learning_rate": 9.739772962874867e-05, "loss": 0.0023, "step": 14800 }, { "epoch": 9.749835418038183, "grad_norm": 0.1465664654970169, "learning_rate": 9.739246229760541e-05, "loss": 0.0017, "step": 14810 }, { "epoch": 9.756418696510863, "grad_norm": 0.15352128446102142, "learning_rate": 9.738718978369376e-05, "loss": 0.0025, "step": 14820 }, { "epoch": 9.763001974983542, "grad_norm": 0.0967302918434143, "learning_rate": 9.738191208759025e-05, "loss": 0.0047, "step": 14830 }, { "epoch": 9.76958525345622, "grad_norm": 0.14880768954753876, "learning_rate": 9.73766292098721e-05, "loss": 0.0034, "step": 14840 }, { "epoch": 9.776168531928901, "grad_norm": 0.1334373652935028, "learning_rate": 9.737134115111699e-05, "loss": 0.0027, "step": 14850 }, { "epoch": 9.78275181040158, "grad_norm": 0.10401709377765656, "learning_rate": 9.736604791190323e-05, "loss": 0.0039, "step": 14860 }, { "epoch": 9.78933508887426, "grad_norm": 0.13231685757637024, "learning_rate": 9.73607494928097e-05, "loss": 0.0057, "step": 14870 }, { "epoch": 9.795918367346939, "grad_norm": 0.14299190044403076, "learning_rate": 9.735544589441581e-05, "loss": 0.0037, "step": 14880 }, { "epoch": 9.802501645819618, "grad_norm": 0.1256221979856491, "learning_rate": 9.735013711730154e-05, "loss": 0.0037, "step": 14890 }, { "epoch": 9.809084924292298, "grad_norm": 0.21058781445026398, "learning_rate": 9.734482316204747e-05, "loss": 0.0024, "step": 14900 }, { "epoch": 9.815668202764977, "grad_norm": 0.12457385659217834, "learning_rate": 9.733950402923473e-05, "loss": 0.0019, "step": 14910 }, { "epoch": 9.822251481237656, "grad_norm": 0.17469444870948792, "learning_rate": 9.7334179719445e-05, "loss": 0.0066, "step": 14920 }, { "epoch": 9.828834759710336, "grad_norm": 0.15024563670158386, "learning_rate": 9.732885023326053e-05, "loss": 0.0044, "step": 14930 }, { "epoch": 9.835418038183015, "grad_norm": 0.14250801503658295, "learning_rate": 9.732351557126418e-05, "loss": 0.0029, "step": 14940 }, { "epoch": 9.842001316655695, "grad_norm": 0.12797997891902924, "learning_rate": 9.731817573403929e-05, "loss": 0.0067, "step": 14950 }, { "epoch": 9.848584595128374, "grad_norm": 0.1746576726436615, "learning_rate": 9.731283072216985e-05, "loss": 0.0029, "step": 14960 }, { "epoch": 9.855167873601053, "grad_norm": 0.12796930968761444, "learning_rate": 9.730748053624039e-05, "loss": 0.0049, "step": 14970 }, { "epoch": 9.861751152073733, "grad_norm": 0.10550738126039505, "learning_rate": 9.730212517683598e-05, "loss": 0.0033, "step": 14980 }, { "epoch": 9.868334430546412, "grad_norm": 0.21375171840190887, "learning_rate": 9.729676464454228e-05, "loss": 0.0031, "step": 14990 }, { "epoch": 9.874917709019092, "grad_norm": 0.12915988266468048, "learning_rate": 9.72913989399455e-05, "loss": 0.0027, "step": 15000 }, { "epoch": 9.881500987491771, "grad_norm": 0.10147684067487717, "learning_rate": 9.728602806363242e-05, "loss": 0.0028, "step": 15010 }, { "epoch": 9.88808426596445, "grad_norm": 0.12336605787277222, "learning_rate": 9.728065201619043e-05, "loss": 0.0026, "step": 15020 }, { "epoch": 9.89466754443713, "grad_norm": 0.11433868855237961, "learning_rate": 9.727527079820742e-05, "loss": 0.0057, "step": 15030 }, { "epoch": 9.901250822909809, "grad_norm": 0.14294905960559845, "learning_rate": 9.726988441027186e-05, "loss": 0.0031, "step": 15040 }, { "epoch": 9.907834101382488, "grad_norm": 0.18759769201278687, "learning_rate": 9.726449285297281e-05, "loss": 0.0026, "step": 15050 }, { "epoch": 9.914417379855168, "grad_norm": 0.14776794612407684, "learning_rate": 9.72590961268999e-05, "loss": 0.0042, "step": 15060 }, { "epoch": 9.921000658327847, "grad_norm": 0.22177553176879883, "learning_rate": 9.725369423264328e-05, "loss": 0.0032, "step": 15070 }, { "epoch": 9.927583936800527, "grad_norm": 0.0834411010146141, "learning_rate": 9.72482871707937e-05, "loss": 0.0027, "step": 15080 }, { "epoch": 9.934167215273206, "grad_norm": 0.14563776552677155, "learning_rate": 9.724287494194247e-05, "loss": 0.0023, "step": 15090 }, { "epoch": 9.940750493745885, "grad_norm": 0.13955265283584595, "learning_rate": 9.723745754668147e-05, "loss": 0.0028, "step": 15100 }, { "epoch": 9.947333772218565, "grad_norm": 0.11536721885204315, "learning_rate": 9.723203498560313e-05, "loss": 0.0021, "step": 15110 }, { "epoch": 9.953917050691244, "grad_norm": 0.10777350515127182, "learning_rate": 9.722660725930046e-05, "loss": 0.0024, "step": 15120 }, { "epoch": 9.960500329163924, "grad_norm": 0.08521030843257904, "learning_rate": 9.722117436836702e-05, "loss": 0.0016, "step": 15130 }, { "epoch": 9.967083607636603, "grad_norm": 0.25590845942497253, "learning_rate": 9.721573631339696e-05, "loss": 0.005, "step": 15140 }, { "epoch": 9.973666886109282, "grad_norm": 0.09746973216533661, "learning_rate": 9.721029309498494e-05, "loss": 0.0028, "step": 15150 }, { "epoch": 9.980250164581962, "grad_norm": 0.11933630704879761, "learning_rate": 9.720484471372627e-05, "loss": 0.005, "step": 15160 }, { "epoch": 9.98683344305464, "grad_norm": 0.10521294176578522, "learning_rate": 9.719939117021673e-05, "loss": 0.0035, "step": 15170 }, { "epoch": 9.993416721527321, "grad_norm": 0.12911196053028107, "learning_rate": 9.719393246505275e-05, "loss": 0.0034, "step": 15180 }, { "epoch": 10.0, "grad_norm": 0.1692739874124527, "learning_rate": 9.718846859883128e-05, "loss": 0.0036, "step": 15190 }, { "epoch": 10.006583278472679, "grad_norm": 0.08036725223064423, "learning_rate": 9.718299957214982e-05, "loss": 0.0038, "step": 15200 }, { "epoch": 10.01316655694536, "grad_norm": 0.10568491369485855, "learning_rate": 9.717752538560646e-05, "loss": 0.0024, "step": 15210 }, { "epoch": 10.019749835418038, "grad_norm": 0.17889145016670227, "learning_rate": 9.717204603979986e-05, "loss": 0.0028, "step": 15220 }, { "epoch": 10.026333113890718, "grad_norm": 0.1264137625694275, "learning_rate": 9.716656153532922e-05, "loss": 0.0036, "step": 15230 }, { "epoch": 10.032916392363397, "grad_norm": 0.10848905891180038, "learning_rate": 9.716107187279434e-05, "loss": 0.0028, "step": 15240 }, { "epoch": 10.039499670836076, "grad_norm": 0.1847030520439148, "learning_rate": 9.715557705279555e-05, "loss": 0.0033, "step": 15250 }, { "epoch": 10.046082949308756, "grad_norm": 0.09012522548437119, "learning_rate": 9.715007707593372e-05, "loss": 0.0026, "step": 15260 }, { "epoch": 10.052666227781435, "grad_norm": 0.17811860144138336, "learning_rate": 9.714457194281036e-05, "loss": 0.0038, "step": 15270 }, { "epoch": 10.059249506254115, "grad_norm": 0.12462268024682999, "learning_rate": 9.713906165402751e-05, "loss": 0.0019, "step": 15280 }, { "epoch": 10.065832784726794, "grad_norm": 0.1381605565547943, "learning_rate": 9.713354621018774e-05, "loss": 0.0045, "step": 15290 }, { "epoch": 10.072416063199473, "grad_norm": 0.12458297610282898, "learning_rate": 9.712802561189422e-05, "loss": 0.0029, "step": 15300 }, { "epoch": 10.078999341672153, "grad_norm": 0.12490264326334, "learning_rate": 9.712249985975069e-05, "loss": 0.0036, "step": 15310 }, { "epoch": 10.085582620144832, "grad_norm": 0.12197179347276688, "learning_rate": 9.71169689543614e-05, "loss": 0.0023, "step": 15320 }, { "epoch": 10.09216589861751, "grad_norm": 0.14768095314502716, "learning_rate": 9.711143289633123e-05, "loss": 0.0044, "step": 15330 }, { "epoch": 10.098749177090191, "grad_norm": 0.13682861626148224, "learning_rate": 9.710589168626561e-05, "loss": 0.0024, "step": 15340 }, { "epoch": 10.10533245556287, "grad_norm": 0.08216214925050735, "learning_rate": 9.710034532477048e-05, "loss": 0.003, "step": 15350 }, { "epoch": 10.11191573403555, "grad_norm": 0.12094544619321823, "learning_rate": 9.709479381245239e-05, "loss": 0.0025, "step": 15360 }, { "epoch": 10.118499012508229, "grad_norm": 0.08196157962083817, "learning_rate": 9.708923714991847e-05, "loss": 0.0045, "step": 15370 }, { "epoch": 10.125082290980908, "grad_norm": 0.09161180257797241, "learning_rate": 9.708367533777638e-05, "loss": 0.0027, "step": 15380 }, { "epoch": 10.131665569453588, "grad_norm": 0.0969768613576889, "learning_rate": 9.707810837663431e-05, "loss": 0.0029, "step": 15390 }, { "epoch": 10.138248847926267, "grad_norm": 0.14186930656433105, "learning_rate": 9.707253626710113e-05, "loss": 0.0033, "step": 15400 }, { "epoch": 10.144832126398947, "grad_norm": 0.0886678695678711, "learning_rate": 9.706695900978613e-05, "loss": 0.0027, "step": 15410 }, { "epoch": 10.151415404871626, "grad_norm": 0.2785549461841583, "learning_rate": 9.706137660529926e-05, "loss": 0.0045, "step": 15420 }, { "epoch": 10.157998683344305, "grad_norm": 0.18256273865699768, "learning_rate": 9.705578905425101e-05, "loss": 0.0038, "step": 15430 }, { "epoch": 10.164581961816985, "grad_norm": 0.14723773300647736, "learning_rate": 9.705019635725241e-05, "loss": 0.0034, "step": 15440 }, { "epoch": 10.171165240289664, "grad_norm": 0.2125595360994339, "learning_rate": 9.704459851491508e-05, "loss": 0.0051, "step": 