{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5825129609133803, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0005825129609133803, "grad_norm": 16.363956451416016, "learning_rate": 6.000000000000001e-07, "loss": 0.8717, "step": 10 }, { "epoch": 0.0011650259218267606, "grad_norm": 13.027254104614258, "learning_rate": 1.2666666666666667e-06, "loss": 0.9411, "step": 20 }, { "epoch": 0.001747538882740141, "grad_norm": 8.33905029296875, "learning_rate": 1.9333333333333336e-06, "loss": 0.6971, "step": 30 }, { "epoch": 0.0023300518436535213, "grad_norm": 3.80228590965271, "learning_rate": 2.6e-06, "loss": 0.4438, "step": 40 }, { "epoch": 0.0029125648045669017, "grad_norm": 3.3239755630493164, "learning_rate": 3.2666666666666666e-06, "loss": 0.2813, "step": 50 }, { "epoch": 0.003495077765480282, "grad_norm": 2.004411458969116, "learning_rate": 3.9333333333333335e-06, "loss": 0.2361, "step": 60 }, { "epoch": 0.004077590726393662, "grad_norm": 2.5876970291137695, "learning_rate": 4.6e-06, "loss": 0.1895, "step": 70 }, { "epoch": 0.0046601036873070425, "grad_norm": 2.054030656814575, "learning_rate": 5.266666666666667e-06, "loss": 0.1678, "step": 80 }, { "epoch": 0.005242616648220423, "grad_norm": 2.5571675300598145, "learning_rate": 5.933333333333334e-06, "loss": 0.1363, "step": 90 }, { "epoch": 0.005825129609133803, "grad_norm": 2.476600408554077, "learning_rate": 6.6e-06, "loss": 0.1123, "step": 100 }, { "epoch": 0.006407642570047184, "grad_norm": 1.5339707136154175, "learning_rate": 7.266666666666668e-06, "loss": 0.1119, "step": 110 }, { "epoch": 0.006990155530960564, "grad_norm": 1.620774269104004, "learning_rate": 7.933333333333334e-06, "loss": 0.091, "step": 120 }, { "epoch": 0.007572668491873944, "grad_norm": 1.8176084756851196, "learning_rate": 8.599999999999999e-06, "loss": 0.098, "step": 130 }, { "epoch": 0.008155181452787324, "grad_norm": 1.9751254320144653, "learning_rate": 9.266666666666667e-06, "loss": 0.0937, "step": 140 }, { "epoch": 0.008737694413700705, "grad_norm": 2.1015803813934326, "learning_rate": 9.933333333333334e-06, "loss": 0.0909, "step": 150 }, { "epoch": 0.009320207374614085, "grad_norm": 1.4475528001785278, "learning_rate": 1.06e-05, "loss": 0.0868, "step": 160 }, { "epoch": 0.009902720335527465, "grad_norm": 1.1314893960952759, "learning_rate": 1.1266666666666667e-05, "loss": 0.0783, "step": 170 }, { "epoch": 0.010485233296440846, "grad_norm": 1.3180276155471802, "learning_rate": 1.1933333333333333e-05, "loss": 0.0701, "step": 180 }, { "epoch": 0.011067746257354226, "grad_norm": 1.5609411001205444, "learning_rate": 1.2600000000000001e-05, "loss": 0.0751, "step": 190 }, { "epoch": 0.011650259218267607, "grad_norm": 1.6207369565963745, "learning_rate": 1.3266666666666666e-05, "loss": 0.072, "step": 200 }, { "epoch": 0.012232772179180987, "grad_norm": 1.1604082584381104, "learning_rate": 1.3933333333333334e-05, "loss": 0.0822, "step": 210 }, { "epoch": 0.012815285140094368, "grad_norm": 1.1570616960525513, "learning_rate": 1.4599999999999999e-05, "loss": 0.077, "step": 220 }, { "epoch": 0.013397798101007748, "grad_norm": 1.152485728263855, "learning_rate": 1.5266666666666667e-05, "loss": 0.0611, "step": 230 }, { "epoch": 0.013980311061921128, "grad_norm": 2.1177918910980225, "learning_rate": 1.5933333333333332e-05, "loss": 0.0761, "step": 240 }, { "epoch": 0.014562824022834509, "grad_norm": 1.538111686706543, "learning_rate": 1.66e-05, "loss": 0.0659, "step": 250 }, { "epoch": 0.015145336983747888, "grad_norm": 1.2610279321670532, "learning_rate": 1.726666666666667e-05, "loss": 0.059, "step": 260 }, { "epoch": 0.015727849944661268, "grad_norm": 1.6613271236419678, "learning_rate": 1.7933333333333337e-05, "loss": 0.064, "step": 270 }, { "epoch": 0.01631036290557465, "grad_norm": 1.3400022983551025, "learning_rate": 1.86e-05, "loss": 0.0624, "step": 280 }, { "epoch": 0.01689287586648803, "grad_norm": 0.9490658640861511, "learning_rate": 1.926666666666667e-05, "loss": 0.0664, "step": 290 }, { "epoch": 0.01747538882740141, "grad_norm": 1.3591103553771973, "learning_rate": 1.9933333333333334e-05, "loss": 0.0605, "step": 300 }, { "epoch": 0.01805790178831479, "grad_norm": 1.059062123298645, "learning_rate": 2.06e-05, "loss": 0.0577, "step": 310 }, { "epoch": 0.01864041474922817, "grad_norm": 1.0266015529632568, "learning_rate": 2.1266666666666667e-05, "loss": 0.054, "step": 320 }, { "epoch": 0.01922292771014155, "grad_norm": 1.577591061592102, "learning_rate": 2.1933333333333332e-05, "loss": 0.0588, "step": 330 }, { "epoch": 0.01980544067105493, "grad_norm": 1.5641520023345947, "learning_rate": 2.26e-05, "loss": 0.0664, "step": 340 }, { "epoch": 0.02038795363196831, "grad_norm": 1.0901275873184204, "learning_rate": 2.326666666666667e-05, "loss": 0.048, "step": 350 }, { "epoch": 0.020970466592881692, "grad_norm": 1.6798902750015259, "learning_rate": 2.3933333333333337e-05, "loss": 0.0597, "step": 360 }, { "epoch": 0.021552979553795072, "grad_norm": 0.9511136412620544, "learning_rate": 2.46e-05, "loss": 0.0704, "step": 370 }, { "epoch": 0.022135492514708453, "grad_norm": 0.9565727710723877, "learning_rate": 2.5266666666666666e-05, "loss": 0.0477, "step": 380 }, { "epoch": 0.022718005475621833, "grad_norm": 1.4605202674865723, "learning_rate": 2.5933333333333338e-05, "loss": 0.0591, "step": 390 }, { "epoch": 0.023300518436535213, "grad_norm": 1.688125491142273, "learning_rate": 2.6600000000000003e-05, "loss": 0.06, "step": 400 }, { "epoch": 0.023883031397448594, "grad_norm": 1.2992297410964966, "learning_rate": 2.7266666666666668e-05, "loss": 0.0556, "step": 410 }, { "epoch": 0.024465544358361974, "grad_norm": 1.2986173629760742, "learning_rate": 2.7933333333333332e-05, "loss": 0.0538, "step": 420 }, { "epoch": 0.025048057319275355, "grad_norm": 1.1889179944992065, "learning_rate": 2.86e-05, "loss": 0.0621, "step": 430 }, { "epoch": 0.025630570280188735, "grad_norm": 1.1551380157470703, "learning_rate": 2.926666666666667e-05, "loss": 0.0632, "step": 440 }, { "epoch": 0.026213083241102116, "grad_norm": 1.21445894241333, "learning_rate": 2.9933333333333337e-05, "loss": 0.0549, "step": 450 }, { "epoch": 0.026795596202015496, "grad_norm": 0.9449787735939026, "learning_rate": 3.06e-05, "loss": 0.0657, "step": 460 }, { "epoch": 0.027378109162928876, "grad_norm": 0.9851808547973633, "learning_rate": 3.126666666666666e-05, "loss": 0.0492, "step": 470 }, { "epoch": 0.027960622123842257, "grad_norm": 0.7608017921447754, "learning_rate": 3.1933333333333335e-05, "loss": 0.0496, "step": 480 }, { "epoch": 0.028543135084755637, "grad_norm": 1.0604749917984009, "learning_rate": 3.26e-05, "loss": 0.0469, "step": 490 }, { "epoch": 0.029125648045669018, "grad_norm": 0.817071259021759, "learning_rate": 3.326666666666667e-05, "loss": 0.0403, "step": 500 }, { "epoch": 0.029708161006582398, "grad_norm": 1.0470209121704102, "learning_rate": 3.3933333333333336e-05, "loss": 0.0453, "step": 510 }, { "epoch": 0.030290673967495775, "grad_norm": 1.0401445627212524, "learning_rate": 3.46e-05, "loss": 0.0423, "step": 520 }, { "epoch": 0.030873186928409155, "grad_norm": 1.1210923194885254, "learning_rate": 3.526666666666667e-05, "loss": 0.0462, "step": 530 }, { "epoch": 0.031455699889322536, "grad_norm": 1.0154286623001099, "learning_rate": 3.593333333333334e-05, "loss": 0.0426, "step": 540 }, { "epoch": 0.03203821285023592, "grad_norm": 0.725735604763031, "learning_rate": 3.66e-05, "loss": 0.0378, "step": 550 }, { "epoch": 0.0326207258111493, "grad_norm": 0.6761109828948975, "learning_rate": 3.726666666666667e-05, "loss": 0.0392, "step": 560 }, { "epoch": 0.03320323877206268, "grad_norm": 0.8301836848258972, "learning_rate": 3.793333333333334e-05, "loss": 0.0332, "step": 570 }, { "epoch": 0.03378575173297606, "grad_norm": 0.8133602142333984, "learning_rate": 3.86e-05, "loss": 0.0355, "step": 580 }, { "epoch": 0.03436826469388944, "grad_norm": 1.0713573694229126, "learning_rate": 3.926666666666667e-05, "loss": 0.0438, "step": 590 }, { "epoch": 0.03495077765480282, "grad_norm": 1.272591233253479, "learning_rate": 3.993333333333333e-05, "loss": 0.0375, "step": 600 }, { "epoch": 0.0355332906157162, "grad_norm": 0.6690412759780884, "learning_rate": 4.0600000000000004e-05, "loss": 0.0429, "step": 610 }, { "epoch": 0.03611580357662958, "grad_norm": 0.8729681968688965, "learning_rate": 4.126666666666667e-05, "loss": 0.041, "step": 620 }, { "epoch": 0.03669831653754296, "grad_norm": 0.7829736471176147, "learning_rate": 4.1933333333333334e-05, "loss": 0.0353, "step": 630 }, { "epoch": 0.03728082949845634, "grad_norm": 0.564466118812561, "learning_rate": 4.26e-05, "loss": 0.0335, "step": 640 }, { "epoch": 0.037863342459369724, "grad_norm": 0.9312437772750854, "learning_rate": 4.3266666666666664e-05, "loss": 0.0388, "step": 650 }, { "epoch": 0.0384458554202831, "grad_norm": 0.9627359509468079, "learning_rate": 4.3933333333333335e-05, "loss": 0.038, "step": 660 }, { "epoch": 0.039028368381196485, "grad_norm": 1.1554635763168335, "learning_rate": 4.46e-05, "loss": 0.0416, "step": 670 }, { "epoch": 0.03961088134210986, "grad_norm": 1.2683342695236206, "learning_rate": 4.526666666666667e-05, "loss": 0.0425, "step": 680 }, { "epoch": 0.040193394303023246, "grad_norm": 0.7468678951263428, "learning_rate": 4.5933333333333336e-05, "loss": 0.0359, "step": 690 }, { "epoch": 0.04077590726393662, "grad_norm": 0.7981159687042236, "learning_rate": 4.660000000000001e-05, "loss": 0.0387, "step": 700 }, { "epoch": 0.04135842022485, "grad_norm": 0.746918797492981, "learning_rate": 4.726666666666667e-05, "loss": 0.0469, "step": 710 }, { "epoch": 0.041940933185763384, "grad_norm": 0.6890844702720642, "learning_rate": 4.793333333333334e-05, "loss": 0.038, "step": 720 }, { "epoch": 0.04252344614667676, "grad_norm": 1.093159556388855, "learning_rate": 4.86e-05, "loss": 0.0376, "step": 730 }, { "epoch": 0.043105959107590144, "grad_norm": 0.7858609557151794, "learning_rate": 4.926666666666667e-05, "loss": 0.0332, "step": 740 }, { "epoch": 0.04368847206850352, "grad_norm": 1.066215991973877, "learning_rate": 4.993333333333334e-05, "loss": 0.0353, "step": 750 }, { "epoch": 0.044270985029416905, "grad_norm": 0.7255071401596069, "learning_rate": 5.0600000000000003e-05, "loss": 0.0291, "step": 760 }, { "epoch": 0.04485349799033028, "grad_norm": 0.451518177986145, "learning_rate": 5.1266666666666675e-05, "loss": 0.0312, "step": 770 }, { "epoch": 0.045436010951243666, "grad_norm": 0.6643555164337158, "learning_rate": 5.193333333333333e-05, "loss": 0.0314, "step": 780 }, { "epoch": 0.04601852391215704, "grad_norm": 0.5722026824951172, "learning_rate": 5.2600000000000005e-05, "loss": 0.0291, "step": 790 }, { "epoch": 0.04660103687307043, "grad_norm": 0.7939475178718567, "learning_rate": 5.326666666666666e-05, "loss": 0.0368, "step": 800 }, { "epoch": 0.047183549833983804, "grad_norm": 0.8282892107963562, "learning_rate": 5.3933333333333334e-05, "loss": 0.0349, "step": 810 }, { "epoch": 0.04776606279489719, "grad_norm": 1.4124369621276855, "learning_rate": 5.4600000000000006e-05, "loss": 0.0385, "step": 820 }, { "epoch": 0.048348575755810565, "grad_norm": 0.6213952302932739, "learning_rate": 5.5266666666666664e-05, "loss": 0.0395, "step": 830 }, { "epoch": 0.04893108871672395, "grad_norm": 0.5522814393043518, "learning_rate": 5.5933333333333335e-05, "loss": 0.0348, "step": 840 }, { "epoch": 0.049513601677637326, "grad_norm": 0.5925297141075134, "learning_rate": 5.66e-05, "loss": 0.0351, "step": 850 }, { "epoch": 0.05009611463855071, "grad_norm": 0.5774127244949341, "learning_rate": 5.726666666666667e-05, "loss": 0.0306, "step": 860 }, { "epoch": 0.050678627599464086, "grad_norm": 0.7426440119743347, "learning_rate": 5.7933333333333337e-05, "loss": 0.0296, "step": 870 }, { "epoch": 0.05126114056037747, "grad_norm": 1.0238310098648071, "learning_rate": 5.86e-05, "loss": 0.0309, "step": 880 }, { "epoch": 0.05184365352129085, "grad_norm": 0.7471153140068054, "learning_rate": 5.926666666666667e-05, "loss": 0.0294, "step": 890 }, { "epoch": 0.05242616648220423, "grad_norm": 0.8437319993972778, "learning_rate": 5.9933333333333345e-05, "loss": 0.035, "step": 900 }, { "epoch": 0.05300867944311761, "grad_norm": 0.76613849401474, "learning_rate": 6.06e-05, "loss": 0.0288, "step": 910 }, { "epoch": 0.05359119240403099, "grad_norm": 0.601531445980072, "learning_rate": 6.126666666666667e-05, "loss": 0.0377, "step": 920 }, { "epoch": 0.05417370536494437, "grad_norm": 0.5623847246170044, "learning_rate": 6.193333333333333e-05, "loss": 0.0292, "step": 930 }, { "epoch": 0.05475621832585775, "grad_norm": 0.5545303225517273, "learning_rate": 6.26e-05, "loss": 0.0271, "step": 940 }, { "epoch": 0.05533873128677113, "grad_norm": 0.876323938369751, "learning_rate": 6.326666666666667e-05, "loss": 0.0356, "step": 950 }, { "epoch": 0.055921244247684514, "grad_norm": 0.9015308022499084, "learning_rate": 6.393333333333333e-05, "loss": 0.0321, "step": 960 }, { "epoch": 0.05650375720859789, "grad_norm": 0.3498276472091675, "learning_rate": 6.460000000000001e-05, "loss": 0.0327, "step": 970 }, { "epoch": 0.057086270169511275, "grad_norm": 0.645955502986908, "learning_rate": 6.526666666666666e-05, "loss": 0.0327, "step": 980 }, { "epoch": 0.05766878313042465, "grad_norm": 0.7454339265823364, "learning_rate": 6.593333333333334e-05, "loss": 0.0327, "step": 990 }, { "epoch": 0.058251296091338035, "grad_norm": 0.9000582695007324, "learning_rate": 6.66e-05, "loss": 0.0286, "step": 1000 }, { "epoch": 0.05883380905225141, "grad_norm": 0.5894749760627747, "learning_rate": 6.726666666666667e-05, "loss": 0.0266, "step": 1010 }, { "epoch": 0.059416322013164796, "grad_norm": 0.47711148858070374, "learning_rate": 6.793333333333334e-05, "loss": 0.0266, "step": 1020 }, { "epoch": 0.05999883497407817, "grad_norm": 0.39817479252815247, "learning_rate": 6.860000000000001e-05, "loss": 0.0296, "step": 1030 }, { "epoch": 0.06058134793499155, "grad_norm": 0.5175421237945557, "learning_rate": 6.926666666666667e-05, "loss": 0.0254, "step": 1040 }, { "epoch": 0.061163860895904934, "grad_norm": 0.5396063923835754, "learning_rate": 6.993333333333334e-05, "loss": 0.0263, "step": 1050 }, { "epoch": 0.06174637385681831, "grad_norm": 0.7684789896011353, "learning_rate": 7.06e-05, "loss": 0.0308, "step": 1060 }, { "epoch": 0.062328886817731695, "grad_norm": 0.4386548697948456, "learning_rate": 7.126666666666667e-05, "loss": 0.0296, "step": 1070 }, { "epoch": 0.06291139977864507, "grad_norm": 0.602109968662262, "learning_rate": 7.193333333333334e-05, "loss": 0.0307, "step": 1080 }, { "epoch": 0.06349391273955846, "grad_norm": 0.48234033584594727, "learning_rate": 7.26e-05, "loss": 0.0184, "step": 1090 }, { "epoch": 0.06407642570047184, "grad_norm": 0.5120774507522583, "learning_rate": 7.326666666666667e-05, "loss": 0.0243, "step": 1100 }, { "epoch": 0.06465893866138521, "grad_norm": 0.6919184327125549, "learning_rate": 7.393333333333333e-05, "loss": 0.0262, "step": 1110 }, { "epoch": 0.0652414516222986, "grad_norm": 0.8763561844825745, "learning_rate": 7.46e-05, "loss": 0.0283, "step": 1120 }, { "epoch": 0.06582396458321198, "grad_norm": 0.6841975450515747, "learning_rate": 7.526666666666668e-05, "loss": 0.0316, "step": 1130 }, { "epoch": 0.06640647754412536, "grad_norm": 0.4978059232234955, "learning_rate": 7.593333333333334e-05, "loss": 0.0267, "step": 1140 }, { "epoch": 0.06698899050503873, "grad_norm": 0.7343912124633789, "learning_rate": 7.66e-05, "loss": 0.0304, "step": 1150 }, { "epoch": 0.06757150346595212, "grad_norm": 0.9718796014785767, "learning_rate": 7.726666666666667e-05, "loss": 0.0298, "step": 1160 }, { "epoch": 0.0681540164268655, "grad_norm": 0.7595483660697937, "learning_rate": 7.793333333333333e-05, "loss": 0.0252, "step": 1170 }, { "epoch": 0.06873652938777888, "grad_norm": 0.6774091720581055, "learning_rate": 7.860000000000001e-05, "loss": 0.026, "step": 1180 }, { "epoch": 0.06931904234869225, "grad_norm": 0.5943043231964111, "learning_rate": 7.926666666666666e-05, "loss": 0.0254, "step": 1190 }, { "epoch": 0.06990155530960564, "grad_norm": 0.32503804564476013, "learning_rate": 7.993333333333334e-05, "loss": 0.0269, "step": 1200 }, { "epoch": 0.07048406827051902, "grad_norm": 0.3729705214500427, "learning_rate": 8.060000000000001e-05, "loss": 0.0292, "step": 1210 }, { "epoch": 0.0710665812314324, "grad_norm": 0.401933878660202, "learning_rate": 8.126666666666667e-05, "loss": 0.0243, "step": 1220 }, { "epoch": 0.07164909419234577, "grad_norm": 0.6299405694007874, "learning_rate": 8.193333333333334e-05, "loss": 0.0255, "step": 1230 }, { "epoch": 0.07223160715325916, "grad_norm": 0.38478463888168335, "learning_rate": 8.26e-05, "loss": 0.0219, "step": 1240 }, { "epoch": 0.07281412011417254, "grad_norm": 0.4080633819103241, "learning_rate": 8.326666666666667e-05, "loss": 0.0272, "step": 1250 }, { "epoch": 0.07339663307508593, "grad_norm": 0.3903158903121948, "learning_rate": 8.393333333333335e-05, "loss": 0.0224, "step": 1260 }, { "epoch": 0.0739791460359993, "grad_norm": 0.6705195307731628, "learning_rate": 8.46e-05, "loss": 