15450 }, { "epoch": 10.177748518762344, "grad_norm": 0.1273730844259262, "learning_rate": 9.703899552785118e-05, "loss": 0.0055, "step": 15460 }, { "epoch": 10.184331797235023, "grad_norm": 0.15798170864582062, "learning_rate": 9.703338739667346e-05, "loss": 0.0033, "step": 15470 }, { "epoch": 10.190915075707702, "grad_norm": 0.1373198926448822, "learning_rate": 9.70277741219952e-05, "loss": 0.004, "step": 15480 }, { "epoch": 10.197498354180382, "grad_norm": 0.149546816945076, "learning_rate": 9.702215570443027e-05, "loss": 0.0024, "step": 15490 }, { "epoch": 10.204081632653061, "grad_norm": 0.13453859090805054, "learning_rate": 9.701653214459309e-05, "loss": 0.0033, "step": 15500 }, { "epoch": 10.210664911125741, "grad_norm": 0.11023376137018204, "learning_rate": 9.701090344309865e-05, "loss": 0.0025, "step": 15510 }, { "epoch": 10.21724818959842, "grad_norm": 0.172676220536232, "learning_rate": 9.700526960056247e-05, "loss": 0.003, "step": 15520 }, { "epoch": 10.223831468071099, "grad_norm": 0.1117565855383873, "learning_rate": 9.699963061760068e-05, "loss": 0.0014, "step": 15530 }, { "epoch": 10.23041474654378, "grad_norm": 0.15941472351551056, "learning_rate": 9.699398649482997e-05, "loss": 0.0019, "step": 15540 }, { "epoch": 10.236998025016458, "grad_norm": 0.10144273191690445, "learning_rate": 9.698833723286753e-05, "loss": 0.0036, "step": 15550 }, { "epoch": 10.243581303489137, "grad_norm": 0.1587684452533722, "learning_rate": 9.698268283233118e-05, "loss": 0.0033, "step": 15560 }, { "epoch": 10.250164581961817, "grad_norm": 0.11091143637895584, "learning_rate": 9.697702329383929e-05, "loss": 0.0026, "step": 15570 }, { "epoch": 10.256747860434496, "grad_norm": 0.10850033909082413, "learning_rate": 9.697135861801074e-05, "loss": 0.0018, "step": 15580 }, { "epoch": 10.263331138907176, "grad_norm": 0.10011931508779526, "learning_rate": 9.696568880546505e-05, "loss": 0.0038, "step": 15590 }, { "epoch": 10.269914417379855, "grad_norm": 0.1127467229962349, "learning_rate": 9.696001385682223e-05, "loss": 0.002, "step": 15600 }, { "epoch": 10.276497695852534, "grad_norm": 0.08754849433898926, "learning_rate": 9.695433377270291e-05, "loss": 0.0022, "step": 15610 }, { "epoch": 10.283080974325214, "grad_norm": 0.10651049017906189, "learning_rate": 9.694864855372824e-05, "loss": 0.0041, "step": 15620 }, { "epoch": 10.289664252797893, "grad_norm": 0.1769791692495346, "learning_rate": 9.694295820051995e-05, "loss": 0.0068, "step": 15630 }, { "epoch": 10.296247531270573, "grad_norm": 0.1355508267879486, "learning_rate": 9.693726271370032e-05, "loss": 0.0035, "step": 15640 }, { "epoch": 10.302830809743252, "grad_norm": 0.1758781522512436, "learning_rate": 9.693156209389221e-05, "loss": 0.0059, "step": 15650 }, { "epoch": 10.30941408821593, "grad_norm": 0.1289355605840683, "learning_rate": 9.692585634171905e-05, "loss": 0.0079, "step": 15660 }, { "epoch": 10.315997366688611, "grad_norm": 0.10526596009731293, "learning_rate": 9.692014545780476e-05, "loss": 0.0028, "step": 15670 }, { "epoch": 10.32258064516129, "grad_norm": 0.21284706890583038, "learning_rate": 9.691442944277393e-05, "loss": 0.0048, "step": 15680 }, { "epoch": 10.32916392363397, "grad_norm": 0.21005521714687347, "learning_rate": 9.690870829725162e-05, "loss": 0.0039, "step": 15690 }, { "epoch": 10.33574720210665, "grad_norm": 0.13361205160617828, "learning_rate": 9.69029820218635e-05, "loss": 0.0034, "step": 15700 }, { "epoch": 10.342330480579328, "grad_norm": 0.16389797627925873, "learning_rate": 9.689725061723579e-05, "loss": 0.0026, "step": 15710 }, { "epoch": 10.348913759052008, "grad_norm": 0.127926304936409, "learning_rate": 9.689151408399527e-05, "loss": 0.0025, "step": 15720 }, { "epoch": 10.355497037524687, "grad_norm": 0.1129746288061142, "learning_rate": 9.688577242276924e-05, "loss": 0.0027, "step": 15730 }, { "epoch": 10.362080315997368, "grad_norm": 0.09847691655158997, "learning_rate": 9.688002563418566e-05, "loss": 0.0024, "step": 15740 }, { "epoch": 10.368663594470046, "grad_norm": 0.10093290358781815, "learning_rate": 9.687427371887293e-05, "loss": 0.0032, "step": 15750 }, { "epoch": 10.375246872942725, "grad_norm": 0.14787502586841583, "learning_rate": 9.686851667746012e-05, "loss": 0.0024, "step": 15760 }, { "epoch": 10.381830151415405, "grad_norm": 0.07472986727952957, "learning_rate": 9.686275451057677e-05, "loss": 0.0031, "step": 15770 }, { "epoch": 10.388413429888084, "grad_norm": 0.16863051056861877, "learning_rate": 9.685698721885308e-05, "loss": 0.0025, "step": 15780 }, { "epoch": 10.394996708360763, "grad_norm": 0.07984867691993713, "learning_rate": 9.68512148029197e-05, "loss": 0.0034, "step": 15790 }, { "epoch": 10.401579986833443, "grad_norm": 0.0832223892211914, "learning_rate": 9.684543726340791e-05, "loss": 0.002, "step": 15800 }, { "epoch": 10.408163265306122, "grad_norm": 0.2888793349266052, "learning_rate": 9.683965460094952e-05, "loss": 0.0043, "step": 15810 }, { "epoch": 10.414746543778802, "grad_norm": 0.12970487773418427, "learning_rate": 9.683386681617694e-05, "loss": 0.0038, "step": 15820 }, { "epoch": 10.421329822251481, "grad_norm": 0.1093292161822319, "learning_rate": 9.68280739097231e-05, "loss": 0.0026, "step": 15830 }, { "epoch": 10.42791310072416, "grad_norm": 0.11228504776954651, "learning_rate": 9.682227588222148e-05, "loss": 0.0028, "step": 15840 }, { "epoch": 10.43449637919684, "grad_norm": 0.16591253876686096, "learning_rate": 9.681647273430618e-05, "loss": 0.0025, "step": 15850 }, { "epoch": 10.441079657669519, "grad_norm": 0.18718938529491425, "learning_rate": 9.681066446661182e-05, "loss": 0.005, "step": 15860 }, { "epoch": 10.4476629361422, "grad_norm": 0.16160349547863007, "learning_rate": 9.680485107977357e-05, "loss": 0.0032, "step": 15870 }, { "epoch": 10.454246214614878, "grad_norm": 0.14637888967990875, "learning_rate": 9.679903257442716e-05, "loss": 0.0033, "step": 15880 }, { "epoch": 10.460829493087557, "grad_norm": 0.1529206931591034, "learning_rate": 9.679320895120891e-05, "loss": 0.0043, "step": 15890 }, { "epoch": 10.467412771560237, "grad_norm": 0.14626345038414001, "learning_rate": 9.67873802107557e-05, "loss": 0.0103, "step": 15900 }, { "epoch": 10.473996050032916, "grad_norm": 0.15902835130691528, "learning_rate": 9.67815463537049e-05, "loss": 0.0056, "step": 15910 }, { "epoch": 10.480579328505597, "grad_norm": 0.13061124086380005, "learning_rate": 9.677570738069457e-05, "loss": 0.0061, "step": 15920 }, { "epoch": 10.487162606978275, "grad_norm": 0.11949317902326584, "learning_rate": 9.676986329236318e-05, "loss": 0.0038, "step": 15930 }, { "epoch": 10.493745885450954, "grad_norm": 0.1682298481464386, "learning_rate": 9.676401408934987e-05, "loss": 0.0045, "step": 15940 }, { "epoch": 10.500329163923634, "grad_norm": 0.11666899919509888, "learning_rate": 9.675815977229428e-05, "loss": 0.0088, "step": 15950 }, { "epoch": 10.506912442396313, "grad_norm": 0.11398320645093918, "learning_rate": 9.675230034183664e-05, "loss": 0.0085, "step": 15960 }, { "epoch": 10.513495720868992, "grad_norm": 0.09845927357673645, "learning_rate": 9.674643579861773e-05, "loss": 0.0021, "step": 15970 }, { "epoch": 10.520078999341672, "grad_norm": 0.10867223143577576, "learning_rate": 9.674056614327886e-05, "loss": 0.0062, "step": 15980 }, { "epoch": 10.526662277814351, "grad_norm": 0.16895629465579987, "learning_rate": 9.673469137646198e-05, "loss": 0.0033, "step": 15990 }, { "epoch": 10.533245556287032, "grad_norm": 0.14702971279621124, "learning_rate": 9.67288114988095e-05, "loss": 0.0043, "step": 16000 }, { "epoch": 10.53982883475971, "grad_norm": 0.11904754489660263, "learning_rate": 9.672292651096447e-05, "loss": 0.0042, "step": 16010 }, { "epoch": 10.546412113232389, "grad_norm": 0.1475028097629547, "learning_rate": 9.671703641357042e-05, "loss": 0.0048, "step": 16020 }, { "epoch": 10.55299539170507, "grad_norm": 0.18422089517116547, "learning_rate": 9.67111412072715e-05, "loss": 0.0038, "step": 16030 }, { "epoch": 10.559578670177748, "grad_norm": 0.14086811244487762, "learning_rate": 9.670524089271242e-05, "loss": 0.0028, "step": 16040 }, { "epoch": 10.566161948650429, "grad_norm": 0.17404000461101532, "learning_rate": 9.669933547053842e-05, "loss": 0.0043, "step": 16050 }, { "epoch": 10.572745227123107, "grad_norm": 0.11574403941631317, "learning_rate": 9.669342494139531e-05, "loss": 0.0029, "step": 16060 }, { "epoch": 10.579328505595786, "grad_norm": 0.09983602911233902, "learning_rate": 9.668750930592943e-05, "loss": 0.003, "step": 16070 }, { "epoch": 10.585911784068466, "grad_norm": 0.12089818716049194, "learning_rate": 9.668158856478775e-05, "loss": 0.0022, "step": 16080 }, { "epoch": 10.592495062541145, "grad_norm": 0.12322981655597687, "learning_rate": 9.66756627186177e-05, "loss": 0.0025, "step": 16090 }, { "epoch": 10.599078341013826, "grad_norm": 0.10360847413539886, "learning_rate": 9.666973176806737e-05, "loss": 0.0024, "step": 16100 }, { "epoch": 