0.0261, "step": 1270 }, { "epoch": 0.07456165899691268, "grad_norm": 0.5832329392433167, "learning_rate": 8.526666666666667e-05, "loss": 0.0266, "step": 1280 }, { "epoch": 0.07514417195782606, "grad_norm": 0.42376530170440674, "learning_rate": 8.593333333333333e-05, "loss": 0.0286, "step": 1290 }, { "epoch": 0.07572668491873945, "grad_norm": 0.3618919849395752, "learning_rate": 8.66e-05, "loss": 0.0245, "step": 1300 }, { "epoch": 0.07630919787965282, "grad_norm": 0.5320606827735901, "learning_rate": 8.726666666666667e-05, "loss": 0.0248, "step": 1310 }, { "epoch": 0.0768917108405662, "grad_norm": 0.6612217426300049, "learning_rate": 8.793333333333333e-05, "loss": 0.0274, "step": 1320 }, { "epoch": 0.07747422380147959, "grad_norm": 0.6843940019607544, "learning_rate": 8.86e-05, "loss": 0.0249, "step": 1330 }, { "epoch": 0.07805673676239297, "grad_norm": 0.37553220987319946, "learning_rate": 8.926666666666668e-05, "loss": 0.0252, "step": 1340 }, { "epoch": 0.07863924972330634, "grad_norm": 0.5357182025909424, "learning_rate": 8.993333333333334e-05, "loss": 0.0251, "step": 1350 }, { "epoch": 0.07922176268421972, "grad_norm": 0.8621771335601807, "learning_rate": 9.06e-05, "loss": 0.0265, "step": 1360 }, { "epoch": 0.07980427564513311, "grad_norm": 0.5881845355033875, "learning_rate": 9.126666666666667e-05, "loss": 0.0222, "step": 1370 }, { "epoch": 0.08038678860604649, "grad_norm": 0.8289936184883118, "learning_rate": 9.193333333333334e-05, "loss": 0.0261, "step": 1380 }, { "epoch": 0.08096930156695986, "grad_norm": 0.4810025095939636, "learning_rate": 9.260000000000001e-05, "loss": 0.0233, "step": 1390 }, { "epoch": 0.08155181452787325, "grad_norm": 0.9317886233329773, "learning_rate": 9.326666666666667e-05, "loss": 0.0289, "step": 1400 }, { "epoch": 0.08213432748878663, "grad_norm": 0.7096083760261536, "learning_rate": 9.393333333333334e-05, "loss": 0.0227, "step": 1410 }, { "epoch": 0.0827168404497, "grad_norm": 0.5615822672843933, "learning_rate": 9.46e-05, "loss": 0.0293, "step": 1420 }, { "epoch": 0.08329935341061338, "grad_norm": 0.6307222247123718, "learning_rate": 9.526666666666667e-05, "loss": 0.0225, "step": 1430 }, { "epoch": 0.08388186637152677, "grad_norm": 0.44919538497924805, "learning_rate": 9.593333333333334e-05, "loss": 0.0237, "step": 1440 }, { "epoch": 0.08446437933244015, "grad_norm": 0.6162551641464233, "learning_rate": 9.66e-05, "loss": 0.0264, "step": 1450 }, { "epoch": 0.08504689229335352, "grad_norm": 0.5569806098937988, "learning_rate": 9.726666666666667e-05, "loss": 0.0252, "step": 1460 }, { "epoch": 0.0856294052542669, "grad_norm": 0.5218840837478638, "learning_rate": 9.793333333333333e-05, "loss": 0.0272, "step": 1470 }, { "epoch": 0.08621191821518029, "grad_norm": 0.7213922142982483, "learning_rate": 9.86e-05, "loss": 0.0345, "step": 1480 }, { "epoch": 0.08679443117609367, "grad_norm": 0.35891667008399963, "learning_rate": 9.926666666666668e-05, "loss": 0.0255, "step": 1490 }, { "epoch": 0.08737694413700704, "grad_norm": 0.36590370535850525, "learning_rate": 9.993333333333334e-05, "loss": 0.033, "step": 1500 }, { "epoch": 0.08795945709792043, "grad_norm": 0.5863376259803772, "learning_rate": 9.999997539434007e-05, "loss": 0.0236, "step": 1510 }, { "epoch": 0.08854197005883381, "grad_norm": 0.7955098748207092, "learning_rate": 9.999989033776898e-05, "loss": 0.0255, "step": 1520 }, { "epoch": 0.0891244830197472, "grad_norm": 0.45443108677864075, "learning_rate": 9.999974452661641e-05, "loss": 0.0259, "step": 1530 }, { "epoch": 0.08970699598066056, "grad_norm": 0.5477753281593323, "learning_rate": 9.999953796105959e-05, "loss": 0.0254, "step": 1540 }, { "epoch": 0.09028950894157395, "grad_norm": 0.2657114267349243, "learning_rate": 9.999927064134949e-05, "loss": 0.0247, "step": 1550 }, { "epoch": 0.09087202190248733, "grad_norm": 0.4057621955871582, "learning_rate": 9.999894256781095e-05, "loss": 0.0223, "step": 1560 }, { "epoch": 0.09145453486340072, "grad_norm": 0.4122306704521179, "learning_rate": 9.99985537408426e-05, "loss": 0.0216, "step": 1570 }, { "epoch": 0.09203704782431409, "grad_norm": 0.2921590507030487, "learning_rate": 9.999810416091688e-05, "loss": 0.0231, "step": 1580 }, { "epoch": 0.09261956078522747, "grad_norm": 0.33729857206344604, "learning_rate": 9.99975938285801e-05, "loss": 0.0202, "step": 1590 }, { "epoch": 0.09320207374614085, "grad_norm": 0.6590836644172668, "learning_rate": 9.999702274445236e-05, "loss": 0.0209, "step": 1600 }, { "epoch": 0.09378458670705424, "grad_norm": 0.5845506191253662, "learning_rate": 9.999639090922756e-05, "loss": 0.0204, "step": 1610 }, { "epoch": 0.09436709966796761, "grad_norm": 0.37395179271698, "learning_rate": 9.999569832367346e-05, "loss": 0.0196, "step": 1620 }, { "epoch": 0.09494961262888099, "grad_norm": 0.32569620013237, "learning_rate": 9.999494498863162e-05, "loss": 0.0217, "step": 1630 }, { "epoch": 0.09553212558979438, "grad_norm": 0.5724361538887024, "learning_rate": 9.99941309050174e-05, "loss": 0.0223, "step": 1640 }, { "epoch": 0.09611463855070776, "grad_norm": 0.41537967324256897, "learning_rate": 9.999325607381999e-05, "loss": 0.0246, "step": 1650 }, { "epoch": 0.09669715151162113, "grad_norm": 0.473274290561676, "learning_rate": 9.999232049610238e-05, "loss": 0.0181, "step": 1660 }, { "epoch": 0.09727966447253451, "grad_norm": 0.44749128818511963, "learning_rate": 9.999132417300142e-05, "loss": 0.0217, "step": 1670 }, { "epoch": 0.0978621774334479, "grad_norm": 0.342777281999588, "learning_rate": 9.99902671057277e-05, "loss": 0.0212, "step": 1680 }, { "epoch": 0.09844469039436128, "grad_norm": 0.5124594569206238, "learning_rate": 9.998914929556569e-05, "loss": 0.024, "step": 1690 }, { "epoch": 0.09902720335527465, "grad_norm": 0.6373164653778076, "learning_rate": 9.998797074387361e-05, "loss": 0.0239, "step": 1700 }, { "epoch": 0.09960971631618804, "grad_norm": 0.5524207353591919, "learning_rate": 9.99867314520835e-05, "loss": 0.019, "step": 1710 }, { "epoch": 0.10019222927710142, "grad_norm": 0.49288803339004517, "learning_rate": 9.998543142170126e-05, "loss": 0.0223, "step": 1720 }, { "epoch": 0.10077474223801479, "grad_norm": 0.39513063430786133, "learning_rate": 9.99840706543065e-05, "loss": 0.021, "step": 1730 }, { "epoch": 0.10135725519892817, "grad_norm": 0.38166385889053345, "learning_rate": 9.998264915155274e-05, "loss": 0.0197, "step": 1740 }, { "epoch": 0.10193976815984156, "grad_norm": 0.4162251949310303, "learning_rate": 9.998116691516718e-05, "loss": 0.0215, "step": 1750 }, { "epoch": 0.10252228112075494, "grad_norm": 0.4375555217266083, "learning_rate": 9.997962394695091e-05, "loss": 0.0243, "step": 1760 }, { "epoch": 0.10310479408166831, "grad_norm": 0.4904995560646057, "learning_rate": 9.997802024877875e-05, "loss": 0.0222, "step": 1770 }, { "epoch": 0.1036873070425817, "grad_norm": 0.2985130846500397, "learning_rate": 9.99763558225994e-05, "loss": 0.0209, "step": 1780 }, { "epoch": 0.10426982000349508, "grad_norm": 0.43920576572418213, "learning_rate": 9.997463067043526e-05, "loss": 0.0258, "step": 1790 }, { "epoch": 0.10485233296440846, "grad_norm": 0.44892460107803345, "learning_rate": 9.997284479438253e-05, "loss": 0.02, "step": 1800 }, { "epoch": 0.10543484592532183, "grad_norm": 0.48475006222724915, "learning_rate": 9.997099819661127e-05, "loss": 0.0214, "step": 1810 }, { "epoch": 0.10601735888623522, "grad_norm": 0.5260090827941895, "learning_rate": 9.996909087936524e-05, "loss": 0.0217, "step": 1820 }, { "epoch": 0.1065998718471486, "grad_norm": 0.4060119092464447, "learning_rate": 9.996712284496201e-05, "loss": 0.02, "step": 1830 }, { "epoch": 0.10718238480806198, "grad_norm": 0.5117402076721191, "learning_rate": 9.996509409579293e-05, "loss": 0.0205, "step": 1840 }, { "epoch": 0.10776489776897535, "grad_norm": 0.36153656244277954, "learning_rate": 9.996300463432312e-05, "loss": 0.0197, "step": 1850 }, { "epoch": 0.10834741072988874, "grad_norm": 0.5402187705039978, "learning_rate": 9.996085446309148e-05, "loss": 0.0233, "step": 1860 }, { "epoch": 0.10892992369080212, "grad_norm": 0.5023298859596252, "learning_rate": 9.995864358471066e-05, "loss": 0.0234, "step": 1870 }, { "epoch": 0.1095124366517155, "grad_norm": 0.42993468046188354, "learning_rate": 9.99563720018671e-05, "loss": 0.0167, "step": 1880 }, { "epoch": 0.11009494961262888, "grad_norm": 0.4474211633205414, "learning_rate": 9.995403971732098e-05, "loss": 0.0187, "step": 1890 }, { "epoch": 0.11067746257354226, "grad_norm": 0.3085026144981384, "learning_rate": 9.995164673390625e-05, "loss": 0.0238, "step": 1900 }, { "epoch": 0.11125997553445564, "grad_norm": 0.571968674659729, "learning_rate": 9.994919305453059e-05, "loss": 0.0211, "step": 1910 }, { "epoch": 0.11184248849536903, "grad_norm": 0.439012348651886, "learning_rate": 9.994667868217548e-05, "loss": 0.0214, "step": 1920 }, { "epoch": 0.1124250014562824, "grad_norm": 0.33423957228660583, "learning_rate": 9.99441036198961e-05, "loss": 0.0203, "step": 1930 }, { "epoch": 0.11300751441719578, "grad_norm": 0.3952094614505768, "learning_rate": 9.99414678708214e-05, "loss": 0.019, "step": 1940 }, { "epoch": 0.11359002737810917, "grad_norm": 0.4013804793357849, "learning_rate": 9.993877143815407e-05, "loss": 0.0203, "step": 1950 }, { "epoch": 0.11417254033902255, "grad_norm": 0.43464595079421997, "learning_rate": 9.993601432517053e-05, "loss": 0.0172, "step": 1960 }, { "epoch": 0.11475505329993592, "grad_norm": 0.6095263957977295, "learning_rate": 9.993319653522091e-05, "loss": 0.0226, "step": 1970 }, { "epoch": 0.1153375662608493, "grad_norm": 0.4852878451347351, "learning_rate": 9.993031807172911e-05, "loss": 0.024, "step": 1980 }, { "epoch": 0.11592007922176269, "grad_norm": 0.4639914929866791, "learning_rate": 9.992737893819273e-05, "loss": 0.0191, "step": 1990 }, { "epoch": 0.11650259218267607, "grad_norm": 0.2555379271507263, "learning_rate": 9.992437913818312e-05, "loss": 0.0177, "step": 2000 }, { "epoch": 0.11708510514358944, "grad_norm": 0.3941870927810669, "learning_rate": 9.992131867534526e-05, "loss": 0.0171, "step": 2010 }, { "epoch": 0.11766761810450282, "grad_norm": 0.33638933300971985, "learning_rate": 9.991819755339796e-05, "loss": 0.022, "step": 2020 }, { "epoch": 0.11825013106541621, "grad_norm": 0.2429458647966385, "learning_rate": 9.991501577613365e-05, "loss": 0.0214, "step": 2030 }, { "epoch": 0.11883264402632959, "grad_norm": 0.3649020791053772, "learning_rate": 9.99117733474185e-05, "loss": 0.0232, "step": 2040 }, { "epoch": 0.11941515698724296, "grad_norm": 0.22641423344612122, "learning_rate": 9.990847027119234e-05, "loss": 0.0187, "step": 2050 }, { "epoch": 0.11999766994815635, "grad_norm": 0.2840929925441742, "learning_rate": 9.990510655146877e-05, "loss": 0.0162, "step": 2060 }, { "epoch": 0.12058018290906973, "grad_norm": 0.33555904030799866, "learning_rate": 9.990168219233496e-05, "loss": 0.0166, "step": 2070 }, { "epoch": 0.1211626958699831, "grad_norm": 0.3273191750049591, "learning_rate": 9.989819719795188e-05, "loss": 0.0164, "step": 2080 }, { "epoch": 0.12174520883089648, "grad_norm": 0.29209548234939575, "learning_rate": 9.989465157255412e-05, "loss": 0.0167, "step": 2090 }, { "epoch": 0.12232772179180987, "grad_norm": 0.6209645867347717, "learning_rate": 9.989104532044994e-05, "loss": 0.0171, "step": 2100 }, { "epoch": 0.12291023475272325, "grad_norm": 0.5538815259933472, "learning_rate": 9.988737844602128e-05, "loss": 0.0168, "step": 2110 }, { "epoch": 0.12349274771363662, "grad_norm": 0.39251554012298584, "learning_rate": 9.988365095372372e-05, "loss": 0.017, "step": 2120 }, { "epoch": 0.12407526067455, "grad_norm": 0.3591337502002716, "learning_rate": 9.987986284808654e-05, "loss": 0.0169, "step": 2130 }, { "epoch": 0.12465777363546339, "grad_norm": 0.5226990580558777, "learning_rate": 9.987601413371264e-05, "loss": 0.0162, "step": 2140 }, { "epoch": 0.12524028659637676, "grad_norm": 0.4623308479785919, "learning_rate": 9.987210481527855e-05, "loss": 0.0215, "step": 2150 }, { "epoch": 0.12582279955729014, "grad_norm": 0.4177090525627136, "learning_rate": 9.98681348975345e-05, "loss": 0.0181, "step": 2160 }, { "epoch": 0.12640531251820353, "grad_norm": 0.43934598565101624, "learning_rate": 9.986410438530427e-05, "loss": 0.0174, "step": 2170 }, { "epoch": 0.1269878254791169, "grad_norm": 0.4332578480243683, "learning_rate": 9.986001328348534e-05, "loss": 0.02, "step": 2180 }, { "epoch": 0.1275703384400303, "grad_norm": 0.29921674728393555, "learning_rate": 9.985586159704878e-05, "loss": 0.0165, "step": 2190 }, { "epoch": 0.12815285140094368, "grad_norm": 0.3050669729709625, "learning_rate": 9.985164933103929e-05, "loss": 0.0174, "step": 2200 }, { "epoch": 0.12873536436185706, "grad_norm": 0.33571580052375793, "learning_rate": 9.984737649057513e-05, "loss": 0.0146, "step": 2210 }, { "epoch": 0.12931787732277042, "grad_norm": 0.3501420319080353, "learning_rate": 9.984304308084827e-05, "loss": 0.0159, "step": 2220 }, { "epoch": 0.1299003902836838, "grad_norm": 0.350046843290329, "learning_rate": 9.983864910712416e-05, "loss": 0.0189, "step": 2230 }, { "epoch": 0.1304829032445972, "grad_norm": 0.45016878843307495, "learning_rate": 9.98341945747419e-05, "loss": 0.0164, "step": 2240 }, { "epoch": 0.13106541620551057, "grad_norm": 0.19611197710037231, "learning_rate": 9.98296794891142e-05, "loss": 0.0213, "step": 2250 }, { "epoch": 0.13164792916642395, "grad_norm": 0.4170686602592468, "learning_rate": 9.982510385572725e-05, "loss": 0.0177, "step": 2260 }, { "epoch": 0.13223044212733734, "grad_norm": 0.49550408124923706, "learning_rate": 9.982046768014094e-05, "loss": 0.0161, "step": 2270 }, { "epoch": 0.13281295508825072, "grad_norm": 0.4277210533618927, "learning_rate": 9.981577096798863e-05, "loss": 0.0205, "step": 2280 }, { "epoch": 0.1333954680491641, "grad_norm": 0.2560710906982422, "learning_rate": 9.981101372497727e-05, "loss": 0.0155, "step": 2290 }, { "epoch": 0.13397798101007746, "grad_norm": 0.5669534802436829, "learning_rate": 9.980619595688737e-05, "loss": 0.0139, "step": 2300 }, { "epoch": 0.13456049397099085, "grad_norm": 0.31216830015182495, "learning_rate": 9.980131766957295e-05, "loss": 0.0204, "step": 2310 }, { "epoch": 0.13514300693190423, "grad_norm": 0.32178542017936707, "learning_rate": 9.979637886896163e-05, "loss": 0.0124, "step": 2320 }, { "epoch": 0.13572551989281761, "grad_norm": 0.48147812485694885, "learning_rate": 9.979137956105447e-05, "loss": 0.0231, "step": 2330 }, { "epoch": 0.136308032853731, "grad_norm": 0.5376265645027161, "learning_rate": 9.978631975192613e-05, "loss": 0.0174, "step": 2340 }, { "epoch": 0.13689054581464438, "grad_norm": 0.5305449962615967, "learning_rate": 9.978119944772475e-05, "loss": 0.0207, "step": 2350 }, { "epoch": 0.13747305877555777, "grad_norm": 0.5217538475990295, "learning_rate": 9.977601865467197e-05, "loss": 0.0166, "step": 2360 }, { "epoch": 0.13805557173647115, "grad_norm": 0.4606058597564697, "learning_rate": 9.977077737906297e-05, "loss": 0.0174, "step": 2370 }, { "epoch": 0.1386380846973845, "grad_norm": 0.42594361305236816, "learning_rate": 9.976547562726636e-05, "loss": 0.0164, "step": 2380 }, { "epoch": 0.1392205976582979, "grad_norm": 0.26246941089630127, "learning_rate": 9.976011340572429e-05, "loss": 0.0198, "step": 2390 }, { "epoch": 0.13980311061921127, "grad_norm": 0.3303341567516327, "learning_rate": 9.975469072095237e-05, "loss": 0.0178, "step": 2400 }, { "epoch": 0.14038562358012466, "grad_norm": 0.3827970325946808, "learning_rate": 9.974920757953965e-05, "loss": 0.0149, "step": 2410 }, { "epoch": 0.14096813654103804, "grad_norm": 0.22160081565380096, "learning_rate": 9.97436639881487e-05, "loss": 0.0146, "step": 2420 }, { "epoch": 0.14155064950195143, "grad_norm": 0.4541112184524536, "learning_rate": 9.973805995351545e-05, "loss": 0.0146, "step": 2430 }, { "epoch": 0.1421331624628648, "grad_norm": 0.2938708961009979, "learning_rate": 9.973239548244939e-05, "loss": 0.0156, "step": 2440 }, { "epoch": 0.14271567542377817, "grad_norm": 0.24117307364940643, "learning_rate": 9.972667058183333e-05, "loss": 0.0162, "step": 2450 }, { "epoch": 0.14329818838469155, "grad_norm": 0.4317592680454254, "learning_rate": 9.972088525862362e-05, "loss": 0.0182, "step": 2460 }, { "epoch": 0.14388070134560493, "grad_norm": 0.3063121438026428, "learning_rate": 9.971503951984995e-05, "loss": 0.015, "step": 2470 }, { "epoch": 0.14446321430651832, "grad_norm": 0.28073713183403015, "learning_rate": 9.970913337261543e-05, "loss": 0.0158, "step": 2480 }, { "epoch": 0.1450457272674317, "grad_norm": 0.4092753827571869, "learning_rate": 9.97031668240966e-05, "loss": 0.0131, "step": 2490 }, { "epoch": 0.14562824022834508, "grad_norm": 0.4061177968978882, "learning_rate": 9.969713988154339e-05, "loss": 0.0171, "step": 2500 }, { "epoch": 0.14621075318925847, "grad_norm": 0.45309191942214966, "learning_rate": 9.969105255227906e-05, "loss": 0.0157, "step": 2510 }, { "epoch": 0.14679326615017185, "grad_norm": 0.48115354776382446, "learning_rate": 9.968490484370035e-05, "loss": 0.0146, "step": 2520 }, { "epoch": 0.1473757791110852, "grad_norm": 0.3232123851776123, "learning_rate": 9.967869676327726e-05, "loss": 0.0134, "step": 2530 }, { "epoch": 0.1479582920719986, "grad_norm": 0.3205547630786896, "learning_rate": 9.96724283185532e-05, "loss": 0.0181, "step": 2540 }, { "epoch": 0.14854080503291198, "grad_norm": 0.5272349119186401, "learning_rate": 9.966609951714494e-05, "loss": 0.0191, "step": 2550 }, { "epoch": 0.14912331799382536, "grad_norm": 0.2887498736381531, "learning_rate": 9.965971036674255e-05, "loss": 0.0188, "step": 2560 }, { "epoch": 0.14970583095473874, "grad_norm": 0.42976537346839905, "learning_rate": 9.965326087510947e-05, "loss": 0.0181, "step": 2570 }, { "epoch": 0.15028834391565213, "grad_norm": 0.6130855679512024, "learning_rate": 9.964675105008244e-05, "loss": 0.0196, "step": 2580 }, { "epoch": 0.1508708568765655, "grad_norm": 0.6346950531005859, "learning_rate": 9.964018089957147e-05, "loss": 0.0196, "step": 2590 }, { "epoch": 0.1514533698374789, "grad_norm": 0.539539635181427, "learning_rate": 9.963355043155996e-05, "loss": 0.0165, "step": 2600 }, { "epoch": 0.15203588279839225, "grad_norm": 0.3816796839237213, "learning_rate": 9.962685965410455e-05, "loss": 0.0156, "step": 2610 }, { "epoch": 0.15261839575930564, "grad_norm": 0.409084677696228, "learning_rate": 9.962010857533514e-05, "loss": 0.0169, "step": 2620 }, { "epoch": 0.15320090872021902, "grad_norm": 0.2978951036930084, "learning_rate": 9.961329720345493e-05, "loss": 0.0162, "step": 2630 }, { "epoch": 0.1537834216811324, "grad_norm": 0.370525062084198, "learning_rate": 9.96064255467404e-05, "loss": 0.0136, "step": 2640 }, { "epoch": 0.1543659346420458, "grad_norm": 0.4469973146915436, "learning_rate": 9.959949361354126e-05, "loss": 0.0193, "step": 2650 }, { "epoch": 0.15494844760295917, "grad_norm": 0.2943956255912781, "learning_rate": 9.959250141228045e-05, "loss": 0.0184, "step": 2660 }, { "epoch": 0.15553096056387256, "grad_norm": 0.22035115957260132, "learning_rate": 9.958544895145414e-05, "loss": 0.0168, "step": 2670 }, { "epoch": 0.15611347352478594, "grad_norm": 0.340677410364151, "learning_rate": 9.957833623963177e-05, "loss": 0.0182, "step": 2680 }, { "epoch": 0.1566959864856993, "grad_norm": 0.4560448229312897, "learning_rate": 9.957116328545593e-05, "loss": 0.0163, "step": 2690 }, { "epoch": 0.15727849944661268, "grad_norm": 0.3833587169647217, "learning_rate": 9.956393009764244e-05, "loss": 0.0188, "step": 2700 }, { "epoch": 0.15786101240752606, "grad_norm": 0.20285329222679138, "learning_rate": 9.955663668498032e-05, "loss": 0.016, "step": 2710 }, { "epoch": 0.15844352536843945, "grad_norm": 0.4248286485671997, "learning_rate": 9.954928305633173e-05, "loss": 0.0141, "step": 2720 }, { "epoch": 0.15902603832935283, "grad_norm": 0.34970593452453613, "learning_rate": 9.954186922063204e-05, "loss": 0.0171, "step": 2730 }, { "epoch": 0.15960855129026622, "grad_norm": 0.33577001094818115, "learning_rate": 9.953439518688974e-05, "loss": 0.0179, "step": 2740 }, { "epoch": 0.1601910642511796, "grad_norm": 0.3365096151828766, "learning_rate": 9.952686096418652e-05, "loss": 0.0137, "step": 2750 }, { "epoch": 0.16077357721209298, "grad_norm": 0.2108718603849411, "learning_rate": 9.951926656167714e-05, "loss": 0.0172, "step": 2760 }, { "epoch": 0.16135609017300634, "grad_norm": 0.4207550883293152, "learning_rate": 9.951161198858953e-05, "loss": 0.0162, "step": 2770 }, { "epoch": 0.16193860313391972, "grad_norm": 0.4917376935482025, "learning_rate": 9.950389725422471e-05, "loss": 0.0155, "step": 2780 }, { "epoch": 0.1625211160948331, "grad_norm": 0.35805121064186096, "learning_rate": 9.949612236795682e-05, "loss": 0.0128, "step": 2790 }, { "epoch": 0.1631036290557465, "grad_norm": 0.44710585474967957, "learning_rate": 9.948828733923305e-05, "loss": 0.0156, "step": 2800 }, { "epoch": 0.16368614201665987, "grad_norm": 0.2824556827545166, "learning_rate": 9.948039217757374e-05, "loss": 0.0161, "step": 2810 }, { "epoch": 0.16426865497757326, "grad_norm": 0.483117014169693, "learning_rate": 9.947243689257225e-05, "loss": 0.0153, "step": 2820 }, { "epoch": 0.16485116793848664, "grad_norm": 0.3247731328010559, "learning_rate": 9.946442149389497e-05, "loss": 0.0172, "step": 2830 }, { "epoch": 0.1654336808994, "grad_norm": 0.3819945752620697, "learning_rate": 9.945634599128139e-05, "loss": 0.0177, "step": 2840 }, { "epoch": 0.16601619386031338, "grad_norm": 0.4045076370239258, "learning_rate": 9.944821039454402e-05, "loss": 0.0154, "step": 2850 }, { "epoch": 0.16659870682122677, "grad_norm": 0.4406282901763916, "learning_rate": 9.944001471356835e-05, "loss": 0.0145, "step": 2860 }, { "epoch": 0.16718121978214015, "grad_norm": 0.28208911418914795, "learning_rate": 9.94317589583129e-05, "loss": 0.0138, "step": 2870 }, { "epoch": 0.16776373274305353, "grad_norm": 0.35577231645584106, "learning_rate": 9.942344313880922e-05, "loss": 0.0165, "step": 2880 }, { "epoch": 0.16834624570396692, "grad_norm": 0.28937041759490967, "learning_rate": 9.941506726516179e-05, "loss": 0.0155, "step": 2890 }, { "epoch": 0.1689287586648803, "grad_norm": 0.39862045645713806, "learning_rate": 9.94066313475481e-05, "loss": 0.0166, "step": 2900 }, { "epoch": 0.16951127162579369, "grad_norm": 0.3496819734573364, "learning_rate": 9.939813539621857e-05, "loss": 0.0202, "step": 2910 }, { "epoch": 0.17009378458670704, "grad_norm": 0.396270751953125, "learning_rate": 9.93895794214966e-05, "loss": 0.0164, "step": 2920 }, { "epoch": 0.17067629754762043, "grad_norm": 0.39987051486968994, "learning_rate": 9.938096343377852e-05, "loss": 0.016, "step": 2930 }, { "epoch": 0.1712588105085338, "grad_norm": 0.33410680294036865, "learning_rate": 9.937228744353353e-05, "loss": 0.0166, "step": 2940 }, { "epoch": 0.1718413234694472, "grad_norm": 0.27986234426498413, "learning_rate": 9.936355146130379e-05, "loss": 0.0142, "step": 2950 }, { "epoch": 0.17242383643036058, "grad_norm": 0.3302169740200043, "learning_rate": 9.935475549770435e-05, "loss": 0.0174, "step": 2960 }, { "epoch": 0.17300634939127396, "grad_norm": 0.41995158791542053, "learning_rate": 9.934589956342315e-05, "loss": 0.0153, "step": 2970 }, { "epoch": 0.17358886235218735, "grad_norm": 0.3632122576236725, "learning_rate": 9.933698366922093e-05, "loss": 0.0132, "step": 2980 }, { "epoch": 0.17417137531310073, "grad_norm": 0.3555792272090912, "learning_rate": 9.93280078259314e-05, "loss": 0.0121, "step": 2990 }, { "epoch": 0.17475388827401409, "grad_norm": 0.32381510734558105, "learning_rate": 9.931897204446104e-05, "loss": 0.0158, "step": 3000 }, { "epoch": 0.17533640123492747, "grad_norm": 0.32888510823249817, "learning_rate": 9.930987633578915e-05, "loss": 0.0146, "step": 3010 }, { "epoch": 0.17591891419584085, "grad_norm": 0.39167606830596924, "learning_rate": 9.93007207109679e-05, "loss": 0.0187, "step": 3020 }, { "epoch": 0.17650142715675424, "grad_norm": 0.634312093257904, "learning_rate": 9.929150518112224e-05, "loss": 0.0163, "step": 3030 }, { "epoch": 0.17708394011766762, "grad_norm": 0.347868949174881, "learning_rate": 9.928222975744991e-05, "loss": 0.0141, "step": 3040 }, { "epoch": 0.177666453078581, "grad_norm": 0.37778526544570923, "learning_rate": 9.92728944512214e-05, "loss": 0.0152, "step": 3050 }, { "epoch": 0.1782489660394944, "grad_norm": 0.3820255398750305, "learning_rate": 9.926349927378001e-05, "loss": 0.0153, "step": 3060 }, { "epoch": 0.17883147900040777, "grad_norm": 0.4140007495880127, "learning_rate": 9.925404423654174e-05, "loss": 0.0192, "step": 3070 }, { "epoch": 0.17941399196132113, "grad_norm": 0.46465981006622314, "learning_rate": 9.924452935099537e-05, "loss": 0.0165, "step": 3080 }, { "epoch": 0.1799965049222345, "grad_norm": 0.3738744556903839, "learning_rate": 9.92349546287024e-05, "loss": 0.0159, "step": 3090 }, { "epoch": 0.1805790178831479, "grad_norm": 0.39833900332450867, "learning_rate": 9.9225320081297e-05, "loss": 0.0164, "step": 3100 }, { "epoch": 0.18116153084406128, "grad_norm": 0.5395527482032776, "learning_rate": 9.921562572048606e-05, "loss": 0.018, "step": 3110 }, { "epoch": 0.18174404380497466, "grad_norm": 0.4580261707305908, "learning_rate": 9.920587155804913e-05, "loss": 0.0159, "step": 3120 }, { "epoch": 0.18232655676588805, "grad_norm": 0.42417433857917786, "learning_rate": 9.919605760583845e-05, "loss": 0.0187, "step": 3130 }, { "epoch": 0.18290906972680143, "grad_norm": 0.33866560459136963, "learning_rate": 9.91861838757789e-05, "loss": 0.0164, "step": 3140 }, { "epoch": 0.1834915826877148, "grad_norm": 0.4014289677143097, "learning_rate": 9.917625037986798e-05, "loss": 0.0162, "step": 3150 }, { "epoch": 0.18407409564862817, "grad_norm": 0.2678658664226532, "learning_rate": 9.916625713017583e-05, "loss": 0.0142, "step": 3160 }, { "epoch": 0.18465660860954156, "grad_norm": 0.24174220860004425, "learning_rate": 9.915620413884519e-05, "loss": 0.0155, "step": 3170 }, { "epoch": 0.18523912157045494, "grad_norm": 0.46617671847343445, "learning_rate": 9.914609141809139e-05, "loss": 0.0162, "step": 3180 }, { "epoch": 0.18582163453136832, "grad_norm": 0.4024803638458252, "learning_rate": 9.913591898020235e-05, "loss": 0.0172, "step": 3190 }, { "epoch": 0.1864041474922817, "grad_norm": 0.6289913654327393, "learning_rate": 9.912568683753853e-05, "loss": 0.0209, "step": 3200 }, { "epoch": 0.1869866604531951, "grad_norm": 0.346633642911911, "learning_rate": 9.911539500253295e-05, "loss": 0.0158, "step": 3210 }, { "epoch": 0.18756917341410848, "grad_norm": 0.37537068128585815, "learning_rate": 9.910504348769118e-05, "loss": 0.0148, "step": 3220 }, { "epoch": 0.18815168637502183, "grad_norm": 0.4973176121711731, "learning_rate": 9.909463230559127e-05, "loss": 0.0142, "step": 3230 }, { "epoch": 0.18873419933593522, "grad_norm": 0.3805164396762848, "learning_rate": 9.908416146888376e-05, "loss": 0.0115, "step": 3240 }, { "epoch": 0.1893167122968486, "grad_norm": 0.39441925287246704, "learning_rate": 9.907363099029175e-05, "loss": 0.0128, "step": 3250 }, { "epoch": 0.18989922525776198, "grad_norm": 0.2503442168235779, "learning_rate": 9.906304088261073e-05, "loss": 0.0206, "step": 3260 }, { "epoch": 0.19048173821867537, "grad_norm": 0.29791730642318726, "learning_rate": 9.905239115870872e-05, "loss": 0.0156, "step": 3270 }, { "epoch": 0.19106425117958875, "grad_norm": 0.3576847016811371, "learning_rate": 9.90416818315261e-05, "loss": 0.0129, "step": 3280 }, { "epoch": 0.19164676414050213, "grad_norm": 0.4002215266227722, "learning_rate": 9.903091291407573e-05, "loss": 0.015, "step": 3290 }, { "epoch": 0.19222927710141552, "grad_norm": 0.44520682096481323, "learning_rate": 9.902008441944286e-05, "loss": 0.014, "step": 3300 }, { "epoch": 0.19281179006232887, "grad_norm": 0.48155078291893005, "learning_rate": 9.900919636078512e-05, "loss": 0.0141, "step": 3310 }, { "epoch": 0.19339430302324226, "grad_norm": 0.22092358767986298, "learning_rate": 9.899824875133255e-05, "loss": 0.014, "step": 3320 }, { "epoch": 0.19397681598415564, "grad_norm": 0.48741286993026733, "learning_rate": 9.898724160438749e-05, "loss": 0.0142, "step": 3330 }, { "epoch": 0.19455932894506903, "grad_norm": 0.34430861473083496, "learning_rate": 9.89761749333247e-05, "loss": 0.0138, "step": 3340 }, { "epoch": 0.1951418419059824, "grad_norm": 0.3429625928401947, "learning_rate": 9.896504875159122e-05, "loss": 0.017, "step": 3350 }, { "epoch": 0.1957243548668958, "grad_norm": 0.297616571187973, "learning_rate": 9.89538630727064e-05, "loss": 0.0131, "step": 3360 }, { "epoch": 0.19630686782780918, "grad_norm": 0.39675647020339966, "learning_rate": 9.894261791026189e-05, "loss": 0.0207, "step": 3370 }, { "epoch": 0.19688938078872256, "grad_norm": 0.29995664954185486, "learning_rate": 9.893131327792165e-05, "loss": 0.0162, "step": 3380 }, { "epoch": 0.19747189374963592, "grad_norm": 0.33954164385795593, "learning_rate": 9.891994918942182e-05, "loss": 0.0161, "step": 3390 }, { "epoch": 0.1980544067105493, "grad_norm": 0.5150769948959351, "learning_rate": 9.890852565857092e-05, "loss": 0.0167, "step": 3400 }, { "epoch": 0.1986369196714627, "grad_norm": 0.5345335006713867, "learning_rate": 9.889704269924954e-05, "loss": 0.0161, "step": 3410 }, { "epoch": 0.19921943263237607, "grad_norm": 0.30202922224998474, "learning_rate": 9.888550032541059e-05, "loss": 0.0187, "step": 3420 }, { "epoch": 0.19980194559328945, "grad_norm": 0.3381350636482239, "learning_rate": 9.887389855107916e-05, "loss": 0.0146, "step": 3430 }, { "epoch": 0.20038445855420284, "grad_norm": 0.5011172890663147, "learning_rate": 9.886223739035248e-05, "loss": 0.0169, "step": 3440 }, { "epoch": 0.20096697151511622, "grad_norm": 0.269399493932724, "learning_rate": 9.885051685739997e-05, "loss": 0.0137, "step": 3450 }, { "epoch": 0.20154948447602958, "grad_norm": 0.5072921514511108, "learning_rate": 9.883873696646316e-05, "loss": 0.0187, "step": 3460 }, { "epoch": 0.20213199743694296, "grad_norm": 0.24462972581386566, "learning_rate": 9.882689773185575e-05, "loss": 0.0137, "step": 3470 }, { "epoch": 0.20271451039785635, "grad_norm": 0.3352923095226288, "learning_rate": 9.881499916796353e-05, "loss": 0.0146, "step": 3480 }, { "epoch": 0.20329702335876973, "grad_norm": 0.28809764981269836, "learning_rate": 9.880304128924434e-05, "loss": 0.0144, "step": 3490 }, { "epoch": 0.2038795363196831, "grad_norm": 0.4793824553489685, "learning_rate": 9.879102411022817e-05, "loss": 0.0179, "step": 3500 }, { "epoch": 0.2044620492805965, "grad_norm": 0.25081145763397217, "learning_rate": 9.877894764551703e-05, "loss": 0.0133, "step": 3510 }, { "epoch": 0.20504456224150988, "grad_norm": 0.2768265902996063, "learning_rate": 9.876681190978494e-05, "loss": 0.0121, "step": 3520 }, { "epoch": 0.20562707520242327, "grad_norm": 0.42935487627983093, "learning_rate": 9.875461691777797e-05, "loss": 0.0143, "step": 3530 }, { "epoch": 0.20620958816333662, "grad_norm": 0.32817867398262024, "learning_rate": 9.874236268431417e-05, "loss": 0.0162, "step": 3540 }, { "epoch": 0.20679210112425, "grad_norm": 0.3461551368236542, "learning_rate": 9.873004922428361e-05, "loss": 0.016, "step": 3550 }, { "epoch": 0.2073746140851634, "grad_norm": 0.5130866765975952, "learning_rate": 9.871767655264829e-05, "loss": 0.0157, "step": 3560 }, { "epoch": 0.20795712704607677, "grad_norm": 0.34219974279403687, "learning_rate": 9.87052446844422e-05, "loss": 0.0181, "step": 3570 }, { "epoch": 0.20853964000699016, "grad_norm": 0.3319697380065918, "learning_rate": 9.869275363477122e-05, "loss": 0.0159, "step": 3580 }, { "epoch": 0.20912215296790354, "grad_norm": 0.37241482734680176, "learning_rate": 9.868020341881312e-05, "loss": 0.0132, "step": 3590 }, { "epoch": 0.20970466592881692, "grad_norm": 0.34658288955688477, "learning_rate": 9.866759405181765e-05, "loss": 0.0163, "step": 3600 }, { "epoch": 0.2102871788897303, "grad_norm": 0.24060270190238953, "learning_rate": 9.865492554910633e-05, "loss": 0.0127, "step": 3610 }, { "epoch": 0.21086969185064366, "grad_norm": 0.46318456530570984, "learning_rate": 9.864219792607262e-05, "loss": 0.016, "step": 3620 }, { "epoch": 0.21145220481155705, "grad_norm": 0.30252811312675476, "learning_rate": 9.862941119818177e-05, "loss": 0.0168, "step": 3630 }, { "epoch": 0.21203471777247043, "grad_norm": 0.39923396706581116, "learning_rate": 9.861656538097086e-05, "loss": 0.0165, "step": 3640 }, { "epoch": 0.21261723073338382, "grad_norm": 0.3751615881919861, "learning_rate": 9.860366049004877e-05, "loss": 0.0188, "step": 3650 }, { "epoch": 0.2131997436942972, "grad_norm": 0.4151049852371216, "learning_rate": 9.859069654109615e-05, "loss": 0.0136, "step": 3660 }, { "epoch": 0.21378225665521058, "grad_norm": 0.49764227867126465, "learning_rate": 9.857767354986545e-05, "loss": 0.0154, "step": 3670 }, { "epoch": 0.21436476961612397, "grad_norm": 0.32790184020996094, "learning_rate": 9.856459153218078e-05, "loss": 0.0201, "step": 3680 }, { "epoch": 0.21494728257703735, "grad_norm": 0.3363580107688904, "learning_rate": 9.855145050393808e-05, "loss": 0.0183, "step": 3690 }, { "epoch": 0.2155297955379507, "grad_norm": 0.36230507493019104, "learning_rate": 9.85382504811049e-05, "loss": 0.0165, "step": 3700 }, { "epoch": 0.2161123084988641, "grad_norm": 0.3681294023990631, "learning_rate": 9.852499147972054e-05, "loss": 0.0198, "step": 3710 }, { "epoch": 0.21669482145977748, "grad_norm": 0.27238598465919495, "learning_rate": 9.851167351589592e-05, "loss": 0.0142, "step": 3720 }, { "epoch": 0.21727733442069086, "grad_norm": 0.20715878903865814, "learning_rate": 9.849829660581363e-05, "loss": 0.0153, "step": 3730 }, { "epoch": 0.21785984738160424, "grad_norm": 0.3693470358848572, "learning_rate": 9.848486076572787e-05, "loss": 0.0181, "step": 3740 }, { "epoch": 0.21844236034251763, "grad_norm": 0.4169458746910095, "learning_rate": 9.847136601196446e-05, "loss": 0.0138, "step": 3750 }, { "epoch": 0.219024873303431, "grad_norm": 0.4135362505912781, "learning_rate": 9.845781236092078e-05, "loss": 0.0146, "step": 3760 }, { "epoch": 0.2196073862643444, "grad_norm": 0.40244174003601074, "learning_rate": 