10.605661619486504, "grad_norm": 0.15599210560321808, "learning_rate": 9.666379571378534e-05, "loss": 0.0028, "step": 16110 }, { "epoch": 10.612244897959183, "grad_norm": 0.16357773542404175, "learning_rate": 9.665785455642076e-05, "loss": 0.003, "step": 16120 }, { "epoch": 10.618828176431863, "grad_norm": 0.12985104322433472, "learning_rate": 9.665190829662337e-05, "loss": 0.0038, "step": 16130 }, { "epoch": 10.625411454904542, "grad_norm": 0.10305918753147125, "learning_rate": 9.664595693504342e-05, "loss": 0.0017, "step": 16140 }, { "epoch": 10.631994733377223, "grad_norm": 0.09795692563056946, "learning_rate": 9.664000047233175e-05, "loss": 0.0025, "step": 16150 }, { "epoch": 10.638578011849901, "grad_norm": 0.12221822142601013, "learning_rate": 9.663403890913976e-05, "loss": 0.0031, "step": 16160 }, { "epoch": 10.64516129032258, "grad_norm": 0.1045532375574112, "learning_rate": 9.662807224611938e-05, "loss": 0.0028, "step": 16170 }, { "epoch": 10.65174456879526, "grad_norm": 0.0884002298116684, "learning_rate": 9.662210048392311e-05, "loss": 0.0023, "step": 16180 }, { "epoch": 10.65832784726794, "grad_norm": 0.13131257891654968, "learning_rate": 9.661612362320405e-05, "loss": 0.0026, "step": 16190 }, { "epoch": 10.66491112574062, "grad_norm": 0.1579866260290146, "learning_rate": 9.661014166461579e-05, "loss": 0.0027, "step": 16200 }, { "epoch": 10.671494404213298, "grad_norm": 0.16512525081634521, "learning_rate": 9.66041546088125e-05, "loss": 0.0026, "step": 16210 }, { "epoch": 10.678077682685977, "grad_norm": 0.08105803281068802, "learning_rate": 9.659816245644895e-05, "loss": 0.0028, "step": 16220 }, { "epoch": 10.684660961158658, "grad_norm": 0.0860104188323021, "learning_rate": 9.65921652081804e-05, "loss": 0.0022, "step": 16230 }, { "epoch": 10.691244239631336, "grad_norm": 0.08805694431066513, "learning_rate": 9.658616286466271e-05, "loss": 0.0034, "step": 16240 }, { "epoch": 10.697827518104015, "grad_norm": 0.11462710052728653, "learning_rate": 9.65801554265523e-05, "loss": 0.0022, "step": 16250 }, { "epoch": 10.704410796576695, "grad_norm": 0.0732785314321518, "learning_rate": 9.657414289450612e-05, "loss": 0.0015, "step": 16260 }, { "epoch": 10.710994075049374, "grad_norm": 0.1536480188369751, "learning_rate": 9.656812526918171e-05, "loss": 0.0028, "step": 16270 }, { "epoch": 10.717577353522055, "grad_norm": 0.10468407720327377, "learning_rate": 9.656210255123712e-05, "loss": 0.0041, "step": 16280 }, { "epoch": 10.724160631994733, "grad_norm": 0.16834744811058044, "learning_rate": 9.6556074741331e-05, "loss": 0.0029, "step": 16290 }, { "epoch": 10.730743910467412, "grad_norm": 0.09658673405647278, "learning_rate": 9.655004184012256e-05, "loss": 0.0037, "step": 16300 }, { "epoch": 10.737327188940093, "grad_norm": 0.18123260140419006, "learning_rate": 9.654400384827152e-05, "loss": 0.0037, "step": 16310 }, { "epoch": 10.743910467412771, "grad_norm": 0.10468707978725433, "learning_rate": 9.653796076643818e-05, "loss": 0.0027, "step": 16320 }, { "epoch": 10.750493745885452, "grad_norm": 0.1499488800764084, "learning_rate": 9.653191259528344e-05, "loss": 0.0023, "step": 16330 }, { "epoch": 10.75707702435813, "grad_norm": 0.18210500478744507, "learning_rate": 9.65258593354687e-05, "loss": 0.0032, "step": 16340 }, { "epoch": 10.763660302830809, "grad_norm": 0.19157780706882477, "learning_rate": 9.651980098765591e-05, "loss": 0.0031, "step": 16350 }, { "epoch": 10.77024358130349, "grad_norm": 0.11963818222284317, "learning_rate": 9.651373755250765e-05, "loss": 0.0033, "step": 16360 }, { "epoch": 10.776826859776168, "grad_norm": 0.10565931349992752, "learning_rate": 9.650766903068697e-05, "loss": 0.0029, "step": 16370 }, { "epoch": 10.783410138248849, "grad_norm": 0.1245986744761467, "learning_rate": 9.650159542285753e-05, "loss": 0.0023, "step": 16380 }, { "epoch": 10.789993416721527, "grad_norm": 0.19217826426029205, "learning_rate": 9.649551672968353e-05, "loss": 0.0038, "step": 16390 }, { "epoch": 10.796576695194206, "grad_norm": 0.1583852469921112, "learning_rate": 9.648943295182973e-05, "loss": 0.0025, "step": 16400 }, { "epoch": 10.803159973666887, "grad_norm": 0.1700221300125122, "learning_rate": 9.648334408996144e-05, "loss": 0.0026, "step": 16410 }, { "epoch": 10.809743252139565, "grad_norm": 0.14200282096862793, "learning_rate": 9.647725014474452e-05, "loss": 0.0073, "step": 16420 }, { "epoch": 10.816326530612244, "grad_norm": 0.13857600092887878, "learning_rate": 9.64711511168454e-05, "loss": 0.0048, "step": 16430 }, { "epoch": 10.822909809084925, "grad_norm": 0.13938231766223907, "learning_rate": 9.646504700693108e-05, "loss": 0.0024, "step": 16440 }, { "epoch": 10.829493087557603, "grad_norm": 0.09454157948493958, "learning_rate": 9.645893781566907e-05, "loss": 0.0034, "step": 16450 }, { "epoch": 10.836076366030284, "grad_norm": 0.14099504053592682, "learning_rate": 9.645282354372744e-05, "loss": 0.0041, "step": 16460 }, { "epoch": 10.842659644502962, "grad_norm": 0.155318483710289, "learning_rate": 9.644670419177491e-05, "loss": 0.0027, "step": 16470 }, { "epoch": 10.849242922975641, "grad_norm": 0.12236496061086655, "learning_rate": 9.644057976048062e-05, "loss": 0.0028, "step": 16480 }, { "epoch": 10.855826201448322, "grad_norm": 0.10719477385282516, "learning_rate": 9.643445025051435e-05, "loss": 0.0025, "step": 16490 }, { "epoch": 10.862409479921, "grad_norm": 0.07888966798782349, "learning_rate": 9.642831566254641e-05, "loss": 0.0024, "step": 16500 }, { "epoch": 10.86899275839368, "grad_norm": 0.10415275394916534, "learning_rate": 9.642217599724769e-05, "loss": 0.0031, "step": 16510 }, { "epoch": 10.87557603686636, "grad_norm": 0.08183840662240982, "learning_rate": 9.64160312552896e-05, "loss": 0.0027, "step": 16520 }, { "epoch": 10.882159315339038, "grad_norm": 0.19453145563602448, "learning_rate": 9.64098814373441e-05, "loss": 0.0042, "step": 16530 }, { "epoch": 10.888742593811719, "grad_norm": 0.14911939203739166, "learning_rate": 9.640372654408374e-05, "loss": 0.0085, "step": 16540 }, { "epoch": 10.895325872284397, "grad_norm": 0.11736586689949036, "learning_rate": 9.639756657618162e-05, "loss": 0.0054, "step": 16550 }, { "epoch": 10.901909150757078, "grad_norm": 0.1890140175819397, "learning_rate": 9.639140153431138e-05, "loss": 0.0031, "step": 16560 }, { "epoch": 10.908492429229756, "grad_norm": 0.17033985257148743, "learning_rate": 9.638523141914721e-05, "loss": 0.0047, "step": 16570 }, { "epoch": 10.915075707702435, "grad_norm": 0.16564054787158966, "learning_rate": 9.637905623136388e-05, "loss": 0.0024, "step": 16580 }, { "epoch": 10.921658986175116, "grad_norm": 0.22322309017181396, "learning_rate": 9.637287597163669e-05, "loss": 0.0029, "step": 16590 }, { "epoch": 10.928242264647794, "grad_norm": 0.15496188402175903, "learning_rate": 9.63666906406415e-05, "loss": 0.002, "step": 16600 }, { "epoch": 10.934825543120475, "grad_norm": 0.09715162962675095, "learning_rate": 9.636050023905473e-05, "loss": 0.0039, "step": 16610 }, { "epoch": 10.941408821593154, "grad_norm": 0.11376439779996872, "learning_rate": 9.635430476755336e-05, "loss": 0.0075, "step": 16620 }, { "epoch": 10.947992100065832, "grad_norm": 0.16445519030094147, "learning_rate": 9.63481042268149e-05, "loss": 0.005, "step": 16630 }, { "epoch": 10.954575378538513, "grad_norm": 0.15176072716712952, "learning_rate": 9.634189861751745e-05, "loss": 0.0031, "step": 16640 }, { "epoch": 10.961158657011191, "grad_norm": 0.14566351473331451, "learning_rate": 9.633568794033967e-05, "loss": 0.0031, "step": 16650 }, { "epoch": 10.967741935483872, "grad_norm": 0.09591031819581985, "learning_rate": 9.63294721959607e-05, "loss": 0.0034, "step": 16660 }, { "epoch": 10.97432521395655, "grad_norm": 0.19433479011058807, "learning_rate": 9.63232513850603e-05, "loss": 0.0036, "step": 16670 }, { "epoch": 10.98090849242923, "grad_norm": 0.10979446768760681, "learning_rate": 9.631702550831878e-05, "loss": 0.002, "step": 16680 }, { "epoch": 10.98749177090191, "grad_norm": 0.18959882855415344, "learning_rate": 9.631079456641698e-05, "loss": 0.004, "step": 16690 }, { "epoch": 10.994075049374588, "grad_norm": 0.11260781437158585, "learning_rate": 9.630455856003632e-05, "loss": 0.0023, "step": 16700 }, { "epoch": 11.000658327847267, "grad_norm": 0.10758821666240692, "learning_rate": 9.629831748985876e-05, "loss": 0.0029, "step": 16710 }, { "epoch": 11.007241606319948, "grad_norm": 0.12162449955940247, "learning_rate": 9.629207135656679e-05, "loss": 0.0025, "step": 16720 }, { "epoch": 11.013824884792626, "grad_norm": 0.09082021564245224, "learning_rate": 9.628582016084353e-05, "loss": 0.0018, "step": 16730 }, { "epoch": 11.020408163265307, "grad_norm": 0.11014358699321747, "learning_rate": 9.627956390337254e-05, "loss": 0.0016, "step": 16740 }, { "epoch": 11.026991441737986, "grad_norm": 0.07644966244697571, "learning_rate": 9.627330258483802e-05, "loss": 0.0024, "step": 16750 }, { "epoch": 11.033574720210664, "grad_norm": 0.08015231788158417, "learning_rate": 9.62670362059247e-05, "loss": 0.003, "step": 16760 }, { "epoch": 11.040157998683345, "grad_norm": 0.09217200428247452, "learning_rate": 9.626076476731786e-05, "loss": 0.0025, "step": 16770 }, { "epoch": 11.046741277156023, "grad_norm": 0.08850143104791641, "learning_rate": 9.625448826970336e-05, "loss": 0.0041, "step": 16780 }, { "epoch": 11.053324555628704, "grad_norm": 0.1618192046880722, "learning_rate": 9.624820671376755e-05, "loss": 0.0028, "step": 16790 }, { "epoch": 11.059907834101383, "grad_norm": 0.08958488702774048, "learning_rate": 9.62419201001974e-05, "loss": 0.0021, "step": 16800 }, { "epoch": 11.066491112574061, "grad_norm": 0.10807262361049652, "learning_rate": 9.623562842968037e-05, "loss": 0.0025, "step": 16810 }, { "epoch": 11.073074391046742, "grad_norm": 0.15803654491901398, "learning_rate": 9.622933170290454e-05, "loss": 0.0026, "step": 16820 }, { "epoch": 11.07965766951942, "grad_norm": 0.10931402444839478, "learning_rate": 9.622302992055849e-05, "loss": 0.0027, "step": 16830 }, { "epoch": 11.086240947992101, "grad_norm": 0.08856100589036942, "learning_rate": 9.62167230833314e-05, "loss": 0.0027, "step": 16840 }, { "epoch": 11.09282422646478, "grad_norm": 0.11392121016979218, "learning_rate": 9.621041119191295e-05, "loss": 0.0018, "step": 16850 }, { "epoch": 11.099407504937458, "grad_norm": 0.08718714863061905, "learning_rate": 9.620409424699342e-05, "loss": 0.0027, "step": 16860 }, { "epoch": 11.105990783410139, "grad_norm": 0.11326928436756134, "learning_rate": 9.619777224926359e-05, "loss": 0.0018, "step": 16870 }, { "epoch": 11.112574061882817, "grad_norm": 0.1599523425102234, "learning_rate": 9.619144519941485e-05, "loss": 0.0029, "step": 16880 }, { "epoch": 11.119157340355496, "grad_norm": 0.09211993217468262, "learning_rate": 9.618511309813912e-05, "loss": 0.0021, "step": 16890 }, { "epoch": 11.125740618828177, "grad_norm": 0.19001148641109467, "learning_rate": 9.617877594612886e-05, "loss": 0.0023, "step": 16900 }, { "epoch": 11.132323897300855, "grad_norm": 0.10113160312175751, "learning_rate": 9.617243374407707e-05, "loss": 0.0018, "step": 16910 }, { "epoch": 11.138907175773536, "grad_norm": 0.133889302611351, "learning_rate": 9.616608649267736e-05, "loss": 0.0021, "step": 16920 }, { "epoch": 11.145490454246215, "grad_norm": 0.09217289090156555, "learning_rate": 9.615973419262385e-05, "loss": 0.0023, "step": 16930 }, { "epoch": 11.152073732718893, "grad_norm": 0.13893038034439087, "learning_rate": 9.615337684461119e-05, "loss": 0.0016, "step": 16940 }, { "epoch": 11.158657011191574, "grad_norm": 0.1257384866476059, "learning_rate": 9.614701444933465e-05, "loss": 0.0024, "step": 16950 }, { "epoch": 11.165240289664252, "grad_norm": 0.18623808026313782, "learning_rate": 9.614064700748997e-05, "loss": 0.0084, "step": 16960 }, { "epoch": 11.171823568136933, "grad_norm": 0.10422549396753311, "learning_rate": 9.613427451977352e-05, "loss": 0.0021, "step": 16970 }, { "epoch": 11.178406846609612, "grad_norm": 0.15108440816402435, "learning_rate": 9.612789698688216e-05, "loss": 0.0057, "step": 16980 }, { "epoch": 11.18499012508229, "grad_norm": 0.10465966165065765, "learning_rate": 9.612151440951334e-05, "loss": 0.0022, "step": 16990 }, { "epoch": 11.19157340355497, "grad_norm": 0.1550467163324356, "learning_rate": 9.611512678836506e-05, "loss": 0.0027, "step": 17000 }, { "epoch": 11.19815668202765, "grad_norm": 0.11843186616897583, "learning_rate": 9.610873412413584e-05, "loss": 0.0035, "step": 17010 }, { "epoch": 11.20473996050033, "grad_norm": 0.12927711009979248, "learning_rate": 9.610233641752476e-05, "loss": 0.003, "step": 17020 }, { "epoch": 11.211323238973009, "grad_norm": 0.1826525777578354, "learning_rate": 9.609593366923151e-05, "loss": 0.0031, "step": 17030 }, { "epoch": 11.217906517445687, "grad_norm": 0.13722099363803864, "learning_rate": 9.608952587995625e-05, "loss": 0.0028, "step": 17040 }, { "epoch": 11.224489795918368, "grad_norm": 0.10018254071474075, "learning_rate": 9.608311305039972e-05, "loss": 0.0024, "step": 17050 }, { "epoch": 11.231073074391047, "grad_norm": 0.14540746808052063, "learning_rate": 9.607669518126326e-05, "loss": 0.0023, "step": 17060 }, { "epoch": 11.237656352863727, "grad_norm": 0.1475183516740799, "learning_rate": 9.607027227324866e-05, "loss": 0.0018, "step": 17070 }, { "epoch": 11.244239631336406, "grad_norm": 0.10400071740150452, "learning_rate": 9.606384432705837e-05, "loss": 0.0022, "step": 17080 }, { "epoch": 11.250822909809084, "grad_norm": 0.08921986073255539, "learning_rate": 9.60574113433953e-05, "loss": 0.0035, "step": 17090 }, { "epoch": 11.257406188281765, "grad_norm": 0.09831489622592926, "learning_rate": 9.6050973322963e-05, "loss": 0.0023, "step": 17100 }, { "epoch": 11.263989466754444, "grad_norm": 0.12723112106323242, "learning_rate": 9.604453026646547e-05, "loss": 0.0019, "step": 17110 }, { "epoch": 11.270572745227122, "grad_norm": 0.08028861880302429, "learning_rate": 9.603808217460735e-05, "loss": 0.0028, "step": 17120 }, { "epoch": 11.277156023699803, "grad_norm": 0.14342691004276276, "learning_rate": 9.603162904809377e-05, "loss": 0.0035, "step": 17130 }, { "epoch": 11.283739302172481, "grad_norm": 0.12297974526882172, "learning_rate": 9.602517088763045e-05, "loss": 0.0041, "step": 17140 }, { "epoch": 11.290322580645162, "grad_norm": 0.11070414632558823, "learning_rate": 9.601870769392365e-05, "loss": 0.0027, "step": 17150 }, { "epoch": 11.29690585911784, "grad_norm": 0.11125894635915756, "learning_rate": 9.601223946768017e-05, "loss": 0.0027, "step": 17160 }, { "epoch": 11.30348913759052, "grad_norm": 0.07118100672960281, "learning_rate": 9.600576620960734e-05, "loss": 0.0015, "step": 17170 }, { "epoch": 11.3100724160632, "grad_norm": 0.15638680756092072, "learning_rate": 9.599928792041308e-05, "loss": 0.0026, "step": 17180 }, { "epoch": 11.316655694535878, "grad_norm": 0.22163055837154388, "learning_rate": 9.599280460080587e-05, "loss": 0.0037, "step": 17190 }, { "epoch": 11.323238973008559, "grad_norm": 0.11759576201438904, "learning_rate": 9.59863162514947e-05, "loss": 0.0031, "step": 17200 }, { "epoch": 11.329822251481238, "grad_norm": 0.14910882711410522, "learning_rate": 9.597982287318911e-05, "loss": 0.0028, "step": 17210 }, { "epoch": 11.336405529953916, "grad_norm": 0.10312371701002121, "learning_rate": 9.597332446659923e-05, "loss": 0.004, "step": 17220 }, { "epoch": 11.342988808426597, "grad_norm": 0.11528109759092331, "learning_rate": 9.59668210324357e-05, "loss": 0.0033, "step": 17230 }, { "epoch": 11.349572086899276, "grad_norm": 0.11919239908456802, "learning_rate": 9.596031257140974e-05, "loss": 0.0033, "step": 17240 }, { "epoch": 11.356155365371956, "grad_norm": 0.12713322043418884, "learning_rate": 9.59537990842331e-05, "loss": 0.0027, "step": 17250 }, { "epoch": 11.362738643844635, "grad_norm": 0.153266042470932, "learning_rate": 9.594728057161806e-05, "loss": 0.0027, "step": 17260 }, { "epoch": 11.369321922317313, "grad_norm": 0.09669490158557892, "learning_rate": 9.594075703427752e-05, "loss": 0.0024, "step": 17270 }, { "epoch": 11.375905200789994, "grad_norm": 0.0749366283416748, "learning_rate": 9.593422847292486e-05, "loss": 0.0017, "step": 17280 }, { "epoch": 11.382488479262673, "grad_norm": 0.10245104879140854, "learning_rate": 9.592769488827402e-05, "loss": 0.0029, "step": 17290 }, { "epoch": 11.389071757735351, "grad_norm": 0.13814356923103333, "learning_rate": 9.592115628103952e-05, "loss": 0.0037, "step": 17300 }, { "epoch": 11.395655036208032, "grad_norm": 0.06844920665025711, "learning_rate": 9.591461265193643e-05, "loss": 0.0023, "step": 17310 }, { "epoch": 11.40223831468071, "grad_norm": 0.17652608454227448, "learning_rate": 9.590806400168032e-05, "loss": 0.0026, "step": 17320 }, { "epoch": 11.408821593153391, "grad_norm": 0.11553140729665756, "learning_rate": 9.590151033098735e-05, "loss": 0.0024, "step": 17330 }, { "epoch": 11.41540487162607, "grad_norm": 0.11591698974370956, "learning_rate": 9.589495164057423e-05, "loss": 0.004, "step": 17340 }, { "epoch": 11.421988150098748, "grad_norm": 0.1857786476612091, "learning_rate": 9.58883879311582e-05, "loss": 0.0027, "step": 17350 }, { "epoch": 11.428571428571429, "grad_norm": 0.0991809144616127, "learning_rate": 9.588181920345705e-05, "loss": 0.0053, "step": 17360 }, { "epoch": 11.435154707044108, "grad_norm": 0.1273217648267746, "learning_rate": 9.587524545818913e-05, "loss": 0.0023, "step": 17370 }, { "epoch": 11.441737985516788, "grad_norm": 0.1433270126581192, "learning_rate": 9.586866669607335e-05, "loss": 0.003, "step": 17380 }, { "epoch": 11.448321263989467, "grad_norm": 0.12651756405830383, "learning_rate": 9.586208291782915e-05, "loss": 0.0058, "step": 17390 }, { "epoch": 11.454904542462145, "grad_norm": 0.09153234958648682, "learning_rate": 9.58554941241765e-05, "loss": 0.0022, "step": 17400 }, { "epoch": 11.461487820934826, "grad_norm": 