9.844419982906583e-05, "loss": 0.0203, "step": 3770 }, { "epoch": 0.22018989922525775, "grad_norm": 0.2031620591878891, "learning_rate": 9.843052843294008e-05, "loss": 0.014, "step": 3780 }, { "epoch": 0.22077241218617114, "grad_norm": 0.42439550161361694, "learning_rate": 9.841679818915559e-05, "loss": 0.0137, "step": 3790 }, { "epoch": 0.22135492514708452, "grad_norm": 0.2717705965042114, "learning_rate": 9.840300911439591e-05, "loss": 0.0172, "step": 3800 }, { "epoch": 0.2219374381079979, "grad_norm": 0.21812152862548828, "learning_rate": 9.838916122541603e-05, "loss": 0.0123, "step": 3810 }, { "epoch": 0.2225199510689113, "grad_norm": 0.3498571813106537, "learning_rate": 9.837525453904246e-05, "loss": 0.0164, "step": 3820 }, { "epoch": 0.22310246402982467, "grad_norm": 0.4606907069683075, "learning_rate": 9.836128907217314e-05, "loss": 0.0181, "step": 3830 }, { "epoch": 0.22368497699073805, "grad_norm": 0.47948184609413147, "learning_rate": 9.834726484177743e-05, "loss": 0.0171, "step": 3840 }, { "epoch": 0.2242674899516514, "grad_norm": 0.3982204794883728, "learning_rate": 9.833318186489609e-05, "loss": 0.0165, "step": 3850 }, { "epoch": 0.2248500029125648, "grad_norm": 0.3378174304962158, "learning_rate": 9.831904015864126e-05, "loss": 0.0159, "step": 3860 }, { "epoch": 0.22543251587347818, "grad_norm": 0.3384462893009186, "learning_rate": 9.830483974019645e-05, "loss": 0.0144, "step": 3870 }, { "epoch": 0.22601502883439156, "grad_norm": 0.3724633753299713, "learning_rate": 9.82905806268165e-05, "loss": 0.0135, "step": 3880 }, { "epoch": 0.22659754179530495, "grad_norm": 0.28294652700424194, "learning_rate": 9.82762628358276e-05, "loss": 0.0131, "step": 3890 }, { "epoch": 0.22718005475621833, "grad_norm": 0.35287946462631226, "learning_rate": 9.826188638462718e-05, "loss": 0.013, "step": 3900 }, { "epoch": 0.22776256771713171, "grad_norm": 0.22180771827697754, "learning_rate": 9.824745129068402e-05, "loss": 0.0141, "step": 3910 }, { "epoch": 0.2283450806780451, "grad_norm": 0.3079909086227417, "learning_rate": 9.82329575715381e-05, "loss": 0.0133, "step": 3920 }, { "epoch": 0.22892759363895845, "grad_norm": 0.25783053040504456, "learning_rate": 9.821840524480066e-05, "loss": 0.0165, "step": 3930 }, { "epoch": 0.22951010659987184, "grad_norm": 0.3007711172103882, "learning_rate": 9.820379432815414e-05, "loss": 0.0161, "step": 3940 }, { "epoch": 0.23009261956078522, "grad_norm": 0.39270880818367004, "learning_rate": 9.81891248393522e-05, "loss": 0.0163, "step": 3950 }, { "epoch": 0.2306751325216986, "grad_norm": 0.3316587507724762, "learning_rate": 9.817439679621963e-05, "loss": 0.0142, "step": 3960 }, { "epoch": 0.231257645482612, "grad_norm": 0.38680002093315125, "learning_rate": 9.815961021665243e-05, "loss": 0.0169, "step": 3970 }, { "epoch": 0.23184015844352537, "grad_norm": 0.48907414078712463, "learning_rate": 9.814476511861763e-05, "loss": 0.0143, "step": 3980 }, { "epoch": 0.23242267140443876, "grad_norm": 0.25044673681259155, "learning_rate": 9.812986152015348e-05, "loss": 0.0136, "step": 3990 }, { "epoch": 0.23300518436535214, "grad_norm": 0.32564136385917664, "learning_rate": 9.811489943936922e-05, "loss": 0.0157, "step": 4000 }, { "epoch": 0.2335876973262655, "grad_norm": 0.5283328890800476, "learning_rate": 9.809987889444522e-05, "loss": 0.0151, "step": 4010 }, { "epoch": 0.23417021028717888, "grad_norm": 0.4246067702770233, "learning_rate": 9.808479990363282e-05, "loss": 0.0133, "step": 4020 }, { "epoch": 0.23475272324809227, "grad_norm": 0.3396816849708557, "learning_rate": 9.806966248525445e-05, "loss": 0.0158, "step": 4030 }, { "epoch": 0.23533523620900565, "grad_norm": 0.10669705271720886, "learning_rate": 9.805446665770348e-05, "loss": 0.014, "step": 4040 }, { "epoch": 0.23591774916991903, "grad_norm": 0.2744532823562622, "learning_rate": 9.803921243944429e-05, "loss": 0.0113, "step": 4050 }, { "epoch": 0.23650026213083242, "grad_norm": 0.403071790933609, "learning_rate": 9.802389984901218e-05, "loss": 0.0167, "step": 4060 }, { "epoch": 0.2370827750917458, "grad_norm": 0.18635526299476624, "learning_rate": 9.80085289050134e-05, "loss": 0.0127, "step": 4070 }, { "epoch": 0.23766528805265918, "grad_norm": 0.4232059121131897, "learning_rate": 9.799309962612508e-05, "loss": 0.013, "step": 4080 }, { "epoch": 0.23824780101357254, "grad_norm": 0.32140085101127625, "learning_rate": 9.797761203109527e-05, "loss": 0.0136, "step": 4090 }, { "epoch": 0.23883031397448592, "grad_norm": 0.3512478470802307, "learning_rate": 9.796206613874283e-05, "loss": 0.0145, "step": 4100 }, { "epoch": 0.2394128269353993, "grad_norm": 0.3515501618385315, "learning_rate": 9.794646196795754e-05, "loss": 0.0138, "step": 4110 }, { "epoch": 0.2399953398963127, "grad_norm": 0.2739296853542328, "learning_rate": 9.793079953769987e-05, "loss": 0.0135, "step": 4120 }, { "epoch": 0.24057785285722608, "grad_norm": 0.27916398644447327, "learning_rate": 9.79150788670012e-05, "loss": 0.0146, "step": 4130 }, { "epoch": 0.24116036581813946, "grad_norm": 0.22873377799987793, "learning_rate": 9.78992999749636e-05, "loss": 0.0143, "step": 4140 }, { "epoch": 0.24174287877905284, "grad_norm": 0.37163299322128296, "learning_rate": 9.788346288075994e-05, "loss": 0.0134, "step": 4150 }, { "epoch": 0.2423253917399662, "grad_norm": 0.44292131066322327, "learning_rate": 9.786756760363373e-05, "loss": 0.0135, "step": 4160 }, { "epoch": 0.24290790470087958, "grad_norm": 0.3150562644004822, "learning_rate": 9.78516141628993e-05, "loss": 0.0164, "step": 4170 }, { "epoch": 0.24349041766179297, "grad_norm": 0.2714001536369324, "learning_rate": 9.783560257794154e-05, "loss": 0.0155, "step": 4180 }, { "epoch": 0.24407293062270635, "grad_norm": 0.32230204343795776, "learning_rate": 9.781953286821603e-05, "loss": 0.0156, "step": 4190 }, { "epoch": 0.24465544358361974, "grad_norm": 0.4852038621902466, "learning_rate": 9.780340505324901e-05, "loss": 0.0136, "step": 4200 }, { "epoch": 0.24523795654453312, "grad_norm": 0.3366095721721649, "learning_rate": 9.778721915263727e-05, "loss": 0.0122, "step": 4210 }, { "epoch": 0.2458204695054465, "grad_norm": 0.30512264370918274, "learning_rate": 9.777097518604824e-05, "loss": 0.0133, "step": 4220 }, { "epoch": 0.2464029824663599, "grad_norm": 0.3890034556388855, "learning_rate": 9.775467317321984e-05, "loss": 0.0119, "step": 4230 }, { "epoch": 0.24698549542727324, "grad_norm": 0.33380448818206787, "learning_rate": 9.773831313396055e-05, "loss": 0.0147, "step": 4240 }, { "epoch": 0.24756800838818663, "grad_norm": 0.3554973006248474, "learning_rate": 9.77218950881494e-05, "loss": 0.015, "step": 4250 }, { "epoch": 0.2481505213491, "grad_norm": 0.2725274860858917, "learning_rate": 9.770541905573583e-05, "loss": 0.0169, "step": 4260 }, { "epoch": 0.2487330343100134, "grad_norm": 0.26896587014198303, "learning_rate": 9.768888505673976e-05, "loss": 0.0143, "step": 4270 }, { "epoch": 0.24931554727092678, "grad_norm": 0.3896014988422394, "learning_rate": 9.767229311125162e-05, "loss": 0.0136, "step": 4280 }, { "epoch": 0.24989806023184016, "grad_norm": 0.224648579955101, "learning_rate": 9.765564323943211e-05, "loss": 0.0128, "step": 4290 }, { "epoch": 0.2504805731927535, "grad_norm": 0.2818295359611511, "learning_rate": 9.763893546151244e-05, "loss": 0.0142, "step": 4300 }, { "epoch": 0.2510630861536669, "grad_norm": 0.2838699519634247, "learning_rate": 9.762216979779412e-05, "loss": 0.0107, "step": 4310 }, { "epoch": 0.2516455991145803, "grad_norm": 0.1902606189250946, "learning_rate": 9.760534626864902e-05, "loss": 0.0123, "step": 4320 }, { "epoch": 0.25222811207549367, "grad_norm": 0.2964578866958618, "learning_rate": 9.758846489451931e-05, "loss": 0.0119, "step": 4330 }, { "epoch": 0.25281062503640706, "grad_norm": 0.33075934648513794, "learning_rate": 9.757152569591748e-05, "loss": 0.0146, "step": 4340 }, { "epoch": 0.25339313799732044, "grad_norm": 0.42020168900489807, "learning_rate": 9.75545286934262e-05, "loss": 0.0133, "step": 4350 }, { "epoch": 0.2539756509582338, "grad_norm": 0.4201684296131134, "learning_rate": 9.753747390769847e-05, "loss": 0.0134, "step": 4360 }, { "epoch": 0.2545581639191472, "grad_norm": 0.2482096552848816, "learning_rate": 9.752036135945744e-05, "loss": 0.0132, "step": 4370 }, { "epoch": 0.2551406768800606, "grad_norm": 0.2185676395893097, "learning_rate": 9.750319106949649e-05, "loss": 0.0112, "step": 4380 }, { "epoch": 0.255723189840974, "grad_norm": 0.4022974669933319, "learning_rate": 9.748596305867913e-05, "loss": 0.02, "step": 4390 }, { "epoch": 0.25630570280188736, "grad_norm": 0.42810171842575073, "learning_rate": 9.746867734793903e-05, "loss": 0.0213, "step": 4400 }, { "epoch": 0.25688821576280074, "grad_norm": 0.7219847440719604, "learning_rate": 9.745133395827993e-05, "loss": 0.0178, "step": 4410 }, { "epoch": 0.2574707287237141, "grad_norm": 0.351438045501709, "learning_rate": 9.743393291077572e-05, "loss": 0.019, "step": 4420 }, { "epoch": 0.2580532416846275, "grad_norm": 0.4004956781864166, "learning_rate": 9.741647422657028e-05, "loss": 0.0179, "step": 4430 }, { "epoch": 0.25863575464554084, "grad_norm": 0.5662891268730164, "learning_rate": 9.739895792687758e-05, "loss": 0.022, "step": 4440 }, { "epoch": 0.2592182676064542, "grad_norm": 0.505552351474762, "learning_rate": 9.738138403298157e-05, "loss": 0.0155, "step": 4450 }, { "epoch": 0.2598007805673676, "grad_norm": 0.4024660587310791, "learning_rate": 9.736375256623619e-05, "loss": 0.0168, "step": 4460 }, { "epoch": 0.260383293528281, "grad_norm": 0.3369954228401184, "learning_rate": 9.734606354806533e-05, "loss": 0.0154, "step": 4470 }, { "epoch": 0.2609658064891944, "grad_norm": 0.3174836337566376, "learning_rate": 9.73283169999628e-05, "loss": 0.0133, "step": 4480 }, { "epoch": 0.26154831945010776, "grad_norm": 0.2838727533817291, "learning_rate": 9.731051294349238e-05, "loss": 0.0152, "step": 4490 }, { "epoch": 0.26213083241102114, "grad_norm": 0.20824427902698517, "learning_rate": 9.729265140028762e-05, "loss": 0.0149, "step": 4500 }, { "epoch": 0.2627133453719345, "grad_norm": 0.4184311032295227, "learning_rate": 9.727473239205201e-05, "loss": 0.0204, "step": 4510 }, { "epoch": 0.2632958583328479, "grad_norm": 0.4343879818916321, "learning_rate": 9.725675594055883e-05, "loss": 0.0164, "step": 4520 }, { "epoch": 0.2638783712937613, "grad_norm": 0.25088533759117126, "learning_rate": 9.723872206765116e-05, "loss": 0.0157, "step": 4530 }, { "epoch": 0.2644608842546747, "grad_norm": 0.28644782304763794, "learning_rate": 9.722063079524185e-05, "loss": 0.0133, "step": 4540 }, { "epoch": 0.26504339721558806, "grad_norm": 0.4774155914783478, "learning_rate": 9.720248214531351e-05, "loss": 0.0143, "step": 4550 }, { "epoch": 0.26562591017650145, "grad_norm": 0.36532267928123474, "learning_rate": 9.718427613991848e-05, "loss": 0.0134, "step": 4560 }, { "epoch": 0.26620842313741483, "grad_norm": 0.2361028641462326, "learning_rate": 9.716601280117873e-05, "loss": 0.0133, "step": 4570 }, { "epoch": 0.2667909360983282, "grad_norm": 0.35117876529693604, "learning_rate": 9.714769215128596e-05, "loss": 0.0145, "step": 4580 }, { "epoch": 0.26737344905924154, "grad_norm": 0.28342658281326294, "learning_rate": 9.712931421250152e-05, "loss": 0.0134, "step": 4590 }, { "epoch": 0.2679559620201549, "grad_norm": 0.2828042805194855, "learning_rate": 9.711087900715627e-05, "loss": 0.013, "step": 4600 }, { "epoch": 0.2685384749810683, "grad_norm": 0.3468429148197174, "learning_rate": 9.709238655765078e-05, "loss": 0.0147, "step": 4610 }, { "epoch": 0.2691209879419817, "grad_norm": 0.27148357033729553, "learning_rate": 9.707383688645511e-05, "loss": 0.0153, "step": 4620 }, { "epoch": 0.2697035009028951, "grad_norm": 0.40684014558792114, "learning_rate": 9.705523001610883e-05, "loss": 0.0145, "step": 4630 }, { "epoch": 0.27028601386380846, "grad_norm": 0.36176109313964844, "learning_rate": 9.703656596922107e-05, "loss": 0.0176, "step": 4640 }, { "epoch": 0.27086852682472184, "grad_norm": 0.3847898840904236, "learning_rate": 9.70178447684704e-05, "loss": 0.0158, "step": 4650 }, { "epoch": 0.27145103978563523, "grad_norm": 0.24188876152038574, "learning_rate": 9.699906643660483e-05, "loss": 0.0202, "step": 4660 }, { "epoch": 0.2720335527465486, "grad_norm": 0.2264142483472824, "learning_rate": 9.698023099644185e-05, "loss": 0.0138, "step": 4670 }, { "epoch": 0.272616065707462, "grad_norm": 0.24442389607429504, "learning_rate": 9.696133847086823e-05, "loss": 0.0107, "step": 4680 }, { "epoch": 0.2731985786683754, "grad_norm": 0.22923371195793152, "learning_rate": 9.694238888284022e-05, "loss": 0.0142, "step": 4690 }, { "epoch": 0.27378109162928876, "grad_norm": 0.269704133272171, "learning_rate": 9.692338225538333e-05, "loss": 0.0169, "step": 4700 }, { "epoch": 0.27436360459020215, "grad_norm": 0.331922709941864, "learning_rate": 9.690431861159241e-05, "loss": 0.0138, "step": 4710 }, { "epoch": 0.27494611755111553, "grad_norm": 0.3346977233886719, "learning_rate": 9.688519797463161e-05, "loss": 0.0153, "step": 4720 }, { "epoch": 0.2755286305120289, "grad_norm": 0.3668387234210968, "learning_rate": 9.686602036773426e-05, "loss": 0.0144, "step": 4730 }, { "epoch": 0.2761111434729423, "grad_norm": 0.43482881784439087, "learning_rate": 9.684678581420302e-05, "loss": 0.015, "step": 4740 }, { "epoch": 0.27669365643385563, "grad_norm": 0.31008052825927734, "learning_rate": 9.682749433740962e-05, "loss": 0.0153, "step": 4750 }, { "epoch": 0.277276169394769, "grad_norm": 0.310477614402771, "learning_rate": 9.680814596079507e-05, "loss": 0.0153, "step": 4760 }, { "epoch": 0.2778586823556824, "grad_norm": 0.3365541398525238, "learning_rate": 9.678874070786945e-05, "loss": 0.0146, "step": 4770 }, { "epoch": 0.2784411953165958, "grad_norm": 0.3239288926124573, "learning_rate": 9.676927860221199e-05, "loss": 0.0162, "step": 4780 }, { "epoch": 0.27902370827750916, "grad_norm": 0.2856786251068115, "learning_rate": 9.674975966747097e-05, "loss": 0.0133, "step": 4790 }, { "epoch": 0.27960622123842255, "grad_norm": 0.346770703792572, "learning_rate": 9.673018392736374e-05, "loss": 0.0149, "step": 4800 }, { "epoch": 0.28018873419933593, "grad_norm": 0.34028011560440063, "learning_rate": 9.671055140567667e-05, "loss": 0.0133, "step": 4810 }, { "epoch": 0.2807712471602493, "grad_norm": 0.32853594422340393, "learning_rate": 9.669086212626511e-05, "loss": 0.0104, "step": 4820 }, { "epoch": 0.2813537601211627, "grad_norm": 0.3057480454444885, "learning_rate": 9.667111611305341e-05, "loss": 0.0151, "step": 4830 }, { "epoch": 0.2819362730820761, "grad_norm": 0.314474880695343, "learning_rate": 9.665131339003486e-05, "loss": 0.0197, "step": 4840 }, { "epoch": 0.28251878604298947, "grad_norm": 0.35622647404670715, "learning_rate": 9.663145398127158e-05, "loss": 0.014, "step": 4850 }, { "epoch": 0.28310129900390285, "grad_norm": 0.5124862194061279, "learning_rate": 9.661153791089467e-05, "loss": 0.0138, "step": 4860 }, { "epoch": 0.28368381196481623, "grad_norm": 0.4098750948905945, "learning_rate": 9.659156520310402e-05, "loss": 0.0136, "step": 4870 }, { "epoch": 0.2842663249257296, "grad_norm": 0.24630773067474365, "learning_rate": 9.657153588216835e-05, "loss": 0.016, "step": 4880 }, { "epoch": 0.284848837886643, "grad_norm": 0.45266249775886536, "learning_rate": 9.655144997242516e-05, "loss": 0.0131, "step": 4890 }, { "epoch": 0.28543135084755633, "grad_norm": 0.3983039855957031, "learning_rate": 9.653130749828075e-05, "loss": 0.0165, "step": 4900 }, { "epoch": 0.2860138638084697, "grad_norm": 0.35880064964294434, "learning_rate": 9.65111084842101e-05, "loss": 0.0151, "step": 4910 }, { "epoch": 0.2865963767693831, "grad_norm": 0.25483396649360657, "learning_rate": 9.649085295475695e-05, "loss": 0.0126, "step": 4920 }, { "epoch": 0.2871788897302965, "grad_norm": 0.3345501720905304, "learning_rate": 9.647054093453365e-05, "loss": 0.0148, "step": 4930 }, { "epoch": 0.28776140269120987, "grad_norm": 0.44244930148124695, "learning_rate": 9.645017244822123e-05, "loss": 0.0136, "step": 4940 }, { "epoch": 0.28834391565212325, "grad_norm": 0.38924461603164673, "learning_rate": 9.642974752056931e-05, "loss": 0.0148, "step": 4950 }, { "epoch": 0.28892642861303663, "grad_norm": 0.2436865270137787, "learning_rate": 9.640926617639613e-05, "loss": 0.0121, "step": 4960 }, { "epoch": 0.28950894157395, "grad_norm": 0.43023666739463806, "learning_rate": 9.638872844058843e-05, "loss": 0.0187, "step": 4970 }, { "epoch": 0.2900914545348634, "grad_norm": 0.32516413927078247, "learning_rate": 9.63681343381015e-05, "loss": 0.0172, "step": 4980 }, { "epoch": 0.2906739674957768, "grad_norm": 0.2989577054977417, "learning_rate": 9.634748389395914e-05, "loss": 0.0123, "step": 4990 }, { "epoch": 0.29125648045669017, "grad_norm": 0.27005958557128906, "learning_rate": 9.632677713325353e-05, "loss": 0.015, "step": 5000 }, { "epoch": 0.29183899341760355, "grad_norm": 0.3099139630794525, "learning_rate": 9.63060140811454e-05, "loss": 0.0168, "step": 5010 }, { "epoch": 0.29242150637851694, "grad_norm": 0.34511786699295044, "learning_rate": 