0.06976069509983063, "learning_rate": 9.584890031583596e-05, "loss": 0.0022, "step": 17410 }, { "epoch": 11.468071099407505, "grad_norm": 0.07782725989818573, "learning_rate": 9.584230149352861e-05, "loss": 0.0016, "step": 17420 }, { "epoch": 11.474654377880185, "grad_norm": 0.1182716116309166, "learning_rate": 9.58356976579761e-05, "loss": 0.0023, "step": 17430 }, { "epoch": 11.481237656352864, "grad_norm": 0.08439533412456512, "learning_rate": 9.58290888099006e-05, "loss": 0.002, "step": 17440 }, { "epoch": 11.487820934825542, "grad_norm": 0.16203516721725464, "learning_rate": 9.582247495002486e-05, "loss": 0.0033, "step": 17450 }, { "epoch": 11.494404213298223, "grad_norm": 0.11023887246847153, "learning_rate": 9.581585607907214e-05, "loss": 0.0035, "step": 17460 }, { "epoch": 11.500987491770902, "grad_norm": 0.07707885652780533, "learning_rate": 9.580923219776628e-05, "loss": 0.0018, "step": 17470 }, { "epoch": 11.507570770243582, "grad_norm": 0.13383297622203827, "learning_rate": 9.580260330683167e-05, "loss": 0.0038, "step": 17480 }, { "epoch": 11.51415404871626, "grad_norm": 0.12913091480731964, "learning_rate": 9.579596940699322e-05, "loss": 0.0034, "step": 17490 }, { "epoch": 11.52073732718894, "grad_norm": 0.12115701287984848, "learning_rate": 9.578933049897643e-05, "loss": 0.002, "step": 17500 }, { "epoch": 11.52732060566162, "grad_norm": 0.08611238747835159, "learning_rate": 9.578268658350728e-05, "loss": 0.0023, "step": 17510 }, { "epoch": 11.533903884134299, "grad_norm": 0.23324602842330933, "learning_rate": 9.577603766131235e-05, "loss": 0.0035, "step": 17520 }, { "epoch": 11.54048716260698, "grad_norm": 0.13333247601985931, "learning_rate": 9.576938373311878e-05, "loss": 0.0028, "step": 17530 }, { "epoch": 11.547070441079658, "grad_norm": 0.15441559255123138, "learning_rate": 9.576272479965421e-05, "loss": 0.0026, "step": 17540 }, { "epoch": 11.553653719552337, "grad_norm": 0.1675134152173996, "learning_rate": 9.575606086164687e-05, "loss": 0.0053, "step": 17550 }, { "epoch": 11.560236998025017, "grad_norm": 0.17737972736358643, "learning_rate": 9.57493919198255e-05, "loss": 0.0028, "step": 17560 }, { "epoch": 11.566820276497696, "grad_norm": 0.22449170053005219, "learning_rate": 9.57427179749194e-05, "loss": 0.0059, "step": 17570 }, { "epoch": 11.573403554970374, "grad_norm": 0.231312558054924, "learning_rate": 9.573603902765846e-05, "loss": 0.0033, "step": 17580 }, { "epoch": 11.579986833443055, "grad_norm": 0.10202602297067642, "learning_rate": 9.572935507877304e-05, "loss": 0.004, "step": 17590 }, { "epoch": 11.586570111915734, "grad_norm": 0.1417197734117508, "learning_rate": 9.57226661289941e-05, "loss": 0.0043, "step": 17600 }, { "epoch": 11.593153390388414, "grad_norm": 0.11091620475053787, "learning_rate": 9.571597217905315e-05, "loss": 0.0055, "step": 17610 }, { "epoch": 11.599736668861093, "grad_norm": 0.11134813725948334, "learning_rate": 9.57092732296822e-05, "loss": 0.0025, "step": 17620 }, { "epoch": 11.606319947333771, "grad_norm": 0.10720554739236832, "learning_rate": 9.570256928161385e-05, "loss": 0.0025, "step": 17630 }, { "epoch": 11.612903225806452, "grad_norm": 0.19860213994979858, "learning_rate": 9.569586033558126e-05, "loss": 0.0049, "step": 17640 }, { "epoch": 11.61948650427913, "grad_norm": 0.14914530515670776, "learning_rate": 9.568914639231807e-05, "loss": 0.0027, "step": 17650 }, { "epoch": 11.626069782751811, "grad_norm": 0.1456700712442398, "learning_rate": 9.568242745255852e-05, "loss": 0.0029, "step": 17660 }, { "epoch": 11.63265306122449, "grad_norm": 0.0915641337633133, "learning_rate": 9.567570351703739e-05, "loss": 0.002, "step": 17670 }, { "epoch": 11.639236339697169, "grad_norm": 0.10023776441812515, "learning_rate": 9.566897458649001e-05, "loss": 0.0023, "step": 17680 }, { "epoch": 11.645819618169849, "grad_norm": 0.15704162418842316, "learning_rate": 9.566224066165221e-05, "loss": 0.0027, "step": 17690 }, { "epoch": 11.652402896642528, "grad_norm": 0.09104430675506592, "learning_rate": 9.565550174326043e-05, "loss": 0.0036, "step": 17700 }, { "epoch": 11.658986175115208, "grad_norm": 0.11438848823308945, "learning_rate": 9.564875783205162e-05, "loss": 0.0035, "step": 17710 }, { "epoch": 11.665569453587887, "grad_norm": 0.09507377445697784, "learning_rate": 9.564200892876328e-05, "loss": 0.0036, "step": 17720 }, { "epoch": 11.672152732060566, "grad_norm": 0.11350542306900024, "learning_rate": 9.563525503413348e-05, "loss": 0.0027, "step": 17730 }, { "epoch": 11.678736010533246, "grad_norm": 0.09029316157102585, "learning_rate": 9.562849614890079e-05, "loss": 0.0017, "step": 17740 }, { "epoch": 11.685319289005925, "grad_norm": 0.17291457951068878, "learning_rate": 9.562173227380436e-05, "loss": 0.0026, "step": 17750 }, { "epoch": 11.691902567478603, "grad_norm": 0.14838126301765442, "learning_rate": 9.561496340958389e-05, "loss": 0.0042, "step": 17760 }, { "epoch": 11.698485845951284, "grad_norm": 0.15644460916519165, "learning_rate": 9.560818955697959e-05, "loss": 0.0036, "step": 17770 }, { "epoch": 11.705069124423963, "grad_norm": 0.12381359189748764, "learning_rate": 9.560141071673228e-05, "loss": 0.0024, "step": 17780 }, { "epoch": 11.711652402896643, "grad_norm": 0.13199445605278015, "learning_rate": 9.559462688958323e-05, "loss": 0.0022, "step": 17790 }, { "epoch": 11.718235681369322, "grad_norm": 0.16483323276042938, "learning_rate": 9.558783807627434e-05, "loss": 0.0069, "step": 17800 }, { "epoch": 11.724818959842, "grad_norm": 0.16913248598575592, "learning_rate": 9.558104427754801e-05, "loss": 0.0026, "step": 17810 }, { "epoch": 11.731402238314681, "grad_norm": 0.21817107498645782, "learning_rate": 9.557424549414722e-05, "loss": 0.0029, "step": 17820 }, { "epoch": 11.73798551678736, "grad_norm": 0.1665235459804535, "learning_rate": 9.556744172681546e-05, "loss": 0.0027, "step": 17830 }, { "epoch": 11.74456879526004, "grad_norm": 0.09344521909952164, "learning_rate": 9.556063297629677e-05, "loss": 0.0026, "step": 17840 }, { "epoch": 11.751152073732719, "grad_norm": 0.1622515469789505, "learning_rate": 9.555381924333578e-05, "loss": 0.003, "step": 17850 }, { "epoch": 11.757735352205398, "grad_norm": 0.10238167643547058, "learning_rate": 9.554700052867758e-05, "loss": 0.0017, "step": 17860 }, { "epoch": 11.764318630678078, "grad_norm": 0.09087339043617249, "learning_rate": 9.554017683306789e-05, "loss": 0.0021, "step": 17870 }, { "epoch": 11.770901909150757, "grad_norm": 0.10424017906188965, "learning_rate": 9.553334815725294e-05, "loss": 0.0026, "step": 17880 }, { "epoch": 11.777485187623437, "grad_norm": 0.1799791008234024, "learning_rate": 9.552651450197949e-05, "loss": 0.0039, "step": 17890 }, { "epoch": 11.784068466096116, "grad_norm": 0.14281269907951355, "learning_rate": 9.551967586799486e-05, "loss": 0.0031, "step": 17900 }, { "epoch": 11.790651744568795, "grad_norm": 0.10231756418943405, "learning_rate": 9.551283225604692e-05, "loss": 0.0032, "step": 17910 }, { "epoch": 11.797235023041475, "grad_norm": 0.1436399668455124, "learning_rate": 9.550598366688406e-05, "loss": 0.0033, "step": 17920 }, { "epoch": 11.803818301514154, "grad_norm": 0.12053346633911133, "learning_rate": 9.549913010125526e-05, "loss": 0.0017, "step": 17930 }, { "epoch": 11.810401579986834, "grad_norm": 0.12495830655097961, "learning_rate": 9.549227155990999e-05, "loss": 0.0019, "step": 17940 }, { "epoch": 11.816984858459513, "grad_norm": 0.08927366137504578, "learning_rate": 9.548540804359828e-05, "loss": 0.0022, "step": 17950 }, { "epoch": 11.823568136932192, "grad_norm": 0.10375919938087463, "learning_rate": 9.547853955307077e-05, "loss": 0.0082, "step": 17960 }, { "epoch": 11.830151415404872, "grad_norm": 0.18340592086315155, "learning_rate": 9.547166608907853e-05, "loss": 0.0032, "step": 17970 }, { "epoch": 11.83673469387755, "grad_norm": 0.1102958545088768, "learning_rate": 9.546478765237326e-05, "loss": 0.0023, "step": 17980 }, { "epoch": 11.843317972350231, "grad_norm": 0.2963999807834625, "learning_rate": 9.545790424370715e-05, "loss": 0.0046, "step": 17990 }, { "epoch": 11.84990125082291, "grad_norm": 0.20254844427108765, "learning_rate": 9.5451015863833e-05, "loss": 0.0047, "step": 18000 }, { "epoch": 11.856484529295589, "grad_norm": 0.19598489999771118, "learning_rate": 9.544412251350408e-05, "loss": 0.0033, "step": 18010 }, { "epoch": 11.86306780776827, "grad_norm": 0.10930831730365753, "learning_rate": 9.543722419347422e-05, "loss": 0.004, "step": 18020 }, { "epoch": 11.869651086240948, "grad_norm": 0.13726691901683807, "learning_rate": 9.543032090449788e-05, "loss": 0.0026, "step": 18030 }, { "epoch": 11.876234364713627, "grad_norm": 0.13084785640239716, "learning_rate": 9.542341264732992e-05, "loss": 0.0023, "step": 18040 }, { "epoch": 11.882817643186307, "grad_norm": 0.1491815745830536, "learning_rate": 9.541649942272585e-05, "loss": 0.0032, "step": 18050 }, { "epoch": 11.889400921658986, "grad_norm": 0.10758651793003082, "learning_rate": 9.54095812314417e-05, "loss": 0.0016, "step": 18060 }, { "epoch": 11.895984200131666, "grad_norm": 0.08503386378288269, "learning_rate": 9.540265807423401e-05, "loss": 0.0024, "step": 18070 }, { "epoch": 11.902567478604345, "grad_norm": 0.13282455503940582, "learning_rate": 9.53957299518599e-05, "loss": 0.002, "step": 18080 }, { "epoch": 11.909150757077024, "grad_norm": 0.12733857333660126, "learning_rate": 9.5388796865077e-05, "loss": 0.0018, "step": 18090 }, { "epoch": 11.915734035549704, "grad_norm": 0.10286642611026764, "learning_rate": 9.538185881464353e-05, "loss": 0.0028, "step": 18100 }, { "epoch": 11.922317314022383, "grad_norm": 0.1202264353632927, "learning_rate": 9.537491580131821e-05, "loss": 0.0048, "step": 18110 }, { "epoch": 11.928900592495063, "grad_norm": 0.12090671807527542, "learning_rate": 9.53679678258603e-05, "loss": 0.0018, "step": 18120 }, { "epoch": 11.935483870967742, "grad_norm": 0.09492264688014984, "learning_rate": 9.536101488902966e-05, "loss": 0.0025, "step": 18130 }, { "epoch": 11.94206714944042, "grad_norm": 0.14405220746994019, "learning_rate": 9.535405699158663e-05, "loss": 0.0028, "step": 18140 }, { "epoch": 11.948650427913101, "grad_norm": 0.10153106600046158, "learning_rate": 9.53470941342921e-05, "loss": 0.0067, "step": 18150 }, { "epoch": 11.95523370638578, "grad_norm": 0.1073550209403038, "learning_rate": 9.534012631790756e-05, "loss": 0.0026, "step": 18160 }, { "epoch": 11.96181698485846, "grad_norm": 0.11912664026021957, "learning_rate": 9.533315354319494e-05, "loss": 0.0025, "step": 18170 }, { "epoch": 11.968400263331139, "grad_norm": 0.08964882791042328, "learning_rate": 9.532617581091682e-05, "loss": 0.0032, "step": 18180 }, { "epoch": 11.974983541803818, "grad_norm": 0.10715022683143616, "learning_rate": 9.531919312183629e-05, "loss": 0.0032, "step": 18190 }, { "epoch": 11.981566820276498, "grad_norm": 0.10744226723909378, "learning_rate": 9.531220547671688e-05, "loss": 0.003, "step": 18200 }, { "epoch": 11.988150098749177, "grad_norm": 0.08053044974803925, "learning_rate": 9.530521287632285e-05, "loss": 0.0031, "step": 18210 }, { "epoch": 11.994733377221856, "grad_norm": 0.10906914621591568, "learning_rate": 9.529821532141884e-05, "loss": 0.0034, "step": 18220 }, { "epoch": 12.001316655694536, "grad_norm": 0.08367866277694702, "learning_rate": 9.52912128127701e-05, "loss": 0.0032, "step": 18230 }, { "epoch": 12.007899934167215, "grad_norm": 0.13774879276752472, "learning_rate": 9.528420535114244e-05, "loss": 0.0024, "step": 18240 }, { "epoch": 12.014483212639895, "grad_norm": 0.08979972451925278, "learning_rate": 9.527719293730215e-05, "loss": 0.0014, "step": 18250 }, { "epoch": 12.021066491112574, "grad_norm": 0.09756061434745789, "learning_rate": 9.527017557201611e-05, "loss": 0.0015, "step": 18260 }, { "epoch": 12.027649769585253, "grad_norm": 0.13029813766479492, "learning_rate": 9.526315325605176e-05, "loss": 0.0022, "step": 18270 }, { "epoch": 12.034233048057933, "grad_norm": 0.1502656489610672, "learning_rate": 9.525612599017699e-05, "loss": 0.0026, "step": 18280 }, { "epoch": 12.040816326530612, "grad_norm": 0.1759401112794876, "learning_rate": 9.524909377516033e-05, "loss": 0.0038, "step": 18290 }, { "epoch": 12.047399605003292, "grad_norm": 0.20121333003044128, "learning_rate": 9.524205661177081e-05, "loss": 0.0032, "step": 18300 }, { "epoch": 12.053982883475971, "grad_norm": 0.17049451172351837, "learning_rate": 9.523501450077801e-05, "loss": 0.0029, "step": 18310 }, { "epoch": 12.06056616194865, "grad_norm": 0.10607939958572388, "learning_rate": 9.522796744295202e-05, "loss": 0.0033, "step": 18320 }, { "epoch": 12.06714944042133, "grad_norm": 0.1439928412437439, "learning_rate": 9.522091543906352e-05, "loss": 0.003, "step": 18330 }, { "epoch": 12.073732718894009, "grad_norm": 0.0955275446176529, "learning_rate": 9.521385848988369e-05, "loss": 0.0047, "step": 18340 }, { "epoch": 12.08031599736669, "grad_norm": 0.0813475102186203, "learning_rate": 9.520679659618428e-05, "loss": 0.0035, "step": 18350 }, { "epoch": 12.086899275839368, "grad_norm": 0.21108512580394745, "learning_rate": 9.519972975873754e-05, "loss": 0.003, "step": 18360 }, { "epoch": 12.093482554312047, "grad_norm": 0.15041764080524445, "learning_rate": 9.519265797831633e-05, "loss": 0.0025, "step": 18370 }, { "epoch": 12.100065832784727, "grad_norm": 0.12994876503944397, "learning_rate": 9.518558125569399e-05, "loss": 0.0027, "step": 18380 }, { "epoch": 12.106649111257406, "grad_norm": 0.1764506846666336, "learning_rate": 9.517849959164442e-05, "loss": 0.0039, "step": 18390 }, { "epoch": 12.113232389730086, "grad_norm": 0.10398194938898087, "learning_rate": 9.517141298694205e-05, "loss": 0.0025, "step": 18400 }, { "epoch": 12.119815668202765, "grad_norm": 0.11376842856407166, "learning_rate": 9.516432144236188e-05, "loss": 0.0031, "step": 18410 }, { "epoch": 12.126398946675444, "grad_norm": 0.13111646473407745, "learning_rate": 9.515722495867941e-05, "loss": 0.0021, "step": 18420 }, { "epoch": 12.132982225148124, "grad_norm": 0.12602680921554565, "learning_rate": 9.515012353667072e-05, "loss": 0.0035, "step": 18430 }, { "epoch": 12.139565503620803, "grad_norm": 0.08189423382282257, "learning_rate": 9.51430171771124e-05, "loss": 0.0052, "step": 18440 }, { "epoch": 12.146148782093482, "grad_norm": 0.10938972979784012, "learning_rate": 9.513590588078159e-05, "loss": 0.0031, "step": 18450 }, { "epoch": 12.152732060566162, "grad_norm": 0.14898234605789185, "learning_rate": 9.512878964845597e-05, "loss": 0.0031, "step": 18460 }, { "epoch": 12.15931533903884, "grad_norm": 0.1454971879720688, "learning_rate": 9.512166848091377e-05, "loss": 0.003, "step": 18470 }, { "epoch": 12.165898617511521, "grad_norm": 0.13793344795703888, "learning_rate": 9.511454237893376e-05, "loss": 0.0023, "step": 18480 }, { "epoch": 12.1724818959842, "grad_norm": 0.1374698430299759, "learning_rate": 9.51074113432952e-05, "loss": 0.0055, "step": 18490 }, { "epoch": 12.179065174456879, "grad_norm": 0.1690942347049713, "learning_rate": 9.510027537477797e-05, "loss": 0.0028, "step": 18500 }, { "epoch": 12.18564845292956, "grad_norm": 0.16033835709095, "learning_rate": 9.509313447416242e-05, "loss": 0.0029, "step": 18510 }, { "epoch": 12.192231731402238, "grad_norm": 0.1512623280286789, "learning_rate": 9.508598864222949e-05, "loss": 0.0025, "step": 18520 }, { "epoch": 12.198815009874918, "grad_norm": 0.15129555761814117, "learning_rate": 9.507883787976062e-05, "loss": 0.004, "step": 18530 }, { "epoch": 12.205398288347597, "grad_norm": 0.13943232595920563, "learning_rate": 9.507168218753781e-05, "loss": 0.003, "step": 18540 }, { "epoch": 12.211981566820276, "grad_norm": 0.14735379815101624, "learning_rate": 9.506452156634362e-05, "loss": 0.0022, "step": 18550 }, { "epoch": 12.218564845292956, "grad_norm": 0.09361173957586288, "learning_rate": 9.505735601696109e-05, "loss": 0.0029, "step": 18560 }, { "epoch": 12.225148123765635, "grad_norm": 0.13204509019851685, "learning_rate": 9.505018554017385e-05, "loss": 0.0034, "step": 18570 }, { "epoch": 12.231731402238315, "grad_norm": 0.10607992112636566, "learning_rate": 9.504301013676604e-05, "loss": 0.0025, "step": 18580 }, { "epoch": 12.238314680710994, "grad_norm": 0.210678830742836, "learning_rate": 9.503582980752238e-05, "loss": 0.0066, "step": 18590 }, { "epoch": 12.244897959183673, "grad_norm": 0.1658446192741394, "learning_rate": 9.502864455322809e-05, "loss": 0.003, "step": 18600 }, { "epoch": 12.251481237656353, "grad_norm": 0.12180709093809128, "learning_rate": 9.502145437466891e-05, "loss": 0.0024, "step": 18610 }, { "epoch": 12.258064516129032, "grad_norm": 0.11873852461576462, "learning_rate": 9.501425927263116e-05, "loss": 0.0041, "step": 18620 }, { "epoch": 12.264647794601713, "grad_norm": 0.13221876323223114, "learning_rate": 9.500705924790172e-05, "loss": 0.0029, "step": 18630 }, { "epoch": 12.271231073074391, "grad_norm": 0.0796675831079483, "learning_rate": 9.499985430126794e-05, "loss": 0.0037, "step": 18640 }, { "epoch": 12.27781435154707, "grad_norm": 0.1285850554704666, "learning_rate": 9.499264443351775e-05, "loss": 0.0023, "step": 18650 }, { "epoch": 12.28439763001975, "grad_norm": 0.20803846418857574, "learning_rate": 9.498542964543961e-05, "loss": 0.0047, "step": 18660 }, { "epoch": 12.290980908492429, "grad_norm": 0.08679748326539993, "learning_rate": 9.497820993782252e-05, "loss": 0.003, "step": 18670 }, { "epoch": 12.297564186965108, "grad_norm": 0.17172764241695404, "learning_rate": 9.497098531145601e-05, "loss": 0.0023, "step": 18680 }, { "epoch": 12.304147465437788, "grad_norm": 0.1238497719168663, "learning_rate": 9.496375576713017e-05, "loss": 0.0028, "step": 18690 }, { "epoch": 12.310730743910467, "grad_norm": 0.0894543007016182, "learning_rate": 9.49565213056356e-05, "loss": 0.0018, "step": 18700 }, { "epoch": 12.317314022383147, "grad_norm": 0.11284058541059494, "learning_rate": 