9.628519476286379e-05, "loss": 0.0165, "step": 5020 }, { "epoch": 0.2930040193394303, "grad_norm": 0.3462317883968353, "learning_rate": 9.626431920370612e-05, "loss": 0.0159, "step": 5030 }, { "epoch": 0.2935865323003437, "grad_norm": 0.2514205276966095, "learning_rate": 9.624338742903819e-05, "loss": 0.0127, "step": 5040 }, { "epoch": 0.2941690452612571, "grad_norm": 0.3226291239261627, "learning_rate": 9.622239946429406e-05, "loss": 0.0142, "step": 5050 }, { "epoch": 0.2947515582221704, "grad_norm": 0.330447256565094, "learning_rate": 9.620135533497609e-05, "loss": 0.0146, "step": 5060 }, { "epoch": 0.2953340711830838, "grad_norm": 0.28347858786582947, "learning_rate": 9.61802550666549e-05, "loss": 0.0103, "step": 5070 }, { "epoch": 0.2959165841439972, "grad_norm": 0.2465979903936386, "learning_rate": 9.615909868496928e-05, "loss": 0.0156, "step": 5080 }, { "epoch": 0.29649909710491057, "grad_norm": 0.31589576601982117, "learning_rate": 9.613788621562622e-05, "loss": 0.015, "step": 5090 }, { "epoch": 0.29708161006582395, "grad_norm": 0.4845690131187439, "learning_rate": 9.611661768440093e-05, "loss": 0.0163, "step": 5100 }, { "epoch": 0.29766412302673734, "grad_norm": 0.33092549443244934, "learning_rate": 9.609529311713661e-05, "loss": 0.0145, "step": 5110 }, { "epoch": 0.2982466359876507, "grad_norm": 0.29073378443717957, "learning_rate": 9.607391253974466e-05, "loss": 0.0173, "step": 5120 }, { "epoch": 0.2988291489485641, "grad_norm": 0.38316112756729126, "learning_rate": 9.605247597820448e-05, "loss": 0.0116, "step": 5130 }, { "epoch": 0.2994116619094775, "grad_norm": 0.2966541647911072, "learning_rate": 9.603098345856354e-05, "loss": 0.0132, "step": 5140 }, { "epoch": 0.2999941748703909, "grad_norm": 0.28130510449409485, "learning_rate": 9.600943500693725e-05, "loss": 0.0106, "step": 5150 }, { "epoch": 0.30057668783130426, "grad_norm": 0.2975674271583557, "learning_rate": 9.598783064950902e-05, "loss": 0.015, "step": 5160 }, { "epoch": 0.30115920079221764, "grad_norm": 0.2338527888059616, "learning_rate": 9.596617041253018e-05, "loss": 0.0096, "step": 5170 }, { "epoch": 0.301741713753131, "grad_norm": 0.3399932384490967, "learning_rate": 9.594445432231996e-05, "loss": 0.0115, "step": 5180 }, { "epoch": 0.3023242267140444, "grad_norm": 0.29272010922431946, "learning_rate": 9.592268240526547e-05, "loss": 0.0137, "step": 5190 }, { "epoch": 0.3029067396749578, "grad_norm": 0.31858593225479126, "learning_rate": 9.590085468782162e-05, "loss": 0.0163, "step": 5200 }, { "epoch": 0.3034892526358711, "grad_norm": 0.38448017835617065, "learning_rate": 9.587897119651116e-05, "loss": 0.0128, "step": 5210 }, { "epoch": 0.3040717655967845, "grad_norm": 0.22529268264770508, "learning_rate": 9.585703195792459e-05, "loss": 0.0129, "step": 5220 }, { "epoch": 0.3046542785576979, "grad_norm": 0.3622327148914337, "learning_rate": 9.583503699872016e-05, "loss": 0.0131, "step": 5230 }, { "epoch": 0.3052367915186113, "grad_norm": 0.2763410210609436, "learning_rate": 9.581298634562381e-05, "loss": 0.0116, "step": 5240 }, { "epoch": 0.30581930447952466, "grad_norm": 0.2970474064350128, "learning_rate": 9.579088002542917e-05, "loss": 0.0126, "step": 5250 }, { "epoch": 0.30640181744043804, "grad_norm": 0.2559505105018616, "learning_rate": 9.57687180649975e-05, "loss": 0.0146, "step": 5260 }, { "epoch": 0.3069843304013514, "grad_norm": 0.28003329038619995, "learning_rate": 9.574650049125768e-05, "loss": 0.0148, "step": 5270 }, { "epoch": 0.3075668433622648, "grad_norm": 0.26014652848243713, "learning_rate": 9.572422733120614e-05, "loss": 0.0149, "step": 5280 }, { "epoch": 0.3081493563231782, "grad_norm": 0.3492480516433716, "learning_rate": 9.570189861190689e-05, "loss": 0.0148, "step": 5290 }, { "epoch": 0.3087318692840916, "grad_norm": 0.400206595659256, "learning_rate": 9.56795143604914e-05, "loss": 0.0142, "step": 5300 }, { "epoch": 0.30931438224500496, "grad_norm": 0.34377798438072205, "learning_rate": 9.565707460415869e-05, "loss": 0.0136, "step": 5310 }, { "epoch": 0.30989689520591834, "grad_norm": 0.34064170718193054, "learning_rate": 9.563457937017515e-05, "loss": 0.013, "step": 5320 }, { "epoch": 0.3104794081668317, "grad_norm": 0.22578756511211395, "learning_rate": 9.56120286858746e-05, "loss": 0.0122, "step": 5330 }, { "epoch": 0.3110619211277451, "grad_norm": 0.2612018585205078, "learning_rate": 9.558942257865829e-05, "loss": 0.0143, "step": 5340 }, { "epoch": 0.3116444340886585, "grad_norm": 0.26774755120277405, "learning_rate": 9.556676107599472e-05, "loss": 0.0125, "step": 5350 }, { "epoch": 0.3122269470495719, "grad_norm": 0.21939046680927277, "learning_rate": 9.554404420541978e-05, "loss": 0.0125, "step": 5360 }, { "epoch": 0.3128094600104852, "grad_norm": 0.34214887022972107, "learning_rate": 9.55212719945366e-05, "loss": 0.0147, "step": 5370 }, { "epoch": 0.3133919729713986, "grad_norm": 0.27684247493743896, "learning_rate": 9.549844447101559e-05, "loss": 0.014, "step": 5380 }, { "epoch": 0.313974485932312, "grad_norm": 0.31412631273269653, "learning_rate": 9.547556166259433e-05, "loss": 0.0136, "step": 5390 }, { "epoch": 0.31455699889322536, "grad_norm": 0.2898443341255188, "learning_rate": 9.545262359707756e-05, "loss": 0.0131, "step": 5400 }, { "epoch": 0.31513951185413874, "grad_norm": 0.24868787825107574, "learning_rate": 9.542963030233724e-05, "loss": 0.0142, "step": 5410 }, { "epoch": 0.3157220248150521, "grad_norm": 0.31491348147392273, "learning_rate": 9.540658180631237e-05, "loss": 0.0133, "step": 5420 }, { "epoch": 0.3163045377759655, "grad_norm": 0.20581558346748352, "learning_rate": 9.538347813700904e-05, "loss": 0.0158, "step": 5430 }, { "epoch": 0.3168870507368789, "grad_norm": 0.2847931385040283, "learning_rate": 9.536031932250036e-05, "loss": 0.0129, "step": 5440 }, { "epoch": 0.3174695636977923, "grad_norm": 0.33832216262817383, "learning_rate": 9.533710539092653e-05, "loss": 0.015, "step": 5450 }, { "epoch": 0.31805207665870566, "grad_norm": 0.28036195039749146, "learning_rate": 9.531383637049464e-05, "loss": 0.011, "step": 5460 }, { "epoch": 0.31863458961961905, "grad_norm": 0.30270060896873474, "learning_rate": 9.529051228947875e-05, "loss": 0.0124, "step": 5470 }, { "epoch": 0.31921710258053243, "grad_norm": 0.34320932626724243, "learning_rate": 9.52671331762198e-05, "loss": 0.0129, "step": 5480 }, { "epoch": 0.3197996155414458, "grad_norm": 0.40417709946632385, "learning_rate": 9.524369905912565e-05, "loss": 0.0109, "step": 5490 }, { "epoch": 0.3203821285023592, "grad_norm": 0.28732940554618835, "learning_rate": 9.522020996667092e-05, "loss": 0.0107, "step": 5500 }, { "epoch": 0.3209646414632726, "grad_norm": 0.22060438990592957, "learning_rate": 9.519666592739709e-05, "loss": 0.0113, "step": 5510 }, { "epoch": 0.32154715442418597, "grad_norm": 0.23726588487625122, "learning_rate": 9.517306696991241e-05, "loss": 0.0102, "step": 5520 }, { "epoch": 0.3221296673850993, "grad_norm": 0.34877848625183105, "learning_rate": 9.51494131228918e-05, "loss": 0.0201, "step": 5530 }, { "epoch": 0.3227121803460127, "grad_norm": 0.20062492787837982, "learning_rate": 9.512570441507695e-05, "loss": 0.014, "step": 5540 }, { "epoch": 0.32329469330692606, "grad_norm": 0.3137679100036621, "learning_rate": 9.510194087527615e-05, "loss": 0.011, "step": 5550 }, { "epoch": 0.32387720626783945, "grad_norm": 0.4614037275314331, "learning_rate": 9.507812253236435e-05, "loss": 0.0147, "step": 5560 }, { "epoch": 0.32445971922875283, "grad_norm": 0.2742924392223358, "learning_rate": 9.505424941528309e-05, "loss": 0.0129, "step": 5570 }, { "epoch": 0.3250422321896662, "grad_norm": 0.2699301242828369, "learning_rate": 9.503032155304046e-05, "loss": 0.013, "step": 5580 }, { "epoch": 0.3256247451505796, "grad_norm": 0.4203640818595886, "learning_rate": 9.500633897471106e-05, "loss": 0.0138, "step": 5590 }, { "epoch": 0.326207258111493, "grad_norm": 0.37082532048225403, "learning_rate": 9.498230170943596e-05, "loss": 0.0141, "step": 5600 }, { "epoch": 0.32678977107240637, "grad_norm": 0.4199288785457611, "learning_rate": 9.495820978642275e-05, "loss": 0.0169, "step": 5610 }, { "epoch": 0.32737228403331975, "grad_norm": 0.46307772397994995, "learning_rate": 9.493406323494535e-05, "loss": 0.0135, "step": 5620 }, { "epoch": 0.32795479699423313, "grad_norm": 0.3684145510196686, "learning_rate": 9.490986208434413e-05, "loss": 0.0127, "step": 5630 }, { "epoch": 0.3285373099551465, "grad_norm": 0.3294019401073456, "learning_rate": 9.488560636402577e-05, "loss": 0.0167, "step": 5640 }, { "epoch": 0.3291198229160599, "grad_norm": 0.3000147342681885, "learning_rate": 9.486129610346321e-05, "loss": 0.0119, "step": 5650 }, { "epoch": 0.3297023358769733, "grad_norm": 0.23165956139564514, "learning_rate": 9.483693133219576e-05, "loss": 0.0113, "step": 5660 }, { "epoch": 0.33028484883788667, "grad_norm": 0.23021915555000305, "learning_rate": 9.481251207982888e-05, "loss": 0.0125, "step": 5670 }, { "epoch": 0.3308673617988, "grad_norm": 0.2385621815919876, "learning_rate": 9.47880383760343e-05, "loss": 0.0139, "step": 5680 }, { "epoch": 0.3314498747597134, "grad_norm": 0.2613489031791687, "learning_rate": 9.476351025054983e-05, "loss": 0.0164, "step": 5690 }, { "epoch": 0.33203238772062676, "grad_norm": 0.23035407066345215, "learning_rate": 9.473892773317952e-05, "loss": 0.0114, "step": 5700 }, { "epoch": 0.33261490068154015, "grad_norm": 0.2664507329463959, "learning_rate": 9.471429085379338e-05, "loss": 0.0183, "step": 5710 }, { "epoch": 0.33319741364245353, "grad_norm": 0.32637298107147217, "learning_rate": 9.468959964232757e-05, "loss": 0.018, "step": 5720 }, { "epoch": 0.3337799266033669, "grad_norm": 0.20291414856910706, "learning_rate": 9.466485412878425e-05, "loss": 0.0201, "step": 5730 }, { "epoch": 0.3343624395642803, "grad_norm": 0.4235072433948517, "learning_rate": 9.464005434323154e-05, "loss": 0.0132, "step": 5740 }, { "epoch": 0.3349449525251937, "grad_norm": 0.3145645260810852, "learning_rate": 9.461520031580352e-05, "loss": 0.013, "step": 5750 }, { "epoch": 0.33552746548610707, "grad_norm": 0.2606288492679596, "learning_rate": 9.459029207670019e-05, "loss": 0.0136, "step": 5760 }, { "epoch": 0.33610997844702045, "grad_norm": 0.3538146913051605, "learning_rate": 9.456532965618737e-05, "loss": 0.0135, "step": 5770 }, { "epoch": 0.33669249140793384, "grad_norm": 0.3229866921901703, "learning_rate": 9.454031308459681e-05, "loss": 0.0134, "step": 5780 }, { "epoch": 0.3372750043688472, "grad_norm": 0.28633013367652893, "learning_rate": 9.451524239232595e-05, "loss": 0.0148, "step": 5790 }, { "epoch": 0.3378575173297606, "grad_norm": 0.3700575828552246, "learning_rate": 9.449011760983809e-05, "loss": 0.0139, "step": 5800 }, { "epoch": 0.338440030290674, "grad_norm": 0.17904630303382874, "learning_rate": 9.446493876766218e-05, "loss": 0.0108, "step": 5810 }, { "epoch": 0.33902254325158737, "grad_norm": 0.4766163229942322, "learning_rate": 9.44397058963929e-05, "loss": 0.0118, "step": 5820 }, { "epoch": 0.33960505621250076, "grad_norm": 0.45538008213043213, "learning_rate": 9.441441902669056e-05, "loss": 0.0177, "step": 5830 }, { "epoch": 0.3401875691734141, "grad_norm": 0.4499160051345825, "learning_rate": 9.43890781892811e-05, "loss": 0.0163, "step": 5840 }, { "epoch": 0.34077008213432747, "grad_norm": 0.2667197287082672, "learning_rate": 9.436368341495603e-05, "loss": 0.0201, "step": 5850 }, { "epoch": 0.34135259509524085, "grad_norm": 0.2519628405570984, "learning_rate": 9.43382347345724e-05, "loss": 0.0139, "step": 5860 }, { "epoch": 0.34193510805615424, "grad_norm": 0.27849137783050537, "learning_rate": 9.431273217905273e-05, "loss": 0.0125, "step": 5870 }, { "epoch": 0.3425176210170676, "grad_norm": 0.3382633626461029, "learning_rate": 9.428717577938504e-05, "loss": 0.0184, "step": 5880 }, { "epoch": 0.343100133977981, "grad_norm": 0.32172083854675293, "learning_rate": 9.426156556662276e-05, "loss": 0.0118, "step": 5890 }, { "epoch": 0.3436826469388944, "grad_norm": 0.36967912316322327, "learning_rate": 9.423590157188474e-05, "loss": 0.0156, "step": 5900 }, { "epoch": 0.34426515989980777, "grad_norm": 0.33469265699386597, "learning_rate": 9.421018382635513e-05, "loss": 0.014, "step": 5910 }, { "epoch": 0.34484767286072115, "grad_norm": 0.30567115545272827, "learning_rate": 9.418441236128343e-05, "loss": 0.0149, "step": 5920 }, { "epoch": 0.34543018582163454, "grad_norm": 0.19063973426818848, "learning_rate": 9.41585872079844e-05, "loss": 0.0133, "step": 5930 }, { "epoch": 0.3460126987825479, "grad_norm": 0.4304380416870117, "learning_rate": 9.413270839783802e-05, "loss": 0.0125, "step": 5940 }, { "epoch": 0.3465952117434613, "grad_norm": 0.2692255973815918, "learning_rate": 9.41067759622895e-05, "loss": 0.0128, "step": 5950 }, { "epoch": 0.3471777247043747, "grad_norm": 0.2566480338573456, "learning_rate": 9.408078993284917e-05, "loss": 0.0128, "step": 5960 }, { "epoch": 0.3477602376652881, "grad_norm": 0.30677348375320435, "learning_rate": 9.405475034109254e-05, "loss": 0.0111, "step": 5970 }, { "epoch": 0.34834275062620146, "grad_norm": 0.31435564160346985, "learning_rate": 9.402865721866015e-05, "loss": 0.0129, "step": 5980 }, { "epoch": 0.3489252635871148, "grad_norm": 0.3343210816383362, "learning_rate": 9.400251059725762e-05, "loss": 0.0104, "step": 5990 }, { "epoch": 0.34950777654802817, "grad_norm": 0.3017050623893738, "learning_rate": 9.397631050865554e-05, "loss": 0.0133, "step": 6000 }, { "epoch": 0.35009028950894155, "grad_norm": 0.20941412448883057, "learning_rate": 9.395005698468949e-05, "loss": 0.0098, "step": 6010 }, { "epoch": 0.35067280246985494, "grad_norm": 0.29110249876976013, "learning_rate": 9.392375005726e-05, "loss": 0.0119, "step": 6020 }, { "epoch": 0.3512553154307683, "grad_norm": 0.38054490089416504, "learning_rate": 9.389738975833242e-05, "loss": 0.01, "step": 6030 }, { "epoch": 0.3518378283916817, "grad_norm": 0.24182026088237762, "learning_rate": 9.387097611993707e-05, "loss": 0.0146, "step": 6040 }, { "epoch": 0.3524203413525951, "grad_norm": 0.20831067860126495, "learning_rate": 9.384450917416894e-05, "loss": 0.0107, "step": 6050 }, { "epoch": 0.3530028543135085, "grad_norm": 0.17184917628765106, "learning_rate": 9.381798895318792e-05, "loss": 0.0117, "step": 6060 }, { "epoch": 0.35358536727442186, "grad_norm": 0.30182504653930664, "learning_rate": 9.379141548921853e-05, "loss": 0.0108, "step": 6070 }, { "epoch": 0.35416788023533524, "grad_norm": 0.2855520248413086, "learning_rate": 9.376478881455009e-05, "loss": 0.0151, "step": 6080 }, { "epoch": 0.3547503931962486, "grad_norm": 0.29148533940315247, "learning_rate": 9.373810896153647e-05, "loss": 0.013, "step": 6090 }, { "epoch": 0.355332906157162, "grad_norm": 0.3114771246910095, "learning_rate": 9.371137596259623e-05, "loss": 0.0101, "step": 6100 }, { "epoch": 0.3559154191180754, "grad_norm": 0.2501264214515686, "learning_rate": 9.368458985021248e-05, "loss": 0.0101, "step": 6110 }, { "epoch": 0.3564979320789888, "grad_norm": 0.2828197479248047, "learning_rate": 9.365775065693287e-05, "loss": 0.0147, "step": 6120 }, { "epoch": 0.35708044503990216, "grad_norm": 0.26596835255622864, "learning_rate": 9.363085841536957e-05, "loss": 0.0135, "step": 6130 }, { "epoch": 0.35766295800081555, "grad_norm": 0.20203420519828796, "learning_rate": 9.360391315819917e-05, "loss": 0.0101, "step": 6140 }, { "epoch": 0.3582454709617289, "grad_norm": 0.30759403109550476, "learning_rate": 9.357691491816269e-05, "loss": 0.0102, "step": 6150 }, { "epoch": 0.35882798392264226, "grad_norm": 0.22800108790397644, "learning_rate": 9.354986372806557e-05, "loss": 0.0137, "step": 6160 }, { "epoch": 0.35941049688355564, "grad_norm": 0.21857386827468872, "learning_rate": 9.352275962077752e-05, "loss": 0.0134, "step": 6170 }, { "epoch": 0.359993009844469, "grad_norm": 0.40877413749694824, "learning_rate": 9.349560262923262e-05, "loss": 0.0144, "step": 6180 }, { "epoch": 0.3605755228053824, "grad_norm": 0.45766687393188477, "learning_rate": 9.346839278642913e-05, "loss": 0.0147, "step": 6190 }, { "epoch": 0.3611580357662958, "grad_norm": 0.31452977657318115, "learning_rate": 9.344113012542963e-05, "loss": 0.0129, "step": 6200 }, { "epoch": 0.3617405487272092, "grad_norm": 0.2539997696876526, "learning_rate": 9.341381467936079e-05, "loss": 0.0115, "step": 6210 }, { "epoch": 0.36232306168812256, "grad_norm": 0.26139014959335327, "learning_rate": 9.338644648141346e-05, "loss": 0.0116, "step": 6220 }, { "epoch": 0.36290557464903594, "grad_norm": 0.1639491766691208, "learning_rate": 9.335902556484257e-05, "loss": 0.0134, "step": 6230 }, { "epoch": 0.36348808760994933, "grad_norm": 0.2698086202144623, "learning_rate": 9.333155196296712e-05, "loss": 0.0119, "step": 6240 }, { "epoch": 0.3640706005708627, "grad_norm": 0.38294240832328796, "learning_rate": 9.330402570917016e-05, "loss": 0.0126, "step": 6250 }, { "epoch": 0.3646531135317761, "grad_norm": 0.36223292350769043, "learning_rate": 9.327644683689865e-05, "loss": 0.0138, "step": 6260 }, { "epoch": 0.3652356264926895, "grad_norm": 0.30007612705230713, "learning_rate": 9.324881537966354e-05, "loss": 0.0124, "step": 6270 }, { "epoch": 0.36581813945360286, "grad_norm": 0.23643670976161957, "learning_rate": 9.322113137103964e-05, "loss": 0.0119, "step": 6280 }, { "epoch": 0.36640065241451625, "grad_norm": 0.24124610424041748, "learning_rate": 9.319339484466564e-05, "loss": 0.0144, "step": 6290 }, { "epoch": 0.3669831653754296, "grad_norm": 0.2288600355386734, "learning_rate": 9.316560583424403e-05, "loss": 0.0109, "step": 6300 }, { "epoch": 0.36756567833634296, "grad_norm": 0.26642534136772156, "learning_rate": 9.313776437354108e-05, "loss": 0.0138, "step": 6310 }, { "epoch": 0.36814819129725634, "grad_norm": 0.26386332511901855, "learning_rate": 9.31098704963868e-05, "loss": 0.0106, "step": 6320 }, { "epoch": 0.36873070425816973, "grad_norm": 0.29608970880508423, "learning_rate": 9.308192423667485e-05, "loss": 0.0101, "step": 6330 }, { "epoch": 0.3693132172190831, "grad_norm": 0.3042888939380646, "learning_rate": 9.305392562836262e-05, "loss": 0.0114, "step": 6340 }, { "epoch": 0.3698957301799965, "grad_norm": 0.30546969175338745, "learning_rate": 9.3025874705471e-05, "loss": 0.0114, "step": 6350 }, { "epoch": 0.3704782431409099, "grad_norm": 0.3741450607776642, "learning_rate": 9.299777150208457e-05, "loss": 0.0133, "step": 6360 }, { "epoch": 0.37106075610182326, "grad_norm": 0.3162546157836914, "learning_rate": 9.296961605235133e-05, "loss": 0.0138, "step": 6370 }, { "epoch": 0.37164326906273665, "grad_norm": 0.3369394838809967, "learning_rate": 9.29414083904828e-05, "loss": 0.0128, "step": 6380 }, { "epoch": 0.37222578202365003, "grad_norm": 0.2751036286354065, "learning_rate": 9.291314855075397e-05, "loss": 0.0126, "step": 6390 }, { "epoch": 0.3728082949845634, "grad_norm": 0.19377347826957703, "learning_rate": 9.288483656750322e-05, "loss": 0.0118, "step": 6400 }, { "epoch": 0.3733908079454768, "grad_norm": 0.31796208024024963, "learning_rate": 9.285647247513225e-05, "loss": 0.0115, "step": 6410 }, { "epoch": 0.3739733209063902, "grad_norm": 0.1682463437318802, "learning_rate": 9.282805630810614e-05, "loss": 0.0124, "step": 6420 }, { "epoch": 0.37455583386730357, "grad_norm": 0.3048756718635559, "learning_rate": 9.279958810095317e-05, "loss": 0.0088, "step": 6430 }, { "epoch": 0.37513834682821695, "grad_norm": 0.2317032665014267, "learning_rate": 9.277106788826494e-05, "loss": 0.0123, "step": 6440 }, { "epoch": 0.37572085978913033, "grad_norm": 0.19936993718147278, "learning_rate": 9.274249570469618e-05, "loss": 0.0089, "step": 6450 }, { "epoch": 0.37630337275004366, "grad_norm": 0.23571209609508514, "learning_rate": 9.271387158496476e-05, "loss": 0.0087, "step": 6460 }, { "epoch": 0.37688588571095705, "grad_norm": 0.30482497811317444, "learning_rate": 9.268519556385173e-05, "loss": 0.0117, "step": 6470 }, { "epoch": 0.37746839867187043, "grad_norm": 0.3876536190509796, "learning_rate": 9.265646767620113e-05, "loss": 0.0101, "step": 6480 }, { "epoch": 0.3780509116327838, "grad_norm": 0.37219473719596863, "learning_rate": 9.262768795692006e-05, "loss": 0.0119, "step": 6490 }, { "epoch": 0.3786334245936972, "grad_norm": 0.35567301511764526, "learning_rate": 9.259885644097861e-05, "loss": 0.0111, "step": 6500 }, { "epoch": 0.3792159375546106, "grad_norm": 0.46365585923194885, "learning_rate": 9.256997316340976e-05, "loss": 0.0106, "step": 6510 }, { "epoch": 0.37979845051552397, "grad_norm": 0.3406113088130951, "learning_rate": 9.254103815930943e-05, "loss": 0.0145, "step": 6520 }, { "epoch": 0.38038096347643735, "grad_norm": 0.30493253469467163, "learning_rate": 9.251205146383636e-05, "loss": 0.014, "step": 6530 }, { "epoch": 0.38096347643735073, "grad_norm": 0.21430696547031403, "learning_rate": 9.248301311221216e-05, "loss": 0.0121, "step": 6540 }, { "epoch": 0.3815459893982641, "grad_norm": 0.3540270924568176, "learning_rate": 9.245392313972115e-05, "loss": 0.0128, "step": 6550 }, { "epoch": 0.3821285023591775, "grad_norm": 0.2746662199497223, "learning_rate": 9.242478158171037e-05, "loss": 0.0132, "step": 6560 }, { "epoch": 0.3827110153200909, "grad_norm": 0.3532452881336212, "learning_rate": 9.239558847358957e-05, "loss": 0.0148, "step": 6570 }, { "epoch": 0.38329352828100427, "grad_norm": 0.2984524667263031, "learning_rate": 9.236634385083114e-05, "loss": 0.0162, "step": 6580 }, { "epoch": 0.38387604124191765, "grad_norm": 0.3702338933944702, "learning_rate": 9.233704774897005e-05, "loss": 0.0126, "step": 6590 }, { "epoch": 0.38445855420283104, "grad_norm": 0.38487353920936584, "learning_rate": 9.230770020360382e-05, "loss": 0.0136, "step": 6600 }, { "epoch": 0.38504106716374437, "grad_norm": 0.43931037187576294, "learning_rate": 9.227830125039247e-05, "loss": 0.0134, "step": 6610 }, { "epoch": 0.38562358012465775, "grad_norm": 0.20062223076820374, "learning_rate": 9.224885092505853e-05, "loss": 0.0122, "step": 6620 }, { "epoch": 0.38620609308557113, "grad_norm": 0.40552452206611633, "learning_rate": 9.221934926338691e-05, "loss": 0.0165, "step": 6630 }, { "epoch": 0.3867886060464845, "grad_norm": 0.41163545846939087, "learning_rate": 9.21897963012249e-05, "loss": 0.0118, "step": 6640 }, { "epoch": 0.3873711190073979, "grad_norm": 0.22466735541820526, "learning_rate": 9.216019207448217e-05, "loss": 0.0106, "step": 6650 }, { "epoch": 0.3879536319683113, "grad_norm": 0.23277828097343445, "learning_rate": 9.21305366191306e-05, "loss": 0.0109, "step": 6660 }, { "epoch": 0.38853614492922467, "grad_norm": 0.33124229311943054, "learning_rate": 9.21008299712044e-05, "loss": 0.0117, "step": 6670 }, { "epoch": 0.38911865789013805, "grad_norm": 0.2248697578907013, "learning_rate": 9.207107216679993e-05, "loss": 0.0125, "step": 6680 }, { "epoch": 0.38970117085105144, "grad_norm": 0.31560787558555603, "learning_rate": 9.204126324207574e-05, "loss": 0.0123, "step": 6690 }, { "epoch": 0.3902836838119648, "grad_norm": 0.37756285071372986, "learning_rate": 9.201140323325248e-05, "loss": 0.0137, "step": 6700 }, { "epoch": 0.3908661967728782, "grad_norm": 0.3497389554977417, "learning_rate": 9.198149217661287e-05, "loss": 0.0128, "step": 6710 }, { "epoch": 0.3914487097337916, "grad_norm": 0.3866865634918213, "learning_rate": 9.195153010850167e-05, "loss": 0.0131, "step": 6720 }, { "epoch": 0.392031222694705, "grad_norm": 0.5670121908187866, "learning_rate": 9.192151706532562e-05, "loss": 0.0152, "step": 6730 }, { "epoch": 0.39261373565561836, "grad_norm": 0.5314503312110901, "learning_rate": 9.189145308355337e-05, "loss": 0.0169, "step": 6740 }, { "epoch": 0.39319624861653174, "grad_norm": 0.4168036878108978, "learning_rate": 9.186133819971555e-05, "loss": 0.013, "step": 6750 }, { "epoch": 0.3937787615774451, "grad_norm": 0.36590367555618286, "learning_rate": 9.183117245040455e-05, "loss": 0.0121, "step": 6760 }, { "epoch": 0.39436127453835845, "grad_norm": 0.29301807284355164, "learning_rate": 9.180095587227458e-05, "loss": 0.0108, "step": 6770 }, { "epoch": 0.39494378749927184, "grad_norm": 0.45780086517333984, "learning_rate": 9.177068850204167e-05, "loss": 0.0124, "step": 6780 }, { "epoch": 0.3955263004601852, "grad_norm": 0.2961581349372864, "learning_rate": 9.174037037648351e-05, "loss": 0.0117, "step": 6790 }, { "epoch": 0.3961088134210986, "grad_norm": 0.27854636311531067, "learning_rate": 9.171000153243947e-05, "loss": 0.016, "step": 6800 }, { "epoch": 0.396691326382012, "grad_norm": 0.2911928594112396, "learning_rate": 9.167958200681057e-05, "loss": 0.0095, "step": 6810 }, { "epoch": 0.3972738393429254, "grad_norm": 0.28199344873428345, "learning_rate": 9.164911183655943e-05, "loss": 0.0112, "step": 6820 }, { "epoch": 0.39785635230383876, "grad_norm": 0.23138020932674408, "learning_rate": 9.161859105871013e-05, "loss": 0.0099, "step": 6830 }, { "epoch": 0.39843886526475214, "grad_norm": 0.2517688572406769, "learning_rate": 9.158801971034832e-05, "loss": 0.0112, "step": 6840 }, { "epoch": 0.3990213782256655, "grad_norm": 0.2718874514102936, "learning_rate": 9.155739782862107e-05, "loss": 0.0084, "step": 6850 }, { "epoch": 0.3996038911865789, "grad_norm": 0.42391732335090637, "learning_rate": 9.152672545073687e-05, "loss": 0.0142, "step": 6860 }, { "epoch": 0.4001864041474923, "grad_norm": 0.23340462148189545, "learning_rate": 9.149600261396551e-05, "loss": 0.0129, "step": 6870 }, { "epoch": 0.4007689171084057, "grad_norm": 0.24849651753902435, "learning_rate": 9.146522935563817e-05, "loss": 0.0165, "step": 6880 }, { "epoch": 0.40135143006931906, "grad_norm": 0.29082268476486206, "learning_rate": 9.143440571314723e-05, "loss": 0.013, "step": 6890 }, { "epoch": 0.40193394303023244, "grad_norm": 0.3280594050884247, "learning_rate": 9.140353172394636e-05, "loss": 0.0112, "step": 6900 }, { "epoch": 0.4025164559911458, "grad_norm": 0.44642722606658936, "learning_rate": 9.137260742555032e-05, "loss": 0.0145, "step": 6910 }, { "epoch": 0.40309896895205916, "grad_norm": 0.2751993238925934, "learning_rate": 9.13416328555351e-05, "loss": 0.0174, "step": 6920 }, { "epoch": 0.40368148191297254, "grad_norm": 0.2839924693107605, "learning_rate": 9.131060805153769e-05, "loss": 0.011, "step": 6930 }, { "epoch": 0.4042639948738859, "grad_norm": 0.26174360513687134, "learning_rate": 9.127953305125618e-05, "loss": 0.011, "step": 6940 }, { "epoch": 0.4048465078347993, "grad_norm": 0.22545038163661957, "learning_rate": 9.124840789244958e-05, "loss": 0.0166, "step": 6950 }, { "epoch": 0.4054290207957127, "grad_norm": 0.22442230582237244, "learning_rate": 9.121723261293794e-05, "loss": 0.0121, "step": 6960 }, { "epoch": 0.4060115337566261, "grad_norm": 0.3519812524318695, "learning_rate": 9.118600725060214e-05, "loss": 0.0138, "step": 6970 }, { "epoch": 0.40659404671753946, "grad_norm": 0.2668507397174835, "learning_rate": 9.115473184338392e-05, "loss": 0.0107, "step": 6980 }, { "epoch": 0.40717655967845284, "grad_norm": 0.284104585647583, "learning_rate": 9.11234064292859e-05, "loss": 0.0135, "step": 6990 }, { "epoch": 0.4077590726393662, "grad_norm": 0.3436839282512665, "learning_rate": 9.109203104637137e-05, "loss": 0.0126, "step": 7000 }, { "epoch": 0.4083415856002796, "grad_norm": 0.2720262408256531, "learning_rate": 9.10606057327644e-05, "loss": 0.0122, "step": 7010 }, { "epoch": 0.408924098561193, "grad_norm": 0.21050935983657837, "learning_rate": 9.10291305266497e-05, "loss": 0.0107, "step": 7020 }, { "epoch": 0.4095066115221064, "grad_norm": 0.33217161893844604, "learning_rate": 9.099760546627261e-05, "loss": 0.0137, "step": 7030 }, { "epoch": 0.41008912448301976, "grad_norm": 0.18960629403591156, "learning_rate": 9.096603058993907e-05, "loss": 0.0144, "step": 7040 }, { "epoch": 0.41067163744393315, "grad_norm": 0.5246853828430176, "learning_rate": 9.093440593601552e-05, "loss": 0.0174, "step": 7050 }, { "epoch": 0.41125415040484653, "grad_norm": 0.27294328808784485, "learning_rate": 9.090273154292889e-05, "loss": 0.0124, "step": 7060 }, { "epoch": 0.4118366633657599, "grad_norm": 0.30159419775009155, "learning_rate": 9.087100744916656e-05, "loss": 0.0133, "step": 7070 }, { "epoch": 0.41241917632667324, "grad_norm": 0.21936914324760437, "learning_rate": 9.08392336932763e-05, "loss": 0.0097, "step": 7080 }, { "epoch": 0.4130016892875866, "grad_norm": 0.37547966837882996, "learning_rate": 9.080741031386619e-05, "loss": 0.012, "step": 7090 }, { "epoch": 0.4135842022485, "grad_norm": 0.4171476662158966, "learning_rate": 9.077553734960468e-05, "loss": 0.012, "step": 7100 }, { "epoch": 0.4141667152094134, "grad_norm": 0.2655220925807953, "learning_rate": 9.074361483922041e-05, "loss": 0.0113, "step": 7110 }, { "epoch": 0.4147492281703268, "grad_norm": 0.3095017671585083, "learning_rate": 9.071164282150224e-05, "loss": 0.0126, "step": 7120 }, { "epoch": 0.41533174113124016, "grad_norm": 0.3188718259334564, "learning_rate": 9.067962133529919e-05, "loss": 0.0122, "step": 7130 }, { "epoch": 0.41591425409215355, "grad_norm": 0.36818554997444153, "learning_rate": 9.064755041952036e-05, "loss": 0.0157, "step": 7140 }, { "epoch": 0.41649676705306693, "grad_norm": 0.19709156453609467, "learning_rate": 9.061543011313497e-05, "loss": 0.015, "step": 7150 }, { "epoch": 0.4170792800139803, "grad_norm": 0.2513802647590637, "learning_rate": 9.05832604551722e-05, "loss": 0.0156, "step": 7160 }, { "epoch": 0.4176617929748937, "grad_norm": 0.25856226682662964, "learning_rate": 9.055104148472124e-05, "loss": 0.0117, "step": 7170 }, { "epoch": 0.4182443059358071, "grad_norm": 0.21729667484760284, "learning_rate": 9.051877324093114e-05, "loss": 0.0123, "step": 7180 }, { "epoch": 0.41882681889672047, "grad_norm": 0.2332047075033188, "learning_rate": 9.04864557630109e-05, "loss": 0.0144, "step": 7190 }, { "epoch": 0.41940933185763385, "grad_norm": 0.24884118139743805, "learning_rate": 9.045408909022928e-05, "loss": 0.014, "step": 7200 }, { "epoch": 0.41999184481854723, "grad_norm": 0.39284127950668335, "learning_rate": 9.042167326191484e-05, "loss": 0.0098, "step": 7210 }, { "epoch": 0.4205743577794606, "grad_norm": 0.29154670238494873, "learning_rate": 9.038920831745587e-05, "loss": 0.0142, "step": 7220 }, { "epoch": 0.421156870740374, "grad_norm": 0.23585759103298187, "learning_rate": 9.035669429630036e-05, "loss": 0.0113, "step": 7230 }, { "epoch": 0.42173938370128733, "grad_norm": 0.1617414355278015, "learning_rate": 9.032413123795588e-05, "loss": 0.0103, "step": 7240 }, { "epoch": 0.4223218966622007, "grad_norm": 0.3884938359260559, "learning_rate": 9.029151918198962e-05, "loss": 0.0115, "step": 7250 }, { "epoch": 0.4229044096231141, "grad_norm": 0.2697516083717346, "learning_rate": 9.025885816802833e-05, "loss": 0.0117, "step": 7260 }, { "epoch": 0.4234869225840275, "grad_norm": 0.3080799877643585, "learning_rate": 9.02261482357582e-05, "loss": 0.0121, "step": 7270 }, { "epoch": 0.42406943554494086, "grad_norm": 0.2844873070716858, "learning_rate": 9.019338942492485e-05, "loss": 0.0112, "step": 7280 }, { "epoch": 0.42465194850585425, "grad_norm": 0.44505390524864197, "learning_rate": 9.01605817753334e-05, "loss": 0.0124, "step": 7290 }, { "epoch": 0.42523446146676763, "grad_norm": 0.2775717079639435, "learning_rate": 9.012772532684818e-05, "loss": 0.0114, "step": 7300 }, { "epoch": 0.425816974427681, "grad_norm": 0.2966673970222473, "learning_rate": 9.00948201193929e-05, "loss": 0.0145, "step": 7310 }, { "epoch": 0.4263994873885944, "grad_norm": 0.2556905746459961, "learning_rate": 9.006186619295047e-05, "loss": 0.012, "step": 7320 }, { "epoch": 0.4269820003495078, "grad_norm": 0.24772870540618896, "learning_rate": 9.002886358756305e-05, "loss": 0.0134, "step": 7330 }, { "epoch": 0.42756451331042117, "grad_norm": 0.29119929671287537, "learning_rate": 8.999581234333188e-05, "loss": 0.0111, "step": 7340 }, { "epoch": 0.42814702627133455, "grad_norm": 0.26602277159690857, "learning_rate": 8.996271250041734e-05, "loss": 0.0101, "step": 7350 }, { "epoch": 0.42872953923224794, "grad_norm": 0.2989645302295685, "learning_rate": 8.99295640990389e-05, "loss": 0.0106, "step": 7360 }, { "epoch": 0.4293120521931613, "grad_norm": 0.2931496798992157, "learning_rate": 8.989636717947494e-05, "loss": 0.0114, "step": 7370 }, { "epoch": 0.4298945651540747, "grad_norm": 0.25160327553749084, "learning_rate": 8.986312178206291e-05, "loss": 0.0132, "step": 7380 }, { "epoch": 0.43047707811498803, "grad_norm": 0.19953225553035736, "learning_rate": 8.982982794719904e-05, "loss": 0.0138, "step": 7390 }, { "epoch": 0.4310595910759014, "grad_norm": 0.2953711152076721, "learning_rate": 8.979648571533852e-05, "loss": 0.0125, "step": 7400 }, { "epoch": 0.4316421040368148, "grad_norm": 0.4054790139198303, "learning_rate": 8.976309512699528e-05, "loss": 0.0125, "step": 7410 }, { "epoch": 0.4322246169977282, "grad_norm": 0.22115576267242432, "learning_rate": 8.972965622274205e-05, "loss": 0.0122, "step": 7420 }, { "epoch": 0.43280712995864157, "grad_norm": 0.33209818601608276, "learning_rate": 8.969616904321026e-05, "loss": 0.0117, "step": 7430 }, { "epoch": 0.43338964291955495, "grad_norm": 0.22817012667655945, "learning_rate": 8.966263362908998e-05, "loss": 0.013, "step": 7440 }, { "epoch": 0.43397215588046834, "grad_norm": 0.3235332667827606, "learning_rate": 8.962905002112989e-05, "loss": 0.0096, "step": 7450 }, { "epoch": 0.4345546688413817, "grad_norm": 0.1812901496887207, "learning_rate": 8.959541826013724e-05, "loss": 0.0131, "step": 7460 }, { "epoch": 0.4351371818022951, "grad_norm": 0.388325959444046, "learning_rate": 8.956173838697781e-05, "loss": 0.0116, "step": 7470 }, { "epoch": 0.4357196947632085, "grad_norm": 0.35594281554222107, "learning_rate": 8.952801044257581e-05, "loss": 0.0121, "step": 7480 }, { "epoch": 0.43630220772412187, "grad_norm": 0.289470911026001, "learning_rate": 8.949423446791387e-05, "loss": 0.0134, "step": 7490 }, { "epoch": 0.43688472068503525, "grad_norm": 0.3230888545513153, "learning_rate": 8.9460410504033e-05, "loss": 0.0115, "step": 7500 }, { "epoch": 0.43746723364594864, "grad_norm": 0.20693032443523407, "learning_rate": 8.942653859203248e-05, "loss": 0.0103, "step": 7510 }, { "epoch": 0.438049746606862, "grad_norm": 0.17195037007331848, "learning_rate": 8.93926187730699e-05, "loss": 0.0134, "step": 7520 }, { "epoch": 0.4386322595677754, "grad_norm": 0.1913907527923584, "learning_rate": 8.935865108836102e-05, "loss": 0.0118, "step": 7530 }, { "epoch": 0.4392147725286888, "grad_norm": 0.23225539922714233, "learning_rate": 8.932463557917981e-05, "loss": 0.0101, "step": 7540 }, { "epoch": 0.4397972854896021, "grad_norm": 0.2920669913291931, "learning_rate": 8.929057228685829e-05, "loss": 0.0099, "step": 7550 }, { "epoch": 0.4403797984505155, "grad_norm": 0.34749093651771545, "learning_rate": 8.925646125278658e-05, "loss": 0.0116, "step": 7560 }, { "epoch": 0.4409623114114289, "grad_norm": 0.3100294768810272, "learning_rate": 8.92223025184128e-05, "loss": 0.0112, "step": 7570 }, { "epoch": 0.44154482437234227, "grad_norm": 0.3635091185569763, "learning_rate": 8.918809612524303e-05, "loss": 0.0147, "step": 7580 }, { "epoch": 0.44212733733325565, "grad_norm": 0.2399435192346573, "learning_rate": 8.91538421148413e-05, "loss": 0.0112, "step": 7590 }, { "epoch": 0.44270985029416904, "grad_norm": 0.25012388825416565, "learning_rate": 8.91195405288294e-05, "loss": 0.0128, "step": 7600 }, { "epoch": 0.4432923632550824, "grad_norm": 0.48543089628219604, "learning_rate": 8.908519140888705e-05, "loss": 0.0127, "step": 7610 }, { "epoch": 0.4438748762159958, "grad_norm": 0.21692843735218048, "learning_rate": 8.905079479675164e-05, "loss": 0.01, "step": 7620 }, { "epoch": 0.4444573891769092, "grad_norm": 0.23970955610275269, "learning_rate": 8.901635073421831e-05, "loss": 0.0109, "step": 7630 }, { "epoch": 0.4450399021378226, "grad_norm": 0.20894008874893188, "learning_rate": 8.898185926313983e-05, "loss": 0.0117, "step": 7640 }, { "epoch": 0.44562241509873596, "grad_norm": 0.2764657735824585, "learning_rate": 8.894732042542659e-05, "loss": 0.0141, "step": 7650 }, { "epoch": 0.44620492805964934, "grad_norm": 0.31210198998451233, "learning_rate": 8.891273426304654e-05, "loss": 0.0114, "step": 7660 }, { "epoch": 0.4467874410205627, "grad_norm": 0.3830687701702118, "learning_rate": 8.887810081802515e-05, "loss": 0.0112, "step": 7670 }, { "epoch": 0.4473699539814761, "grad_norm": 0.3342027962207794, "learning_rate": 8.884342013244529e-05, "loss": 0.0112, "step": 7680 }, { "epoch": 0.4479524669423895, "grad_norm": 0.27329325675964355, "learning_rate": 8.880869224844725e-05, "loss": 0.0099, "step": 7690 }, { "epoch": 0.4485349799033028, "grad_norm": 0.35470640659332275, "learning_rate": 8.877391720822875e-05, "loss": 0.0102, "step": 7700 }, { "epoch": 0.4491174928642162, "grad_norm": 0.3574216663837433, "learning_rate": 8.873909505404468e-05, "loss": 0.0119, "step": 7710 }, { "epoch": 0.4497000058251296, "grad_norm": 0.3599697947502136, "learning_rate": 8.870422582820726e-05, "loss": 0.0092, "step": 7720 }, { "epoch": 0.450282518786043, "grad_norm": 0.2701171338558197, "learning_rate": 8.866930957308588e-05, "loss": 0.0138, "step": 7730 }, { "epoch": 0.45086503174695636, "grad_norm": 0.31447646021842957, "learning_rate": 8.86343463311071e-05, "loss": 0.0102, "step": 7740 }, { "epoch": 0.45144754470786974, "grad_norm": 0.30923134088516235, "learning_rate": 8.859933614475452e-05, "loss": 0.0119, "step": 7750 }, { "epoch": 0.4520300576687831, "grad_norm": 0.1891762614250183, "learning_rate": 8.85642790565689e-05, "loss": 0.0103, "step": 7760 }, { "epoch": 0.4526125706296965, "grad_norm": 0.2857305705547333, "learning_rate": 8.852917510914782e-05, "loss": 0.0096, "step": 7770 }, { "epoch": 0.4531950835906099, "grad_norm": 0.2828695774078369, "learning_rate": 8.84940243451459e-05, "loss": 0.0129, "step": 7780 }, { "epoch": 0.4537775965515233, "grad_norm": 0.24774837493896484, "learning_rate": 8.845882680727469e-05, "loss": 0.0102, "step": 7790 }, { "epoch": 0.45436010951243666, "grad_norm": 0.21391414105892181, "learning_rate": 8.842358253830246e-05, "loss": 0.0087, "step": 7800 }, { "epoch": 0.45494262247335004, "grad_norm": 0.23724718391895294, "learning_rate": 8.838829158105434e-05, "loss": 0.0104, "step": 7810 }, { "epoch": 0.45552513543426343, "grad_norm": 0.26340359449386597, "learning_rate": 8.835295397841217e-05, "loss": 0.0142, "step": 7820 }, { "epoch": 0.4561076483951768, "grad_norm": 0.32494908571243286, "learning_rate": 8.831756977331447e-05, "loss": 0.0152, "step": 7830 }, { "epoch": 0.4566901613560902, "grad_norm": 0.42112550139427185, "learning_rate": 8.828213900875638e-05, "loss": 0.0124, "step": 7840 }, { "epoch": 0.4572726743170036, "grad_norm": 0.3547781705856323, "learning_rate": 8.824666172778964e-05, "loss": 0.0107, "step": 7850 }, { "epoch": 0.4578551872779169, "grad_norm": 0.32153594493865967, "learning_rate": 8.821113797352245e-05, "loss": 0.0091, "step": 7860 }, { "epoch": 0.4584377002388303, "grad_norm": 0.25656503438949585, "learning_rate": 8.817556778911956e-05, "loss": 0.0109, "step": 7870 }, { "epoch": 0.4590202131997437, "grad_norm": 0.2758748233318329, "learning_rate": 8.81399512178021e-05, "loss": 0.0119, "step": 7880 }, { "epoch": 0.45960272616065706, "grad_norm": 0.25740793347358704, "learning_rate": 8.810428830284752e-05, "loss": 0.0116, "step": 7890 }, { "epoch": 0.46018523912157044, "grad_norm": 0.23661264777183533, "learning_rate": 8.806857908758967e-05, "loss": 0.0109, "step": 7900 }, { "epoch": 0.46076775208248383, "grad_norm": 0.33345550298690796, "learning_rate": 8.80328236154186e-05, "loss": 0.0129, "step": 7910 }, { "epoch": 0.4613502650433972, "grad_norm": 0.15988755226135254, "learning_rate": 8.799702192978056e-05, "loss": 0.0085, "step": 7920 }, { "epoch": 0.4619327780043106, "grad_norm": 0.2677052319049835, "learning_rate": 8.7961174074178e-05, "loss": 0.0112, "step": 7930 }, { "epoch": 0.462515290965224, "grad_norm": 0.16391333937644958, "learning_rate": 8.792528009216942e-05, "loss": 0.0108, "step": 7940 }, { "epoch": 0.46309780392613736, "grad_norm": 0.3000504672527313, "learning_rate": 8.788934002736943e-05, "loss": 0.0159, "step": 7950 }, { "epoch": 0.46368031688705075, "grad_norm": 0.2887212038040161, "learning_rate": 8.785335392344857e-05, "loss": 0.0157, "step": 7960 }, { "epoch": 0.46426282984796413, "grad_norm": 0.34124699234962463, "learning_rate": 8.781732182413336e-05, "loss": 0.0132, "step": 7970 }, { "epoch": 0.4648453428088775, "grad_norm": 0.3974616229534149, "learning_rate": 8.778124377320618e-05, "loss": 0.0195, "step": 7980 }, { "epoch": 0.4654278557697909, "grad_norm": 0.3733963668346405, "learning_rate": 8.774511981450529e-05, "loss": 0.0132, "step": 7990 }, { "epoch": 0.4660103687307043, "grad_norm": 0.29485955834388733, "learning_rate": 8.770894999192467e-05, "loss": 0.0139, "step": 8000 }, { "epoch": 0.4665928816916176, "grad_norm": 0.3780854642391205, "learning_rate": 8.767273434941412e-05, "loss": 0.0154, "step": 8010 }, { "epoch": 0.467175394652531, "grad_norm": 0.2440689653158188, "learning_rate": 8.7636472930979e-05, "loss": 0.0133, "step": 8020 }, { "epoch": 0.4677579076134444, "grad_norm": 0.14352597296237946, "learning_rate": 8.76001657806804e-05, "loss": 0.0122, "step": 8030 }, { "epoch": 0.46834042057435776, "grad_norm": 0.2142922729253769, "learning_rate": 8.756381294263491e-05, "loss": 0.0096, "step": 8040 }, { "epoch": 0.46892293353527115, "grad_norm": 0.3044984042644501, "learning_rate": 8.752741446101464e-05, "loss": 0.009, "step": 8050 }, { "epoch": 0.46950544649618453, "grad_norm": 0.34259000420570374, "learning_rate": 8.749097038004722e-05, "loss": 0.0127, "step": 8060 }, { "epoch": 0.4700879594570979, "grad_norm": 0.213138148188591, "learning_rate": 8.745448074401561e-05, "loss": 0.0098, "step": 8070 }, { "epoch": 0.4706704724180113, "grad_norm": 0.3469372093677521, "learning_rate": 8.741794559725818e-05, "loss": 0.0117, "step": 8080 }, { "epoch": 0.4712529853789247, "grad_norm": 0.2347586750984192, "learning_rate": 8.738136498416856e-05, "loss": 0.0087, "step": 8090 }, { "epoch": 0.47183549833983807, "grad_norm": 0.30541104078292847, "learning_rate": 8.734473894919564e-05, "loss": 0.0083, "step": 8100 }, { "epoch": 0.47241801130075145, "grad_norm": 0.227009117603302, "learning_rate": 8.730806753684353e-05, "loss": 0.0086, "step": 8110 }, { "epoch": 0.47300052426166483, "grad_norm": 0.12994255125522614, "learning_rate": 8.727135079167144e-05, "loss": 0.0101, "step": 8120 }, { "epoch": 0.4735830372225782, "grad_norm": 0.14903749525547028, "learning_rate": 8.723458875829368e-05, "loss": 0.0071, "step": 8130 }, { "epoch": 0.4741655501834916, "grad_norm": 0.2669014632701874, "learning_rate": 8.719778148137958e-05, "loss": 0.0118, "step": 8140 }, { "epoch": 0.474748063144405, "grad_norm": 0.2108674794435501, "learning_rate": 8.716092900565346e-05, "loss": 0.0107, "step": 8150 }, { "epoch": 0.47533057610531837, "grad_norm": 0.17570310831069946, "learning_rate": 8.712403137589456e-05, "loss": 0.0113, "step": 8160 }, { "epoch": 0.4759130890662317, "grad_norm": 0.23179100453853607, "learning_rate": 8.708708863693697e-05, "loss": 0.0103, "step": 8170 }, { "epoch": 0.4764956020271451, "grad_norm": 0.20041556656360626, "learning_rate": 8.70501008336696e-05, "loss": 0.0117, "step": 8180 }, { "epoch": 0.47707811498805847, "grad_norm": 0.3331192433834076, "learning_rate": 8.701306801103611e-05, "loss": 0.0125, "step": 8190 }, { "epoch": 0.47766062794897185, "grad_norm": 0.3115651309490204, "learning_rate": 8.697599021403489e-05, "loss": 0.0115, "step": 8200 }, { "epoch": 0.47824314090988523, "grad_norm": 0.18313908576965332, "learning_rate": 8.693886748771896e-05, "loss": 0.0118, "step": 8210 }, { "epoch": 0.4788256538707986, "grad_norm": 0.3065651059150696, "learning_rate": 8.690169987719593e-05, "loss": 0.0101, "step": 8220 }, { "epoch": 0.479408166831712, "grad_norm": 0.3847108781337738, "learning_rate": 8.686448742762791e-05, "loss": 0.0106, "step": 8230 }, { "epoch": 0.4799906797926254, "grad_norm": 0.27982595562934875, "learning_rate": 8.682723018423161e-05, "loss": 0.0084, "step": 8240 }, { "epoch": 0.48057319275353877, "grad_norm": 0.2973509728908539, "learning_rate": 8.678992819227804e-05, "loss": 0.0105, "step": 8250 }, { "epoch": 0.48115570571445215, "grad_norm": 0.22349682450294495, "learning_rate": 8.675258149709265e-05, "loss": 0.0096, "step": 8260 }, { "epoch": 0.48173821867536554, "grad_norm": 0.33618637919425964, "learning_rate": 8.67151901440552e-05, "loss": 0.0148, "step": 8270 }, { "epoch": 0.4823207316362789, "grad_norm": 0.43434813618659973, "learning_rate": 8.667775417859971e-05, "loss": 0.0145, "step": 8280 }, { "epoch": 0.4829032445971923, "grad_norm": 0.26474690437316895, "learning_rate": 8.664027364621441e-05, "loss": 0.0126, "step": 8290 }, { "epoch": 0.4834857575581057, "grad_norm": 0.22505629062652588, "learning_rate": 8.660274859244168e-05, "loss": 0.0091, "step": 8300 }, { "epoch": 0.4840682705190191, "grad_norm": 0.33556726574897766, "learning_rate": 8.656517906287799e-05, "loss": 0.0151, "step": 8310 }, { "epoch": 0.4846507834799324, "grad_norm": 0.3187960386276245, "learning_rate": 8.652756510317387e-05, "loss": 0.0121, "step": 8320 }, { "epoch": 0.4852332964408458, "grad_norm": 0.2117210030555725, "learning_rate": 8.648990675903382e-05, "loss": 0.0113, "step": 8330 }, { "epoch": 0.48581580940175917, "grad_norm": 0.3122064769268036, "learning_rate": 8.64522040762163e-05, "loss": 0.0121, "step": 8340 }, { "epoch": 0.48639832236267255, "grad_norm": 0.28426116704940796, "learning_rate": 8.64144571005336e-05, "loss": 0.0129, "step": 8350 }, { "epoch": 0.48698083532358594, "grad_norm": 0.3730778992176056, "learning_rate": 8.637666587785184e-05, "loss": 0.0127, "step": 8360 }, { "epoch": 0.4875633482844993, "grad_norm": 0.3090813457965851, "learning_rate": 8.633883045409096e-05, "loss": 0.0111, "step": 8370 }, { "epoch": 0.4881458612454127, "grad_norm": 0.23262611031532288, "learning_rate": 8.630095087522457e-05, "loss": 0.0102, "step": 8380 }, { "epoch": 0.4887283742063261, "grad_norm": 0.23348644375801086, "learning_rate": 8.62630271872799e-05, "loss": 0.0112, "step": 8390 }, { "epoch": 0.48931088716723947, "grad_norm": 0.23587965965270996, "learning_rate": 8.62250594363378e-05, "loss": 0.008, "step": 8400 }, { "epoch": 0.48989340012815286, "grad_norm": 0.37986883521080017, "learning_rate": 8.61870476685327e-05, "loss": 0.0109, "step": 8410 }, { "epoch": 0.49047591308906624, "grad_norm": 0.4875761866569519, "learning_rate": 8.614899193005248e-05, "loss": 0.0115, "step": 8420 }, { "epoch": 0.4910584260499796, "grad_norm": 0.2998208999633789, "learning_rate": 8.611089226713842e-05, "loss": 0.01, "step": 8430 }, { "epoch": 0.491640939010893, "grad_norm": 0.20100432634353638, "learning_rate": 8.607274872608521e-05, "loss": 0.0099, "step": 8440 }, { "epoch": 0.4922234519718064, "grad_norm": 0.3055593967437744, "learning_rate": 8.603456135324088e-05, "loss": 0.0096, "step": 8450 }, { "epoch": 0.4928059649327198, "grad_norm": 0.22969605028629303, "learning_rate": 8.599633019500665e-05, "loss": 0.009, "step": 8460 }, { "epoch": 0.49338847789363316, "grad_norm": 0.2377668172121048, "learning_rate": 8.595805529783702e-05, "loss": 0.0095, "step": 8470 }, { "epoch": 0.4939709908545465, "grad_norm": 0.3378452956676483, "learning_rate": 8.591973670823958e-05, "loss": 0.0145, "step": 8480 }, { "epoch": 0.49455350381545987, "grad_norm": 0.33285266160964966, "learning_rate": 8.588137447277503e-05, "loss": 0.0125, "step": 8490 }, { "epoch": 0.49513601677637326, "grad_norm": 0.32284048199653625, "learning_rate": 8.584296863805709e-05, "loss": 0.009, "step": 8500 }, { "epoch": 0.49571852973728664, "grad_norm": 0.3326222002506256, "learning_rate": 8.580451925075248e-05, "loss": 0.0119, "step": 8510 }, { "epoch": 0.4963010426982, "grad_norm": 0.21906715631484985, "learning_rate": 8.576602635758085e-05, "loss": 0.0102, "step": 8520 }, { "epoch": 0.4968835556591134, "grad_norm": 0.20839254558086395, "learning_rate": 8.572749000531469e-05, "loss": 0.0086, "step": 8530 }, { "epoch": 0.4974660686200268, "grad_norm": 0.34974223375320435, "learning_rate": 8.568891024077926e-05, "loss": 0.0096, "step": 8540 }, { "epoch": 0.4980485815809402, "grad_norm": 0.35410234332084656, "learning_rate": 8.565028711085265e-05, "loss": 0.0122, "step": 8550 }, { "epoch": 0.49863109454185356, "grad_norm": 0.33047378063201904, "learning_rate": 8.561162066246561e-05, "loss": 0.0107, "step": 8560 }, { "epoch": 0.49921360750276694, "grad_norm": 0.3621072471141815, "learning_rate": 8.557291094260152e-05, "loss": 0.0127, "step": 8570 }, { "epoch": 0.4997961204636803, "grad_norm": 0.26117265224456787, "learning_rate": 8.553415799829629e-05, "loss": 0.0107, "step": 8580 }, { "epoch": 0.5003786334245937, "grad_norm": 0.16914372146129608, "learning_rate": 8.549536187663847e-05, "loss": 0.0084, "step": 8590 }, { "epoch": 0.500961146385507, "grad_norm": 0.2097463309764862, "learning_rate": 8.545652262476897e-05, "loss": 0.013, "step": 8600 }, { "epoch": 0.5015436593464204, "grad_norm": 0.2855682671070099, "learning_rate": 8.541764028988115e-05, "loss": 0.0113, "step": 8610 }, { "epoch": 0.5021261723073338, "grad_norm": 0.21892935037612915, "learning_rate": 8.537871491922071e-05, "loss": 0.0117, "step": 8620 }, { "epoch": 0.5027086852682472, "grad_norm": 0.26727095246315, "learning_rate": 8.533974656008566e-05, "loss": 0.0112, "step": 8630 }, { "epoch": 0.5032911982291606, "grad_norm": 0.23189052939414978, "learning_rate": 8.530073525982621e-05, "loss": 0.009, "step": 8640 }, { "epoch": 0.503873711190074, "grad_norm": 0.20634156465530396, "learning_rate": 8.526168106584476e-05, "loss": 0.0129, "step": 8650 }, { "epoch": 0.5044562241509873, "grad_norm": 0.3063119947910309, "learning_rate": 8.522258402559586e-05, "loss": 0.0116, "step": 8660 }, { "epoch": 0.5050387371119007, "grad_norm": 0.3161621391773224, "learning_rate": 8.518344418658613e-05, "loss": 0.0114, "step": 8670 }, { "epoch": 0.5056212500728141, "grad_norm": 0.23569059371948242, "learning_rate": 8.51442615963741e-05, "loss": 0.0133, "step": 8680 }, { "epoch": 0.5062037630337275, "grad_norm": 0.24308477342128754, "learning_rate": 8.510503630257034e-05, "loss": 0.0104, "step": 8690 }, { "epoch": 0.5067862759946409, "grad_norm": 0.303740531206131, "learning_rate": 8.50657683528373e-05, "loss": 0.0096, "step": 8700 }, { "epoch": 0.5073687889555543, "grad_norm": 0.22782230377197266, "learning_rate": 8.502645779488922e-05, "loss": 0.0086, "step": 8710 }, { "epoch": 0.5079513019164676, "grad_norm": 0.3413344919681549, "learning_rate": 8.498710467649213e-05, "loss": 0.01, "step": 8720 }, { "epoch": 0.508533814877381, "grad_norm": 0.2716794013977051, "learning_rate": 8.494770904546381e-05, "loss": 0.0123, "step": 8730 }, { "epoch": 0.5091163278382944, "grad_norm": 0.31317803263664246, "learning_rate": 8.490827094967363e-05, "loss": 0.0114, "step": 8740 }, { "epoch": 0.5096988407992078, "grad_norm": 0.25290921330451965, "learning_rate": 8.486879043704263e-05, "loss": 0.0112, "step": 8750 }, { "epoch": 0.5102813537601212, "grad_norm": 0.3568704128265381, "learning_rate": 8.482926755554332e-05, "loss": 0.0113, "step": 8760 }, { "epoch": 0.5108638667210346, "grad_norm": 0.30301928520202637, "learning_rate": 8.478970235319975e-05, "loss": 0.01, "step": 8770 }, { "epoch": 0.511446379681948, "grad_norm": 0.40317073464393616, "learning_rate": 8.475009487808737e-05, "loss": 0.014, "step": 8780 }, { "epoch": 0.5120288926428613, "grad_norm": 0.38880282640457153, "learning_rate": 8.471044517833298e-05, "loss": 0.0131, "step": 8790 }, { "epoch": 0.5126114056037747, "grad_norm": 0.4629809856414795, "learning_rate": 8.467075330211474e-05, "loss": 0.011, "step": 8800 }, { "epoch": 0.5131939185646881, "grad_norm": 0.20804764330387115, "learning_rate": 8.463101929766198e-05, "loss": 0.0138, "step": 8810 }, { "epoch": 0.5137764315256015, "grad_norm": 0.45887529850006104, "learning_rate": 8.459124321325528e-05, "loss": 0.0132, "step": 8820 }, { "epoch": 0.5143589444865149, "grad_norm": 0.360636830329895, "learning_rate": 8.455142509722634e-05, "loss": 0.0119, "step": 8830 }, { "epoch": 0.5149414574474283, "grad_norm": 0.38743698596954346, "learning_rate": 8.451156499795792e-05, "loss": 0.01, "step": 8840 }, { "epoch": 0.5155239704083416, "grad_norm": 0.3410528302192688, "learning_rate": 8.447166296388381e-05, "loss": 0.0118, "step": 8850 }, { "epoch": 0.516106483369255, "grad_norm": 0.22457276284694672, "learning_rate": 8.443171904348872e-05, "loss": 0.0119, "step": 8860 }, { "epoch": 0.5166889963301683, "grad_norm": 0.23236508667469025, "learning_rate": 8.43917332853083e-05, "loss": 0.0093, "step": 8870 }, { "epoch": 0.5172715092910817, "grad_norm": 0.2859998643398285, "learning_rate": 8.435170573792901e-05, "loss": 0.0105, "step": 8880 }, { "epoch": 0.5178540222519951, "grad_norm": 0.25001728534698486, "learning_rate": 8.431163644998808e-05, "loss": 0.0094, "step": 8890 }, { "epoch": 0.5184365352129084, "grad_norm": 0.3896948993206024, "learning_rate": 8.42715254701735e-05, "loss": 0.0143, "step": 8900 }, { "epoch": 0.5190190481738218, "grad_norm": 0.27904146909713745, "learning_rate": 8.42313728472239e-05, "loss": 0.0089, "step": 8910 }, { "epoch": 0.5196015611347352, "grad_norm": 0.25483042001724243, "learning_rate": 8.419117862992844e-05, "loss": 0.0096, "step": 8920 }, { "epoch": 0.5201840740956486, "grad_norm": 0.2366785705089569, "learning_rate": 8.415094286712694e-05, "loss": 0.0083, "step": 8930 }, { "epoch": 0.520766587056562, "grad_norm": 0.2771153450012207, "learning_rate": 8.411066560770965e-05, "loss": 0.009, "step": 8940 }, { "epoch": 0.5213491000174754, "grad_norm": 0.22708876430988312, "learning_rate": 8.407034690061721e-05, "loss": 0.011, "step": 8950 }, { "epoch": 0.5219316129783887, "grad_norm": 0.32733431458473206, "learning_rate": 8.402998679484067e-05, "loss": 0.0115, "step": 8960 }, { "epoch": 0.5225141259393021, "grad_norm": 0.2673744857311249, "learning_rate": 8.398958533942135e-05, "loss": 0.0074, "step": 8970 }, { "epoch": 0.5230966389002155, "grad_norm": 0.2505076825618744, "learning_rate": 8.394914258345083e-05, "loss": 0.0122, "step": 8980 }, { "epoch": 0.5236791518611289, "grad_norm": 0.42512375116348267, "learning_rate": 8.390865857607089e-05, "loss": 0.0114, "step": 8990 }, { "epoch": 0.5242616648220423, "grad_norm": 0.4811088740825653, "learning_rate": 8.386813336647339e-05, "loss": 0.015, "step": 9000 }, { "epoch": 0.5248441777829557, "grad_norm": 0.2737468481063843, "learning_rate": 8.38275670039003e-05, "loss": 0.0108, "step": 9010 }, { "epoch": 0.525426690743869, "grad_norm": 0.21187454462051392, "learning_rate": 8.378695953764356e-05, "loss": 0.0085, "step": 9020 }, { "epoch": 0.5260092037047824, "grad_norm": 0.30845895409584045, "learning_rate": 8.374631101704509e-05, "loss": 0.0099, "step": 9030 }, { "epoch": 0.5265917166656958, "grad_norm": 0.20114067196846008, "learning_rate": 8.370562149149665e-05, "loss": 0.009, "step": 9040 }, { "epoch": 0.5271742296266092, "grad_norm": 0.22177201509475708, "learning_rate": 8.366489101043989e-05, "loss": 0.0095, "step": 9050 }, { "epoch": 0.5277567425875226, "grad_norm": 0.19219940900802612, "learning_rate": 8.362411962336614e-05, "loss": 0.0095, "step": 9060 }, { "epoch": 0.528339255548436, "grad_norm": 0.44997864961624146, "learning_rate": 8.358330737981651e-05, "loss": 0.0105, "step": 9070 }, { "epoch": 0.5289217685093494, "grad_norm": 0.2769324481487274, "learning_rate": 8.35424543293817e-05, "loss": 0.0116, "step": 9080 }, { "epoch": 0.5295042814702627, "grad_norm": 0.28238800168037415, "learning_rate": 8.350156052170206e-05, "loss": 0.0101, "step": 9090 }, { "epoch": 0.5300867944311761, "grad_norm": 0.2641945481300354, "learning_rate": 8.346062600646738e-05, "loss": 0.0097, "step": 9100 }, { "epoch": 0.5306693073920895, "grad_norm": 0.27481865882873535, "learning_rate": 8.341965083341697e-05, "loss": 0.0134, "step": 9110 }, { "epoch": 0.5312518203530029, "grad_norm": 0.3104117512702942, "learning_rate": 8.337863505233953e-05, "loss": 0.0105, "step": 9120 }, { "epoch": 0.5318343333139163, "grad_norm": 0.38113418221473694, "learning_rate": 8.33375787130731e-05, "loss": 0.0083, "step": 9130 }, { "epoch": 0.5324168462748297, "grad_norm": 0.21746212244033813, "learning_rate": 8.3296481865505e-05, "loss": 0.0106, "step": 9140 }, { "epoch": 0.532999359235743, "grad_norm": 0.3854799270629883, "learning_rate": 8.325534455957181e-05, "loss": 0.0113, "step": 9150 }, { "epoch": 0.5335818721966564, "grad_norm": 0.3229290544986725, "learning_rate": 8.321416684525917e-05, "loss": 0.01, "step": 9160 }, { "epoch": 0.5341643851575698, "grad_norm": 0.28699108958244324, "learning_rate": 8.317294877260193e-05, "loss": 0.0116, "step": 9170 }, { "epoch": 0.5347468981184831, "grad_norm": 0.38508835434913635, "learning_rate": 8.313169039168394e-05, "loss": 0.0116, "step": 9180 }, { "epoch": 0.5353294110793965, "grad_norm": 0.19626887142658234, "learning_rate": 8.3090391752638e-05, "loss": 0.0087, "step": 9190 }, { "epoch": 0.5359119240403099, "grad_norm": 0.21657134592533112, "learning_rate": 8.304905290564586e-05, "loss": 0.0115, "step": 9200 }, { "epoch": 0.5364944370012232, "grad_norm": 0.3363923132419586, "learning_rate": 8.300767390093813e-05, "loss": 0.0107, "step": 9210 }, { "epoch": 0.5370769499621366, "grad_norm": 0.3121386170387268, "learning_rate": 8.296625478879417e-05, "loss": 0.0158, "step": 9220 }, { "epoch": 0.53765946292305, "grad_norm": 0.371357262134552, "learning_rate": 8.292479561954214e-05, "loss": 0.0112, "step": 9230 }, { "epoch": 0.5382419758839634, "grad_norm": 0.37682169675827026, "learning_rate": 8.288329644355883e-05, "loss": 0.0114, "step": 9240 }, { "epoch": 0.5388244888448768, "grad_norm": 0.28736403584480286, "learning_rate": 8.284175731126965e-05, "loss": 0.0134, "step": 9250 }, { "epoch": 0.5394070018057902, "grad_norm": 0.32662689685821533, "learning_rate": 8.280017827314854e-05, "loss": 0.0098, "step": 9260 }, { "epoch": 0.5399895147667035, "grad_norm": 0.1859537810087204, "learning_rate": 8.2758559379718e-05, "loss": 0.0138, "step": 9270 }, { "epoch": 0.5405720277276169, "grad_norm": 0.3454076051712036, "learning_rate": 8.271690068154887e-05, "loss": 0.0105, "step": 9280 }, { "epoch": 0.5411545406885303, "grad_norm": 0.25456947088241577, "learning_rate": 8.26752022292604e-05, "loss": 0.0093, "step": 9290 }, { "epoch": 0.5417370536494437, "grad_norm": 0.3912484049797058, "learning_rate": 8.263346407352018e-05, "loss": 0.0113, "step": 9300 }, { "epoch": 0.5423195666103571, "grad_norm": 0.289750337600708, "learning_rate": 8.259168626504395e-05, "loss": 0.0113, "step": 9310 }, { "epoch": 0.5429020795712705, "grad_norm": 0.21054647862911224, "learning_rate": 8.254986885459572e-05, "loss": 0.0113, "step": 9320 }, { "epoch": 0.5434845925321838, "grad_norm": 0.2637708783149719, "learning_rate": 8.250801189298757e-05, "loss": 0.0115, "step": 9330 }, { "epoch": 0.5440671054930972, "grad_norm": 0.27710995078086853, "learning_rate": 8.246611543107967e-05, "loss": 0.0113, "step": 9340 }, { "epoch": 0.5446496184540106, "grad_norm": 0.2379639893770218, "learning_rate": 8.242417951978018e-05, "loss": 0.0098, "step": 9350 }, { "epoch": 0.545232131414924, "grad_norm": 0.2423040270805359, "learning_rate": 8.238220421004517e-05, "loss": 0.0124, "step": 9360 }, { "epoch": 0.5458146443758374, "grad_norm": 0.31561440229415894, "learning_rate": 8.234018955287859e-05, "loss": 0.0177, "step": 9370 }, { "epoch": 0.5463971573367508, "grad_norm": 0.4590918719768524, "learning_rate": 8.229813559933224e-05, "loss": 0.0155, "step": 9380 }, { "epoch": 0.5469796702976641, "grad_norm": 0.40657955408096313, "learning_rate": 8.22560424005056e-05, "loss": 0.0149, "step": 9390 }, { "epoch": 0.5475621832585775, "grad_norm": 0.3211154043674469, "learning_rate": 8.22139100075459e-05, "loss": 0.0092, "step": 9400 }, { "epoch": 0.5481446962194909, "grad_norm": 0.2968524694442749, "learning_rate": 8.217173847164799e-05, "loss": 0.0114, "step": 9410 }, { "epoch": 0.5487272091804043, "grad_norm": 0.3226020336151123, "learning_rate": 8.212952784405422e-05, "loss": 0.0115, "step": 9420 }, { "epoch": 0.5493097221413177, "grad_norm": 0.250797837972641, "learning_rate": 8.208727817605451e-05, "loss": 0.0097, "step": 9430 }, { "epoch": 0.5498922351022311, "grad_norm": 0.2678021788597107, "learning_rate": 8.204498951898618e-05, "loss": 0.0134, "step": 9440 }, { "epoch": 0.5504747480631444, "grad_norm": 0.1832326352596283, "learning_rate": 8.200266192423395e-05, "loss": 0.009, "step": 9450 }, { "epoch": 0.5510572610240578, "grad_norm": 0.3206072449684143, "learning_rate": 8.196029544322981e-05, "loss": 0.0112, "step": 9460 }, { "epoch": 0.5516397739849712, "grad_norm": 0.17319335043430328, "learning_rate": 8.191789012745305e-05, "loss": 0.0106, "step": 9470 }, { "epoch": 0.5522222869458846, "grad_norm": 0.2627083659172058, "learning_rate": 8.187544602843014e-05, "loss": 0.009, "step": 9480 }, { "epoch": 0.5528047999067979, "grad_norm": 0.2164127081632614, "learning_rate": 8.183296319773466e-05, "loss": 0.0112, "step": 9490 }, { "epoch": 0.5533873128677113, "grad_norm": 0.41261550784111023, "learning_rate": 8.179044168698721e-05, "loss": 0.0136, "step": 9500 }, { "epoch": 0.5539698258286246, "grad_norm": 0.29752349853515625, "learning_rate": 8.174788154785548e-05, "loss": 0.011, "step": 9510 }, { "epoch": 0.554552338789538, "grad_norm": 0.32278183102607727, "learning_rate": 8.170528283205403e-05, "loss": 0.0134, "step": 9520 }, { "epoch": 0.5551348517504514, "grad_norm": 0.22981145977973938, "learning_rate": 8.166264559134435e-05, "loss": 0.0114, "step": 9530 }, { "epoch": 0.5557173647113648, "grad_norm": 0.3100016415119171, "learning_rate": 8.161996987753467e-05, "loss": 0.0139, "step": 9540 }, { "epoch": 0.5562998776722782, "grad_norm": 0.3087938725948334, "learning_rate": 8.157725574248e-05, "loss": 0.0135, "step": 9550 }, { "epoch": 0.5568823906331916, "grad_norm": 0.21534736454486847, "learning_rate": 8.153450323808209e-05, "loss": 0.0109, "step": 9560 }, { "epoch": 0.5574649035941049, "grad_norm": 0.38279926776885986, "learning_rate": 8.149171241628923e-05, "loss": 0.0126, "step": 9570 }, { "epoch": 0.5580474165550183, "grad_norm": 0.3262600600719452, "learning_rate": 8.144888332909631e-05, "loss": 0.0123, "step": 9580 }, { "epoch": 0.5586299295159317, "grad_norm": 0.17854773998260498, "learning_rate": 8.14060160285447e-05, "loss": 0.0083, "step": 9590 }, { "epoch": 0.5592124424768451, "grad_norm": 0.2897261083126068, "learning_rate": 8.136311056672223e-05, "loss": 0.0123, "step": 9600 }, { "epoch": 0.5597949554377585, "grad_norm": 0.42957374453544617, "learning_rate": 8.132016699576308e-05, "loss": 0.012, "step": 9610 }, { "epoch": 0.5603774683986719, "grad_norm": 0.4951796233654022, "learning_rate": 8.127718536784771e-05, "loss": 0.0115, "step": 9620 }, { "epoch": 0.5609599813595852, "grad_norm": 0.44225791096687317, "learning_rate": 8.12341657352029e-05, "loss": 0.0136, "step": 9630 }, { "epoch": 0.5615424943204986, "grad_norm": 0.14812079071998596, "learning_rate": 8.119110815010151e-05, "loss": 0.0087, "step": 9640 }, { "epoch": 0.562125007281412, "grad_norm": 0.46010565757751465, "learning_rate": 8.114801266486261e-05, "loss": 0.0136, "step": 9650 }, { "epoch": 0.5627075202423254, "grad_norm": 0.48483937978744507, "learning_rate": 8.110487933185122e-05, "loss": 0.0107, "step": 9660 }, { "epoch": 0.5632900332032388, "grad_norm": 0.27445095777511597, "learning_rate": 8.106170820347847e-05, "loss": 0.0112, "step": 9670 }, { "epoch": 0.5638725461641522, "grad_norm": 0.4026414752006531, "learning_rate": 8.101849933220132e-05, "loss": 0.0126, "step": 9680 }, { "epoch": 0.5644550591250656, "grad_norm": 0.24285726249217987, "learning_rate": 8.097525277052264e-05, "loss": 0.009, "step": 9690 }, { "epoch": 0.5650375720859789, "grad_norm": 0.35967713594436646, "learning_rate": 8.093196857099106e-05, "loss": 0.0092, "step": 9700 }, { "epoch": 0.5656200850468923, "grad_norm": 0.1624266505241394, "learning_rate": 8.088864678620096e-05, "loss": 0.0116, "step": 9710 }, { "epoch": 0.5662025980078057, "grad_norm": 0.19601480662822723, "learning_rate": 8.084528746879243e-05, "loss": 0.0108, "step": 9720 }, { "epoch": 0.5667851109687191, "grad_norm": 0.30270540714263916, "learning_rate": 8.080189067145108e-05, "loss": 0.011, "step": 9730 }, { "epoch": 0.5673676239296325, "grad_norm": 0.16536946594715118, "learning_rate": 8.075845644690813e-05, "loss": 0.0104, "step": 9740 }, { "epoch": 0.5679501368905459, "grad_norm": 0.29546692967414856, "learning_rate": 8.07149848479403e-05, "loss": 0.0122, "step": 9750 }, { "epoch": 0.5685326498514592, "grad_norm": 0.3476848304271698, "learning_rate": 8.067147592736962e-05, "loss": 0.0098, "step": 9760 }, { "epoch": 0.5691151628123726, "grad_norm": 0.16247175633907318, "learning_rate": 8.062792973806357e-05, "loss": 0.0087, "step": 9770 }, { "epoch": 0.569697675773286, "grad_norm": 0.3933895230293274, "learning_rate": 8.058434633293485e-05, "loss": 0.0086, "step": 9780 }, { "epoch": 0.5702801887341994, "grad_norm": 0.2233133465051651, "learning_rate": 8.054072576494143e-05, "loss": 0.01, "step": 9790 }, { "epoch": 0.5708627016951127, "grad_norm": 0.3780955970287323, "learning_rate": 8.04970680870864e-05, "loss": 0.0076, "step": 9800 }, { "epoch": 0.571445214656026, "grad_norm": 0.30459094047546387, "learning_rate": 8.045337335241793e-05, "loss": 0.0109, "step": 9810 }, { "epoch": 0.5720277276169394, "grad_norm": 0.4079815447330475, "learning_rate": 8.040964161402932e-05, "loss": 0.0093, "step": 9820 }, { "epoch": 0.5726102405778528, "grad_norm": 0.35216766595840454, "learning_rate": 8.036587292505869e-05, "loss": 0.015, "step": 9830 }, { "epoch": 0.5731927535387662, "grad_norm": 0.483693927526474, "learning_rate": 8.032206733868912e-05, "loss": 0.0133, "step": 9840 }, { "epoch": 0.5737752664996796, "grad_norm": 0.4553718864917755, "learning_rate": 8.027822490814858e-05, "loss": 0.0159, "step": 9850 }, { "epoch": 0.574357779460593, "grad_norm": 0.31747058033943176, "learning_rate": 8.023434568670972e-05, "loss": 0.0147, "step": 9860 }, { "epoch": 0.5749402924215063, "grad_norm": 0.30035489797592163, "learning_rate": 8.019042972768992e-05, "loss": 0.0104, "step": 9870 }, { "epoch": 0.5755228053824197, "grad_norm": 0.2849234640598297, "learning_rate": 8.014647708445124e-05, "loss": 0.0109, "step": 9880 }, { "epoch": 0.5761053183433331, "grad_norm": 0.4712446331977844, "learning_rate": 8.010248781040027e-05, "loss": 0.0132, "step": 9890 }, { "epoch": 0.5766878313042465, "grad_norm": 0.38158559799194336, "learning_rate": 8.005846195898814e-05, "loss": 0.0108, "step": 9900 }, { "epoch": 0.5772703442651599, "grad_norm": 0.3007536828517914, "learning_rate": 8.00143995837104e-05, "loss": 0.0112, "step": 9910 }, { "epoch": 0.5778528572260733, "grad_norm": 0.2956334948539734, "learning_rate": 7.997030073810698e-05, "loss": 0.0114, "step": 9920 }, { "epoch": 0.5784353701869867, "grad_norm": 0.18578597903251648, "learning_rate": 7.992616547576217e-05, "loss": 0.0084, "step": 9930 }, { "epoch": 0.5790178831479, "grad_norm": 0.26690977811813354, "learning_rate": 7.988199385030445e-05, "loss": 0.0128, "step": 9940 }, { "epoch": 0.5796003961088134, "grad_norm": 0.3819132447242737, "learning_rate": 7.98377859154065e-05, "loss": 0.0129, "step": 9950 }, { "epoch": 0.5801829090697268, "grad_norm": 0.2008133828639984, "learning_rate": 7.979354172478516e-05, "loss": 0.0088, "step": 9960 }, { "epoch": 0.5807654220306402, "grad_norm": 0.20811396837234497, "learning_rate": 7.974926133220127e-05, "loss": 0.0099, "step": 9970 }, { "epoch": 0.5813479349915536, "grad_norm": 0.23433077335357666, "learning_rate": 7.970494479145967e-05, "loss": 0.0098, "step": 9980 }, { "epoch": 0.581930447952467, "grad_norm": 0.4400034248828888, "learning_rate": 7.966059215640919e-05, "loss": 0.0127, "step": 9990 }, { "epoch": 0.5825129609133803, "grad_norm": 0.2626042366027832, "learning_rate": 7.961620348094241e-05, "loss": 0.0115, "step": 10000 } ], "logging_steps": 10, "max_steps": 30000, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 10000, "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": 32, "trial_name": null, "trial_params": null }