9.494928192776342e-05, "loss": 0.0029, "step": 18710 }, { "epoch": 12.323897300855826, "grad_norm": 0.09485975652933121, "learning_rate": 9.494203763430538e-05, "loss": 0.0061, "step": 18720 }, { "epoch": 12.330480579328505, "grad_norm": 0.10810939222574234, "learning_rate": 9.493478842605366e-05, "loss": 0.0041, "step": 18730 }, { "epoch": 12.337063857801185, "grad_norm": 0.1311211735010147, "learning_rate": 9.492753430380105e-05, "loss": 0.0056, "step": 18740 }, { "epoch": 12.343647136273864, "grad_norm": 0.09832682460546494, "learning_rate": 9.492027526834083e-05, "loss": 0.0035, "step": 18750 }, { "epoch": 12.350230414746544, "grad_norm": 0.1391357183456421, "learning_rate": 9.491301132046684e-05, "loss": 0.005, "step": 18760 }, { "epoch": 12.356813693219223, "grad_norm": 0.23004606366157532, "learning_rate": 9.490574246097345e-05, "loss": 0.0037, "step": 18770 }, { "epoch": 12.363396971691902, "grad_norm": 0.1978398710489273, "learning_rate": 9.48984686906556e-05, "loss": 0.0043, "step": 18780 }, { "epoch": 12.369980250164582, "grad_norm": 0.14150449633598328, "learning_rate": 9.489119001030871e-05, "loss": 0.0024, "step": 18790 }, { "epoch": 12.376563528637261, "grad_norm": 0.1406874656677246, "learning_rate": 9.488390642072878e-05, "loss": 0.0026, "step": 18800 }, { "epoch": 12.383146807109942, "grad_norm": 0.14570222795009613, "learning_rate": 9.48766179227123e-05, "loss": 0.0032, "step": 18810 }, { "epoch": 12.38973008558262, "grad_norm": 0.15959936380386353, "learning_rate": 9.486932451705636e-05, "loss": 0.0028, "step": 18820 }, { "epoch": 12.396313364055299, "grad_norm": 0.13625626266002655, "learning_rate": 9.486202620455857e-05, "loss": 0.004, "step": 18830 }, { "epoch": 12.40289664252798, "grad_norm": 0.14743521809577942, "learning_rate": 9.485472298601704e-05, "loss": 0.0028, "step": 18840 }, { "epoch": 12.409479921000658, "grad_norm": 0.10440659523010254, "learning_rate": 9.484741486223043e-05, "loss": 0.0029, "step": 18850 }, { "epoch": 12.416063199473339, "grad_norm": 0.13500529527664185, "learning_rate": 9.484010183399797e-05, "loss": 0.0028, "step": 18860 }, { "epoch": 12.422646477946017, "grad_norm": 0.1286812126636505, "learning_rate": 9.483278390211938e-05, "loss": 0.0014, "step": 18870 }, { "epoch": 12.429229756418696, "grad_norm": 0.22951281070709229, "learning_rate": 9.482546106739496e-05, "loss": 0.0062, "step": 18880 }, { "epoch": 12.435813034891376, "grad_norm": 0.2535831034183502, "learning_rate": 9.48181333306255e-05, "loss": 0.0039, "step": 18890 }, { "epoch": 12.442396313364055, "grad_norm": 0.13207672536373138, "learning_rate": 9.481080069261237e-05, "loss": 0.0023, "step": 18900 }, { "epoch": 12.448979591836734, "grad_norm": 0.16355089843273163, "learning_rate": 9.480346315415745e-05, "loss": 0.0044, "step": 18910 }, { "epoch": 12.455562870309414, "grad_norm": 0.18526603281497955, "learning_rate": 9.479612071606314e-05, "loss": 0.0031, "step": 18920 }, { "epoch": 12.462146148782093, "grad_norm": 0.13252174854278564, "learning_rate": 9.478877337913244e-05, "loss": 0.0027, "step": 18930 }, { "epoch": 12.468729427254774, "grad_norm": 0.14358651638031006, "learning_rate": 9.478142114416881e-05, "loss": 0.0037, "step": 18940 }, { "epoch": 12.475312705727452, "grad_norm": 0.1786671131849289, "learning_rate": 9.47740640119763e-05, "loss": 0.0025, "step": 18950 }, { "epoch": 12.481895984200131, "grad_norm": 0.11152368783950806, "learning_rate": 9.476670198335947e-05, "loss": 0.0031, "step": 18960 }, { "epoch": 12.488479262672811, "grad_norm": 0.09797266125679016, "learning_rate": 9.47593350591234e-05, "loss": 0.0024, "step": 18970 }, { "epoch": 12.49506254114549, "grad_norm": 0.10991983115673065, "learning_rate": 9.475196324007376e-05, "loss": 0.0041, "step": 18980 }, { "epoch": 12.50164581961817, "grad_norm": 0.12090910226106644, "learning_rate": 9.474458652701669e-05, "loss": 0.0025, "step": 18990 }, { "epoch": 12.50822909809085, "grad_norm": 0.24536234140396118, "learning_rate": 9.473720492075892e-05, "loss": 0.0077, "step": 19000 }, { "epoch": 12.514812376563528, "grad_norm": 0.1156945452094078, "learning_rate": 9.472981842210768e-05, "loss": 0.0042, "step": 19010 }, { "epoch": 12.521395655036208, "grad_norm": 0.20286080241203308, "learning_rate": 9.472242703187074e-05, "loss": 0.0029, "step": 19020 }, { "epoch": 12.527978933508887, "grad_norm": 0.11372341960668564, "learning_rate": 9.471503075085643e-05, "loss": 0.0054, "step": 19030 }, { "epoch": 12.534562211981568, "grad_norm": 0.20432215929031372, "learning_rate": 9.470762957987359e-05, "loss": 0.0041, "step": 19040 }, { "epoch": 12.541145490454246, "grad_norm": 0.12096273899078369, "learning_rate": 9.470022351973158e-05, "loss": 0.0025, "step": 19050 }, { "epoch": 12.547728768926925, "grad_norm": 0.317565381526947, "learning_rate": 9.469281257124034e-05, "loss": 0.003, "step": 19060 }, { "epoch": 12.554312047399605, "grad_norm": 0.18214555084705353, "learning_rate": 9.46853967352103e-05, "loss": 0.0058, "step": 19070 }, { "epoch": 12.560895325872284, "grad_norm": 0.13554145395755768, "learning_rate": 9.467797601245246e-05, "loss": 0.003, "step": 19080 }, { "epoch": 12.567478604344963, "grad_norm": 0.1286960393190384, "learning_rate": 9.467055040377834e-05, "loss": 0.0036, "step": 19090 }, { "epoch": 12.574061882817643, "grad_norm": 0.16922630369663239, "learning_rate": 9.466311990999999e-05, "loss": 0.005, "step": 19100 }, { "epoch": 12.580645161290322, "grad_norm": 0.18171653151512146, "learning_rate": 9.465568453193e-05, "loss": 0.0028, "step": 19110 }, { "epoch": 12.587228439763003, "grad_norm": 0.11307086050510406, "learning_rate": 9.464824427038148e-05, "loss": 0.0052, "step": 19120 }, { "epoch": 12.593811718235681, "grad_norm": 0.1320609301328659, "learning_rate": 9.46407991261681e-05, "loss": 0.0023, "step": 19130 }, { "epoch": 12.60039499670836, "grad_norm": 0.19940872490406036, "learning_rate": 9.463334910010404e-05, "loss": 0.0051, "step": 19140 }, { "epoch": 12.60697827518104, "grad_norm": 0.1785893440246582, "learning_rate": 9.462589419300403e-05, "loss": 0.0039, "step": 19150 }, { "epoch": 12.61356155365372, "grad_norm": 0.15741395950317383, "learning_rate": 9.461843440568333e-05, "loss": 0.0047, "step": 19160 }, { "epoch": 12.6201448321264, "grad_norm": 0.1520237922668457, "learning_rate": 9.461096973895773e-05, "loss": 0.0028, "step": 19170 }, { "epoch": 12.626728110599078, "grad_norm": 0.0975327119231224, "learning_rate": 9.460350019364355e-05, "loss": 0.0015, "step": 19180 }, { "epoch": 12.633311389071757, "grad_norm": 0.09717284888029099, "learning_rate": 9.459602577055764e-05, "loss": 0.0026, "step": 19190 }, { "epoch": 12.639894667544437, "grad_norm": 0.1653132289648056, "learning_rate": 9.45885464705174e-05, "loss": 0.0032, "step": 19200 }, { "epoch": 12.646477946017116, "grad_norm": 0.1953120231628418, "learning_rate": 9.458106229434076e-05, "loss": 0.0049, "step": 19210 }, { "epoch": 12.653061224489797, "grad_norm": 0.22593478858470917, "learning_rate": 9.457357324284617e-05, "loss": 0.0044, "step": 19220 }, { "epoch": 12.659644502962475, "grad_norm": 0.13242577016353607, "learning_rate": 9.456607931685262e-05, "loss": 0.0038, "step": 19230 }, { "epoch": 12.666227781435154, "grad_norm": 0.11239619553089142, "learning_rate": 9.455858051717965e-05, "loss": 0.0038, "step": 19240 }, { "epoch": 12.672811059907835, "grad_norm": 0.11607124656438828, "learning_rate": 9.45510768446473e-05, "loss": 0.0047, "step": 19250 }, { "epoch": 12.679394338380513, "grad_norm": 0.11191558837890625, "learning_rate": 9.454356830007618e-05, "loss": 0.0029, "step": 19260 }, { "epoch": 12.685977616853194, "grad_norm": 0.11724695563316345, "learning_rate": 9.45360548842874e-05, "loss": 0.0026, "step": 19270 }, { "epoch": 12.692560895325872, "grad_norm": 0.11181312799453735, "learning_rate": 9.452853659810261e-05, "loss": 0.0021, "step": 19280 }, { "epoch": 12.699144173798551, "grad_norm": 0.11795350909233093, "learning_rate": 9.452101344234401e-05, "loss": 0.0027, "step": 19290 }, { "epoch": 12.705727452271232, "grad_norm": 0.20854082703590393, "learning_rate": 9.451348541783431e-05, "loss": 0.0051, "step": 19300 }, { "epoch": 12.71231073074391, "grad_norm": 0.08039818704128265, "learning_rate": 9.450595252539678e-05, "loss": 0.002, "step": 19310 }, { "epoch": 12.71889400921659, "grad_norm": 0.11525449901819229, "learning_rate": 9.449841476585518e-05, "loss": 0.0033, "step": 19320 }, { "epoch": 12.72547728768927, "grad_norm": 0.09309639036655426, "learning_rate": 9.449087214003384e-05, "loss": 0.0024, "step": 19330 }, { "epoch": 12.732060566161948, "grad_norm": 0.09867465496063232, "learning_rate": 9.448332464875765e-05, "loss": 0.0034, "step": 19340 }, { "epoch": 12.738643844634629, "grad_norm": 0.16885614395141602, "learning_rate": 9.447577229285192e-05, "loss": 0.0028, "step": 19350 }, { "epoch": 12.745227123107307, "grad_norm": 0.10974617302417755, "learning_rate": 9.446821507314261e-05, "loss": 0.0038, "step": 19360 }, { "epoch": 12.751810401579986, "grad_norm": 0.13216155767440796, "learning_rate": 9.446065299045617e-05, "loss": 0.0022, "step": 19370 }, { "epoch": 12.758393680052666, "grad_norm": 0.07207121700048447, "learning_rate": 9.445308604561955e-05, "loss": 0.0017, "step": 19380 }, { "epoch": 12.764976958525345, "grad_norm": 0.07916229963302612, "learning_rate": 9.444551423946028e-05, "loss": 0.0024, "step": 19390 }, { "epoch": 12.771560236998026, "grad_norm": 0.11109692603349686, "learning_rate": 9.443793757280638e-05, "loss": 0.0018, "step": 19400 }, { "epoch": 12.778143515470704, "grad_norm": 0.10249415040016174, "learning_rate": 9.443035604648646e-05, "loss": 0.0024, "step": 19410 }, { "epoch": 12.784726793943383, "grad_norm": 0.10273155570030212, "learning_rate": 9.44227696613296e-05, "loss": 0.0031, "step": 19420 }, { "epoch": 12.791310072416064, "grad_norm": 0.093709796667099, "learning_rate": 9.441517841816542e-05, "loss": 0.0046, "step": 19430 }, { "epoch": 12.797893350888742, "grad_norm": 0.18133923411369324, "learning_rate": 9.440758231782413e-05, "loss": 0.0036, "step": 19440 }, { "epoch": 12.804476629361423, "grad_norm": 0.08957184106111526, "learning_rate": 9.439998136113639e-05, "loss": 0.003, "step": 19450 }, { "epoch": 12.811059907834101, "grad_norm": 0.1259177327156067, "learning_rate": 9.439237554893344e-05, "loss": 0.0051, "step": 19460 }, { "epoch": 12.81764318630678, "grad_norm": 0.0978337898850441, "learning_rate": 9.438476488204705e-05, "loss": 0.0033, "step": 19470 }, { "epoch": 12.82422646477946, "grad_norm": 0.10396566987037659, "learning_rate": 9.43771493613095e-05, "loss": 0.002, "step": 19480 }, { "epoch": 12.83080974325214, "grad_norm": 0.13242875039577484, "learning_rate": 9.436952898755362e-05, "loss": 0.0026, "step": 19490 }, { "epoch": 12.83739302172482, "grad_norm": 0.09364724159240723, "learning_rate": 9.436190376161276e-05, "loss": 0.0025, "step": 19500 }, { "epoch": 12.843976300197498, "grad_norm": 0.10661923885345459, "learning_rate": 9.43542736843208e-05, "loss": 0.0016, "step": 19510 }, { "epoch": 12.850559578670177, "grad_norm": 0.12070319801568985, "learning_rate": 9.434663875651216e-05, "loss": 0.0023, "step": 19520 }, { "epoch": 12.857142857142858, "grad_norm": 0.11466336250305176, "learning_rate": 9.433899897902177e-05, "loss": 0.003, "step": 19530 }, { "epoch": 12.863726135615536, "grad_norm": 0.09226647019386292, "learning_rate": 9.433135435268511e-05, "loss": 0.0018, "step": 19540 }, { "epoch": 12.870309414088215, "grad_norm": 0.10435882210731506, "learning_rate": 9.432370487833819e-05, "loss": 0.0023, "step": 19550 }, { "epoch": 12.876892692560896, "grad_norm": 0.10851921886205673, "learning_rate": 9.431605055681756e-05, "loss": 0.0018, "step": 19560 }, { "epoch": 12.883475971033574, "grad_norm": 0.12032008916139603, "learning_rate": 9.430839138896026e-05, "loss": 0.002, "step": 19570 }, { "epoch": 12.890059249506255, "grad_norm": 0.12643669545650482, "learning_rate": 9.43007273756039e-05, "loss": 0.0053, "step": 19580 }, { "epoch": 12.896642527978933, "grad_norm": 0.11737558990716934, "learning_rate": 9.429305851758658e-05, "loss": 0.0022, "step": 19590 }, { "epoch": 12.903225806451612, "grad_norm": 0.1050499826669693, "learning_rate": 9.428538481574699e-05, "loss": 0.0021, "step": 19600 }, { "epoch": 12.909809084924293, "grad_norm": 0.08668049424886703, "learning_rate": 9.42777062709243e-05, "loss": 0.0029, "step": 19610 }, { "epoch": 12.916392363396971, "grad_norm": 0.1717112958431244, "learning_rate": 9.427002288395821e-05, "loss": 0.0036, "step": 19620 }, { "epoch": 12.922975641869652, "grad_norm": 0.08804901689291, "learning_rate": 9.426233465568898e-05, "loss": 0.0019, "step": 19630 }, { "epoch": 12.92955892034233, "grad_norm": 0.13358217477798462, "learning_rate": 9.42546415869574e-05, "loss": 0.0024, "step": 19640 }, { "epoch": 12.93614219881501, "grad_norm": 0.11441519111394882, "learning_rate": 9.424694367860473e-05, "loss": 0.0069, "step": 19650 }, { "epoch": 12.94272547728769, "grad_norm": 0.055728647857904434, "learning_rate": 9.423924093147284e-05, "loss": 0.0021, "step": 19660 }, { "epoch": 12.949308755760368, "grad_norm": 0.13242191076278687, "learning_rate": 9.423153334640407e-05, "loss": 0.0036, "step": 19670 }, { "epoch": 12.955892034233049, "grad_norm": 0.12025835365056992, "learning_rate": 9.42238209242413e-05, "loss": 0.004, "step": 19680 }, { "epoch": 12.962475312705728, "grad_norm": 0.13676030933856964, "learning_rate": 9.421610366582798e-05, "loss": 0.0029, "step": 19690 }, { "epoch": 12.969058591178406, "grad_norm": 0.13448667526245117, "learning_rate": 9.420838157200803e-05, "loss": 0.0055, "step": 19700 }, { "epoch": 12.975641869651087, "grad_norm": 0.1450134515762329, "learning_rate": 9.420065464362594e-05, "loss": 0.0069, "step": 19710 }, { "epoch": 12.982225148123765, "grad_norm": 0.0979725643992424, "learning_rate": 9.419292288152673e-05, "loss": 0.003, "step": 19720 }, { "epoch": 12.988808426596446, "grad_norm": 0.1568930447101593, "learning_rate": 9.418518628655588e-05, "loss": 0.0029, "step": 19730 }, { "epoch": 12.995391705069125, "grad_norm": 0.1774197369813919, "learning_rate": 9.417744485955951e-05, "loss": 0.0037, "step": 19740 }, { "epoch": 13.001974983541803, "grad_norm": 0.11954934895038605, "learning_rate": 9.41696986013842e-05, "loss": 0.0062, "step": 19750 }, { "epoch": 13.008558262014484, "grad_norm": 0.12294644862413406, "learning_rate": 9.416194751287705e-05, "loss": 0.0027, "step": 19760 }, { "epoch": 13.015141540487162, "grad_norm": 0.10507215559482574, "learning_rate": 9.415419159488572e-05, "loss": 0.0032, "step": 19770 }, { "epoch": 13.021724818959841, "grad_norm": 0.1738380640745163, "learning_rate": 9.414643084825837e-05, "loss": 0.0066, "step": 19780 }, { "epoch": 13.028308097432522, "grad_norm": 0.12812957167625427, "learning_rate": 9.413866527384372e-05, "loss": 0.0034, "step": 19790 }, { "epoch": 13.0348913759052, "grad_norm": 0.15281549096107483, "learning_rate": 9.4130894872491e-05, "loss": 0.0054, "step": 19800 }, { "epoch": 13.04147465437788, "grad_norm": 0.1281885951757431, "learning_rate": 9.412311964504998e-05, "loss": 0.0023, "step": 19810 }, { "epoch": 13.04805793285056, "grad_norm": 0.17303074896335602, "learning_rate": 9.411533959237091e-05, "loss": 0.0068, "step": 19820 }, { "epoch": 13.054641211323238, "grad_norm": 0.09921866655349731, "learning_rate": 9.410755471530464e-05, "loss": 0.0066, "step": 19830 }, { "epoch": 13.061224489795919, "grad_norm": 0.13461506366729736, "learning_rate": 9.40997650147025e-05, "loss": 0.0025, "step": 19840 }, { "epoch": 13.067807768268597, "grad_norm": 0.09415294229984283, "learning_rate": 9.409197049141637e-05, "loss": 0.0014, "step": 19850 }, { "epoch": 13.074391046741278, "grad_norm": 0.07992962747812271, "learning_rate": 9.408417114629863e-05, "loss": 0.0019, "step": 19860 }, { "epoch": 13.080974325213957, "grad_norm": 0.07746117562055588, "learning_rate": 9.40763669802022e-05, "loss": 0.005, "step": 19870 }, { "epoch": 13.087557603686635, "grad_norm": 0.08449748158454895, "learning_rate": 9.406855799398056e-05, "loss": 0.0033, "step": 19880 }, { "epoch": 13.094140882159316, "grad_norm": 0.08395242691040039, "learning_rate": 9.406074418848767e-05, "loss": 0.0037, "step": 19890 }, { "epoch": 13.100724160631994, "grad_norm": 0.09927991777658463, "learning_rate": 9.405292556457805e-05, "loss": 0.0039, "step": 19900 }, { "epoch": 13.107307439104675, "grad_norm": 0.1767202913761139, "learning_rate": 9.404510212310671e-05, "loss": 0.0044, "step": 19910 }, { "epoch": 13.113890717577354, "grad_norm": 0.1109730526804924, "learning_rate": 9.403727386492924e-05, "loss": 0.0022, "step": 19920 }, { "epoch": 13.120473996050032, "grad_norm": 0.16811798512935638, "learning_rate": 9.40294407909017e-05, "loss": 0.0024, "step": 19930 }, { "epoch": 13.127057274522713, "grad_norm": 0.20603728294372559, "learning_rate": 9.40216029018807e-05, "loss": 0.0034, "step": 19940 }, { "epoch": 13.133640552995391, "grad_norm": 0.11018000543117523, "learning_rate": 9.401376019872338e-05, "loss": 0.0025, "step": 19950 }, { "epoch": 13.140223831468072, "grad_norm": 0.14023706316947937, "learning_rate": 9.400591268228746e-05, "loss": 0.0048, "step": 19960 }, { "epoch": 13.14680710994075, "grad_norm": 0.14614100754261017, "learning_rate": 9.399806035343106e-05, "loss": 0.0046, "step": 19970 }, { "epoch": 13.15339038841343, "grad_norm": 0.13489915430545807, "learning_rate": 9.399020321301294e-05, "loss": 0.003, "step": 19980 }, { "epoch": 13.15997366688611, "grad_norm": 0.08032411336898804, "learning_rate": 9.398234126189234e-05, "loss": 0.0049, "step": 19990 }, { "epoch": 13.166556945358789, "grad_norm": 0.16438418626785278, "learning_rate": 9.397447450092902e-05, "loss": 0.0032, "step": 20000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 66, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }