{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 30.0, "eval_steps": 500, "global_step": 33750, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.008888888888888889, "grad_norm": 7.635982990264893, "learning_rate": 5.331753554502371e-07, "loss": 0.936, "step": 10 }, { "epoch": 0.017777777777777778, "grad_norm": 8.823503494262695, "learning_rate": 1.1255924170616115e-06, "loss": 0.9998, "step": 20 }, { "epoch": 0.02666666666666667, "grad_norm": 6.84179162979126, "learning_rate": 1.7180094786729859e-06, "loss": 0.8225, "step": 30 }, { "epoch": 0.035555555555555556, "grad_norm": 3.374079465866089, "learning_rate": 2.31042654028436e-06, "loss": 0.5363, "step": 40 }, { "epoch": 0.044444444444444446, "grad_norm": 1.3296486139297485, "learning_rate": 2.902843601895735e-06, "loss": 0.3849, "step": 50 }, { "epoch": 0.05333333333333334, "grad_norm": 1.2770323753356934, "learning_rate": 3.4952606635071093e-06, "loss": 0.2609, "step": 60 }, { "epoch": 0.06222222222222222, "grad_norm": 1.0149893760681152, "learning_rate": 4.087677725118483e-06, "loss": 0.2407, "step": 70 }, { "epoch": 0.07111111111111111, "grad_norm": 0.8106550574302673, "learning_rate": 4.680094786729858e-06, "loss": 0.2192, "step": 80 }, { "epoch": 0.08, "grad_norm": 1.0121684074401855, "learning_rate": 5.272511848341233e-06, "loss": 0.2004, "step": 90 }, { "epoch": 0.08888888888888889, "grad_norm": 1.0550570487976074, "learning_rate": 5.864928909952607e-06, "loss": 0.1774, "step": 100 }, { "epoch": 0.09777777777777778, "grad_norm": 0.930065929889679, "learning_rate": 6.457345971563982e-06, "loss": 0.1798, "step": 110 }, { "epoch": 0.10666666666666667, "grad_norm": 0.7328033447265625, "learning_rate": 7.049763033175356e-06, "loss": 0.1872, "step": 120 }, { "epoch": 0.11555555555555555, "grad_norm": 1.2247834205627441, "learning_rate": 7.64218009478673e-06, "loss": 0.1669, "step": 130 }, { "epoch": 0.12444444444444444, "grad_norm": 1.2973235845565796, "learning_rate": 8.234597156398106e-06, "loss": 0.1701, "step": 140 }, { "epoch": 0.13333333333333333, "grad_norm": 0.8500670194625854, "learning_rate": 8.82701421800948e-06, "loss": 0.1655, "step": 150 }, { "epoch": 0.14222222222222222, "grad_norm": 1.0475229024887085, "learning_rate": 9.419431279620854e-06, "loss": 0.1566, "step": 160 }, { "epoch": 0.1511111111111111, "grad_norm": 0.8827213048934937, "learning_rate": 1.0011848341232228e-05, "loss": 0.1471, "step": 170 }, { "epoch": 0.16, "grad_norm": 0.9696575999259949, "learning_rate": 1.0604265402843602e-05, "loss": 0.1439, "step": 180 }, { "epoch": 0.1688888888888889, "grad_norm": 1.0812995433807373, "learning_rate": 1.1196682464454977e-05, "loss": 0.1297, "step": 190 }, { "epoch": 0.17777777777777778, "grad_norm": 0.7758895754814148, "learning_rate": 1.178909952606635e-05, "loss": 0.1168, "step": 200 }, { "epoch": 0.18666666666666668, "grad_norm": 1.3302476406097412, "learning_rate": 1.2381516587677726e-05, "loss": 0.1168, "step": 210 }, { "epoch": 0.19555555555555557, "grad_norm": 0.8562203049659729, "learning_rate": 1.2973933649289099e-05, "loss": 0.101, "step": 220 }, { "epoch": 0.20444444444444446, "grad_norm": 1.0701277256011963, "learning_rate": 1.3566350710900474e-05, "loss": 0.1031, "step": 230 }, { "epoch": 0.21333333333333335, "grad_norm": 0.7784116864204407, "learning_rate": 1.415876777251185e-05, "loss": 0.0897, "step": 240 }, { "epoch": 0.2222222222222222, "grad_norm": 1.1919257640838623, "learning_rate": 1.4751184834123225e-05, "loss": 0.0975, "step": 250 }, { "epoch": 0.2311111111111111, "grad_norm": 0.8140836358070374, "learning_rate": 1.5343601895734597e-05, "loss": 0.091, "step": 260 }, { "epoch": 0.24, "grad_norm": 0.6603021621704102, "learning_rate": 1.5936018957345972e-05, "loss": 0.0834, "step": 270 }, { "epoch": 0.24888888888888888, "grad_norm": 0.875921368598938, "learning_rate": 1.6528436018957347e-05, "loss": 0.077, "step": 280 }, { "epoch": 0.2577777777777778, "grad_norm": 0.9051255583763123, "learning_rate": 1.7120853080568722e-05, "loss": 0.079, "step": 290 }, { "epoch": 0.26666666666666666, "grad_norm": 0.7191042900085449, "learning_rate": 1.7713270142180098e-05, "loss": 0.0694, "step": 300 }, { "epoch": 0.27555555555555555, "grad_norm": 0.9481140375137329, "learning_rate": 1.830568720379147e-05, "loss": 0.0721, "step": 310 }, { "epoch": 0.28444444444444444, "grad_norm": 0.8378931879997253, "learning_rate": 1.8898104265402845e-05, "loss": 0.0861, "step": 320 }, { "epoch": 0.29333333333333333, "grad_norm": 0.7944449782371521, "learning_rate": 1.949052132701422e-05, "loss": 0.075, "step": 330 }, { "epoch": 0.3022222222222222, "grad_norm": 0.8207589387893677, "learning_rate": 2.0082938388625595e-05, "loss": 0.0778, "step": 340 }, { "epoch": 0.3111111111111111, "grad_norm": 1.059295654296875, "learning_rate": 2.067535545023697e-05, "loss": 0.0755, "step": 350 }, { "epoch": 0.32, "grad_norm": 0.9660500288009644, "learning_rate": 2.1267772511848342e-05, "loss": 0.074, "step": 360 }, { "epoch": 0.3288888888888889, "grad_norm": 0.7071358561515808, "learning_rate": 2.1860189573459718e-05, "loss": 0.0749, "step": 370 }, { "epoch": 0.3377777777777778, "grad_norm": 0.717815101146698, "learning_rate": 2.2452606635071093e-05, "loss": 0.07, "step": 380 }, { "epoch": 0.3466666666666667, "grad_norm": 0.5262020230293274, "learning_rate": 2.3045023696682465e-05, "loss": 0.0668, "step": 390 }, { "epoch": 0.35555555555555557, "grad_norm": 0.8403083682060242, "learning_rate": 2.363744075829384e-05, "loss": 0.062, "step": 400 }, { "epoch": 0.36444444444444446, "grad_norm": 0.7108381986618042, "learning_rate": 2.4229857819905215e-05, "loss": 0.0664, "step": 410 }, { "epoch": 0.37333333333333335, "grad_norm": 0.7727111577987671, "learning_rate": 2.4822274881516587e-05, "loss": 0.0715, "step": 420 }, { "epoch": 0.38222222222222224, "grad_norm": 0.8929181694984436, "learning_rate": 2.5414691943127962e-05, "loss": 0.0636, "step": 430 }, { "epoch": 0.39111111111111113, "grad_norm": 1.0163711309432983, "learning_rate": 2.6007109004739337e-05, "loss": 0.0665, "step": 440 }, { "epoch": 0.4, "grad_norm": 1.0706408023834229, "learning_rate": 2.659952606635071e-05, "loss": 0.0679, "step": 450 }, { "epoch": 0.4088888888888889, "grad_norm": 0.9639104008674622, "learning_rate": 2.7191943127962088e-05, "loss": 0.0667, "step": 460 }, { "epoch": 0.4177777777777778, "grad_norm": 0.6514188051223755, "learning_rate": 2.778436018957346e-05, "loss": 0.0563, "step": 470 }, { "epoch": 0.4266666666666667, "grad_norm": 0.7542102336883545, "learning_rate": 2.837677725118483e-05, "loss": 0.0583, "step": 480 }, { "epoch": 0.43555555555555553, "grad_norm": 0.6054924130439758, "learning_rate": 2.896919431279621e-05, "loss": 0.0799, "step": 490 }, { "epoch": 0.4444444444444444, "grad_norm": 1.0176395177841187, "learning_rate": 2.9561611374407582e-05, "loss": 0.0591, "step": 500 }, { "epoch": 0.4533333333333333, "grad_norm": 0.8142734169960022, "learning_rate": 3.015402843601896e-05, "loss": 0.0658, "step": 510 }, { "epoch": 0.4622222222222222, "grad_norm": 0.7942567467689514, "learning_rate": 3.0746445497630336e-05, "loss": 0.0633, "step": 520 }, { "epoch": 0.4711111111111111, "grad_norm": 0.8154179453849792, "learning_rate": 3.133886255924171e-05, "loss": 0.0635, "step": 530 }, { "epoch": 0.48, "grad_norm": 0.6405626535415649, "learning_rate": 3.193127962085308e-05, "loss": 0.0629, "step": 540 }, { "epoch": 0.4888888888888889, "grad_norm": 0.7488952875137329, "learning_rate": 3.252369668246446e-05, "loss": 0.074, "step": 550 }, { "epoch": 0.49777777777777776, "grad_norm": 0.6593071222305298, "learning_rate": 3.311611374407583e-05, "loss": 0.0606, "step": 560 }, { "epoch": 0.5066666666666667, "grad_norm": 1.0396125316619873, "learning_rate": 3.37085308056872e-05, "loss": 0.062, "step": 570 }, { "epoch": 0.5155555555555555, "grad_norm": 1.141627311706543, "learning_rate": 3.430094786729858e-05, "loss": 0.0554, "step": 580 }, { "epoch": 0.5244444444444445, "grad_norm": 0.6764670014381409, "learning_rate": 3.489336492890995e-05, "loss": 0.0502, "step": 590 }, { "epoch": 0.5333333333333333, "grad_norm": 0.6383158564567566, "learning_rate": 3.5485781990521324e-05, "loss": 0.0652, "step": 600 }, { "epoch": 0.5422222222222223, "grad_norm": 0.6914287209510803, "learning_rate": 3.60781990521327e-05, "loss": 0.0592, "step": 610 }, { "epoch": 0.5511111111111111, "grad_norm": 0.6463748216629028, "learning_rate": 3.6670616113744075e-05, "loss": 0.0546, "step": 620 }, { "epoch": 0.56, "grad_norm": 0.831497073173523, "learning_rate": 3.726303317535545e-05, "loss": 0.0614, "step": 630 }, { "epoch": 0.5688888888888889, "grad_norm": 0.753236711025238, "learning_rate": 3.7855450236966825e-05, "loss": 0.0544, "step": 640 }, { "epoch": 0.5777777777777777, "grad_norm": 1.071272850036621, "learning_rate": 3.84478672985782e-05, "loss": 0.0605, "step": 650 }, { "epoch": 0.5866666666666667, "grad_norm": 0.6668487787246704, "learning_rate": 3.9040284360189576e-05, "loss": 0.0492, "step": 660 }, { "epoch": 0.5955555555555555, "grad_norm": 0.7504937648773193, "learning_rate": 3.963270142180095e-05, "loss": 0.0506, "step": 670 }, { "epoch": 0.6044444444444445, "grad_norm": 0.7830075025558472, "learning_rate": 4.022511848341232e-05, "loss": 0.0561, "step": 680 }, { "epoch": 0.6133333333333333, "grad_norm": 0.8854774832725525, "learning_rate": 4.08175355450237e-05, "loss": 0.0565, "step": 690 }, { "epoch": 0.6222222222222222, "grad_norm": 0.7671079635620117, "learning_rate": 4.140995260663507e-05, "loss": 0.0583, "step": 700 }, { "epoch": 0.6311111111111111, "grad_norm": 0.7122657895088196, "learning_rate": 4.200236966824645e-05, "loss": 0.065, "step": 710 }, { "epoch": 0.64, "grad_norm": 0.7132411003112793, "learning_rate": 4.259478672985782e-05, "loss": 0.0646, "step": 720 }, { "epoch": 0.6488888888888888, "grad_norm": 0.7070056796073914, "learning_rate": 4.318720379146919e-05, "loss": 0.0571, "step": 730 }, { "epoch": 0.6577777777777778, "grad_norm": 0.457063764333725, "learning_rate": 4.377962085308057e-05, "loss": 0.0556, "step": 740 }, { "epoch": 0.6666666666666666, "grad_norm": 0.4275483191013336, "learning_rate": 4.437203791469194e-05, "loss": 0.0465, "step": 750 }, { "epoch": 0.6755555555555556, "grad_norm": 0.7410489916801453, "learning_rate": 4.496445497630332e-05, "loss": 0.0551, "step": 760 }, { "epoch": 0.6844444444444444, "grad_norm": 0.7652745246887207, "learning_rate": 4.555687203791469e-05, "loss": 0.0619, "step": 770 }, { "epoch": 0.6933333333333334, "grad_norm": 0.6531593799591064, "learning_rate": 4.6149289099526065e-05, "loss": 0.0562, "step": 780 }, { "epoch": 0.7022222222222222, "grad_norm": 0.9065576195716858, "learning_rate": 4.6741706161137444e-05, "loss": 0.062, "step": 790 }, { "epoch": 0.7111111111111111, "grad_norm": 0.6093670725822449, "learning_rate": 4.7334123222748816e-05, "loss": 0.0608, "step": 800 }, { "epoch": 0.72, "grad_norm": 0.7139759063720703, "learning_rate": 4.7926540284360194e-05, "loss": 0.0622, "step": 810 }, { "epoch": 0.7288888888888889, "grad_norm": 0.6038567423820496, "learning_rate": 4.8518957345971566e-05, "loss": 0.0563, "step": 820 }, { "epoch": 0.7377777777777778, "grad_norm": 0.7094252109527588, "learning_rate": 4.911137440758294e-05, "loss": 0.055, "step": 830 }, { "epoch": 0.7466666666666667, "grad_norm": 0.7277010083198547, "learning_rate": 4.9703791469194316e-05, "loss": 0.062, "step": 840 }, { "epoch": 0.7555555555555555, "grad_norm": 0.626179039478302, "learning_rate": 5.029620853080569e-05, "loss": 0.0561, "step": 850 }, { "epoch": 0.7644444444444445, "grad_norm": 0.6517444252967834, "learning_rate": 5.088862559241706e-05, "loss": 0.051, "step": 860 }, { "epoch": 0.7733333333333333, "grad_norm": 0.6472556591033936, "learning_rate": 5.148104265402843e-05, "loss": 0.0522, "step": 870 }, { "epoch": 0.7822222222222223, "grad_norm": 0.6448968648910522, "learning_rate": 5.207345971563982e-05, "loss": 0.0487, "step": 880 }, { "epoch": 0.7911111111111111, "grad_norm": 0.7059515118598938, "learning_rate": 5.266587677725119e-05, "loss": 0.0523, "step": 890 }, { "epoch": 0.8, "grad_norm": 0.5950090289115906, "learning_rate": 5.325829383886256e-05, "loss": 0.0467, "step": 900 }, { "epoch": 0.8088888888888889, "grad_norm": 0.8738225698471069, "learning_rate": 5.385071090047393e-05, "loss": 0.0532, "step": 910 }, { "epoch": 0.8177777777777778, "grad_norm": 0.8248814940452576, "learning_rate": 5.4443127962085305e-05, "loss": 0.0604, "step": 920 }, { "epoch": 0.8266666666666667, "grad_norm": 0.598600447177887, "learning_rate": 5.503554502369669e-05, "loss": 0.0497, "step": 930 }, { "epoch": 0.8355555555555556, "grad_norm": 0.642088770866394, "learning_rate": 5.562796208530806e-05, "loss": 0.0592, "step": 940 }, { "epoch": 0.8444444444444444, "grad_norm": 0.6585528254508972, "learning_rate": 5.6220379146919434e-05, "loss": 0.0557, "step": 950 }, { "epoch": 0.8533333333333334, "grad_norm": 0.7627522945404053, "learning_rate": 5.6812796208530806e-05, "loss": 0.0627, "step": 960 }, { "epoch": 0.8622222222222222, "grad_norm": 0.6394461393356323, "learning_rate": 5.740521327014218e-05, "loss": 0.0638, "step": 970 }, { "epoch": 0.8711111111111111, "grad_norm": 0.6414893269538879, "learning_rate": 5.799763033175356e-05, "loss": 0.0552, "step": 980 }, { "epoch": 0.88, "grad_norm": 0.7108924984931946, "learning_rate": 5.8590047393364935e-05, "loss": 0.0594, "step": 990 }, { "epoch": 0.8888888888888888, "grad_norm": 0.7174683809280396, "learning_rate": 5.918246445497631e-05, "loss": 0.0543, "step": 1000 }, { "epoch": 0.8977777777777778, "grad_norm": 0.6943054795265198, "learning_rate": 5.977488151658768e-05, "loss": 0.0576, "step": 1010 }, { "epoch": 0.9066666666666666, "grad_norm": 0.5762858986854553, "learning_rate": 6.036729857819905e-05, "loss": 0.0573, "step": 1020 }, { "epoch": 0.9155555555555556, "grad_norm": 0.8038665652275085, "learning_rate": 6.0959715639810436e-05, "loss": 0.0568, "step": 1030 }, { "epoch": 0.9244444444444444, "grad_norm": 0.5898231863975525, "learning_rate": 6.15521327014218e-05, "loss": 0.0502, "step": 1040 }, { "epoch": 0.9333333333333333, "grad_norm": 0.8258023858070374, "learning_rate": 6.214454976303317e-05, "loss": 0.0493, "step": 1050 }, { "epoch": 0.9422222222222222, "grad_norm": 0.6160328388214111, "learning_rate": 6.273696682464454e-05, "loss": 0.0549, "step": 1060 }, { "epoch": 0.9511111111111111, "grad_norm": 0.6316657066345215, "learning_rate": 6.332938388625592e-05, "loss": 0.0531, "step": 1070 }, { "epoch": 0.96, "grad_norm": 0.5872039794921875, "learning_rate": 6.39218009478673e-05, "loss": 0.0451, "step": 1080 }, { "epoch": 0.9688888888888889, "grad_norm": 0.6453844904899597, "learning_rate": 6.451421800947867e-05, "loss": 0.0573, "step": 1090 }, { "epoch": 0.9777777777777777, "grad_norm": 0.7083536982536316, "learning_rate": 6.510663507109005e-05, "loss": 0.0487, "step": 1100 }, { "epoch": 0.9866666666666667, "grad_norm": 0.5306820869445801, "learning_rate": 6.569905213270142e-05, "loss": 0.0478, "step": 1110 }, { "epoch": 0.9955555555555555, "grad_norm": 0.5878357291221619, "learning_rate": 6.629146919431279e-05, "loss": 0.0507, "step": 1120 }, { "epoch": 1.0044444444444445, "grad_norm": 0.7111451625823975, "learning_rate": 6.688388625592417e-05, "loss": 0.0516, "step": 1130 }, { "epoch": 1.0133333333333334, "grad_norm": 0.8390199542045593, "learning_rate": 6.747630331753555e-05, "loss": 0.0524, "step": 1140 }, { "epoch": 1.0222222222222221, "grad_norm": 0.7699342370033264, "learning_rate": 6.806872037914692e-05, "loss": 0.0489, "step": 1150 }, { "epoch": 1.031111111111111, "grad_norm": 0.5783478617668152, "learning_rate": 6.866113744075829e-05, "loss": 0.0586, "step": 1160 }, { "epoch": 1.04, "grad_norm": 0.6112527251243591, "learning_rate": 6.925355450236966e-05, "loss": 0.0553, "step": 1170 }, { "epoch": 1.048888888888889, "grad_norm": 0.3951888978481293, "learning_rate": 6.984597156398105e-05, "loss": 0.0569, "step": 1180 }, { "epoch": 1.0577777777777777, "grad_norm": 0.9480990767478943, "learning_rate": 7.043838862559242e-05, "loss": 0.0529, "step": 1190 }, { "epoch": 1.0666666666666667, "grad_norm": 0.5764167904853821, "learning_rate": 7.103080568720379e-05, "loss": 0.0508, "step": 1200 }, { "epoch": 1.0755555555555556, "grad_norm": 0.4282839000225067, "learning_rate": 7.162322274881516e-05, "loss": 0.0523, "step": 1210 }, { "epoch": 1.0844444444444445, "grad_norm": 0.7515479326248169, "learning_rate": 7.221563981042653e-05, "loss": 0.0519, "step": 1220 }, { "epoch": 1.0933333333333333, "grad_norm": 0.5480772852897644, "learning_rate": 7.280805687203792e-05, "loss": 0.0551, "step": 1230 }, { "epoch": 1.1022222222222222, "grad_norm": 0.5177034735679626, "learning_rate": 7.340047393364929e-05, "loss": 0.0443, "step": 1240 }, { "epoch": 1.1111111111111112, "grad_norm": 0.6852489113807678, "learning_rate": 7.399289099526066e-05, "loss": 0.0509, "step": 1250 }, { "epoch": 1.12, "grad_norm": 0.5949726104736328, "learning_rate": 7.458530805687204e-05, "loss": 0.0652, "step": 1260 }, { "epoch": 1.1288888888888888, "grad_norm": 0.7931841611862183, "learning_rate": 7.517772511848341e-05, "loss": 0.051, "step": 1270 }, { "epoch": 1.1377777777777778, "grad_norm": 0.6654778122901917, "learning_rate": 7.577014218009479e-05, "loss": 0.0537, "step": 1280 }, { "epoch": 1.1466666666666667, "grad_norm": 0.7448049783706665, "learning_rate": 7.636255924170616e-05, "loss": 0.0434, "step": 1290 }, { "epoch": 1.1555555555555554, "grad_norm": 0.5791196823120117, "learning_rate": 7.695497630331754e-05, "loss": 0.0494, "step": 1300 }, { "epoch": 1.1644444444444444, "grad_norm": 0.5269730091094971, "learning_rate": 7.754739336492891e-05, "loss": 0.053, "step": 1310 }, { "epoch": 1.1733333333333333, "grad_norm": 0.6474888920783997, "learning_rate": 7.813981042654028e-05, "loss": 0.0507, "step": 1320 }, { "epoch": 1.1822222222222223, "grad_norm": 0.4114304184913635, "learning_rate": 7.873222748815167e-05, "loss": 0.0521, "step": 1330 }, { "epoch": 1.1911111111111112, "grad_norm": 0.5078536868095398, "learning_rate": 7.932464454976304e-05, "loss": 0.0606, "step": 1340 }, { "epoch": 1.2, "grad_norm": 0.651603639125824, "learning_rate": 7.991706161137441e-05, "loss": 0.0502, "step": 1350 }, { "epoch": 1.208888888888889, "grad_norm": 0.5146484375, "learning_rate": 8.050947867298578e-05, "loss": 0.0575, "step": 1360 }, { "epoch": 1.2177777777777778, "grad_norm": 0.6116874814033508, "learning_rate": 8.110189573459715e-05, "loss": 0.0498, "step": 1370 }, { "epoch": 1.2266666666666666, "grad_norm": 0.6122986078262329, "learning_rate": 8.169431279620854e-05, "loss": 0.0599, "step": 1380 }, { "epoch": 1.2355555555555555, "grad_norm": 0.5883094668388367, "learning_rate": 8.228672985781991e-05, "loss": 0.0547, "step": 1390 }, { "epoch": 1.2444444444444445, "grad_norm": 0.7432698607444763, "learning_rate": 8.287914691943128e-05, "loss": 0.05, "step": 1400 }, { "epoch": 1.2533333333333334, "grad_norm": 0.4993435740470886, "learning_rate": 8.347156398104265e-05, "loss": 0.0527, "step": 1410 }, { "epoch": 1.2622222222222224, "grad_norm": 0.8369204998016357, "learning_rate": 8.406398104265403e-05, "loss": 0.0612, "step": 1420 }, { "epoch": 1.271111111111111, "grad_norm": 0.621932864189148, "learning_rate": 8.465639810426541e-05, "loss": 0.0563, "step": 1430 }, { "epoch": 1.28, "grad_norm": 0.6532228589057922, "learning_rate": 8.524881516587678e-05, "loss": 0.0555, "step": 1440 }, { "epoch": 1.2888888888888888, "grad_norm": 0.5963582992553711, "learning_rate": 8.584123222748816e-05, "loss": 0.0551, "step": 1450 }, { "epoch": 1.2977777777777777, "grad_norm": 0.7781577110290527, "learning_rate": 8.643364928909953e-05, "loss": 0.0575, "step": 1460 }, { "epoch": 1.3066666666666666, "grad_norm": 0.4843505620956421, "learning_rate": 8.70260663507109e-05, "loss": 0.054, "step": 1470 }, { "epoch": 1.3155555555555556, "grad_norm": 0.5443546772003174, "learning_rate": 8.761848341232228e-05, "loss": 0.056, "step": 1480 }, { "epoch": 1.3244444444444445, "grad_norm": 0.7557148933410645, "learning_rate": 8.821090047393366e-05, "loss": 0.0549, "step": 1490 }, { "epoch": 1.3333333333333333, "grad_norm": 0.6067489385604858, "learning_rate": 8.880331753554503e-05, "loss": 0.0447, "step": 1500 }, { "epoch": 1.3422222222222222, "grad_norm": 0.4741000831127167, "learning_rate": 8.93957345971564e-05, "loss": 0.0556, "step": 1510 }, { "epoch": 1.3511111111111112, "grad_norm": 0.7074664831161499, "learning_rate": 8.998815165876777e-05, "loss": 0.0579, "step": 1520 }, { "epoch": 1.3599999999999999, "grad_norm": 0.6011343002319336, "learning_rate": 9.058056872037916e-05, "loss": 0.0569, "step": 1530 }, { "epoch": 1.3688888888888888, "grad_norm": 0.7246229648590088, "learning_rate": 9.117298578199053e-05, "loss": 0.055, "step": 1540 }, { "epoch": 1.3777777777777778, "grad_norm": 0.6248283386230469, "learning_rate": 9.17654028436019e-05, "loss": 0.0461, "step": 1550 }, { "epoch": 1.3866666666666667, "grad_norm": 0.6015632152557373, "learning_rate": 9.235781990521327e-05, "loss": 0.049, "step": 1560 }, { "epoch": 1.3955555555555557, "grad_norm": 0.6176114082336426, "learning_rate": 9.295023696682464e-05, "loss": 0.0535, "step": 1570 }, { "epoch": 1.4044444444444444, "grad_norm": 0.3943069577217102, "learning_rate": 9.354265402843603e-05, "loss": 0.0472, "step": 1580 }, { "epoch": 1.4133333333333333, "grad_norm": 0.5434098243713379, "learning_rate": 9.41350710900474e-05, "loss": 0.0437, "step": 1590 }, { "epoch": 1.4222222222222223, "grad_norm": 0.5428091883659363, "learning_rate": 9.472748815165877e-05, "loss": 0.0434, "step": 1600 }, { "epoch": 1.431111111111111, "grad_norm": 0.6058080792427063, "learning_rate": 9.531990521327015e-05, "loss": 0.0472, "step": 1610 }, { "epoch": 1.44, "grad_norm": 0.7003351449966431, "learning_rate": 9.591232227488152e-05, "loss": 0.0406, "step": 1620 }, { "epoch": 1.448888888888889, "grad_norm": 0.5199935436248779, "learning_rate": 9.65047393364929e-05, "loss": 0.045, "step": 1630 }, { "epoch": 1.4577777777777778, "grad_norm": 0.8328996896743774, "learning_rate": 9.709715639810427e-05, "loss": 0.0481, "step": 1640 }, { "epoch": 1.4666666666666668, "grad_norm": 0.7033499479293823, "learning_rate": 9.768957345971565e-05, "loss": 0.048, "step": 1650 }, { "epoch": 1.4755555555555555, "grad_norm": 0.4052645266056061, "learning_rate": 9.828199052132702e-05, "loss": 0.0505, "step": 1660 }, { "epoch": 1.4844444444444445, "grad_norm": 0.6453437805175781, "learning_rate": 9.887440758293839e-05, "loss": 0.0455, "step": 1670 }, { "epoch": 1.4933333333333334, "grad_norm": 0.49197080731391907, "learning_rate": 9.946682464454978e-05, "loss": 0.0499, "step": 1680 }, { "epoch": 1.5022222222222221, "grad_norm": 0.5326331257820129, "learning_rate": 9.999999975997387e-05, "loss": 0.0514, "step": 1690 }, { "epoch": 1.511111111111111, "grad_norm": 0.6397197842597961, "learning_rate": 9.999997095684032e-05, "loss": 0.051, "step": 1700 }, { "epoch": 1.52, "grad_norm": 0.47068774700164795, "learning_rate": 9.99998941485112e-05, "loss": 0.0509, "step": 1710 }, { "epoch": 1.528888888888889, "grad_norm": 0.5211347937583923, "learning_rate": 9.99997693350603e-05, "loss": 0.0507, "step": 1720 }, { "epoch": 1.537777777777778, "grad_norm": 0.4757305383682251, "learning_rate": 9.99995965166074e-05, "loss": 0.0535, "step": 1730 }, { "epoch": 1.5466666666666666, "grad_norm": 0.7017833590507507, "learning_rate": 9.999937569331847e-05, "loss": 0.0502, "step": 1740 }, { "epoch": 1.5555555555555556, "grad_norm": 0.5903001427650452, "learning_rate": 9.99991068654055e-05, "loss": 0.0592, "step": 1750 }, { "epoch": 1.5644444444444443, "grad_norm": 0.5562971830368042, "learning_rate": 9.999879003312659e-05, "loss": 0.0512, "step": 1760 }, { "epoch": 1.5733333333333333, "grad_norm": 0.5284918546676636, "learning_rate": 9.999842519678594e-05, "loss": 0.048, "step": 1770 }, { "epoch": 1.5822222222222222, "grad_norm": 0.7244512438774109, "learning_rate": 9.999801235673383e-05, "loss": 0.0583, "step": 1780 }, { "epoch": 1.5911111111111111, "grad_norm": 0.59028559923172, "learning_rate": 9.999755151336663e-05, "loss": 0.0506, "step": 1790 }, { "epoch": 1.6, "grad_norm": 0.6866387128829956, "learning_rate": 9.999704266712679e-05, "loss": 0.0496, "step": 1800 }, { "epoch": 1.608888888888889, "grad_norm": 0.5830965638160706, "learning_rate": 9.999648581850287e-05, "loss": 0.0552, "step": 1810 }, { "epoch": 1.6177777777777778, "grad_norm": 0.4100777208805084, "learning_rate": 9.99958809680295e-05, "loss": 0.0462, "step": 1820 }, { "epoch": 1.6266666666666667, "grad_norm": 0.46138760447502136, "learning_rate": 9.999522811628738e-05, "loss": 0.062, "step": 1830 }, { "epoch": 1.6355555555555554, "grad_norm": 0.494722843170166, "learning_rate": 9.999452726390335e-05, "loss": 0.0487, "step": 1840 }, { "epoch": 1.6444444444444444, "grad_norm": 0.49877458810806274, "learning_rate": 9.999377841155028e-05, "loss": 0.045, "step": 1850 }, { "epoch": 1.6533333333333333, "grad_norm": 0.4711073935031891, "learning_rate": 9.999298155994715e-05, "loss": 0.0433, "step": 1860 }, { "epoch": 1.6622222222222223, "grad_norm": 0.4654456675052643, "learning_rate": 9.999213670985901e-05, "loss": 0.0457, "step": 1870 }, { "epoch": 1.6711111111111112, "grad_norm": 0.5897451043128967, "learning_rate": 9.9991243862097e-05, "loss": 0.0441, "step": 1880 }, { "epoch": 1.6800000000000002, "grad_norm": 0.5073348879814148, "learning_rate": 9.999030301751838e-05, "loss": 0.0445, "step": 1890 }, { "epoch": 1.6888888888888889, "grad_norm": 0.5675478577613831, "learning_rate": 9.998931417702644e-05, "loss": 0.0423, "step": 1900 }, { "epoch": 1.6977777777777778, "grad_norm": 0.4709509015083313, "learning_rate": 9.998827734157057e-05, "loss": 0.046, "step": 1910 }, { "epoch": 1.7066666666666666, "grad_norm": 0.5216087102890015, "learning_rate": 9.998719251214623e-05, "loss": 0.043, "step": 1920 }, { "epoch": 1.7155555555555555, "grad_norm": 0.4784368574619293, "learning_rate": 9.998605968979499e-05, "loss": 0.0414, "step": 1930 }, { "epoch": 1.7244444444444444, "grad_norm": 0.42035412788391113, "learning_rate": 9.998487887560446e-05, "loss": 0.052, "step": 1940 }, { "epoch": 1.7333333333333334, "grad_norm": 0.4229920506477356, "learning_rate": 9.998365007070837e-05, "loss": 0.0437, "step": 1950 }, { "epoch": 1.7422222222222223, "grad_norm": 0.38644281029701233, "learning_rate": 9.998237327628647e-05, "loss": 0.0501, "step": 1960 }, { "epoch": 1.751111111111111, "grad_norm": 0.4330987334251404, "learning_rate": 9.998104849356464e-05, "loss": 0.0365, "step": 1970 }, { "epoch": 1.76, "grad_norm": 0.568157434463501, "learning_rate": 9.997967572381479e-05, "loss": 0.0453, "step": 1980 }, { "epoch": 1.7688888888888887, "grad_norm": 0.42765119671821594, "learning_rate": 9.997825496835494e-05, "loss": 0.0513, "step": 1990 }, { "epoch": 1.7777777777777777, "grad_norm": 0.4584512412548065, "learning_rate": 9.997678622854915e-05, "loss": 0.0513, "step": 2000 }, { "epoch": 1.7866666666666666, "grad_norm": 0.4704533517360687, "learning_rate": 9.997526950580757e-05, "loss": 0.0476, "step": 2010 }, { "epoch": 1.7955555555555556, "grad_norm": 0.4793166518211365, "learning_rate": 9.997370480158641e-05, "loss": 0.0489, "step": 2020 }, { "epoch": 1.8044444444444445, "grad_norm": 0.5848549008369446, "learning_rate": 9.997209211738796e-05, "loss": 0.0531, "step": 2030 }, { "epoch": 1.8133333333333335, "grad_norm": 0.7921404838562012, "learning_rate": 9.997043145476054e-05, "loss": 0.0544, "step": 2040 }, { "epoch": 1.8222222222222222, "grad_norm": 0.5397977828979492, "learning_rate": 9.996872281529858e-05, "loss": 0.0511, "step": 2050 }, { "epoch": 1.8311111111111111, "grad_norm": 0.5546301007270813, "learning_rate": 9.996696620064256e-05, "loss": 0.0468, "step": 2060 }, { "epoch": 1.8399999999999999, "grad_norm": 0.48175743222236633, "learning_rate": 9.996516161247902e-05, "loss": 0.0498, "step": 2070 }, { "epoch": 1.8488888888888888, "grad_norm": 0.4818541705608368, "learning_rate": 9.99633090525405e-05, "loss": 0.0491, "step": 2080 }, { "epoch": 1.8577777777777778, "grad_norm": 0.6182060241699219, "learning_rate": 9.996140852260572e-05, "loss": 0.0516, "step": 2090 }, { "epoch": 1.8666666666666667, "grad_norm": 0.5543986558914185, "learning_rate": 9.995946002449932e-05, "loss": 0.0495, "step": 2100 }, { "epoch": 1.8755555555555556, "grad_norm": 0.46864646673202515, "learning_rate": 9.995746356009212e-05, "loss": 0.0449, "step": 2110 }, { "epoch": 1.8844444444444446, "grad_norm": 0.5903393030166626, "learning_rate": 9.995541913130089e-05, "loss": 0.0509, "step": 2120 }, { "epoch": 1.8933333333333333, "grad_norm": 0.5846372842788696, "learning_rate": 9.995332674008853e-05, "loss": 0.0502, "step": 2130 }, { "epoch": 1.9022222222222223, "grad_norm": 0.5359988808631897, "learning_rate": 9.995118638846394e-05, "loss": 0.0482, "step": 2140 }, { "epoch": 1.911111111111111, "grad_norm": 0.5447133183479309, "learning_rate": 9.994899807848207e-05, "loss": 0.05, "step": 2150 }, { "epoch": 1.92, "grad_norm": 0.701386570930481, "learning_rate": 9.994676181224394e-05, "loss": 0.0416, "step": 2160 }, { "epoch": 1.9288888888888889, "grad_norm": 0.4216717779636383, "learning_rate": 9.99444775918966e-05, "loss": 0.045, "step": 2170 }, { "epoch": 1.9377777777777778, "grad_norm": 0.5127856731414795, "learning_rate": 9.994214541963314e-05, "loss": 0.0439, "step": 2180 }, { "epoch": 1.9466666666666668, "grad_norm": 0.48860904574394226, "learning_rate": 9.993976529769268e-05, "loss": 0.0466, "step": 2190 }, { "epoch": 1.9555555555555557, "grad_norm": 0.6821528673171997, "learning_rate": 9.993733722836038e-05, "loss": 0.0497, "step": 2200 }, { "epoch": 1.9644444444444444, "grad_norm": 0.7485427856445312, "learning_rate": 9.993486121396747e-05, "loss": 0.0483, "step": 2210 }, { "epoch": 1.9733333333333334, "grad_norm": 0.4416391849517822, "learning_rate": 9.993233725689115e-05, "loss": 0.0552, "step": 2220 }, { "epoch": 1.982222222222222, "grad_norm": 0.3782356083393097, "learning_rate": 9.99297653595547e-05, "loss": 0.0376, "step": 2230 }, { "epoch": 1.991111111111111, "grad_norm": 0.4678187072277069, "learning_rate": 9.992714552442741e-05, "loss": 0.0429, "step": 2240 }, { "epoch": 2.0, "grad_norm": 0.9699333906173706, "learning_rate": 9.992447775402459e-05, "loss": 0.0466, "step": 2250 }, { "epoch": 2.008888888888889, "grad_norm": 0.7402211427688599, "learning_rate": 9.992176205090757e-05, "loss": 0.0483, "step": 2260 }, { "epoch": 2.017777777777778, "grad_norm": 0.6028904914855957, "learning_rate": 9.991899841768373e-05, "loss": 0.0446, "step": 2270 }, { "epoch": 2.026666666666667, "grad_norm": 0.39548084139823914, "learning_rate": 9.991618685700643e-05, "loss": 0.0436, "step": 2280 }, { "epoch": 2.0355555555555553, "grad_norm": 0.49818339943885803, "learning_rate": 9.991332737157508e-05, "loss": 0.0489, "step": 2290 }, { "epoch": 2.0444444444444443, "grad_norm": 0.44063156843185425, "learning_rate": 9.991041996413507e-05, "loss": 0.0477, "step": 2300 }, { "epoch": 2.0533333333333332, "grad_norm": 0.5084669589996338, "learning_rate": 9.99074646374778e-05, "loss": 0.047, "step": 2310 }, { "epoch": 2.062222222222222, "grad_norm": 0.3018571734428406, "learning_rate": 9.990446139444072e-05, "loss": 0.0409, "step": 2320 }, { "epoch": 2.071111111111111, "grad_norm": 0.46813327074050903, "learning_rate": 9.990141023790725e-05, "loss": 0.0425, "step": 2330 }, { "epoch": 2.08, "grad_norm": 0.3539779484272003, "learning_rate": 9.989831117080678e-05, "loss": 0.0475, "step": 2340 }, { "epoch": 2.088888888888889, "grad_norm": 0.5678339600563049, "learning_rate": 9.989516419611483e-05, "loss": 0.0519, "step": 2350 }, { "epoch": 2.097777777777778, "grad_norm": 0.6110586524009705, "learning_rate": 9.989196931685272e-05, "loss": 0.0448, "step": 2360 }, { "epoch": 2.1066666666666665, "grad_norm": 0.4977332353591919, "learning_rate": 9.988872653608794e-05, "loss": 0.0412, "step": 2370 }, { "epoch": 2.1155555555555554, "grad_norm": 0.498484343290329, "learning_rate": 9.988543585693386e-05, "loss": 0.0402, "step": 2380 }, { "epoch": 2.1244444444444444, "grad_norm": 0.4707137644290924, "learning_rate": 9.988209728254991e-05, "loss": 0.0462, "step": 2390 }, { "epoch": 2.1333333333333333, "grad_norm": 0.5397852063179016, "learning_rate": 9.987871081614144e-05, "loss": 0.0504, "step": 2400 }, { "epoch": 2.1422222222222222, "grad_norm": 0.4427320659160614, "learning_rate": 9.987527646095983e-05, "loss": 0.0459, "step": 2410 }, { "epoch": 2.151111111111111, "grad_norm": 0.36918312311172485, "learning_rate": 9.987179422030238e-05, "loss": 0.0549, "step": 2420 }, { "epoch": 2.16, "grad_norm": 0.46088486909866333, "learning_rate": 9.986826409751247e-05, "loss": 0.0427, "step": 2430 }, { "epoch": 2.168888888888889, "grad_norm": 0.5316842198371887, "learning_rate": 9.986468609597934e-05, "loss": 0.0381, "step": 2440 }, { "epoch": 2.1777777777777776, "grad_norm": 0.43516552448272705, "learning_rate": 9.986106021913824e-05, "loss": 0.0399, "step": 2450 }, { "epoch": 2.1866666666666665, "grad_norm": 0.532513439655304, "learning_rate": 9.985738647047044e-05, "loss": 0.0433, "step": 2460 }, { "epoch": 2.1955555555555555, "grad_norm": 0.32671698927879333, "learning_rate": 9.985366485350307e-05, "loss": 0.0423, "step": 2470 }, { "epoch": 2.2044444444444444, "grad_norm": 0.5176675915718079, "learning_rate": 9.98498953718093e-05, "loss": 0.0445, "step": 2480 }, { "epoch": 2.2133333333333334, "grad_norm": 0.4066351652145386, "learning_rate": 9.984607802900821e-05, "loss": 0.0418, "step": 2490 }, { "epoch": 2.2222222222222223, "grad_norm": 0.4666891098022461, "learning_rate": 9.984221282876486e-05, "loss": 0.0406, "step": 2500 }, { "epoch": 2.2311111111111113, "grad_norm": 0.40268608927726746, "learning_rate": 9.983829977479025e-05, "loss": 0.0405, "step": 2510 }, { "epoch": 2.24, "grad_norm": 0.5873245596885681, "learning_rate": 9.983433887084132e-05, "loss": 0.0434, "step": 2520 }, { "epoch": 2.2488888888888887, "grad_norm": 0.4484671950340271, "learning_rate": 9.983033012072093e-05, "loss": 0.0404, "step": 2530 }, { "epoch": 2.2577777777777777, "grad_norm": 0.5311611294746399, "learning_rate": 9.982627352827791e-05, "loss": 0.0483, "step": 2540 }, { "epoch": 2.2666666666666666, "grad_norm": 0.6491085290908813, "learning_rate": 9.982216909740702e-05, "loss": 0.044, "step": 2550 }, { "epoch": 2.2755555555555556, "grad_norm": 0.3876890540122986, "learning_rate": 9.981801683204895e-05, "loss": 0.0433, "step": 2560 }, { "epoch": 2.2844444444444445, "grad_norm": 0.6076945066452026, "learning_rate": 9.981381673619028e-05, "loss": 0.0467, "step": 2570 }, { "epoch": 2.2933333333333334, "grad_norm": 0.5941410660743713, "learning_rate": 9.980956881386356e-05, "loss": 0.0409, "step": 2580 }, { "epoch": 2.3022222222222224, "grad_norm": 0.3566558063030243, "learning_rate": 9.980527306914725e-05, "loss": 0.0418, "step": 2590 }, { "epoch": 2.311111111111111, "grad_norm": 0.5357877612113953, "learning_rate": 9.980092950616568e-05, "loss": 0.042, "step": 2600 }, { "epoch": 2.32, "grad_norm": 0.37787678837776184, "learning_rate": 9.979653812908916e-05, "loss": 0.035, "step": 2610 }, { "epoch": 2.328888888888889, "grad_norm": 0.5237792134284973, "learning_rate": 9.979209894213386e-05, "loss": 0.0365, "step": 2620 }, { "epoch": 2.3377777777777777, "grad_norm": 0.4795762598514557, "learning_rate": 9.978761194956184e-05, "loss": 0.0399, "step": 2630 }, { "epoch": 2.3466666666666667, "grad_norm": 0.5068551301956177, "learning_rate": 9.978307715568111e-05, "loss": 0.0407, "step": 2640 }, { "epoch": 2.3555555555555556, "grad_norm": 0.4115622043609619, "learning_rate": 9.977849456484553e-05, "loss": 0.0425, "step": 2650 }, { "epoch": 2.3644444444444446, "grad_norm": 0.3817325830459595, "learning_rate": 9.977386418145489e-05, "loss": 0.0333, "step": 2660 }, { "epoch": 2.3733333333333335, "grad_norm": 0.26332467794418335, "learning_rate": 9.976918600995482e-05, "loss": 0.0472, "step": 2670 }, { "epoch": 2.3822222222222225, "grad_norm": 0.41719910502433777, "learning_rate": 9.976446005483684e-05, "loss": 0.0421, "step": 2680 }, { "epoch": 2.391111111111111, "grad_norm": 0.48341479897499084, "learning_rate": 9.975968632063839e-05, "loss": 0.0366, "step": 2690 }, { "epoch": 2.4, "grad_norm": 0.3212173879146576, "learning_rate": 9.975486481194274e-05, "loss": 0.043, "step": 2700 }, { "epoch": 2.408888888888889, "grad_norm": 0.3700253665447235, "learning_rate": 9.974999553337904e-05, "loss": 0.0381, "step": 2710 }, { "epoch": 2.417777777777778, "grad_norm": 0.39644119143486023, "learning_rate": 9.974507848962231e-05, "loss": 0.0396, "step": 2720 }, { "epoch": 2.4266666666666667, "grad_norm": 0.4125751256942749, "learning_rate": 9.974011368539342e-05, "loss": 0.0415, "step": 2730 }, { "epoch": 2.4355555555555557, "grad_norm": 0.2982364892959595, "learning_rate": 9.973510112545911e-05, "loss": 0.0485, "step": 2740 }, { "epoch": 2.4444444444444446, "grad_norm": 0.4623163640499115, "learning_rate": 9.973004081463197e-05, "loss": 0.036, "step": 2750 }, { "epoch": 2.453333333333333, "grad_norm": 0.39890429377555847, "learning_rate": 9.97249327577704e-05, "loss": 0.0498, "step": 2760 }, { "epoch": 2.462222222222222, "grad_norm": 0.4336410164833069, "learning_rate": 9.97197769597787e-05, "loss": 0.0462, "step": 2770 }, { "epoch": 2.471111111111111, "grad_norm": 0.46421414613723755, "learning_rate": 9.971457342560694e-05, "loss": 0.0418, "step": 2780 }, { "epoch": 2.48, "grad_norm": 0.5289233326911926, "learning_rate": 9.970932216025108e-05, "loss": 0.0434, "step": 2790 }, { "epoch": 2.488888888888889, "grad_norm": 0.42652541399002075, "learning_rate": 9.970402316875287e-05, "loss": 0.0384, "step": 2800 }, { "epoch": 2.497777777777778, "grad_norm": 0.3700074851512909, "learning_rate": 9.96986764561999e-05, "loss": 0.04, "step": 2810 }, { "epoch": 2.506666666666667, "grad_norm": 0.42075055837631226, "learning_rate": 9.969328202772559e-05, "loss": 0.0408, "step": 2820 }, { "epoch": 2.5155555555555553, "grad_norm": 0.47732624411582947, "learning_rate": 9.968783988850913e-05, "loss": 0.0494, "step": 2830 }, { "epoch": 2.5244444444444447, "grad_norm": 0.691931962966919, "learning_rate": 9.968235004377555e-05, "loss": 0.0472, "step": 2840 }, { "epoch": 2.533333333333333, "grad_norm": 0.5139862298965454, "learning_rate": 9.967681249879567e-05, "loss": 0.0413, "step": 2850 }, { "epoch": 2.542222222222222, "grad_norm": 0.5603801012039185, "learning_rate": 9.967122725888614e-05, "loss": 0.0406, "step": 2860 }, { "epoch": 2.551111111111111, "grad_norm": 0.47628894448280334, "learning_rate": 9.966559432940934e-05, "loss": 0.0394, "step": 2870 }, { "epoch": 2.56, "grad_norm": 0.35130104422569275, "learning_rate": 9.965991371577348e-05, "loss": 0.037, "step": 2880 }, { "epoch": 2.568888888888889, "grad_norm": 0.5226022005081177, "learning_rate": 9.965418542343254e-05, "loss": 0.0454, "step": 2890 }, { "epoch": 2.5777777777777775, "grad_norm": 0.4336131513118744, "learning_rate": 9.96484094578863e-05, "loss": 0.0441, "step": 2900 }, { "epoch": 2.586666666666667, "grad_norm": 0.4621409773826599, "learning_rate": 9.964258582468026e-05, "loss": 0.044, "step": 2910 }, { "epoch": 2.5955555555555554, "grad_norm": 0.5120822787284851, "learning_rate": 9.963671452940576e-05, "loss": 0.0468, "step": 2920 }, { "epoch": 2.6044444444444443, "grad_norm": 0.4062204658985138, "learning_rate": 9.963079557769978e-05, "loss": 0.038, "step": 2930 }, { "epoch": 2.6133333333333333, "grad_norm": 0.4003904461860657, "learning_rate": 9.962482897524521e-05, "loss": 0.0391, "step": 2940 }, { "epoch": 2.6222222222222222, "grad_norm": 0.6013824939727783, "learning_rate": 9.961881472777058e-05, "loss": 0.04, "step": 2950 }, { "epoch": 2.631111111111111, "grad_norm": 0.45361050963401794, "learning_rate": 9.96127528410502e-05, "loss": 0.0422, "step": 2960 }, { "epoch": 2.64, "grad_norm": 0.3957284092903137, "learning_rate": 9.96066433209041e-05, "loss": 0.0431, "step": 2970 }, { "epoch": 2.648888888888889, "grad_norm": 0.7024198770523071, "learning_rate": 9.960048617319808e-05, "loss": 0.043, "step": 2980 }, { "epoch": 2.6577777777777776, "grad_norm": 0.46282336115837097, "learning_rate": 9.959428140384362e-05, "loss": 0.0357, "step": 2990 }, { "epoch": 2.6666666666666665, "grad_norm": 0.4048396646976471, "learning_rate": 9.958802901879798e-05, "loss": 0.0362, "step": 3000 }, { "epoch": 2.6755555555555555, "grad_norm": 0.44506096839904785, "learning_rate": 9.958172902406407e-05, "loss": 0.0417, "step": 3010 }, { "epoch": 2.6844444444444444, "grad_norm": 0.43336114287376404, "learning_rate": 9.957538142569056e-05, "loss": 0.0424, "step": 3020 }, { "epoch": 2.6933333333333334, "grad_norm": 0.43466073274612427, "learning_rate": 9.95689862297718e-05, "loss": 0.0429, "step": 3030 }, { "epoch": 2.7022222222222223, "grad_norm": 0.32737648487091064, "learning_rate": 9.956254344244787e-05, "loss": 0.0377, "step": 3040 }, { "epoch": 2.7111111111111112, "grad_norm": 0.3841151297092438, "learning_rate": 9.955605306990448e-05, "loss": 0.0465, "step": 3050 }, { "epoch": 2.7199999999999998, "grad_norm": 0.3416834771633148, "learning_rate": 9.954951511837309e-05, "loss": 0.0354, "step": 3060 }, { "epoch": 2.728888888888889, "grad_norm": 0.34268301725387573, "learning_rate": 9.954292959413081e-05, "loss": 0.044, "step": 3070 }, { "epoch": 2.7377777777777776, "grad_norm": 0.35896554589271545, "learning_rate": 9.953629650350045e-05, "loss": 0.0341, "step": 3080 }, { "epoch": 2.7466666666666666, "grad_norm": 0.38793715834617615, "learning_rate": 9.952961585285043e-05, "loss": 0.0395, "step": 3090 }, { "epoch": 2.7555555555555555, "grad_norm": 0.4323543608188629, "learning_rate": 9.952288764859491e-05, "loss": 0.0369, "step": 3100 }, { "epoch": 2.7644444444444445, "grad_norm": 0.39903882145881653, "learning_rate": 9.951611189719366e-05, "loss": 0.0374, "step": 3110 }, { "epoch": 2.7733333333333334, "grad_norm": 0.39197343587875366, "learning_rate": 9.950928860515209e-05, "loss": 0.0372, "step": 3120 }, { "epoch": 2.7822222222222224, "grad_norm": 0.37400147318840027, "learning_rate": 9.950241777902128e-05, "loss": 0.0318, "step": 3130 }, { "epoch": 2.7911111111111113, "grad_norm": 0.2922102212905884, "learning_rate": 9.949549942539796e-05, "loss": 0.037, "step": 3140 }, { "epoch": 2.8, "grad_norm": 0.48542511463165283, "learning_rate": 9.948853355092446e-05, "loss": 0.0495, "step": 3150 }, { "epoch": 2.8088888888888888, "grad_norm": 0.42923346161842346, "learning_rate": 9.948152016228875e-05, "loss": 0.0445, "step": 3160 }, { "epoch": 2.8177777777777777, "grad_norm": 0.4559933543205261, "learning_rate": 9.94744592662244e-05, "loss": 0.0363, "step": 3170 }, { "epoch": 2.8266666666666667, "grad_norm": 0.46220526099205017, "learning_rate": 9.946735086951064e-05, "loss": 0.0428, "step": 3180 }, { "epoch": 2.8355555555555556, "grad_norm": 0.5164117217063904, "learning_rate": 9.946019497897224e-05, "loss": 0.0416, "step": 3190 }, { "epoch": 2.8444444444444446, "grad_norm": 0.4774520695209503, "learning_rate": 9.945299160147963e-05, "loss": 0.0446, "step": 3200 }, { "epoch": 2.8533333333333335, "grad_norm": 0.48979097604751587, "learning_rate": 9.944574074394879e-05, "loss": 0.0397, "step": 3210 }, { "epoch": 2.862222222222222, "grad_norm": 0.34886378049850464, "learning_rate": 9.943844241334128e-05, "loss": 0.0404, "step": 3220 }, { "epoch": 2.871111111111111, "grad_norm": 0.43473726511001587, "learning_rate": 9.94310966166643e-05, "loss": 0.0481, "step": 3230 }, { "epoch": 2.88, "grad_norm": 0.5907363295555115, "learning_rate": 9.942370336097057e-05, "loss": 0.0407, "step": 3240 }, { "epoch": 2.888888888888889, "grad_norm": 0.4862346053123474, "learning_rate": 9.941626265335837e-05, "loss": 0.0392, "step": 3250 }, { "epoch": 2.897777777777778, "grad_norm": 0.4517713189125061, "learning_rate": 9.940877450097158e-05, "loss": 0.0373, "step": 3260 }, { "epoch": 2.9066666666666667, "grad_norm": 0.41444066166877747, "learning_rate": 9.940123891099959e-05, "loss": 0.0371, "step": 3270 }, { "epoch": 2.9155555555555557, "grad_norm": 0.4749915599822998, "learning_rate": 9.939365589067736e-05, "loss": 0.0357, "step": 3280 }, { "epoch": 2.924444444444444, "grad_norm": 0.33119961619377136, "learning_rate": 9.938602544728539e-05, "loss": 0.0342, "step": 3290 }, { "epoch": 2.9333333333333336, "grad_norm": 0.43718770146369934, "learning_rate": 9.93783475881497e-05, "loss": 0.0392, "step": 3300 }, { "epoch": 2.942222222222222, "grad_norm": 0.5572205185890198, "learning_rate": 9.937062232064182e-05, "loss": 0.035, "step": 3310 }, { "epoch": 2.951111111111111, "grad_norm": 0.6048962473869324, "learning_rate": 9.936284965217883e-05, "loss": 0.0418, "step": 3320 }, { "epoch": 2.96, "grad_norm": 0.36902597546577454, "learning_rate": 9.93550295902233e-05, "loss": 0.0431, "step": 3330 }, { "epoch": 2.968888888888889, "grad_norm": 0.3841482996940613, "learning_rate": 9.93471621422833e-05, "loss": 0.033, "step": 3340 }, { "epoch": 2.977777777777778, "grad_norm": 0.5407798886299133, "learning_rate": 9.933924731591242e-05, "loss": 0.0419, "step": 3350 }, { "epoch": 2.986666666666667, "grad_norm": 0.4398995637893677, "learning_rate": 9.93312851187097e-05, "loss": 0.0357, "step": 3360 }, { "epoch": 2.9955555555555557, "grad_norm": 0.4541158080101013, "learning_rate": 9.93232755583197e-05, "loss": 0.0391, "step": 3370 }, { "epoch": 3.0044444444444443, "grad_norm": 0.4378274381160736, "learning_rate": 9.931521864243243e-05, "loss": 0.0407, "step": 3380 }, { "epoch": 3.013333333333333, "grad_norm": 0.4993390440940857, "learning_rate": 9.930711437878334e-05, "loss": 0.0372, "step": 3390 }, { "epoch": 3.022222222222222, "grad_norm": 0.4270539879798889, "learning_rate": 9.929896277515341e-05, "loss": 0.0376, "step": 3400 }, { "epoch": 3.031111111111111, "grad_norm": 0.5044734477996826, "learning_rate": 9.929076383936901e-05, "loss": 0.0384, "step": 3410 }, { "epoch": 3.04, "grad_norm": 0.41071414947509766, "learning_rate": 9.9282517579302e-05, "loss": 0.0373, "step": 3420 }, { "epoch": 3.048888888888889, "grad_norm": 0.4080047905445099, "learning_rate": 9.927422400286962e-05, "loss": 0.0403, "step": 3430 }, { "epoch": 3.057777777777778, "grad_norm": 0.4767039716243744, "learning_rate": 9.926588311803457e-05, "loss": 0.0377, "step": 3440 }, { "epoch": 3.066666666666667, "grad_norm": 0.3855324387550354, "learning_rate": 9.9257494932805e-05, "loss": 0.0468, "step": 3450 }, { "epoch": 3.0755555555555554, "grad_norm": 0.2699227035045624, "learning_rate": 9.924905945523442e-05, "loss": 0.0334, "step": 3460 }, { "epoch": 3.0844444444444443, "grad_norm": 0.33927956223487854, "learning_rate": 9.92405766934218e-05, "loss": 0.0342, "step": 3470 }, { "epoch": 3.0933333333333333, "grad_norm": 0.43876272439956665, "learning_rate": 9.923204665551142e-05, "loss": 0.0367, "step": 3480 }, { "epoch": 3.102222222222222, "grad_norm": 0.3528018891811371, "learning_rate": 9.922346934969306e-05, "loss": 0.0374, "step": 3490 }, { "epoch": 3.111111111111111, "grad_norm": 0.37652161717414856, "learning_rate": 9.92148447842018e-05, "loss": 0.039, "step": 3500 }, { "epoch": 3.12, "grad_norm": 0.32494619488716125, "learning_rate": 9.920617296731813e-05, "loss": 0.0425, "step": 3510 }, { "epoch": 3.128888888888889, "grad_norm": 0.4178219437599182, "learning_rate": 9.919745390736792e-05, "loss": 0.0398, "step": 3520 }, { "epoch": 3.137777777777778, "grad_norm": 0.6598771810531616, "learning_rate": 9.918868761272238e-05, "loss": 0.0433, "step": 3530 }, { "epoch": 3.1466666666666665, "grad_norm": 0.5390450954437256, "learning_rate": 9.917987409179801e-05, "loss": 0.0379, "step": 3540 }, { "epoch": 3.1555555555555554, "grad_norm": 0.36635622382164, "learning_rate": 9.917101335305678e-05, "loss": 0.036, "step": 3550 }, { "epoch": 3.1644444444444444, "grad_norm": 0.33388134837150574, "learning_rate": 9.91621054050059e-05, "loss": 0.0414, "step": 3560 }, { "epoch": 3.1733333333333333, "grad_norm": 0.4057382643222809, "learning_rate": 9.915315025619792e-05, "loss": 0.0381, "step": 3570 }, { "epoch": 3.1822222222222223, "grad_norm": 0.425436794757843, "learning_rate": 9.914414791523073e-05, "loss": 0.0424, "step": 3580 }, { "epoch": 3.1911111111111112, "grad_norm": 0.4099542796611786, "learning_rate": 9.913509839074753e-05, "loss": 0.0393, "step": 3590 }, { "epoch": 3.2, "grad_norm": 0.40293025970458984, "learning_rate": 9.912600169143677e-05, "loss": 0.0347, "step": 3600 }, { "epoch": 3.2088888888888887, "grad_norm": 0.4650910198688507, "learning_rate": 9.911685782603226e-05, "loss": 0.0351, "step": 3610 }, { "epoch": 3.2177777777777776, "grad_norm": 0.3463270366191864, "learning_rate": 9.910766680331307e-05, "loss": 0.0454, "step": 3620 }, { "epoch": 3.2266666666666666, "grad_norm": 0.3617923855781555, "learning_rate": 9.909842863210353e-05, "loss": 0.0391, "step": 3630 }, { "epoch": 3.2355555555555555, "grad_norm": 0.501361072063446, "learning_rate": 9.908914332127327e-05, "loss": 0.0401, "step": 3640 }, { "epoch": 3.2444444444444445, "grad_norm": 0.591056764125824, "learning_rate": 9.907981087973711e-05, "loss": 0.0356, "step": 3650 }, { "epoch": 3.2533333333333334, "grad_norm": 0.3857851028442383, "learning_rate": 9.907043131645522e-05, "loss": 0.0405, "step": 3660 }, { "epoch": 3.2622222222222224, "grad_norm": 0.5137868523597717, "learning_rate": 9.906100464043293e-05, "loss": 0.0404, "step": 3670 }, { "epoch": 3.2711111111111113, "grad_norm": 0.4093244969844818, "learning_rate": 9.905153086072086e-05, "loss": 0.0426, "step": 3680 }, { "epoch": 3.2800000000000002, "grad_norm": 0.4638502895832062, "learning_rate": 9.90420099864148e-05, "loss": 0.0471, "step": 3690 }, { "epoch": 3.2888888888888888, "grad_norm": 0.4333135485649109, "learning_rate": 9.903244202665581e-05, "loss": 0.0347, "step": 3700 }, { "epoch": 3.2977777777777777, "grad_norm": 0.4751374423503876, "learning_rate": 9.90228269906301e-05, "loss": 0.0431, "step": 3710 }, { "epoch": 3.3066666666666666, "grad_norm": 0.33454063534736633, "learning_rate": 9.901316488756914e-05, "loss": 0.0373, "step": 3720 }, { "epoch": 3.3155555555555556, "grad_norm": 0.43450284004211426, "learning_rate": 9.900345572674954e-05, "loss": 0.0401, "step": 3730 }, { "epoch": 3.3244444444444445, "grad_norm": 0.44553816318511963, "learning_rate": 9.899369951749312e-05, "loss": 0.0378, "step": 3740 }, { "epoch": 3.3333333333333335, "grad_norm": 0.42485615611076355, "learning_rate": 9.898389626916685e-05, "loss": 0.0366, "step": 3750 }, { "epoch": 3.3422222222222224, "grad_norm": 0.5780754089355469, "learning_rate": 9.897404599118287e-05, "loss": 0.0414, "step": 3760 }, { "epoch": 3.351111111111111, "grad_norm": 0.46000218391418457, "learning_rate": 9.89641486929985e-05, "loss": 0.0377, "step": 3770 }, { "epoch": 3.36, "grad_norm": 0.48487913608551025, "learning_rate": 9.895420438411616e-05, "loss": 0.0375, "step": 3780 }, { "epoch": 3.368888888888889, "grad_norm": 0.3728860914707184, "learning_rate": 9.894421307408341e-05, "loss": 0.0447, "step": 3790 }, { "epoch": 3.3777777777777778, "grad_norm": 0.29233887791633606, "learning_rate": 9.893417477249299e-05, "loss": 0.0436, "step": 3800 }, { "epoch": 3.3866666666666667, "grad_norm": 0.3768678903579712, "learning_rate": 9.892408948898267e-05, "loss": 0.038, "step": 3810 }, { "epoch": 3.3955555555555557, "grad_norm": 0.2894018590450287, "learning_rate": 9.891395723323542e-05, "loss": 0.0352, "step": 3820 }, { "epoch": 3.4044444444444446, "grad_norm": 0.5104938745498657, "learning_rate": 9.890377801497925e-05, "loss": 0.0366, "step": 3830 }, { "epoch": 3.413333333333333, "grad_norm": 0.3009335398674011, "learning_rate": 9.889355184398727e-05, "loss": 0.0312, "step": 3840 }, { "epoch": 3.422222222222222, "grad_norm": 0.4489968419075012, "learning_rate": 9.888327873007768e-05, "loss": 0.0316, "step": 3850 }, { "epoch": 3.431111111111111, "grad_norm": 0.5225479006767273, "learning_rate": 9.887295868311373e-05, "loss": 0.045, "step": 3860 }, { "epoch": 3.44, "grad_norm": 0.34820154309272766, "learning_rate": 9.886259171300374e-05, "loss": 0.0341, "step": 3870 }, { "epoch": 3.448888888888889, "grad_norm": 0.365832656621933, "learning_rate": 9.88521778297011e-05, "loss": 0.0363, "step": 3880 }, { "epoch": 3.457777777777778, "grad_norm": 0.39247530698776245, "learning_rate": 9.884171704320422e-05, "loss": 0.0306, "step": 3890 }, { "epoch": 3.466666666666667, "grad_norm": 0.4535060524940491, "learning_rate": 9.883120936355654e-05, "loss": 0.037, "step": 3900 }, { "epoch": 3.4755555555555557, "grad_norm": 0.35170453786849976, "learning_rate": 9.882065480084654e-05, "loss": 0.0334, "step": 3910 }, { "epoch": 3.4844444444444447, "grad_norm": 0.41468361020088196, "learning_rate": 9.88100533652077e-05, "loss": 0.0314, "step": 3920 }, { "epoch": 3.493333333333333, "grad_norm": 0.43466314673423767, "learning_rate": 9.879940506681852e-05, "loss": 0.0363, "step": 3930 }, { "epoch": 3.502222222222222, "grad_norm": 0.4654955267906189, "learning_rate": 9.878870991590246e-05, "loss": 0.0364, "step": 3940 }, { "epoch": 3.511111111111111, "grad_norm": 0.5055062174797058, "learning_rate": 9.877796792272796e-05, "loss": 0.0383, "step": 3950 }, { "epoch": 3.52, "grad_norm": 0.3462908864021301, "learning_rate": 9.876717909760851e-05, "loss": 0.0367, "step": 3960 }, { "epoch": 3.528888888888889, "grad_norm": 0.355025053024292, "learning_rate": 9.875634345090247e-05, "loss": 0.0434, "step": 3970 }, { "epoch": 3.537777777777778, "grad_norm": 0.381779283285141, "learning_rate": 9.87454609930132e-05, "loss": 0.0406, "step": 3980 }, { "epoch": 3.546666666666667, "grad_norm": 0.3355260193347931, "learning_rate": 9.873453173438902e-05, "loss": 0.0451, "step": 3990 }, { "epoch": 3.5555555555555554, "grad_norm": 0.31535857915878296, "learning_rate": 9.872355568552312e-05, "loss": 0.0388, "step": 4000 }, { "epoch": 3.5644444444444443, "grad_norm": 0.38463279604911804, "learning_rate": 9.871253285695368e-05, "loss": 0.0385, "step": 4010 }, { "epoch": 3.5733333333333333, "grad_norm": 0.3574258089065552, "learning_rate": 9.870146325926377e-05, "loss": 0.0361, "step": 4020 }, { "epoch": 3.582222222222222, "grad_norm": 0.47130995988845825, "learning_rate": 9.869034690308134e-05, "loss": 0.0444, "step": 4030 }, { "epoch": 3.591111111111111, "grad_norm": 0.36850252747535706, "learning_rate": 9.867918379907928e-05, "loss": 0.036, "step": 4040 }, { "epoch": 3.6, "grad_norm": 0.4545682966709137, "learning_rate": 9.866797395797531e-05, "loss": 0.043, "step": 4050 }, { "epoch": 3.608888888888889, "grad_norm": 0.4097519516944885, "learning_rate": 9.865671739053206e-05, "loss": 0.0351, "step": 4060 }, { "epoch": 3.6177777777777775, "grad_norm": 0.40340396761894226, "learning_rate": 9.864541410755702e-05, "loss": 0.0332, "step": 4070 }, { "epoch": 3.626666666666667, "grad_norm": 0.3582291603088379, "learning_rate": 9.86340641199025e-05, "loss": 0.044, "step": 4080 }, { "epoch": 3.6355555555555554, "grad_norm": 0.3430737555027008, "learning_rate": 9.862266743846568e-05, "loss": 0.0317, "step": 4090 }, { "epoch": 3.6444444444444444, "grad_norm": 0.445867657661438, "learning_rate": 9.861122407418857e-05, "loss": 0.0427, "step": 4100 }, { "epoch": 3.6533333333333333, "grad_norm": 0.4102824032306671, "learning_rate": 9.859973403805802e-05, "loss": 0.0339, "step": 4110 }, { "epoch": 3.6622222222222223, "grad_norm": 0.4015534818172455, "learning_rate": 9.858819734110563e-05, "loss": 0.0341, "step": 4120 }, { "epoch": 3.671111111111111, "grad_norm": 0.4085457921028137, "learning_rate": 9.857661399440785e-05, "loss": 0.034, "step": 4130 }, { "epoch": 3.68, "grad_norm": 0.3532131314277649, "learning_rate": 9.856498400908588e-05, "loss": 0.0382, "step": 4140 }, { "epoch": 3.688888888888889, "grad_norm": 0.5730438232421875, "learning_rate": 9.855330739630575e-05, "loss": 0.0465, "step": 4150 }, { "epoch": 3.6977777777777776, "grad_norm": 0.4944850504398346, "learning_rate": 9.854158416727822e-05, "loss": 0.0393, "step": 4160 }, { "epoch": 3.7066666666666666, "grad_norm": 0.4614255130290985, "learning_rate": 9.852981433325882e-05, "loss": 0.036, "step": 4170 }, { "epoch": 3.7155555555555555, "grad_norm": 0.3927440345287323, "learning_rate": 9.85179979055478e-05, "loss": 0.0376, "step": 4180 }, { "epoch": 3.7244444444444444, "grad_norm": 0.42971909046173096, "learning_rate": 9.850613489549019e-05, "loss": 0.0359, "step": 4190 }, { "epoch": 3.7333333333333334, "grad_norm": 0.3831291198730469, "learning_rate": 9.849422531447569e-05, "loss": 0.038, "step": 4200 }, { "epoch": 3.7422222222222223, "grad_norm": 0.41654643416404724, "learning_rate": 9.848226917393876e-05, "loss": 0.0346, "step": 4210 }, { "epoch": 3.7511111111111113, "grad_norm": 0.3389831781387329, "learning_rate": 9.847026648535856e-05, "loss": 0.04, "step": 4220 }, { "epoch": 3.76, "grad_norm": 0.545345664024353, "learning_rate": 9.84582172602589e-05, "loss": 0.043, "step": 4230 }, { "epoch": 3.7688888888888887, "grad_norm": 0.49265724420547485, "learning_rate": 9.84461215102083e-05, "loss": 0.0369, "step": 4240 }, { "epoch": 3.7777777777777777, "grad_norm": 0.34559866786003113, "learning_rate": 9.843397924681994e-05, "loss": 0.0378, "step": 4250 }, { "epoch": 3.7866666666666666, "grad_norm": 0.3168230354785919, "learning_rate": 9.842179048175167e-05, "loss": 0.0356, "step": 4260 }, { "epoch": 3.7955555555555556, "grad_norm": 0.4163779020309448, "learning_rate": 9.840955522670599e-05, "loss": 0.0401, "step": 4270 }, { "epoch": 3.8044444444444445, "grad_norm": 0.2749326825141907, "learning_rate": 9.839727349342998e-05, "loss": 0.0352, "step": 4280 }, { "epoch": 3.8133333333333335, "grad_norm": 0.4113200604915619, "learning_rate": 9.838494529371544e-05, "loss": 0.0419, "step": 4290 }, { "epoch": 3.822222222222222, "grad_norm": 0.2543994188308716, "learning_rate": 9.837257063939869e-05, "loss": 0.0372, "step": 4300 }, { "epoch": 3.8311111111111114, "grad_norm": 0.47850728034973145, "learning_rate": 9.83601495423607e-05, "loss": 0.0387, "step": 4310 }, { "epoch": 3.84, "grad_norm": 0.4750727713108063, "learning_rate": 9.834768201452705e-05, "loss": 0.0321, "step": 4320 }, { "epoch": 3.848888888888889, "grad_norm": 0.35429051518440247, "learning_rate": 9.833516806786782e-05, "loss": 0.0408, "step": 4330 }, { "epoch": 3.8577777777777778, "grad_norm": 0.43823400139808655, "learning_rate": 9.832260771439775e-05, "loss": 0.035, "step": 4340 }, { "epoch": 3.8666666666666667, "grad_norm": 0.3537866473197937, "learning_rate": 9.831000096617606e-05, "loss": 0.0392, "step": 4350 }, { "epoch": 3.8755555555555556, "grad_norm": 0.38295942544937134, "learning_rate": 9.829734783530658e-05, "loss": 0.0345, "step": 4360 }, { "epoch": 3.8844444444444446, "grad_norm": 0.2717559039592743, "learning_rate": 9.82846483339376e-05, "loss": 0.0367, "step": 4370 }, { "epoch": 3.8933333333333335, "grad_norm": 0.43603354692459106, "learning_rate": 9.8271902474262e-05, "loss": 0.0392, "step": 4380 }, { "epoch": 3.902222222222222, "grad_norm": 0.4227178990840912, "learning_rate": 9.825911026851709e-05, "loss": 0.0351, "step": 4390 }, { "epoch": 3.911111111111111, "grad_norm": 0.3930717408657074, "learning_rate": 9.824627172898478e-05, "loss": 0.0417, "step": 4400 }, { "epoch": 3.92, "grad_norm": 0.40021997690200806, "learning_rate": 9.823338686799137e-05, "loss": 0.0401, "step": 4410 }, { "epoch": 3.928888888888889, "grad_norm": 0.41574376821517944, "learning_rate": 9.822045569790769e-05, "loss": 0.0345, "step": 4420 }, { "epoch": 3.937777777777778, "grad_norm": 0.4086770713329315, "learning_rate": 9.820747823114901e-05, "loss": 0.0365, "step": 4430 }, { "epoch": 3.9466666666666668, "grad_norm": 0.39606887102127075, "learning_rate": 9.819445448017505e-05, "loss": 0.0407, "step": 4440 }, { "epoch": 3.9555555555555557, "grad_norm": 0.383430540561676, "learning_rate": 9.818138445748999e-05, "loss": 0.0379, "step": 4450 }, { "epoch": 3.964444444444444, "grad_norm": 0.4091561734676361, "learning_rate": 9.816826817564239e-05, "loss": 0.0395, "step": 4460 }, { "epoch": 3.9733333333333336, "grad_norm": 0.41599440574645996, "learning_rate": 9.815510564722527e-05, "loss": 0.0408, "step": 4470 }, { "epoch": 3.982222222222222, "grad_norm": 0.3528141379356384, "learning_rate": 9.814189688487603e-05, "loss": 0.0343, "step": 4480 }, { "epoch": 3.991111111111111, "grad_norm": 0.6037723422050476, "learning_rate": 9.812864190127647e-05, "loss": 0.0321, "step": 4490 }, { "epoch": 4.0, "grad_norm": 0.7627918720245361, "learning_rate": 9.811534070915271e-05, "loss": 0.0436, "step": 4500 }, { "epoch": 4.0088888888888885, "grad_norm": 0.40366944670677185, "learning_rate": 9.810199332127536e-05, "loss": 0.0404, "step": 4510 }, { "epoch": 4.017777777777778, "grad_norm": 0.44338035583496094, "learning_rate": 9.808859975045924e-05, "loss": 0.0314, "step": 4520 }, { "epoch": 4.026666666666666, "grad_norm": 0.23564976453781128, "learning_rate": 9.807516000956362e-05, "loss": 0.0415, "step": 4530 }, { "epoch": 4.035555555555556, "grad_norm": 0.48682844638824463, "learning_rate": 9.806167411149203e-05, "loss": 0.0438, "step": 4540 }, { "epoch": 4.044444444444444, "grad_norm": 0.4615076184272766, "learning_rate": 9.804814206919236e-05, "loss": 0.0311, "step": 4550 }, { "epoch": 4.053333333333334, "grad_norm": 0.35988786816596985, "learning_rate": 9.803456389565678e-05, "loss": 0.0364, "step": 4560 }, { "epoch": 4.062222222222222, "grad_norm": 0.5323358774185181, "learning_rate": 9.802093960392174e-05, "loss": 0.0348, "step": 4570 }, { "epoch": 4.071111111111111, "grad_norm": 0.46436238288879395, "learning_rate": 9.8007269207068e-05, "loss": 0.0448, "step": 4580 }, { "epoch": 4.08, "grad_norm": 0.3452894985675812, "learning_rate": 9.799355271822055e-05, "loss": 0.0369, "step": 4590 }, { "epoch": 4.088888888888889, "grad_norm": 0.45791852474212646, "learning_rate": 9.797979015054867e-05, "loss": 0.0369, "step": 4600 }, { "epoch": 4.097777777777778, "grad_norm": 0.5014939308166504, "learning_rate": 9.796598151726588e-05, "loss": 0.0336, "step": 4610 }, { "epoch": 4.1066666666666665, "grad_norm": 0.4170559048652649, "learning_rate": 9.795212683162986e-05, "loss": 0.0442, "step": 4620 }, { "epoch": 4.115555555555556, "grad_norm": 0.47332754731178284, "learning_rate": 9.793822610694261e-05, "loss": 0.038, "step": 4630 }, { "epoch": 4.124444444444444, "grad_norm": 0.5621696710586548, "learning_rate": 9.792427935655024e-05, "loss": 0.0386, "step": 4640 }, { "epoch": 4.133333333333334, "grad_norm": 0.58274906873703, "learning_rate": 9.791028659384309e-05, "loss": 0.0354, "step": 4650 }, { "epoch": 4.142222222222222, "grad_norm": 0.4139659106731415, "learning_rate": 9.789624783225571e-05, "loss": 0.0387, "step": 4660 }, { "epoch": 4.151111111111111, "grad_norm": 0.6394178867340088, "learning_rate": 9.788216308526672e-05, "loss": 0.0418, "step": 4670 }, { "epoch": 4.16, "grad_norm": 0.4662797451019287, "learning_rate": 9.7868032366399e-05, "loss": 0.0403, "step": 4680 }, { "epoch": 4.168888888888889, "grad_norm": 0.6108471155166626, "learning_rate": 9.78538556892195e-05, "loss": 0.0385, "step": 4690 }, { "epoch": 4.177777777777778, "grad_norm": 0.49182555079460144, "learning_rate": 9.783963306733928e-05, "loss": 0.0342, "step": 4700 }, { "epoch": 4.1866666666666665, "grad_norm": 0.45262157917022705, "learning_rate": 9.78253645144136e-05, "loss": 0.0406, "step": 4710 }, { "epoch": 4.195555555555556, "grad_norm": 0.4807431995868683, "learning_rate": 9.781105004414171e-05, "loss": 0.0339, "step": 4720 }, { "epoch": 4.204444444444444, "grad_norm": 0.30490773916244507, "learning_rate": 9.779668967026703e-05, "loss": 0.042, "step": 4730 }, { "epoch": 4.213333333333333, "grad_norm": 0.4220482110977173, "learning_rate": 9.778228340657701e-05, "loss": 0.0368, "step": 4740 }, { "epoch": 4.222222222222222, "grad_norm": 0.35848045349121094, "learning_rate": 9.776783126690316e-05, "loss": 0.0343, "step": 4750 }, { "epoch": 4.231111111111111, "grad_norm": 0.3352150619029999, "learning_rate": 9.775333326512105e-05, "loss": 0.0386, "step": 4760 }, { "epoch": 4.24, "grad_norm": 0.36960306763648987, "learning_rate": 9.773878941515026e-05, "loss": 0.0329, "step": 4770 }, { "epoch": 4.248888888888889, "grad_norm": 0.4530479907989502, "learning_rate": 9.772419973095446e-05, "loss": 0.0373, "step": 4780 }, { "epoch": 4.257777777777778, "grad_norm": 0.38365840911865234, "learning_rate": 9.77095642265412e-05, "loss": 0.0392, "step": 4790 }, { "epoch": 4.266666666666667, "grad_norm": 0.33469241857528687, "learning_rate": 9.769488291596213e-05, "loss": 0.0333, "step": 4800 }, { "epoch": 4.275555555555556, "grad_norm": 0.30587515234947205, "learning_rate": 9.768015581331284e-05, "loss": 0.0417, "step": 4810 }, { "epoch": 4.2844444444444445, "grad_norm": 0.34831503033638, "learning_rate": 9.76653829327329e-05, "loss": 0.0412, "step": 4820 }, { "epoch": 4.293333333333333, "grad_norm": 0.3754204511642456, "learning_rate": 9.765056428840577e-05, "loss": 0.0379, "step": 4830 }, { "epoch": 4.302222222222222, "grad_norm": 0.3521517515182495, "learning_rate": 9.763569989455896e-05, "loss": 0.0359, "step": 4840 }, { "epoch": 4.311111111111111, "grad_norm": 0.41145581007003784, "learning_rate": 9.76207897654638e-05, "loss": 0.0332, "step": 4850 }, { "epoch": 4.32, "grad_norm": 0.39777234196662903, "learning_rate": 9.760583391543559e-05, "loss": 0.0407, "step": 4860 }, { "epoch": 4.328888888888889, "grad_norm": 0.3133266866207123, "learning_rate": 9.759083235883349e-05, "loss": 0.0384, "step": 4870 }, { "epoch": 4.337777777777778, "grad_norm": 0.45780590176582336, "learning_rate": 9.757578511006058e-05, "loss": 0.0366, "step": 4880 }, { "epoch": 4.346666666666667, "grad_norm": 0.5303405523300171, "learning_rate": 9.756069218356376e-05, "loss": 0.0388, "step": 4890 }, { "epoch": 4.355555555555555, "grad_norm": 0.3261024057865143, "learning_rate": 9.754555359383386e-05, "loss": 0.0362, "step": 4900 }, { "epoch": 4.364444444444445, "grad_norm": 0.3941381573677063, "learning_rate": 9.75303693554055e-05, "loss": 0.0392, "step": 4910 }, { "epoch": 4.373333333333333, "grad_norm": 0.4001061022281647, "learning_rate": 9.751513948285709e-05, "loss": 0.0334, "step": 4920 }, { "epoch": 4.3822222222222225, "grad_norm": 0.5368455052375793, "learning_rate": 9.749986399081095e-05, "loss": 0.0291, "step": 4930 }, { "epoch": 4.391111111111111, "grad_norm": 0.36943164467811584, "learning_rate": 9.748454289393313e-05, "loss": 0.0345, "step": 4940 }, { "epoch": 4.4, "grad_norm": 0.3861916661262512, "learning_rate": 9.746917620693348e-05, "loss": 0.0402, "step": 4950 }, { "epoch": 4.408888888888889, "grad_norm": 0.4154004752635956, "learning_rate": 9.745376394456562e-05, "loss": 0.042, "step": 4960 }, { "epoch": 4.417777777777777, "grad_norm": 0.36532604694366455, "learning_rate": 9.743830612162695e-05, "loss": 0.0361, "step": 4970 }, { "epoch": 4.426666666666667, "grad_norm": 0.4756120443344116, "learning_rate": 9.742280275295857e-05, "loss": 0.0365, "step": 4980 }, { "epoch": 4.435555555555555, "grad_norm": 0.3702144920825958, "learning_rate": 9.740725385344536e-05, "loss": 0.0353, "step": 4990 }, { "epoch": 4.444444444444445, "grad_norm": 0.4962805211544037, "learning_rate": 9.739165943801587e-05, "loss": 0.0377, "step": 5000 }, { "epoch": 4.453333333333333, "grad_norm": 0.25181570649147034, "learning_rate": 9.737601952164237e-05, "loss": 0.0344, "step": 5010 }, { "epoch": 4.4622222222222225, "grad_norm": 0.3841783404350281, "learning_rate": 9.736033411934083e-05, "loss": 0.0345, "step": 5020 }, { "epoch": 4.471111111111111, "grad_norm": 0.3889508545398712, "learning_rate": 9.734460324617085e-05, "loss": 0.04, "step": 5030 }, { "epoch": 4.48, "grad_norm": 0.3233131170272827, "learning_rate": 9.732882691723572e-05, "loss": 0.0374, "step": 5040 }, { "epoch": 4.488888888888889, "grad_norm": 0.327555775642395, "learning_rate": 9.731300514768238e-05, "loss": 0.0383, "step": 5050 }, { "epoch": 4.497777777777777, "grad_norm": 0.3682672679424286, "learning_rate": 9.729713795270136e-05, "loss": 0.032, "step": 5060 }, { "epoch": 4.506666666666667, "grad_norm": 0.37643906474113464, "learning_rate": 9.728122534752685e-05, "loss": 0.0314, "step": 5070 }, { "epoch": 4.515555555555555, "grad_norm": 0.31632721424102783, "learning_rate": 9.726526734743657e-05, "loss": 0.0324, "step": 5080 }, { "epoch": 4.524444444444445, "grad_norm": 0.32028165459632874, "learning_rate": 9.724926396775192e-05, "loss": 0.0371, "step": 5090 }, { "epoch": 4.533333333333333, "grad_norm": 0.3114984333515167, "learning_rate": 9.723321522383777e-05, "loss": 0.0355, "step": 5100 }, { "epoch": 4.542222222222223, "grad_norm": 0.4615582525730133, "learning_rate": 9.721712113110261e-05, "loss": 0.034, "step": 5110 }, { "epoch": 4.551111111111111, "grad_norm": 0.39241254329681396, "learning_rate": 9.720098170499848e-05, "loss": 0.0372, "step": 5120 }, { "epoch": 4.5600000000000005, "grad_norm": 0.3887083828449249, "learning_rate": 9.718479696102087e-05, "loss": 0.0347, "step": 5130 }, { "epoch": 4.568888888888889, "grad_norm": 0.46469056606292725, "learning_rate": 9.716856691470884e-05, "loss": 0.0305, "step": 5140 }, { "epoch": 4.5777777777777775, "grad_norm": 0.38900402188301086, "learning_rate": 9.715229158164494e-05, "loss": 0.0347, "step": 5150 }, { "epoch": 4.586666666666667, "grad_norm": 0.28600287437438965, "learning_rate": 9.713597097745517e-05, "loss": 0.0327, "step": 5160 }, { "epoch": 4.595555555555555, "grad_norm": 0.44450175762176514, "learning_rate": 9.711960511780903e-05, "loss": 0.0376, "step": 5170 }, { "epoch": 4.604444444444445, "grad_norm": 0.36220648884773254, "learning_rate": 9.710319401841945e-05, "loss": 0.0358, "step": 5180 }, { "epoch": 4.613333333333333, "grad_norm": 0.33447256684303284, "learning_rate": 9.708673769504281e-05, "loss": 0.0349, "step": 5190 }, { "epoch": 4.622222222222222, "grad_norm": 0.4241136610507965, "learning_rate": 9.70702361634789e-05, "loss": 0.0342, "step": 5200 }, { "epoch": 4.631111111111111, "grad_norm": 0.4919908940792084, "learning_rate": 9.70536894395709e-05, "loss": 0.0403, "step": 5210 }, { "epoch": 4.64, "grad_norm": 0.38925617933273315, "learning_rate": 9.703709753920538e-05, "loss": 0.0424, "step": 5220 }, { "epoch": 4.648888888888889, "grad_norm": 0.3211295008659363, "learning_rate": 9.702046047831233e-05, "loss": 0.0336, "step": 5230 }, { "epoch": 4.657777777777778, "grad_norm": 0.29306885600090027, "learning_rate": 9.700377827286505e-05, "loss": 0.0377, "step": 5240 }, { "epoch": 4.666666666666667, "grad_norm": 0.34621700644493103, "learning_rate": 9.698705093888021e-05, "loss": 0.0316, "step": 5250 }, { "epoch": 4.6755555555555555, "grad_norm": 0.323324590921402, "learning_rate": 9.69702784924178e-05, "loss": 0.0346, "step": 5260 }, { "epoch": 4.684444444444445, "grad_norm": 0.48832353949546814, "learning_rate": 9.695346094958108e-05, "loss": 0.0383, "step": 5270 }, { "epoch": 4.693333333333333, "grad_norm": 0.43838462233543396, "learning_rate": 9.693659832651672e-05, "loss": 0.0351, "step": 5280 }, { "epoch": 4.702222222222222, "grad_norm": 0.4823376536369324, "learning_rate": 9.691969063941454e-05, "loss": 0.0316, "step": 5290 }, { "epoch": 4.711111111111111, "grad_norm": 0.34601134061813354, "learning_rate": 9.69027379045077e-05, "loss": 0.0328, "step": 5300 }, { "epoch": 4.72, "grad_norm": 0.3575587570667267, "learning_rate": 9.688574013807261e-05, "loss": 0.0361, "step": 5310 }, { "epoch": 4.728888888888889, "grad_norm": 0.47709450125694275, "learning_rate": 9.68686973564289e-05, "loss": 0.033, "step": 5320 }, { "epoch": 4.737777777777778, "grad_norm": 0.36871641874313354, "learning_rate": 9.685160957593942e-05, "loss": 0.0359, "step": 5330 }, { "epoch": 4.746666666666667, "grad_norm": 0.32967424392700195, "learning_rate": 9.68344768130102e-05, "loss": 0.0371, "step": 5340 }, { "epoch": 4.7555555555555555, "grad_norm": 0.508967399597168, "learning_rate": 9.681729908409052e-05, "loss": 0.035, "step": 5350 }, { "epoch": 4.764444444444445, "grad_norm": 0.34443268179893494, "learning_rate": 9.680007640567277e-05, "loss": 0.0314, "step": 5360 }, { "epoch": 4.773333333333333, "grad_norm": 0.36615681648254395, "learning_rate": 9.678280879429251e-05, "loss": 0.0366, "step": 5370 }, { "epoch": 4.782222222222222, "grad_norm": 0.3773902654647827, "learning_rate": 9.67654962665285e-05, "loss": 0.0318, "step": 5380 }, { "epoch": 4.791111111111111, "grad_norm": 0.31916072964668274, "learning_rate": 9.674813883900251e-05, "loss": 0.0327, "step": 5390 }, { "epoch": 4.8, "grad_norm": 0.33504801988601685, "learning_rate": 9.673073652837952e-05, "loss": 0.0296, "step": 5400 }, { "epoch": 4.808888888888889, "grad_norm": 0.3753975033760071, "learning_rate": 9.671328935136758e-05, "loss": 0.0353, "step": 5410 }, { "epoch": 4.817777777777778, "grad_norm": 0.37190911173820496, "learning_rate": 9.669579732471779e-05, "loss": 0.034, "step": 5420 }, { "epoch": 4.826666666666666, "grad_norm": 0.3664259910583496, "learning_rate": 9.667826046522428e-05, "loss": 0.0325, "step": 5430 }, { "epoch": 4.835555555555556, "grad_norm": 0.36160075664520264, "learning_rate": 9.666067878972434e-05, "loss": 0.0324, "step": 5440 }, { "epoch": 4.844444444444444, "grad_norm": 0.4942307770252228, "learning_rate": 9.664305231509818e-05, "loss": 0.0338, "step": 5450 }, { "epoch": 4.8533333333333335, "grad_norm": 0.432277113199234, "learning_rate": 9.662538105826904e-05, "loss": 0.04, "step": 5460 }, { "epoch": 4.862222222222222, "grad_norm": 0.27758559584617615, "learning_rate": 9.66076650362032e-05, "loss": 0.0301, "step": 5470 }, { "epoch": 4.871111111111111, "grad_norm": 0.36470770835876465, "learning_rate": 9.658990426590988e-05, "loss": 0.0366, "step": 5480 }, { "epoch": 4.88, "grad_norm": 0.5110194683074951, "learning_rate": 9.657209876444128e-05, "loss": 0.0333, "step": 5490 }, { "epoch": 4.888888888888889, "grad_norm": 0.46089035272598267, "learning_rate": 9.655424854889254e-05, "loss": 0.0294, "step": 5500 }, { "epoch": 4.897777777777778, "grad_norm": 0.47193101048469543, "learning_rate": 9.653635363640172e-05, "loss": 0.0391, "step": 5510 }, { "epoch": 4.906666666666666, "grad_norm": 0.3921513557434082, "learning_rate": 9.651841404414984e-05, "loss": 0.0374, "step": 5520 }, { "epoch": 4.915555555555556, "grad_norm": 0.41946712136268616, "learning_rate": 9.650042978936072e-05, "loss": 0.0451, "step": 5530 }, { "epoch": 4.924444444444444, "grad_norm": 0.291852205991745, "learning_rate": 9.648240088930118e-05, "loss": 0.037, "step": 5540 }, { "epoch": 4.933333333333334, "grad_norm": 0.35839247703552246, "learning_rate": 9.646432736128084e-05, "loss": 0.0308, "step": 5550 }, { "epoch": 4.942222222222222, "grad_norm": 0.49170440435409546, "learning_rate": 9.644620922265215e-05, "loss": 0.0381, "step": 5560 }, { "epoch": 4.9511111111111115, "grad_norm": 0.3545382618904114, "learning_rate": 9.642804649081045e-05, "loss": 0.0434, "step": 5570 }, { "epoch": 4.96, "grad_norm": 0.5008895993232727, "learning_rate": 9.640983918319382e-05, "loss": 0.0389, "step": 5580 }, { "epoch": 4.968888888888889, "grad_norm": 0.3140445947647095, "learning_rate": 9.63915873172832e-05, "loss": 0.0381, "step": 5590 }, { "epoch": 4.977777777777778, "grad_norm": 0.38196703791618347, "learning_rate": 9.63732909106023e-05, "loss": 0.0342, "step": 5600 }, { "epoch": 4.986666666666666, "grad_norm": 0.4555117189884186, "learning_rate": 9.635494998071755e-05, "loss": 0.0367, "step": 5610 }, { "epoch": 4.995555555555556, "grad_norm": 0.3690134286880493, "learning_rate": 9.633656454523819e-05, "loss": 0.0331, "step": 5620 }, { "epoch": 5.004444444444444, "grad_norm": 0.48908108472824097, "learning_rate": 9.631813462181613e-05, "loss": 0.0338, "step": 5630 }, { "epoch": 5.013333333333334, "grad_norm": 0.3832654058933258, "learning_rate": 9.629966022814605e-05, "loss": 0.0364, "step": 5640 }, { "epoch": 5.022222222222222, "grad_norm": 0.7008596658706665, "learning_rate": 9.628114138196528e-05, "loss": 0.0361, "step": 5650 }, { "epoch": 5.0311111111111115, "grad_norm": 0.45013728737831116, "learning_rate": 9.626257810105385e-05, "loss": 0.0359, "step": 5660 }, { "epoch": 5.04, "grad_norm": 0.4489860236644745, "learning_rate": 9.624397040323445e-05, "loss": 0.0341, "step": 5670 }, { "epoch": 5.0488888888888885, "grad_norm": 0.3280660808086395, "learning_rate": 9.622531830637241e-05, "loss": 0.0315, "step": 5680 }, { "epoch": 5.057777777777778, "grad_norm": 0.4990573525428772, "learning_rate": 9.62066218283757e-05, "loss": 0.0339, "step": 5690 }, { "epoch": 5.066666666666666, "grad_norm": 0.33039042353630066, "learning_rate": 9.61878809871949e-05, "loss": 0.0309, "step": 5700 }, { "epoch": 5.075555555555556, "grad_norm": 0.425883948802948, "learning_rate": 9.616909580082316e-05, "loss": 0.0331, "step": 5710 }, { "epoch": 5.084444444444444, "grad_norm": 0.43031632900238037, "learning_rate": 9.615026628729622e-05, "loss": 0.0329, "step": 5720 }, { "epoch": 5.093333333333334, "grad_norm": 0.43156570196151733, "learning_rate": 9.613139246469238e-05, "loss": 0.0335, "step": 5730 }, { "epoch": 5.102222222222222, "grad_norm": 0.3823620676994324, "learning_rate": 9.611247435113249e-05, "loss": 0.0421, "step": 5740 }, { "epoch": 5.111111111111111, "grad_norm": 0.31392577290534973, "learning_rate": 9.609351196477991e-05, "loss": 0.0342, "step": 5750 }, { "epoch": 5.12, "grad_norm": 0.44784989953041077, "learning_rate": 9.607450532384051e-05, "loss": 0.0305, "step": 5760 }, { "epoch": 5.128888888888889, "grad_norm": 0.2623730003833771, "learning_rate": 9.605545444656267e-05, "loss": 0.0332, "step": 5770 }, { "epoch": 5.137777777777778, "grad_norm": 0.31600087881088257, "learning_rate": 9.603635935123719e-05, "loss": 0.0374, "step": 5780 }, { "epoch": 5.1466666666666665, "grad_norm": 0.28764626383781433, "learning_rate": 9.601722005619737e-05, "loss": 0.0356, "step": 5790 }, { "epoch": 5.155555555555556, "grad_norm": 0.4513314962387085, "learning_rate": 9.599803657981897e-05, "loss": 0.0364, "step": 5800 }, { "epoch": 5.164444444444444, "grad_norm": 0.5010198354721069, "learning_rate": 9.597880894052006e-05, "loss": 0.0369, "step": 5810 }, { "epoch": 5.173333333333334, "grad_norm": 0.4262983202934265, "learning_rate": 9.595953715676123e-05, "loss": 0.038, "step": 5820 }, { "epoch": 5.182222222222222, "grad_norm": 0.329451322555542, "learning_rate": 9.59402212470454e-05, "loss": 0.0308, "step": 5830 }, { "epoch": 5.191111111111111, "grad_norm": 0.37369006872177124, "learning_rate": 9.592086122991786e-05, "loss": 0.0351, "step": 5840 }, { "epoch": 5.2, "grad_norm": 0.42062315344810486, "learning_rate": 9.590145712396623e-05, "loss": 0.0374, "step": 5850 }, { "epoch": 5.208888888888889, "grad_norm": 0.36420151591300964, "learning_rate": 9.588200894782051e-05, "loss": 0.0366, "step": 5860 }, { "epoch": 5.217777777777778, "grad_norm": 0.49455347657203674, "learning_rate": 9.586251672015296e-05, "loss": 0.0363, "step": 5870 }, { "epoch": 5.226666666666667, "grad_norm": 0.470613569021225, "learning_rate": 9.584298045967816e-05, "loss": 0.0309, "step": 5880 }, { "epoch": 5.235555555555556, "grad_norm": 0.29402562975883484, "learning_rate": 9.582340018515297e-05, "loss": 0.0369, "step": 5890 }, { "epoch": 5.2444444444444445, "grad_norm": 0.38661566376686096, "learning_rate": 9.580377591537648e-05, "loss": 0.049, "step": 5900 }, { "epoch": 5.253333333333333, "grad_norm": 0.4020589292049408, "learning_rate": 9.578410766919005e-05, "loss": 0.0319, "step": 5910 }, { "epoch": 5.262222222222222, "grad_norm": 0.26784682273864746, "learning_rate": 9.576439546547726e-05, "loss": 0.0363, "step": 5920 }, { "epoch": 5.271111111111111, "grad_norm": 0.4629688262939453, "learning_rate": 9.574463932316386e-05, "loss": 0.0364, "step": 5930 }, { "epoch": 5.28, "grad_norm": 0.34414732456207275, "learning_rate": 9.572483926121783e-05, "loss": 0.0364, "step": 5940 }, { "epoch": 5.288888888888889, "grad_norm": 0.44594988226890564, "learning_rate": 9.57049952986493e-05, "loss": 0.0326, "step": 5950 }, { "epoch": 5.297777777777778, "grad_norm": 0.8140583038330078, "learning_rate": 9.568510745451054e-05, "loss": 0.0403, "step": 5960 }, { "epoch": 5.306666666666667, "grad_norm": 0.2755383253097534, "learning_rate": 9.566517574789596e-05, "loss": 0.0346, "step": 5970 }, { "epoch": 5.315555555555555, "grad_norm": 0.47529953718185425, "learning_rate": 9.564520019794207e-05, "loss": 0.0362, "step": 5980 }, { "epoch": 5.3244444444444445, "grad_norm": 0.30277228355407715, "learning_rate": 9.56251808238275e-05, "loss": 0.0346, "step": 5990 }, { "epoch": 5.333333333333333, "grad_norm": 0.3508113622665405, "learning_rate": 9.560511764477293e-05, "loss": 0.0398, "step": 6000 }, { "epoch": 5.342222222222222, "grad_norm": 0.36564114689826965, "learning_rate": 9.558501068004113e-05, "loss": 0.0313, "step": 6010 }, { "epoch": 5.351111111111111, "grad_norm": 0.3424344062805176, "learning_rate": 9.556485994893686e-05, "loss": 0.0346, "step": 6020 }, { "epoch": 5.36, "grad_norm": 0.5820956826210022, "learning_rate": 9.554466547080694e-05, "loss": 0.0366, "step": 6030 }, { "epoch": 5.368888888888889, "grad_norm": 0.3042823374271393, "learning_rate": 9.552442726504016e-05, "loss": 0.0379, "step": 6040 }, { "epoch": 5.377777777777778, "grad_norm": 0.5184895396232605, "learning_rate": 9.550414535106734e-05, "loss": 0.0369, "step": 6050 }, { "epoch": 5.386666666666667, "grad_norm": 0.40114668011665344, "learning_rate": 9.548381974836121e-05, "loss": 0.0329, "step": 6060 }, { "epoch": 5.395555555555555, "grad_norm": 0.30401965975761414, "learning_rate": 9.54634504764365e-05, "loss": 0.0444, "step": 6070 }, { "epoch": 5.404444444444445, "grad_norm": 0.4897063374519348, "learning_rate": 9.544303755484983e-05, "loss": 0.0311, "step": 6080 }, { "epoch": 5.413333333333333, "grad_norm": 0.5148985981941223, "learning_rate": 9.54225810031997e-05, "loss": 0.0297, "step": 6090 }, { "epoch": 5.4222222222222225, "grad_norm": 0.4142674505710602, "learning_rate": 9.540208084112659e-05, "loss": 0.0327, "step": 6100 }, { "epoch": 5.431111111111111, "grad_norm": 0.33650851249694824, "learning_rate": 9.538153708831278e-05, "loss": 0.0369, "step": 6110 }, { "epoch": 5.44, "grad_norm": 0.3085589110851288, "learning_rate": 9.53609497644824e-05, "loss": 0.0371, "step": 6120 }, { "epoch": 5.448888888888889, "grad_norm": 0.3101387917995453, "learning_rate": 9.534031888940146e-05, "loss": 0.0283, "step": 6130 }, { "epoch": 5.457777777777777, "grad_norm": 0.4244759678840637, "learning_rate": 9.531964448287774e-05, "loss": 0.0298, "step": 6140 }, { "epoch": 5.466666666666667, "grad_norm": 0.3815563917160034, "learning_rate": 9.529892656476082e-05, "loss": 0.0383, "step": 6150 }, { "epoch": 5.475555555555555, "grad_norm": 0.320860892534256, "learning_rate": 9.527816515494208e-05, "loss": 0.04, "step": 6160 }, { "epoch": 5.484444444444445, "grad_norm": 0.38909897208213806, "learning_rate": 9.525736027335463e-05, "loss": 0.0306, "step": 6170 }, { "epoch": 5.493333333333333, "grad_norm": 0.3340566158294678, "learning_rate": 9.523651193997336e-05, "loss": 0.0389, "step": 6180 }, { "epoch": 5.502222222222223, "grad_norm": 0.2622203528881073, "learning_rate": 9.521562017481483e-05, "loss": 0.033, "step": 6190 }, { "epoch": 5.511111111111111, "grad_norm": 0.32652702927589417, "learning_rate": 9.519468499793729e-05, "loss": 0.038, "step": 6200 }, { "epoch": 5.52, "grad_norm": 0.3374256193637848, "learning_rate": 9.517370642944073e-05, "loss": 0.0254, "step": 6210 }, { "epoch": 5.528888888888889, "grad_norm": 0.31643080711364746, "learning_rate": 9.515268448946677e-05, "loss": 0.0326, "step": 6220 }, { "epoch": 5.5377777777777775, "grad_norm": 0.2991648018360138, "learning_rate": 9.513161919819864e-05, "loss": 0.0351, "step": 6230 }, { "epoch": 5.546666666666667, "grad_norm": 0.3750928044319153, "learning_rate": 9.511051057586124e-05, "loss": 0.0357, "step": 6240 }, { "epoch": 5.555555555555555, "grad_norm": 0.3720608949661255, "learning_rate": 9.508935864272105e-05, "loss": 0.0331, "step": 6250 }, { "epoch": 5.564444444444445, "grad_norm": 0.431270569562912, "learning_rate": 9.506816341908616e-05, "loss": 0.0391, "step": 6260 }, { "epoch": 5.573333333333333, "grad_norm": 0.5521478652954102, "learning_rate": 9.504692492530614e-05, "loss": 0.0385, "step": 6270 }, { "epoch": 5.582222222222223, "grad_norm": 0.3653121590614319, "learning_rate": 9.502564318177221e-05, "loss": 0.0354, "step": 6280 }, { "epoch": 5.591111111111111, "grad_norm": 0.3907417058944702, "learning_rate": 9.500431820891706e-05, "loss": 0.0325, "step": 6290 }, { "epoch": 5.6, "grad_norm": 0.35111933946609497, "learning_rate": 9.498295002721489e-05, "loss": 0.0325, "step": 6300 }, { "epoch": 5.608888888888889, "grad_norm": 0.3357151448726654, "learning_rate": 9.496153865718137e-05, "loss": 0.0365, "step": 6310 }, { "epoch": 5.6177777777777775, "grad_norm": 0.2726459801197052, "learning_rate": 9.494008411937367e-05, "loss": 0.0297, "step": 6320 }, { "epoch": 5.626666666666667, "grad_norm": 0.31007468700408936, "learning_rate": 9.491858643439038e-05, "loss": 0.0381, "step": 6330 }, { "epoch": 5.635555555555555, "grad_norm": 0.29119375348091125, "learning_rate": 9.489704562287153e-05, "loss": 0.032, "step": 6340 }, { "epoch": 5.644444444444445, "grad_norm": 0.5571187138557434, "learning_rate": 9.487546170549854e-05, "loss": 0.0392, "step": 6350 }, { "epoch": 5.653333333333333, "grad_norm": 0.3037572503089905, "learning_rate": 9.485383470299423e-05, "loss": 0.0359, "step": 6360 }, { "epoch": 5.662222222222223, "grad_norm": 0.429470032453537, "learning_rate": 9.483216463612278e-05, "loss": 0.033, "step": 6370 }, { "epoch": 5.671111111111111, "grad_norm": 0.37073180079460144, "learning_rate": 9.481045152568973e-05, "loss": 0.0308, "step": 6380 }, { "epoch": 5.68, "grad_norm": 0.34142258763313293, "learning_rate": 9.478869539254193e-05, "loss": 0.0337, "step": 6390 }, { "epoch": 5.688888888888889, "grad_norm": 0.48527830839157104, "learning_rate": 9.476689625756752e-05, "loss": 0.0313, "step": 6400 }, { "epoch": 5.697777777777778, "grad_norm": 0.32324475049972534, "learning_rate": 9.474505414169597e-05, "loss": 0.0364, "step": 6410 }, { "epoch": 5.706666666666667, "grad_norm": 0.5245936512947083, "learning_rate": 9.472316906589798e-05, "loss": 0.0305, "step": 6420 }, { "epoch": 5.7155555555555555, "grad_norm": 0.5119931101799011, "learning_rate": 9.470124105118552e-05, "loss": 0.0284, "step": 6430 }, { "epoch": 5.724444444444444, "grad_norm": 0.4225582182407379, "learning_rate": 9.467927011861177e-05, "loss": 0.0357, "step": 6440 }, { "epoch": 5.733333333333333, "grad_norm": 0.4645766615867615, "learning_rate": 9.465725628927111e-05, "loss": 0.0327, "step": 6450 }, { "epoch": 5.742222222222222, "grad_norm": 0.33736416697502136, "learning_rate": 9.463519958429914e-05, "loss": 0.0379, "step": 6460 }, { "epoch": 5.751111111111111, "grad_norm": 0.3285878896713257, "learning_rate": 9.461310002487258e-05, "loss": 0.031, "step": 6470 }, { "epoch": 5.76, "grad_norm": 0.2968837320804596, "learning_rate": 9.459095763220932e-05, "loss": 0.0338, "step": 6480 }, { "epoch": 5.768888888888889, "grad_norm": 0.30555763840675354, "learning_rate": 9.456877242756838e-05, "loss": 0.0278, "step": 6490 }, { "epoch": 5.777777777777778, "grad_norm": 0.2725641429424286, "learning_rate": 9.454654443224986e-05, "loss": 0.0296, "step": 6500 }, { "epoch": 5.786666666666667, "grad_norm": 0.3232053518295288, "learning_rate": 9.452427366759497e-05, "loss": 0.0281, "step": 6510 }, { "epoch": 5.795555555555556, "grad_norm": 0.38599109649658203, "learning_rate": 9.450196015498596e-05, "loss": 0.0299, "step": 6520 }, { "epoch": 5.804444444444444, "grad_norm": 0.4532170295715332, "learning_rate": 9.447960391584614e-05, "loss": 0.0335, "step": 6530 }, { "epoch": 5.8133333333333335, "grad_norm": 0.2966475784778595, "learning_rate": 9.445720497163983e-05, "loss": 0.0298, "step": 6540 }, { "epoch": 5.822222222222222, "grad_norm": 0.43191859126091003, "learning_rate": 9.443476334387236e-05, "loss": 0.0327, "step": 6550 }, { "epoch": 5.831111111111111, "grad_norm": 0.35407161712646484, "learning_rate": 9.441227905409005e-05, "loss": 0.0401, "step": 6560 }, { "epoch": 5.84, "grad_norm": 0.6124844551086426, "learning_rate": 9.438975212388015e-05, "loss": 0.0281, "step": 6570 }, { "epoch": 5.848888888888889, "grad_norm": 0.3583097755908966, "learning_rate": 9.436718257487086e-05, "loss": 0.0266, "step": 6580 }, { "epoch": 5.857777777777778, "grad_norm": 0.37551307678222656, "learning_rate": 9.434457042873133e-05, "loss": 0.033, "step": 6590 }, { "epoch": 5.866666666666667, "grad_norm": 0.26058319211006165, "learning_rate": 9.432191570717155e-05, "loss": 0.0276, "step": 6600 }, { "epoch": 5.875555555555556, "grad_norm": 0.2700827121734619, "learning_rate": 9.429921843194246e-05, "loss": 0.0297, "step": 6610 }, { "epoch": 5.884444444444444, "grad_norm": 0.435099333524704, "learning_rate": 9.427647862483576e-05, "loss": 0.0336, "step": 6620 }, { "epoch": 5.8933333333333335, "grad_norm": 0.3227585554122925, "learning_rate": 9.425369630768409e-05, "loss": 0.0342, "step": 6630 }, { "epoch": 5.902222222222222, "grad_norm": 0.3839111626148224, "learning_rate": 9.423087150236085e-05, "loss": 0.0342, "step": 6640 }, { "epoch": 5.911111111111111, "grad_norm": 0.34203535318374634, "learning_rate": 9.42080042307802e-05, "loss": 0.0375, "step": 6650 }, { "epoch": 5.92, "grad_norm": 0.3292289972305298, "learning_rate": 9.418509451489716e-05, "loss": 0.0317, "step": 6660 }, { "epoch": 5.928888888888888, "grad_norm": 0.3365071713924408, "learning_rate": 9.416214237670741e-05, "loss": 0.0295, "step": 6670 }, { "epoch": 5.937777777777778, "grad_norm": 0.3057865798473358, "learning_rate": 9.413914783824742e-05, "loss": 0.0269, "step": 6680 }, { "epoch": 5.946666666666666, "grad_norm": 0.24801290035247803, "learning_rate": 9.411611092159434e-05, "loss": 0.0331, "step": 6690 }, { "epoch": 5.955555555555556, "grad_norm": 0.30153635144233704, "learning_rate": 9.409303164886603e-05, "loss": 0.025, "step": 6700 }, { "epoch": 5.964444444444444, "grad_norm": 0.4181388318538666, "learning_rate": 9.4069910042221e-05, "loss": 0.0324, "step": 6710 }, { "epoch": 5.973333333333334, "grad_norm": 0.4792896509170532, "learning_rate": 9.40467461238584e-05, "loss": 0.0371, "step": 6720 }, { "epoch": 5.982222222222222, "grad_norm": 0.44467002153396606, "learning_rate": 9.4023539916018e-05, "loss": 0.0309, "step": 6730 }, { "epoch": 5.9911111111111115, "grad_norm": 0.26609718799591064, "learning_rate": 9.40002914409802e-05, "loss": 0.0329, "step": 6740 }, { "epoch": 6.0, "grad_norm": 0.5773561000823975, "learning_rate": 9.397700072106597e-05, "loss": 0.0361, "step": 6750 }, { "epoch": 6.0088888888888885, "grad_norm": 0.28954264521598816, "learning_rate": 9.395366777863683e-05, "loss": 0.0295, "step": 6760 }, { "epoch": 6.017777777777778, "grad_norm": 0.3568158745765686, "learning_rate": 9.393029263609483e-05, "loss": 0.0377, "step": 6770 }, { "epoch": 6.026666666666666, "grad_norm": 0.3227669894695282, "learning_rate": 9.390687531588255e-05, "loss": 0.0371, "step": 6780 }, { "epoch": 6.035555555555556, "grad_norm": 0.3540012240409851, "learning_rate": 9.38834158404831e-05, "loss": 0.0425, "step": 6790 }, { "epoch": 6.044444444444444, "grad_norm": 0.3106261193752289, "learning_rate": 9.385991423241998e-05, "loss": 0.0314, "step": 6800 }, { "epoch": 6.053333333333334, "grad_norm": 0.2594573497772217, "learning_rate": 9.383637051425722e-05, "loss": 0.0323, "step": 6810 }, { "epoch": 6.062222222222222, "grad_norm": 0.3725073039531708, "learning_rate": 9.381278470859922e-05, "loss": 0.0304, "step": 6820 }, { "epoch": 6.071111111111111, "grad_norm": 0.20926804840564728, "learning_rate": 9.378915683809085e-05, "loss": 0.0316, "step": 6830 }, { "epoch": 6.08, "grad_norm": 0.4273773729801178, "learning_rate": 9.376548692541729e-05, "loss": 0.0331, "step": 6840 }, { "epoch": 6.088888888888889, "grad_norm": 0.3843939006328583, "learning_rate": 9.374177499330419e-05, "loss": 0.0356, "step": 6850 }, { "epoch": 6.097777777777778, "grad_norm": 0.3278975784778595, "learning_rate": 9.371802106451743e-05, "loss": 0.0276, "step": 6860 }, { "epoch": 6.1066666666666665, "grad_norm": 0.40987062454223633, "learning_rate": 9.369422516186328e-05, "loss": 0.0343, "step": 6870 }, { "epoch": 6.115555555555556, "grad_norm": 0.3296256959438324, "learning_rate": 9.367038730818832e-05, "loss": 0.0302, "step": 6880 }, { "epoch": 6.124444444444444, "grad_norm": 0.42148610949516296, "learning_rate": 9.364650752637933e-05, "loss": 0.0293, "step": 6890 }, { "epoch": 6.133333333333334, "grad_norm": 0.3069022297859192, "learning_rate": 9.362258583936342e-05, "loss": 0.0312, "step": 6900 }, { "epoch": 6.142222222222222, "grad_norm": 0.3354296386241913, "learning_rate": 9.359862227010791e-05, "loss": 0.036, "step": 6910 }, { "epoch": 6.151111111111111, "grad_norm": 0.3716127276420593, "learning_rate": 9.357461684162034e-05, "loss": 0.0396, "step": 6920 }, { "epoch": 6.16, "grad_norm": 0.30714988708496094, "learning_rate": 9.355056957694842e-05, "loss": 0.0332, "step": 6930 }, { "epoch": 6.168888888888889, "grad_norm": 0.4107031524181366, "learning_rate": 9.352648049918002e-05, "loss": 0.0407, "step": 6940 }, { "epoch": 6.177777777777778, "grad_norm": 0.3528226315975189, "learning_rate": 9.35023496314432e-05, "loss": 0.0323, "step": 6950 }, { "epoch": 6.1866666666666665, "grad_norm": 0.4652569890022278, "learning_rate": 9.347817699690608e-05, "loss": 0.0323, "step": 6960 }, { "epoch": 6.195555555555556, "grad_norm": 0.356037437915802, "learning_rate": 9.345396261877694e-05, "loss": 0.0356, "step": 6970 }, { "epoch": 6.204444444444444, "grad_norm": 0.321561723947525, "learning_rate": 9.342970652030411e-05, "loss": 0.0355, "step": 6980 }, { "epoch": 6.213333333333333, "grad_norm": 0.39457085728645325, "learning_rate": 9.340540872477597e-05, "loss": 0.0291, "step": 6990 }, { "epoch": 6.222222222222222, "grad_norm": 0.3595648407936096, "learning_rate": 9.338106925552095e-05, "loss": 0.0338, "step": 7000 }, { "epoch": 6.231111111111111, "grad_norm": 0.3207085132598877, "learning_rate": 9.335668813590747e-05, "loss": 0.0301, "step": 7010 }, { "epoch": 6.24, "grad_norm": 0.3962297737598419, "learning_rate": 9.333226538934398e-05, "loss": 0.0335, "step": 7020 }, { "epoch": 6.248888888888889, "grad_norm": 0.2896721065044403, "learning_rate": 9.330780103927885e-05, "loss": 0.0258, "step": 7030 }, { "epoch": 6.257777777777778, "grad_norm": 0.4001779556274414, "learning_rate": 9.328329510920039e-05, "loss": 0.0298, "step": 7040 }, { "epoch": 6.266666666666667, "grad_norm": 0.3631964921951294, "learning_rate": 9.32587476226369e-05, "loss": 0.0314, "step": 7050 }, { "epoch": 6.275555555555556, "grad_norm": 0.33581095933914185, "learning_rate": 9.32341586031565e-05, "loss": 0.0308, "step": 7060 }, { "epoch": 6.2844444444444445, "grad_norm": 0.31441444158554077, "learning_rate": 9.32095280743672e-05, "loss": 0.0308, "step": 7070 }, { "epoch": 6.293333333333333, "grad_norm": 0.3388790786266327, "learning_rate": 9.318485605991693e-05, "loss": 0.0277, "step": 7080 }, { "epoch": 6.302222222222222, "grad_norm": 0.30772024393081665, "learning_rate": 9.316014258349337e-05, "loss": 0.0294, "step": 7090 }, { "epoch": 6.311111111111111, "grad_norm": 0.29166531562805176, "learning_rate": 9.313538766882405e-05, "loss": 0.0299, "step": 7100 }, { "epoch": 6.32, "grad_norm": 0.391686350107193, "learning_rate": 9.311059133967626e-05, "loss": 0.0372, "step": 7110 }, { "epoch": 6.328888888888889, "grad_norm": 0.3694501221179962, "learning_rate": 9.308575361985708e-05, "loss": 0.0319, "step": 7120 }, { "epoch": 6.337777777777778, "grad_norm": 0.2531419098377228, "learning_rate": 9.306087453321331e-05, "loss": 0.0319, "step": 7130 }, { "epoch": 6.346666666666667, "grad_norm": 0.4099632799625397, "learning_rate": 9.303595410363146e-05, "loss": 0.0336, "step": 7140 }, { "epoch": 6.355555555555555, "grad_norm": 0.36935877799987793, "learning_rate": 9.301099235503779e-05, "loss": 0.0311, "step": 7150 }, { "epoch": 6.364444444444445, "grad_norm": 0.35985639691352844, "learning_rate": 9.298598931139812e-05, "loss": 0.0336, "step": 7160 }, { "epoch": 6.373333333333333, "grad_norm": 0.3263987898826599, "learning_rate": 9.296094499671805e-05, "loss": 0.0303, "step": 7170 }, { "epoch": 6.3822222222222225, "grad_norm": 0.31828826665878296, "learning_rate": 9.29358594350427e-05, "loss": 0.025, "step": 7180 }, { "epoch": 6.391111111111111, "grad_norm": 0.4107153117656708, "learning_rate": 9.291073265045684e-05, "loss": 0.0301, "step": 7190 }, { "epoch": 6.4, "grad_norm": 0.3776908218860626, "learning_rate": 9.28855646670848e-05, "loss": 0.03, "step": 7200 }, { "epoch": 6.408888888888889, "grad_norm": 0.2795926630496979, "learning_rate": 9.286035550909047e-05, "loss": 0.0369, "step": 7210 }, { "epoch": 6.417777777777777, "grad_norm": 0.2699511647224426, "learning_rate": 9.28351052006773e-05, "loss": 0.03, "step": 7220 }, { "epoch": 6.426666666666667, "grad_norm": 0.35640960931777954, "learning_rate": 9.280981376608819e-05, "loss": 0.0268, "step": 7230 }, { "epoch": 6.435555555555555, "grad_norm": 0.36802738904953003, "learning_rate": 9.27844812296056e-05, "loss": 0.0287, "step": 7240 }, { "epoch": 6.444444444444445, "grad_norm": 0.49969232082366943, "learning_rate": 9.275910761555138e-05, "loss": 0.0292, "step": 7250 }, { "epoch": 6.453333333333333, "grad_norm": 0.3464072644710541, "learning_rate": 9.273369294828687e-05, "loss": 0.0333, "step": 7260 }, { "epoch": 6.4622222222222225, "grad_norm": 0.4520910084247589, "learning_rate": 9.270823725221278e-05, "loss": 0.0329, "step": 7270 }, { "epoch": 6.471111111111111, "grad_norm": 0.3811221122741699, "learning_rate": 9.268274055176926e-05, "loss": 0.0269, "step": 7280 }, { "epoch": 6.48, "grad_norm": 0.32476386427879333, "learning_rate": 9.26572028714358e-05, "loss": 0.0257, "step": 7290 }, { "epoch": 6.488888888888889, "grad_norm": 0.4957089126110077, "learning_rate": 9.263162423573124e-05, "loss": 0.031, "step": 7300 }, { "epoch": 6.497777777777777, "grad_norm": 0.31816568970680237, "learning_rate": 9.260600466921375e-05, "loss": 0.0295, "step": 7310 }, { "epoch": 6.506666666666667, "grad_norm": 0.5138843059539795, "learning_rate": 9.258034419648076e-05, "loss": 0.037, "step": 7320 }, { "epoch": 6.515555555555555, "grad_norm": 0.40023431181907654, "learning_rate": 9.255464284216903e-05, "loss": 0.0318, "step": 7330 }, { "epoch": 6.524444444444445, "grad_norm": 0.37239596247673035, "learning_rate": 9.252890063095456e-05, "loss": 0.0354, "step": 7340 }, { "epoch": 6.533333333333333, "grad_norm": 0.5376574397087097, "learning_rate": 9.250311758755255e-05, "loss": 0.0348, "step": 7350 }, { "epoch": 6.542222222222223, "grad_norm": 0.43184739351272583, "learning_rate": 9.247729373671738e-05, "loss": 0.0326, "step": 7360 }, { "epoch": 6.551111111111111, "grad_norm": 0.2623891532421112, "learning_rate": 9.24514291032427e-05, "loss": 0.0287, "step": 7370 }, { "epoch": 6.5600000000000005, "grad_norm": 0.40383678674697876, "learning_rate": 9.242552371196123e-05, "loss": 0.0339, "step": 7380 }, { "epoch": 6.568888888888889, "grad_norm": 0.3490070700645447, "learning_rate": 9.239957758774485e-05, "loss": 0.0301, "step": 7390 }, { "epoch": 6.5777777777777775, "grad_norm": 0.32680559158325195, "learning_rate": 9.237359075550455e-05, "loss": 0.0279, "step": 7400 }, { "epoch": 6.586666666666667, "grad_norm": 0.3646572232246399, "learning_rate": 9.234756324019042e-05, "loss": 0.0347, "step": 7410 }, { "epoch": 6.595555555555555, "grad_norm": 0.23718124628067017, "learning_rate": 9.23214950667916e-05, "loss": 0.0321, "step": 7420 }, { "epoch": 6.604444444444445, "grad_norm": 0.27027627825737, "learning_rate": 9.229538626033621e-05, "loss": 0.033, "step": 7430 }, { "epoch": 6.613333333333333, "grad_norm": 0.33824318647384644, "learning_rate": 9.226923684589148e-05, "loss": 0.0291, "step": 7440 }, { "epoch": 6.622222222222222, "grad_norm": 0.4109753966331482, "learning_rate": 9.224304684856358e-05, "loss": 0.0367, "step": 7450 }, { "epoch": 6.631111111111111, "grad_norm": 0.21559232473373413, "learning_rate": 9.221681629349761e-05, "loss": 0.0283, "step": 7460 }, { "epoch": 6.64, "grad_norm": 0.3599995970726013, "learning_rate": 9.219054520587766e-05, "loss": 0.0283, "step": 7470 }, { "epoch": 6.648888888888889, "grad_norm": 0.2910494804382324, "learning_rate": 9.216423361092673e-05, "loss": 0.0319, "step": 7480 }, { "epoch": 6.657777777777778, "grad_norm": 0.2808232605457306, "learning_rate": 9.213788153390668e-05, "loss": 0.024, "step": 7490 }, { "epoch": 6.666666666666667, "grad_norm": 0.3430759906768799, "learning_rate": 9.211148900011828e-05, "loss": 0.0308, "step": 7500 }, { "epoch": 6.6755555555555555, "grad_norm": 0.35993894934654236, "learning_rate": 9.20850560349011e-05, "loss": 0.0315, "step": 7510 }, { "epoch": 6.684444444444445, "grad_norm": 0.3576984405517578, "learning_rate": 9.205858266363355e-05, "loss": 0.0269, "step": 7520 }, { "epoch": 6.693333333333333, "grad_norm": 0.4079616367816925, "learning_rate": 9.203206891173284e-05, "loss": 0.0239, "step": 7530 }, { "epoch": 6.702222222222222, "grad_norm": 0.2872537672519684, "learning_rate": 9.200551480465493e-05, "loss": 0.0323, "step": 7540 }, { "epoch": 6.711111111111111, "grad_norm": 0.4455048739910126, "learning_rate": 9.197892036789455e-05, "loss": 0.0311, "step": 7550 }, { "epoch": 6.72, "grad_norm": 0.3817867636680603, "learning_rate": 9.195228562698515e-05, "loss": 0.0305, "step": 7560 }, { "epoch": 6.728888888888889, "grad_norm": 0.3617621958255768, "learning_rate": 9.192561060749882e-05, "loss": 0.0302, "step": 7570 }, { "epoch": 6.737777777777778, "grad_norm": 0.32246387004852295, "learning_rate": 9.189889533504641e-05, "loss": 0.0321, "step": 7580 }, { "epoch": 6.746666666666667, "grad_norm": 0.32514485716819763, "learning_rate": 9.187213983527735e-05, "loss": 0.0313, "step": 7590 }, { "epoch": 6.7555555555555555, "grad_norm": 0.35704755783081055, "learning_rate": 9.184534413387972e-05, "loss": 0.0318, "step": 7600 }, { "epoch": 6.764444444444445, "grad_norm": 0.32508257031440735, "learning_rate": 9.181850825658018e-05, "loss": 0.0312, "step": 7610 }, { "epoch": 6.773333333333333, "grad_norm": 0.39004790782928467, "learning_rate": 9.1791632229144e-05, "loss": 0.0278, "step": 7620 }, { "epoch": 6.782222222222222, "grad_norm": 0.37237173318862915, "learning_rate": 9.176471607737495e-05, "loss": 0.0309, "step": 7630 }, { "epoch": 6.791111111111111, "grad_norm": 0.28705841302871704, "learning_rate": 9.173775982711538e-05, "loss": 0.0289, "step": 7640 }, { "epoch": 6.8, "grad_norm": 0.34220242500305176, "learning_rate": 9.171076350424607e-05, "loss": 0.0265, "step": 7650 }, { "epoch": 6.808888888888889, "grad_norm": 0.41831469535827637, "learning_rate": 9.168372713468633e-05, "loss": 0.0315, "step": 7660 }, { "epoch": 6.817777777777778, "grad_norm": 0.35941070318222046, "learning_rate": 9.16566507443939e-05, "loss": 0.0305, "step": 7670 }, { "epoch": 6.826666666666666, "grad_norm": 0.4006064832210541, "learning_rate": 9.162953435936493e-05, "loss": 0.0338, "step": 7680 }, { "epoch": 6.835555555555556, "grad_norm": 0.310115784406662, "learning_rate": 9.1602378005634e-05, "loss": 0.026, "step": 7690 }, { "epoch": 6.844444444444444, "grad_norm": 0.34289824962615967, "learning_rate": 9.157518170927403e-05, "loss": 0.0282, "step": 7700 }, { "epoch": 6.8533333333333335, "grad_norm": 0.264691025018692, "learning_rate": 9.154794549639631e-05, "loss": 0.0251, "step": 7710 }, { "epoch": 6.862222222222222, "grad_norm": 0.4459560215473175, "learning_rate": 9.152066939315047e-05, "loss": 0.0327, "step": 7720 }, { "epoch": 6.871111111111111, "grad_norm": 0.40657421946525574, "learning_rate": 9.149335342572438e-05, "loss": 0.0309, "step": 7730 }, { "epoch": 6.88, "grad_norm": 0.4232308268547058, "learning_rate": 9.146599762034425e-05, "loss": 0.037, "step": 7740 }, { "epoch": 6.888888888888889, "grad_norm": 0.32039782404899597, "learning_rate": 9.143860200327452e-05, "loss": 0.0322, "step": 7750 }, { "epoch": 6.897777777777778, "grad_norm": 0.2866861820220947, "learning_rate": 9.141116660081781e-05, "loss": 0.0273, "step": 7760 }, { "epoch": 6.906666666666666, "grad_norm": 0.35743480920791626, "learning_rate": 9.138369143931499e-05, "loss": 0.0324, "step": 7770 }, { "epoch": 6.915555555555556, "grad_norm": 0.3774275779724121, "learning_rate": 9.13561765451451e-05, "loss": 0.0353, "step": 7780 }, { "epoch": 6.924444444444444, "grad_norm": 0.3290146589279175, "learning_rate": 9.132862194472528e-05, "loss": 0.0347, "step": 7790 }, { "epoch": 6.933333333333334, "grad_norm": 0.46776407957077026, "learning_rate": 9.130102766451087e-05, "loss": 0.0433, "step": 7800 }, { "epoch": 6.942222222222222, "grad_norm": 0.32841211557388306, "learning_rate": 9.127339373099523e-05, "loss": 0.028, "step": 7810 }, { "epoch": 6.9511111111111115, "grad_norm": 0.44936850666999817, "learning_rate": 9.124572017070982e-05, "loss": 0.0276, "step": 7820 }, { "epoch": 6.96, "grad_norm": 0.28805476427078247, "learning_rate": 9.121800701022415e-05, "loss": 0.0312, "step": 7830 }, { "epoch": 6.968888888888889, "grad_norm": 0.33721235394477844, "learning_rate": 9.119025427614578e-05, "loss": 0.0325, "step": 7840 }, { "epoch": 6.977777777777778, "grad_norm": 0.31499260663986206, "learning_rate": 9.116246199512021e-05, "loss": 0.0268, "step": 7850 }, { "epoch": 6.986666666666666, "grad_norm": 0.28905394673347473, "learning_rate": 9.113463019383093e-05, "loss": 0.0333, "step": 7860 }, { "epoch": 6.995555555555556, "grad_norm": 0.2991236448287964, "learning_rate": 9.110675889899938e-05, "loss": 0.0293, "step": 7870 }, { "epoch": 7.004444444444444, "grad_norm": 0.3039570450782776, "learning_rate": 9.107884813738492e-05, "loss": 0.0269, "step": 7880 }, { "epoch": 7.013333333333334, "grad_norm": 0.26635846495628357, "learning_rate": 9.105089793578479e-05, "loss": 0.0294, "step": 7890 }, { "epoch": 7.022222222222222, "grad_norm": 0.34643521904945374, "learning_rate": 9.102290832103411e-05, "loss": 0.0279, "step": 7900 }, { "epoch": 7.0311111111111115, "grad_norm": 0.33304595947265625, "learning_rate": 9.099487932000582e-05, "loss": 0.0286, "step": 7910 }, { "epoch": 7.04, "grad_norm": 0.34373658895492554, "learning_rate": 9.09668109596107e-05, "loss": 0.0356, "step": 7920 }, { "epoch": 7.0488888888888885, "grad_norm": 0.3285376727581024, "learning_rate": 9.093870326679733e-05, "loss": 0.0241, "step": 7930 }, { "epoch": 7.057777777777778, "grad_norm": 0.32233279943466187, "learning_rate": 9.0910556268552e-05, "loss": 0.0314, "step": 7940 }, { "epoch": 7.066666666666666, "grad_norm": 0.27612194418907166, "learning_rate": 9.088236999189876e-05, "loss": 0.0273, "step": 7950 }, { "epoch": 7.075555555555556, "grad_norm": 0.2875770628452301, "learning_rate": 9.085414446389941e-05, "loss": 0.028, "step": 7960 }, { "epoch": 7.084444444444444, "grad_norm": 0.3895089030265808, "learning_rate": 9.08258797116534e-05, "loss": 0.0279, "step": 7970 }, { "epoch": 7.093333333333334, "grad_norm": 0.2454494833946228, "learning_rate": 9.079757576229784e-05, "loss": 0.0273, "step": 7980 }, { "epoch": 7.102222222222222, "grad_norm": 0.3658406436443329, "learning_rate": 9.076923264300748e-05, "loss": 0.039, "step": 7990 }, { "epoch": 7.111111111111111, "grad_norm": 0.19138087332248688, "learning_rate": 9.074085038099467e-05, "loss": 0.026, "step": 8000 }, { "epoch": 7.12, "grad_norm": 0.2557944357395172, "learning_rate": 9.071242900350935e-05, "loss": 0.0238, "step": 8010 }, { "epoch": 7.128888888888889, "grad_norm": 0.3701525628566742, "learning_rate": 9.0683968537839e-05, "loss": 0.0261, "step": 8020 }, { "epoch": 7.137777777777778, "grad_norm": 0.314574271440506, "learning_rate": 9.065546901130868e-05, "loss": 0.0273, "step": 8030 }, { "epoch": 7.1466666666666665, "grad_norm": 0.44193923473358154, "learning_rate": 9.062693045128085e-05, "loss": 0.0286, "step": 8040 }, { "epoch": 7.155555555555556, "grad_norm": 0.4512566924095154, "learning_rate": 9.059835288515557e-05, "loss": 0.0312, "step": 8050 }, { "epoch": 7.164444444444444, "grad_norm": 0.34740984439849854, "learning_rate": 9.056973634037025e-05, "loss": 0.0343, "step": 8060 }, { "epoch": 7.173333333333334, "grad_norm": 0.3132869303226471, "learning_rate": 9.054108084439976e-05, "loss": 0.0291, "step": 8070 }, { "epoch": 7.182222222222222, "grad_norm": 0.4976303279399872, "learning_rate": 9.051238642475641e-05, "loss": 0.0276, "step": 8080 }, { "epoch": 7.191111111111111, "grad_norm": 0.3932148814201355, "learning_rate": 9.04836531089898e-05, "loss": 0.0248, "step": 8090 }, { "epoch": 7.2, "grad_norm": 0.39199987053871155, "learning_rate": 9.045488092468694e-05, "loss": 0.0261, "step": 8100 }, { "epoch": 7.208888888888889, "grad_norm": 0.30254074931144714, "learning_rate": 9.042606989947212e-05, "loss": 0.0315, "step": 8110 }, { "epoch": 7.217777777777778, "grad_norm": 0.33541417121887207, "learning_rate": 9.039722006100693e-05, "loss": 0.0263, "step": 8120 }, { "epoch": 7.226666666666667, "grad_norm": 0.3093273639678955, "learning_rate": 9.036833143699024e-05, "loss": 0.0268, "step": 8130 }, { "epoch": 7.235555555555556, "grad_norm": 0.22370792925357819, "learning_rate": 9.033940405515815e-05, "loss": 0.0323, "step": 8140 }, { "epoch": 7.2444444444444445, "grad_norm": 0.3396274149417877, "learning_rate": 9.031043794328395e-05, "loss": 0.0302, "step": 8150 }, { "epoch": 7.253333333333333, "grad_norm": 0.31462982296943665, "learning_rate": 9.028143312917815e-05, "loss": 0.04, "step": 8160 }, { "epoch": 7.262222222222222, "grad_norm": 0.3016805350780487, "learning_rate": 9.025238964068838e-05, "loss": 0.0297, "step": 8170 }, { "epoch": 7.271111111111111, "grad_norm": 0.32213258743286133, "learning_rate": 9.022330750569947e-05, "loss": 0.0279, "step": 8180 }, { "epoch": 7.28, "grad_norm": 0.2567034959793091, "learning_rate": 9.019418675213326e-05, "loss": 0.0328, "step": 8190 }, { "epoch": 7.288888888888889, "grad_norm": 0.350227415561676, "learning_rate": 9.016502740794873e-05, "loss": 0.0316, "step": 8200 }, { "epoch": 7.297777777777778, "grad_norm": 0.362481951713562, "learning_rate": 9.01358295011419e-05, "loss": 0.0378, "step": 8210 }, { "epoch": 7.306666666666667, "grad_norm": 0.3739061653614044, "learning_rate": 9.010659305974582e-05, "loss": 0.0264, "step": 8220 }, { "epoch": 7.315555555555555, "grad_norm": 0.2844863533973694, "learning_rate": 9.007731811183051e-05, "loss": 0.0294, "step": 8230 }, { "epoch": 7.3244444444444445, "grad_norm": 0.3338690400123596, "learning_rate": 9.004800468550298e-05, "loss": 0.0349, "step": 8240 }, { "epoch": 7.333333333333333, "grad_norm": 0.4571111798286438, "learning_rate": 9.00186528089072e-05, "loss": 0.0316, "step": 8250 }, { "epoch": 7.342222222222222, "grad_norm": 0.402822345495224, "learning_rate": 8.998926251022403e-05, "loss": 0.0283, "step": 8260 }, { "epoch": 7.351111111111111, "grad_norm": 0.3256841003894806, "learning_rate": 8.995983381767121e-05, "loss": 0.0341, "step": 8270 }, { "epoch": 7.36, "grad_norm": 0.3240208029747009, "learning_rate": 8.993036675950337e-05, "loss": 0.0309, "step": 8280 }, { "epoch": 7.368888888888889, "grad_norm": 0.378725528717041, "learning_rate": 8.9900861364012e-05, "loss": 0.0318, "step": 8290 }, { "epoch": 7.377777777777778, "grad_norm": 0.35005098581314087, "learning_rate": 8.987131765952528e-05, "loss": 0.029, "step": 8300 }, { "epoch": 7.386666666666667, "grad_norm": 0.27951011061668396, "learning_rate": 8.984173567440832e-05, "loss": 0.032, "step": 8310 }, { "epoch": 7.395555555555555, "grad_norm": 0.36754074692726135, "learning_rate": 8.98121154370629e-05, "loss": 0.0249, "step": 8320 }, { "epoch": 7.404444444444445, "grad_norm": 0.26333215832710266, "learning_rate": 8.978245697592756e-05, "loss": 0.0264, "step": 8330 }, { "epoch": 7.413333333333333, "grad_norm": 0.27066925168037415, "learning_rate": 8.975276031947747e-05, "loss": 0.0276, "step": 8340 }, { "epoch": 7.4222222222222225, "grad_norm": 0.3507596552371979, "learning_rate": 8.972302549622457e-05, "loss": 0.0309, "step": 8350 }, { "epoch": 7.431111111111111, "grad_norm": 0.2818746566772461, "learning_rate": 8.969325253471736e-05, "loss": 0.0325, "step": 8360 }, { "epoch": 7.44, "grad_norm": 0.346864253282547, "learning_rate": 8.966344146354104e-05, "loss": 0.0307, "step": 8370 }, { "epoch": 7.448888888888889, "grad_norm": 0.2209518998861313, "learning_rate": 8.963359231131732e-05, "loss": 0.0306, "step": 8380 }, { "epoch": 7.457777777777777, "grad_norm": 0.4508216083049774, "learning_rate": 8.96037051067045e-05, "loss": 0.0267, "step": 8390 }, { "epoch": 7.466666666666667, "grad_norm": 0.2683308720588684, "learning_rate": 8.957377987839743e-05, "loss": 0.0287, "step": 8400 }, { "epoch": 7.475555555555555, "grad_norm": 0.3317680358886719, "learning_rate": 8.954381665512745e-05, "loss": 0.0285, "step": 8410 }, { "epoch": 7.484444444444445, "grad_norm": 0.2929815351963043, "learning_rate": 8.95138154656624e-05, "loss": 0.028, "step": 8420 }, { "epoch": 7.493333333333333, "grad_norm": 0.38005247712135315, "learning_rate": 8.948377633880655e-05, "loss": 0.0325, "step": 8430 }, { "epoch": 7.502222222222223, "grad_norm": 0.3119649887084961, "learning_rate": 8.945369930340059e-05, "loss": 0.0262, "step": 8440 }, { "epoch": 7.511111111111111, "grad_norm": 0.33570432662963867, "learning_rate": 8.942358438832162e-05, "loss": 0.0311, "step": 8450 }, { "epoch": 7.52, "grad_norm": 0.3263564109802246, "learning_rate": 8.93934316224831e-05, "loss": 0.0326, "step": 8460 }, { "epoch": 7.528888888888889, "grad_norm": 0.2770530879497528, "learning_rate": 8.936324103483484e-05, "loss": 0.0341, "step": 8470 }, { "epoch": 7.5377777777777775, "grad_norm": 0.40351635217666626, "learning_rate": 8.933301265436297e-05, "loss": 0.033, "step": 8480 }, { "epoch": 7.546666666666667, "grad_norm": 0.3575883209705353, "learning_rate": 8.930274651008988e-05, "loss": 0.0318, "step": 8490 }, { "epoch": 7.555555555555555, "grad_norm": 0.35939323902130127, "learning_rate": 8.927244263107423e-05, "loss": 0.0313, "step": 8500 }, { "epoch": 7.564444444444445, "grad_norm": 0.2517669200897217, "learning_rate": 8.924210104641091e-05, "loss": 0.0302, "step": 8510 }, { "epoch": 7.573333333333333, "grad_norm": 0.30903786420822144, "learning_rate": 8.921172178523101e-05, "loss": 0.0329, "step": 8520 }, { "epoch": 7.582222222222223, "grad_norm": 0.37116703391075134, "learning_rate": 8.91813048767018e-05, "loss": 0.0322, "step": 8530 }, { "epoch": 7.591111111111111, "grad_norm": 0.27307337522506714, "learning_rate": 8.91508503500267e-05, "loss": 0.0286, "step": 8540 }, { "epoch": 7.6, "grad_norm": 0.22804753482341766, "learning_rate": 8.912035823444523e-05, "loss": 0.0262, "step": 8550 }, { "epoch": 7.608888888888889, "grad_norm": 0.4197857677936554, "learning_rate": 8.908982855923297e-05, "loss": 0.0338, "step": 8560 }, { "epoch": 7.6177777777777775, "grad_norm": 0.3254260718822479, "learning_rate": 8.905926135370164e-05, "loss": 0.0307, "step": 8570 }, { "epoch": 7.626666666666667, "grad_norm": 0.3640803396701813, "learning_rate": 8.902865664719895e-05, "loss": 0.0263, "step": 8580 }, { "epoch": 7.635555555555555, "grad_norm": 0.37134358286857605, "learning_rate": 8.89980144691086e-05, "loss": 0.0309, "step": 8590 }, { "epoch": 7.644444444444445, "grad_norm": 0.2653670310974121, "learning_rate": 8.89673348488503e-05, "loss": 0.0255, "step": 8600 }, { "epoch": 7.653333333333333, "grad_norm": 0.3828783333301544, "learning_rate": 8.893661781587965e-05, "loss": 0.029, "step": 8610 }, { "epoch": 7.662222222222223, "grad_norm": 0.3344173729419708, "learning_rate": 8.890586339968826e-05, "loss": 0.0297, "step": 8620 }, { "epoch": 7.671111111111111, "grad_norm": 0.25384604930877686, "learning_rate": 8.887507162980355e-05, "loss": 0.0264, "step": 8630 }, { "epoch": 7.68, "grad_norm": 0.2763369679450989, "learning_rate": 8.884424253578885e-05, "loss": 0.0255, "step": 8640 }, { "epoch": 7.688888888888889, "grad_norm": 0.4674917459487915, "learning_rate": 8.88133761472433e-05, "loss": 0.0341, "step": 8650 }, { "epoch": 7.697777777777778, "grad_norm": 0.38205230236053467, "learning_rate": 8.878247249380187e-05, "loss": 0.0291, "step": 8660 }, { "epoch": 7.706666666666667, "grad_norm": 0.3934917449951172, "learning_rate": 8.87515316051353e-05, "loss": 0.0317, "step": 8670 }, { "epoch": 7.7155555555555555, "grad_norm": 0.34055739641189575, "learning_rate": 8.872055351095005e-05, "loss": 0.0316, "step": 8680 }, { "epoch": 7.724444444444444, "grad_norm": 0.3755132555961609, "learning_rate": 8.868953824098835e-05, "loss": 0.0363, "step": 8690 }, { "epoch": 7.733333333333333, "grad_norm": 0.34768441319465637, "learning_rate": 8.865848582502809e-05, "loss": 0.0293, "step": 8700 }, { "epoch": 7.742222222222222, "grad_norm": 0.25570908188819885, "learning_rate": 8.862739629288284e-05, "loss": 0.0337, "step": 8710 }, { "epoch": 7.751111111111111, "grad_norm": 0.33595192432403564, "learning_rate": 8.85962696744018e-05, "loss": 0.0249, "step": 8720 }, { "epoch": 7.76, "grad_norm": 0.4172503650188446, "learning_rate": 8.856510599946977e-05, "loss": 0.0252, "step": 8730 }, { "epoch": 7.768888888888889, "grad_norm": 0.3134748637676239, "learning_rate": 8.853390529800712e-05, "loss": 0.0275, "step": 8740 }, { "epoch": 7.777777777777778, "grad_norm": 0.364092618227005, "learning_rate": 8.850266759996979e-05, "loss": 0.0315, "step": 8750 }, { "epoch": 7.786666666666667, "grad_norm": 0.39060091972351074, "learning_rate": 8.847139293534926e-05, "loss": 0.032, "step": 8760 }, { "epoch": 7.795555555555556, "grad_norm": 0.2839621901512146, "learning_rate": 8.844008133417246e-05, "loss": 0.03, "step": 8770 }, { "epoch": 7.804444444444444, "grad_norm": 0.26483967900276184, "learning_rate": 8.840873282650178e-05, "loss": 0.0239, "step": 8780 }, { "epoch": 7.8133333333333335, "grad_norm": 0.23573113977909088, "learning_rate": 8.837734744243506e-05, "loss": 0.0298, "step": 8790 }, { "epoch": 7.822222222222222, "grad_norm": 0.2601010501384735, "learning_rate": 8.834592521210557e-05, "loss": 0.0319, "step": 8800 }, { "epoch": 7.831111111111111, "grad_norm": 0.4041232466697693, "learning_rate": 8.831446616568194e-05, "loss": 0.0316, "step": 8810 }, { "epoch": 7.84, "grad_norm": 0.4015391170978546, "learning_rate": 8.828297033336812e-05, "loss": 0.0344, "step": 8820 }, { "epoch": 7.848888888888889, "grad_norm": 0.33405235409736633, "learning_rate": 8.825143774540341e-05, "loss": 0.0255, "step": 8830 }, { "epoch": 7.857777777777778, "grad_norm": 0.32894036173820496, "learning_rate": 8.821986843206238e-05, "loss": 0.0256, "step": 8840 }, { "epoch": 7.866666666666667, "grad_norm": 0.2824382185935974, "learning_rate": 8.818826242365487e-05, "loss": 0.0368, "step": 8850 }, { "epoch": 7.875555555555556, "grad_norm": 0.24134701490402222, "learning_rate": 8.815661975052598e-05, "loss": 0.029, "step": 8860 }, { "epoch": 7.884444444444444, "grad_norm": 0.2910250425338745, "learning_rate": 8.812494044305593e-05, "loss": 0.0288, "step": 8870 }, { "epoch": 7.8933333333333335, "grad_norm": 0.4239876866340637, "learning_rate": 8.80932245316602e-05, "loss": 0.0303, "step": 8880 }, { "epoch": 7.902222222222222, "grad_norm": 0.2671927213668823, "learning_rate": 8.806147204678939e-05, "loss": 0.0297, "step": 8890 }, { "epoch": 7.911111111111111, "grad_norm": 0.5261624455451965, "learning_rate": 8.802968301892916e-05, "loss": 0.0262, "step": 8900 }, { "epoch": 7.92, "grad_norm": 0.35427501797676086, "learning_rate": 8.799785747860031e-05, "loss": 0.0321, "step": 8910 }, { "epoch": 7.928888888888888, "grad_norm": 0.23362451791763306, "learning_rate": 8.79659954563587e-05, "loss": 0.0263, "step": 8920 }, { "epoch": 7.937777777777778, "grad_norm": 0.35687702894210815, "learning_rate": 8.79340969827952e-05, "loss": 0.0325, "step": 8930 }, { "epoch": 7.946666666666666, "grad_norm": 0.5230968594551086, "learning_rate": 8.790216208853567e-05, "loss": 0.0271, "step": 8940 }, { "epoch": 7.955555555555556, "grad_norm": 0.2734135389328003, "learning_rate": 8.787019080424093e-05, "loss": 0.0243, "step": 8950 }, { "epoch": 7.964444444444444, "grad_norm": 0.22999460995197296, "learning_rate": 8.78381831606068e-05, "loss": 0.0263, "step": 8960 }, { "epoch": 7.973333333333334, "grad_norm": 0.3126041293144226, "learning_rate": 8.780613918836388e-05, "loss": 0.0288, "step": 8970 }, { "epoch": 7.982222222222222, "grad_norm": 0.24423398077487946, "learning_rate": 8.777405891827781e-05, "loss": 0.0312, "step": 8980 }, { "epoch": 7.9911111111111115, "grad_norm": 0.2044190764427185, "learning_rate": 8.774194238114899e-05, "loss": 0.0311, "step": 8990 }, { "epoch": 8.0, "grad_norm": 0.40201127529144287, "learning_rate": 8.77097896078126e-05, "loss": 0.029, "step": 9000 }, { "epoch": 8.008888888888889, "grad_norm": 0.33709120750427246, "learning_rate": 8.76776006291387e-05, "loss": 0.034, "step": 9010 }, { "epoch": 8.017777777777777, "grad_norm": 0.3465985953807831, "learning_rate": 8.764537547603208e-05, "loss": 0.0257, "step": 9020 }, { "epoch": 8.026666666666667, "grad_norm": 0.328377902507782, "learning_rate": 8.761311417943222e-05, "loss": 0.0271, "step": 9030 }, { "epoch": 8.035555555555556, "grad_norm": 0.3562348186969757, "learning_rate": 8.758081677031337e-05, "loss": 0.0266, "step": 9040 }, { "epoch": 8.044444444444444, "grad_norm": 0.39214637875556946, "learning_rate": 8.754848327968439e-05, "loss": 0.0314, "step": 9050 }, { "epoch": 8.053333333333333, "grad_norm": 0.3319479525089264, "learning_rate": 8.751611373858881e-05, "loss": 0.0273, "step": 9060 }, { "epoch": 8.062222222222223, "grad_norm": 0.43139010667800903, "learning_rate": 8.748370817810481e-05, "loss": 0.0329, "step": 9070 }, { "epoch": 8.071111111111112, "grad_norm": 0.29186728596687317, "learning_rate": 8.745126662934505e-05, "loss": 0.029, "step": 9080 }, { "epoch": 8.08, "grad_norm": 0.2697169780731201, "learning_rate": 8.741878912345687e-05, "loss": 0.0274, "step": 9090 }, { "epoch": 8.088888888888889, "grad_norm": 0.1896631270647049, "learning_rate": 8.738627569162202e-05, "loss": 0.0262, "step": 9100 }, { "epoch": 8.097777777777777, "grad_norm": 0.2692280411720276, "learning_rate": 8.735372636505681e-05, "loss": 0.0257, "step": 9110 }, { "epoch": 8.106666666666667, "grad_norm": 0.31216102838516235, "learning_rate": 8.7321141175012e-05, "loss": 0.0247, "step": 9120 }, { "epoch": 8.115555555555556, "grad_norm": 0.26826533675193787, "learning_rate": 8.728852015277276e-05, "loss": 0.0259, "step": 9130 }, { "epoch": 8.124444444444444, "grad_norm": 0.39493873715400696, "learning_rate": 8.72558633296587e-05, "loss": 0.0315, "step": 9140 }, { "epoch": 8.133333333333333, "grad_norm": 0.3448045551776886, "learning_rate": 8.722317073702378e-05, "loss": 0.0284, "step": 9150 }, { "epoch": 8.142222222222221, "grad_norm": 0.24198827147483826, "learning_rate": 8.719044240625627e-05, "loss": 0.0299, "step": 9160 }, { "epoch": 8.151111111111112, "grad_norm": 0.317514568567276, "learning_rate": 8.715767836877884e-05, "loss": 0.0296, "step": 9170 }, { "epoch": 8.16, "grad_norm": 0.39699625968933105, "learning_rate": 8.712487865604835e-05, "loss": 0.0245, "step": 9180 }, { "epoch": 8.168888888888889, "grad_norm": 0.3401624262332916, "learning_rate": 8.709204329955596e-05, "loss": 0.0277, "step": 9190 }, { "epoch": 8.177777777777777, "grad_norm": 0.329136461019516, "learning_rate": 8.705917233082704e-05, "loss": 0.0271, "step": 9200 }, { "epoch": 8.186666666666667, "grad_norm": 0.42991548776626587, "learning_rate": 8.702626578142114e-05, "loss": 0.0265, "step": 9210 }, { "epoch": 8.195555555555556, "grad_norm": 0.3494561016559601, "learning_rate": 8.699332368293203e-05, "loss": 0.0305, "step": 9220 }, { "epoch": 8.204444444444444, "grad_norm": 0.34544336795806885, "learning_rate": 8.696034606698753e-05, "loss": 0.03, "step": 9230 }, { "epoch": 8.213333333333333, "grad_norm": 0.23547375202178955, "learning_rate": 8.692733296524961e-05, "loss": 0.0278, "step": 9240 }, { "epoch": 8.222222222222221, "grad_norm": 0.32988983392715454, "learning_rate": 8.68942844094143e-05, "loss": 0.0285, "step": 9250 }, { "epoch": 8.231111111111112, "grad_norm": 0.3610537350177765, "learning_rate": 8.686120043121165e-05, "loss": 0.0337, "step": 9260 }, { "epoch": 8.24, "grad_norm": 0.31629642844200134, "learning_rate": 8.682808106240575e-05, "loss": 0.0316, "step": 9270 }, { "epoch": 8.248888888888889, "grad_norm": 0.4187433421611786, "learning_rate": 8.679492633479462e-05, "loss": 0.0356, "step": 9280 }, { "epoch": 8.257777777777777, "grad_norm": 0.2880855202674866, "learning_rate": 8.676173628021032e-05, "loss": 0.0328, "step": 9290 }, { "epoch": 8.266666666666667, "grad_norm": 0.43582767248153687, "learning_rate": 8.672851093051872e-05, "loss": 0.0334, "step": 9300 }, { "epoch": 8.275555555555556, "grad_norm": 0.36886680126190186, "learning_rate": 8.669525031761967e-05, "loss": 0.0254, "step": 9310 }, { "epoch": 8.284444444444444, "grad_norm": 0.2937784790992737, "learning_rate": 8.666195447344679e-05, "loss": 0.0267, "step": 9320 }, { "epoch": 8.293333333333333, "grad_norm": 0.3198051452636719, "learning_rate": 8.662862342996759e-05, "loss": 0.027, "step": 9330 }, { "epoch": 8.302222222222222, "grad_norm": 0.3318449854850769, "learning_rate": 8.659525721918335e-05, "loss": 0.0307, "step": 9340 }, { "epoch": 8.311111111111112, "grad_norm": 0.21017032861709595, "learning_rate": 8.656185587312912e-05, "loss": 0.0307, "step": 9350 }, { "epoch": 8.32, "grad_norm": 0.2729027271270752, "learning_rate": 8.652841942387369e-05, "loss": 0.029, "step": 9360 }, { "epoch": 8.328888888888889, "grad_norm": 0.3152466118335724, "learning_rate": 8.649494790351954e-05, "loss": 0.0229, "step": 9370 }, { "epoch": 8.337777777777777, "grad_norm": 0.30758124589920044, "learning_rate": 8.646144134420281e-05, "loss": 0.034, "step": 9380 }, { "epoch": 8.346666666666668, "grad_norm": 0.4215352237224579, "learning_rate": 8.642789977809332e-05, "loss": 0.0326, "step": 9390 }, { "epoch": 8.355555555555556, "grad_norm": 0.48360493779182434, "learning_rate": 8.639432323739445e-05, "loss": 0.0281, "step": 9400 }, { "epoch": 8.364444444444445, "grad_norm": 0.3996516466140747, "learning_rate": 8.636071175434321e-05, "loss": 0.0313, "step": 9410 }, { "epoch": 8.373333333333333, "grad_norm": 0.42908772826194763, "learning_rate": 8.632706536121014e-05, "loss": 0.0301, "step": 9420 }, { "epoch": 8.382222222222222, "grad_norm": 0.3858035206794739, "learning_rate": 8.629338409029927e-05, "loss": 0.0388, "step": 9430 }, { "epoch": 8.391111111111112, "grad_norm": 0.3011014759540558, "learning_rate": 8.625966797394816e-05, "loss": 0.0332, "step": 9440 }, { "epoch": 8.4, "grad_norm": 0.36272507905960083, "learning_rate": 8.622591704452778e-05, "loss": 0.0293, "step": 9450 }, { "epoch": 8.408888888888889, "grad_norm": 0.29575657844543457, "learning_rate": 8.619213133444257e-05, "loss": 0.0339, "step": 9460 }, { "epoch": 8.417777777777777, "grad_norm": 0.3715441823005676, "learning_rate": 8.615831087613034e-05, "loss": 0.0299, "step": 9470 }, { "epoch": 8.426666666666666, "grad_norm": 0.2470073699951172, "learning_rate": 8.612445570206224e-05, "loss": 0.0301, "step": 9480 }, { "epoch": 8.435555555555556, "grad_norm": 0.34828224778175354, "learning_rate": 8.60905658447428e-05, "loss": 0.0294, "step": 9490 }, { "epoch": 8.444444444444445, "grad_norm": 0.30567726492881775, "learning_rate": 8.605664133670982e-05, "loss": 0.029, "step": 9500 }, { "epoch": 8.453333333333333, "grad_norm": 0.19591040909290314, "learning_rate": 8.602268221053436e-05, "loss": 0.0307, "step": 9510 }, { "epoch": 8.462222222222222, "grad_norm": 0.4021168649196625, "learning_rate": 8.598868849882073e-05, "loss": 0.0293, "step": 9520 }, { "epoch": 8.471111111111112, "grad_norm": 0.30323994159698486, "learning_rate": 8.595466023420647e-05, "loss": 0.0245, "step": 9530 }, { "epoch": 8.48, "grad_norm": 0.31285908818244934, "learning_rate": 8.592059744936222e-05, "loss": 0.0263, "step": 9540 }, { "epoch": 8.488888888888889, "grad_norm": 0.30331525206565857, "learning_rate": 8.588650017699185e-05, "loss": 0.0266, "step": 9550 }, { "epoch": 8.497777777777777, "grad_norm": 0.2594868242740631, "learning_rate": 8.58523684498323e-05, "loss": 0.0297, "step": 9560 }, { "epoch": 8.506666666666666, "grad_norm": 0.36788448691368103, "learning_rate": 8.581820230065356e-05, "loss": 0.0236, "step": 9570 }, { "epoch": 8.515555555555556, "grad_norm": 0.35221222043037415, "learning_rate": 8.578400176225876e-05, "loss": 0.0274, "step": 9580 }, { "epoch": 8.524444444444445, "grad_norm": 0.36423736810684204, "learning_rate": 8.574976686748395e-05, "loss": 0.0236, "step": 9590 }, { "epoch": 8.533333333333333, "grad_norm": 0.41025352478027344, "learning_rate": 8.571549764919822e-05, "loss": 0.0274, "step": 9600 }, { "epoch": 8.542222222222222, "grad_norm": 0.3262971043586731, "learning_rate": 8.568119414030357e-05, "loss": 0.0247, "step": 9610 }, { "epoch": 8.551111111111112, "grad_norm": 0.27208971977233887, "learning_rate": 8.564685637373501e-05, "loss": 0.0288, "step": 9620 }, { "epoch": 8.56, "grad_norm": 0.31660622358322144, "learning_rate": 8.561248438246034e-05, "loss": 0.0282, "step": 9630 }, { "epoch": 8.568888888888889, "grad_norm": 0.3890058696269989, "learning_rate": 8.557807819948024e-05, "loss": 0.0246, "step": 9640 }, { "epoch": 8.577777777777778, "grad_norm": 0.35399895906448364, "learning_rate": 8.55436378578283e-05, "loss": 0.0298, "step": 9650 }, { "epoch": 8.586666666666666, "grad_norm": 0.2688673138618469, "learning_rate": 8.550916339057083e-05, "loss": 0.0293, "step": 9660 }, { "epoch": 8.595555555555556, "grad_norm": 0.2690223455429077, "learning_rate": 8.54746548308069e-05, "loss": 0.0266, "step": 9670 }, { "epoch": 8.604444444444445, "grad_norm": 0.30771031975746155, "learning_rate": 8.544011221166834e-05, "loss": 0.0253, "step": 9680 }, { "epoch": 8.613333333333333, "grad_norm": 0.4690385162830353, "learning_rate": 8.540553556631967e-05, "loss": 0.0282, "step": 9690 }, { "epoch": 8.622222222222222, "grad_norm": 0.31280988454818726, "learning_rate": 8.537092492795808e-05, "loss": 0.0275, "step": 9700 }, { "epoch": 8.63111111111111, "grad_norm": 0.2500731945037842, "learning_rate": 8.53362803298134e-05, "loss": 0.0256, "step": 9710 }, { "epoch": 8.64, "grad_norm": 0.31095069646835327, "learning_rate": 8.530160180514809e-05, "loss": 0.0274, "step": 9720 }, { "epoch": 8.648888888888889, "grad_norm": 0.23119139671325684, "learning_rate": 8.526688938725711e-05, "loss": 0.0244, "step": 9730 }, { "epoch": 8.657777777777778, "grad_norm": 0.26702168583869934, "learning_rate": 8.523214310946804e-05, "loss": 0.0267, "step": 9740 }, { "epoch": 8.666666666666666, "grad_norm": 0.2595406174659729, "learning_rate": 8.519736300514094e-05, "loss": 0.0275, "step": 9750 }, { "epoch": 8.675555555555556, "grad_norm": 0.3722926676273346, "learning_rate": 8.516254910766832e-05, "loss": 0.0294, "step": 9760 }, { "epoch": 8.684444444444445, "grad_norm": 0.38656577467918396, "learning_rate": 8.512770145047518e-05, "loss": 0.0251, "step": 9770 }, { "epoch": 8.693333333333333, "grad_norm": 0.30050814151763916, "learning_rate": 8.509282006701888e-05, "loss": 0.0249, "step": 9780 }, { "epoch": 8.702222222222222, "grad_norm": 0.4044692814350128, "learning_rate": 8.505790499078923e-05, "loss": 0.0257, "step": 9790 }, { "epoch": 8.71111111111111, "grad_norm": 0.38244083523750305, "learning_rate": 8.502295625530832e-05, "loss": 0.0313, "step": 9800 }, { "epoch": 8.72, "grad_norm": 0.3544146418571472, "learning_rate": 8.498797389413063e-05, "loss": 0.0346, "step": 9810 }, { "epoch": 8.72888888888889, "grad_norm": 0.33287522196769714, "learning_rate": 8.495295794084284e-05, "loss": 0.0314, "step": 9820 }, { "epoch": 8.737777777777778, "grad_norm": 0.36457380652427673, "learning_rate": 8.491790842906392e-05, "loss": 0.0287, "step": 9830 }, { "epoch": 8.746666666666666, "grad_norm": 0.31409263610839844, "learning_rate": 8.488282539244508e-05, "loss": 0.025, "step": 9840 }, { "epoch": 8.755555555555556, "grad_norm": 0.3508381247520447, "learning_rate": 8.484770886466972e-05, "loss": 0.0255, "step": 9850 }, { "epoch": 8.764444444444445, "grad_norm": 0.26290854811668396, "learning_rate": 8.481255887945334e-05, "loss": 0.0309, "step": 9860 }, { "epoch": 8.773333333333333, "grad_norm": 0.36217108368873596, "learning_rate": 8.477737547054362e-05, "loss": 0.0257, "step": 9870 }, { "epoch": 8.782222222222222, "grad_norm": 0.3703044354915619, "learning_rate": 8.474215867172029e-05, "loss": 0.0283, "step": 9880 }, { "epoch": 8.79111111111111, "grad_norm": 0.26244011521339417, "learning_rate": 8.470690851679517e-05, "loss": 0.0251, "step": 9890 }, { "epoch": 8.8, "grad_norm": 0.3017309606075287, "learning_rate": 8.467162503961208e-05, "loss": 0.0315, "step": 9900 }, { "epoch": 8.80888888888889, "grad_norm": 0.30173957347869873, "learning_rate": 8.463630827404687e-05, "loss": 0.0254, "step": 9910 }, { "epoch": 8.817777777777778, "grad_norm": 0.3225434720516205, "learning_rate": 8.46009582540073e-05, "loss": 0.025, "step": 9920 }, { "epoch": 8.826666666666666, "grad_norm": 0.38454321026802063, "learning_rate": 8.456557501343308e-05, "loss": 0.0249, "step": 9930 }, { "epoch": 8.835555555555555, "grad_norm": 0.34209445118904114, "learning_rate": 8.453015858629583e-05, "loss": 0.0297, "step": 9940 }, { "epoch": 8.844444444444445, "grad_norm": 0.27084240317344666, "learning_rate": 8.449470900659903e-05, "loss": 0.0238, "step": 9950 }, { "epoch": 8.853333333333333, "grad_norm": 0.27318212389945984, "learning_rate": 8.445922630837795e-05, "loss": 0.0219, "step": 9960 }, { "epoch": 8.862222222222222, "grad_norm": 0.3107755184173584, "learning_rate": 8.44237105256997e-05, "loss": 0.0252, "step": 9970 }, { "epoch": 8.87111111111111, "grad_norm": 0.26141607761383057, "learning_rate": 8.438816169266314e-05, "loss": 0.0265, "step": 9980 }, { "epoch": 8.88, "grad_norm": 0.38170087337493896, "learning_rate": 8.435257984339888e-05, "loss": 0.0254, "step": 9990 }, { "epoch": 8.88888888888889, "grad_norm": 0.22981029748916626, "learning_rate": 8.43169650120692e-05, "loss": 0.0261, "step": 10000 }, { "epoch": 8.897777777777778, "grad_norm": 0.367840439081192, "learning_rate": 8.428131723286807e-05, "loss": 0.0233, "step": 10010 }, { "epoch": 8.906666666666666, "grad_norm": 0.3327315151691437, "learning_rate": 8.424563654002104e-05, "loss": 0.0268, "step": 10020 }, { "epoch": 8.915555555555555, "grad_norm": 0.22993861138820648, "learning_rate": 8.420992296778536e-05, "loss": 0.0251, "step": 10030 }, { "epoch": 8.924444444444445, "grad_norm": 0.33790671825408936, "learning_rate": 8.417417655044977e-05, "loss": 0.028, "step": 10040 }, { "epoch": 8.933333333333334, "grad_norm": 0.2989477217197418, "learning_rate": 8.413839732233452e-05, "loss": 0.0245, "step": 10050 }, { "epoch": 8.942222222222222, "grad_norm": 0.3165825605392456, "learning_rate": 8.410258531779148e-05, "loss": 0.0312, "step": 10060 }, { "epoch": 8.95111111111111, "grad_norm": 0.36194103956222534, "learning_rate": 8.406674057120387e-05, "loss": 0.0277, "step": 10070 }, { "epoch": 8.96, "grad_norm": 0.33868658542633057, "learning_rate": 8.403086311698642e-05, "loss": 0.0264, "step": 10080 }, { "epoch": 8.96888888888889, "grad_norm": 0.3958742916584015, "learning_rate": 8.399495298958519e-05, "loss": 0.0268, "step": 10090 }, { "epoch": 8.977777777777778, "grad_norm": 0.37399956583976746, "learning_rate": 8.395901022347769e-05, "loss": 0.0245, "step": 10100 }, { "epoch": 8.986666666666666, "grad_norm": 0.2834714949131012, "learning_rate": 8.392303485317272e-05, "loss": 0.0262, "step": 10110 }, { "epoch": 8.995555555555555, "grad_norm": 0.37519580125808716, "learning_rate": 8.38870269132104e-05, "loss": 0.0217, "step": 10120 }, { "epoch": 9.004444444444445, "grad_norm": 0.41434457898139954, "learning_rate": 8.385098643816212e-05, "loss": 0.0257, "step": 10130 }, { "epoch": 9.013333333333334, "grad_norm": 0.24073714017868042, "learning_rate": 8.381491346263047e-05, "loss": 0.0262, "step": 10140 }, { "epoch": 9.022222222222222, "grad_norm": 0.38471877574920654, "learning_rate": 8.377880802124931e-05, "loss": 0.0258, "step": 10150 }, { "epoch": 9.03111111111111, "grad_norm": 0.3132060468196869, "learning_rate": 8.37426701486836e-05, "loss": 0.0267, "step": 10160 }, { "epoch": 9.04, "grad_norm": 0.32572582364082336, "learning_rate": 8.370649987962951e-05, "loss": 0.0289, "step": 10170 }, { "epoch": 9.04888888888889, "grad_norm": 0.26974397897720337, "learning_rate": 8.367029724881426e-05, "loss": 0.0258, "step": 10180 }, { "epoch": 9.057777777777778, "grad_norm": 0.3145279288291931, "learning_rate": 8.363406229099617e-05, "loss": 0.0315, "step": 10190 }, { "epoch": 9.066666666666666, "grad_norm": 0.37162208557128906, "learning_rate": 8.359779504096455e-05, "loss": 0.0304, "step": 10200 }, { "epoch": 9.075555555555555, "grad_norm": 0.2649388015270233, "learning_rate": 8.356149553353977e-05, "loss": 0.0281, "step": 10210 }, { "epoch": 9.084444444444445, "grad_norm": 0.3943069577217102, "learning_rate": 8.352516380357317e-05, "loss": 0.0306, "step": 10220 }, { "epoch": 9.093333333333334, "grad_norm": 0.31964242458343506, "learning_rate": 8.348879988594697e-05, "loss": 0.0243, "step": 10230 }, { "epoch": 9.102222222222222, "grad_norm": 0.31255823373794556, "learning_rate": 8.345240381557438e-05, "loss": 0.0271, "step": 10240 }, { "epoch": 9.11111111111111, "grad_norm": 0.2642662823200226, "learning_rate": 8.341597562739936e-05, "loss": 0.0252, "step": 10250 }, { "epoch": 9.12, "grad_norm": 0.2754298150539398, "learning_rate": 8.337951535639683e-05, "loss": 0.0315, "step": 10260 }, { "epoch": 9.12888888888889, "grad_norm": 0.3356761634349823, "learning_rate": 8.334302303757244e-05, "loss": 0.0245, "step": 10270 }, { "epoch": 9.137777777777778, "grad_norm": 0.3413902521133423, "learning_rate": 8.330649870596263e-05, "loss": 0.0254, "step": 10280 }, { "epoch": 9.146666666666667, "grad_norm": 0.32140642404556274, "learning_rate": 8.326994239663458e-05, "loss": 0.0309, "step": 10290 }, { "epoch": 9.155555555555555, "grad_norm": 0.2768080234527588, "learning_rate": 8.323335414468616e-05, "loss": 0.03, "step": 10300 }, { "epoch": 9.164444444444445, "grad_norm": 0.3029347062110901, "learning_rate": 8.319673398524591e-05, "loss": 0.0266, "step": 10310 }, { "epoch": 9.173333333333334, "grad_norm": 0.316653311252594, "learning_rate": 8.316008195347302e-05, "loss": 0.0214, "step": 10320 }, { "epoch": 9.182222222222222, "grad_norm": 0.4333246052265167, "learning_rate": 8.312339808455726e-05, "loss": 0.0238, "step": 10330 }, { "epoch": 9.19111111111111, "grad_norm": 0.25465619564056396, "learning_rate": 8.308668241371897e-05, "loss": 0.0277, "step": 10340 }, { "epoch": 9.2, "grad_norm": 0.33712172508239746, "learning_rate": 8.304993497620906e-05, "loss": 0.0273, "step": 10350 }, { "epoch": 9.20888888888889, "grad_norm": 0.2789132297039032, "learning_rate": 8.301315580730887e-05, "loss": 0.0254, "step": 10360 }, { "epoch": 9.217777777777778, "grad_norm": 0.3308916985988617, "learning_rate": 8.297634494233026e-05, "loss": 0.0256, "step": 10370 }, { "epoch": 9.226666666666667, "grad_norm": 0.3375512659549713, "learning_rate": 8.293950241661553e-05, "loss": 0.0276, "step": 10380 }, { "epoch": 9.235555555555555, "grad_norm": 0.2788207530975342, "learning_rate": 8.290262826553732e-05, "loss": 0.0284, "step": 10390 }, { "epoch": 9.244444444444444, "grad_norm": 0.2589036524295807, "learning_rate": 8.286572252449869e-05, "loss": 0.0226, "step": 10400 }, { "epoch": 9.253333333333334, "grad_norm": 0.3000525236129761, "learning_rate": 8.282878522893298e-05, "loss": 0.0269, "step": 10410 }, { "epoch": 9.262222222222222, "grad_norm": 0.3163992464542389, "learning_rate": 8.279181641430388e-05, "loss": 0.033, "step": 10420 }, { "epoch": 9.27111111111111, "grad_norm": 0.29609814286231995, "learning_rate": 8.275481611610529e-05, "loss": 0.0231, "step": 10430 }, { "epoch": 9.28, "grad_norm": 0.22813443839550018, "learning_rate": 8.27177843698614e-05, "loss": 0.0342, "step": 10440 }, { "epoch": 9.28888888888889, "grad_norm": 0.2993294596672058, "learning_rate": 8.26807212111265e-05, "loss": 0.0235, "step": 10450 }, { "epoch": 9.297777777777778, "grad_norm": 0.3490791916847229, "learning_rate": 8.264362667548515e-05, "loss": 0.0318, "step": 10460 }, { "epoch": 9.306666666666667, "grad_norm": 0.22168734669685364, "learning_rate": 8.260650079855195e-05, "loss": 0.0264, "step": 10470 }, { "epoch": 9.315555555555555, "grad_norm": 0.3533547520637512, "learning_rate": 8.256934361597159e-05, "loss": 0.0211, "step": 10480 }, { "epoch": 9.324444444444444, "grad_norm": 0.4388718008995056, "learning_rate": 8.253215516341893e-05, "loss": 0.0273, "step": 10490 }, { "epoch": 9.333333333333334, "grad_norm": 0.2983302175998688, "learning_rate": 8.249493547659869e-05, "loss": 0.0274, "step": 10500 }, { "epoch": 9.342222222222222, "grad_norm": 0.370485782623291, "learning_rate": 8.24576845912457e-05, "loss": 0.0268, "step": 10510 }, { "epoch": 9.351111111111111, "grad_norm": 0.26842427253723145, "learning_rate": 8.242040254312469e-05, "loss": 0.0233, "step": 10520 }, { "epoch": 9.36, "grad_norm": 0.30940890312194824, "learning_rate": 8.23830893680303e-05, "loss": 0.0262, "step": 10530 }, { "epoch": 9.36888888888889, "grad_norm": 0.3233903646469116, "learning_rate": 8.234574510178713e-05, "loss": 0.0239, "step": 10540 }, { "epoch": 9.377777777777778, "grad_norm": 0.38977694511413574, "learning_rate": 8.230836978024951e-05, "loss": 0.0242, "step": 10550 }, { "epoch": 9.386666666666667, "grad_norm": 0.3499661684036255, "learning_rate": 8.227096343930171e-05, "loss": 0.0289, "step": 10560 }, { "epoch": 9.395555555555555, "grad_norm": 0.3802552819252014, "learning_rate": 8.223352611485769e-05, "loss": 0.0237, "step": 10570 }, { "epoch": 9.404444444444444, "grad_norm": 0.32579535245895386, "learning_rate": 8.219605784286121e-05, "loss": 0.0255, "step": 10580 }, { "epoch": 9.413333333333334, "grad_norm": 0.3040288984775543, "learning_rate": 8.215855865928572e-05, "loss": 0.0213, "step": 10590 }, { "epoch": 9.422222222222222, "grad_norm": 0.35177311301231384, "learning_rate": 8.212102860013435e-05, "loss": 0.0278, "step": 10600 }, { "epoch": 9.431111111111111, "grad_norm": 0.2924827039241791, "learning_rate": 8.208346770143988e-05, "loss": 0.0249, "step": 10610 }, { "epoch": 9.44, "grad_norm": 0.3944423794746399, "learning_rate": 8.204587599926472e-05, "loss": 0.0254, "step": 10620 }, { "epoch": 9.448888888888888, "grad_norm": 0.2560625672340393, "learning_rate": 8.200825352970078e-05, "loss": 0.024, "step": 10630 }, { "epoch": 9.457777777777778, "grad_norm": 0.5386123657226562, "learning_rate": 8.197060032886961e-05, "loss": 0.0223, "step": 10640 }, { "epoch": 9.466666666666667, "grad_norm": 0.2913590669631958, "learning_rate": 8.193291643292218e-05, "loss": 0.0279, "step": 10650 }, { "epoch": 9.475555555555555, "grad_norm": 0.2517930567264557, "learning_rate": 8.1895201878039e-05, "loss": 0.0287, "step": 10660 }, { "epoch": 9.484444444444444, "grad_norm": 0.2865137755870819, "learning_rate": 8.185745670042994e-05, "loss": 0.0279, "step": 10670 }, { "epoch": 9.493333333333334, "grad_norm": 0.28178584575653076, "learning_rate": 8.181968093633436e-05, "loss": 0.0288, "step": 10680 }, { "epoch": 9.502222222222223, "grad_norm": 0.23410628736019135, "learning_rate": 8.178187462202092e-05, "loss": 0.0243, "step": 10690 }, { "epoch": 9.511111111111111, "grad_norm": 0.36403676867485046, "learning_rate": 8.174403779378764e-05, "loss": 0.0246, "step": 10700 }, { "epoch": 9.52, "grad_norm": 0.22028061747550964, "learning_rate": 8.170617048796181e-05, "loss": 0.0237, "step": 10710 }, { "epoch": 9.528888888888888, "grad_norm": 0.27803319692611694, "learning_rate": 8.166827274089999e-05, "loss": 0.023, "step": 10720 }, { "epoch": 9.537777777777778, "grad_norm": 0.24052701890468597, "learning_rate": 8.163034458898802e-05, "loss": 0.0275, "step": 10730 }, { "epoch": 9.546666666666667, "grad_norm": 0.34541645646095276, "learning_rate": 8.159238606864084e-05, "loss": 0.0263, "step": 10740 }, { "epoch": 9.555555555555555, "grad_norm": 0.26138612627983093, "learning_rate": 8.155439721630264e-05, "loss": 0.0243, "step": 10750 }, { "epoch": 9.564444444444444, "grad_norm": 0.4534415006637573, "learning_rate": 8.151637806844665e-05, "loss": 0.0286, "step": 10760 }, { "epoch": 9.573333333333334, "grad_norm": 0.3826700747013092, "learning_rate": 8.147832866157526e-05, "loss": 0.0234, "step": 10770 }, { "epoch": 9.582222222222223, "grad_norm": 0.3062479496002197, "learning_rate": 8.144024903221983e-05, "loss": 0.026, "step": 10780 }, { "epoch": 9.591111111111111, "grad_norm": 0.3133678436279297, "learning_rate": 8.140213921694082e-05, "loss": 0.0282, "step": 10790 }, { "epoch": 9.6, "grad_norm": 0.35873696208000183, "learning_rate": 8.136399925232762e-05, "loss": 0.0272, "step": 10800 }, { "epoch": 9.608888888888888, "grad_norm": 0.3182799816131592, "learning_rate": 8.132582917499858e-05, "loss": 0.0289, "step": 10810 }, { "epoch": 9.617777777777778, "grad_norm": 0.29153668880462646, "learning_rate": 8.128762902160096e-05, "loss": 0.0215, "step": 10820 }, { "epoch": 9.626666666666667, "grad_norm": 0.3184148073196411, "learning_rate": 8.124939882881089e-05, "loss": 0.0247, "step": 10830 }, { "epoch": 9.635555555555555, "grad_norm": 0.26010560989379883, "learning_rate": 8.121113863333337e-05, "loss": 0.0257, "step": 10840 }, { "epoch": 9.644444444444444, "grad_norm": 0.2743711471557617, "learning_rate": 8.117284847190218e-05, "loss": 0.0284, "step": 10850 }, { "epoch": 9.653333333333332, "grad_norm": 0.2966720163822174, "learning_rate": 8.113452838127986e-05, "loss": 0.0297, "step": 10860 }, { "epoch": 9.662222222222223, "grad_norm": 0.32659274339675903, "learning_rate": 8.109617839825771e-05, "loss": 0.0271, "step": 10870 }, { "epoch": 9.671111111111111, "grad_norm": 0.29668891429901123, "learning_rate": 8.105779855965572e-05, "loss": 0.0335, "step": 10880 }, { "epoch": 9.68, "grad_norm": 0.32903584837913513, "learning_rate": 8.101938890232254e-05, "loss": 0.0288, "step": 10890 }, { "epoch": 9.688888888888888, "grad_norm": 0.27863362431526184, "learning_rate": 8.098094946313547e-05, "loss": 0.0225, "step": 10900 }, { "epoch": 9.697777777777778, "grad_norm": 0.28241321444511414, "learning_rate": 8.094248027900038e-05, "loss": 0.0298, "step": 10910 }, { "epoch": 9.706666666666667, "grad_norm": 0.26574158668518066, "learning_rate": 8.090398138685169e-05, "loss": 0.0268, "step": 10920 }, { "epoch": 9.715555555555556, "grad_norm": 0.31064605712890625, "learning_rate": 8.086545282365236e-05, "loss": 0.0276, "step": 10930 }, { "epoch": 9.724444444444444, "grad_norm": 0.30327731370925903, "learning_rate": 8.082689462639386e-05, "loss": 0.0257, "step": 10940 }, { "epoch": 9.733333333333333, "grad_norm": 0.26208731532096863, "learning_rate": 8.078830683209607e-05, "loss": 0.0264, "step": 10950 }, { "epoch": 9.742222222222223, "grad_norm": 0.23091182112693787, "learning_rate": 8.07496894778073e-05, "loss": 0.0245, "step": 10960 }, { "epoch": 9.751111111111111, "grad_norm": 0.2932151257991791, "learning_rate": 8.071104260060426e-05, "loss": 0.0302, "step": 10970 }, { "epoch": 9.76, "grad_norm": 0.32203319668769836, "learning_rate": 8.067236623759197e-05, "loss": 0.024, "step": 10980 }, { "epoch": 9.768888888888888, "grad_norm": 0.23526550829410553, "learning_rate": 8.06336604259038e-05, "loss": 0.0246, "step": 10990 }, { "epoch": 9.777777777777779, "grad_norm": 0.2983398735523224, "learning_rate": 8.059492520270135e-05, "loss": 0.0227, "step": 11000 }, { "epoch": 9.786666666666667, "grad_norm": 0.35906434059143066, "learning_rate": 8.05561606051745e-05, "loss": 0.0248, "step": 11010 }, { "epoch": 9.795555555555556, "grad_norm": 0.32534757256507874, "learning_rate": 8.05173666705413e-05, "loss": 0.019, "step": 11020 }, { "epoch": 9.804444444444444, "grad_norm": 0.32276368141174316, "learning_rate": 8.047854343604799e-05, "loss": 0.0319, "step": 11030 }, { "epoch": 9.813333333333333, "grad_norm": 0.31563565135002136, "learning_rate": 8.043969093896893e-05, "loss": 0.0244, "step": 11040 }, { "epoch": 9.822222222222223, "grad_norm": 0.28249841928482056, "learning_rate": 8.040080921660657e-05, "loss": 0.0254, "step": 11050 }, { "epoch": 9.831111111111111, "grad_norm": 0.38400816917419434, "learning_rate": 8.036189830629144e-05, "loss": 0.0264, "step": 11060 }, { "epoch": 9.84, "grad_norm": 0.25360342860221863, "learning_rate": 8.032295824538206e-05, "loss": 0.0337, "step": 11070 }, { "epoch": 9.848888888888888, "grad_norm": 0.37573981285095215, "learning_rate": 8.028398907126494e-05, "loss": 0.0254, "step": 11080 }, { "epoch": 9.857777777777777, "grad_norm": 0.31982266902923584, "learning_rate": 8.024499082135461e-05, "loss": 0.0235, "step": 11090 }, { "epoch": 9.866666666666667, "grad_norm": 0.3561207354068756, "learning_rate": 8.020596353309342e-05, "loss": 0.0301, "step": 11100 }, { "epoch": 9.875555555555556, "grad_norm": 0.29419976472854614, "learning_rate": 8.016690724395167e-05, "loss": 0.03, "step": 11110 }, { "epoch": 9.884444444444444, "grad_norm": 0.40748661756515503, "learning_rate": 8.012782199142747e-05, "loss": 0.0251, "step": 11120 }, { "epoch": 9.893333333333333, "grad_norm": 0.4318525493144989, "learning_rate": 8.008870781304674e-05, "loss": 0.0293, "step": 11130 }, { "epoch": 9.902222222222223, "grad_norm": 0.43829092383384705, "learning_rate": 8.004956474636316e-05, "loss": 0.0249, "step": 11140 }, { "epoch": 9.911111111111111, "grad_norm": 0.2352384626865387, "learning_rate": 8.001039282895822e-05, "loss": 0.0221, "step": 11150 }, { "epoch": 9.92, "grad_norm": 0.362231582403183, "learning_rate": 7.997119209844099e-05, "loss": 0.0247, "step": 11160 }, { "epoch": 9.928888888888888, "grad_norm": 0.27405253052711487, "learning_rate": 7.99319625924483e-05, "loss": 0.0258, "step": 11170 }, { "epoch": 9.937777777777779, "grad_norm": 0.49628233909606934, "learning_rate": 7.989270434864458e-05, "loss": 0.0229, "step": 11180 }, { "epoch": 9.946666666666667, "grad_norm": 0.2253984957933426, "learning_rate": 7.985341740472183e-05, "loss": 0.0236, "step": 11190 }, { "epoch": 9.955555555555556, "grad_norm": 0.2590222656726837, "learning_rate": 7.98141017983996e-05, "loss": 0.0319, "step": 11200 }, { "epoch": 9.964444444444444, "grad_norm": 0.4041983485221863, "learning_rate": 7.977475756742502e-05, "loss": 0.0254, "step": 11210 }, { "epoch": 9.973333333333333, "grad_norm": 0.5051749348640442, "learning_rate": 7.973538474957265e-05, "loss": 0.0293, "step": 11220 }, { "epoch": 9.982222222222223, "grad_norm": 0.29879531264305115, "learning_rate": 7.96959833826445e-05, "loss": 0.0255, "step": 11230 }, { "epoch": 9.991111111111111, "grad_norm": 0.2785200774669647, "learning_rate": 7.965655350446998e-05, "loss": 0.0249, "step": 11240 }, { "epoch": 10.0, "grad_norm": 0.49701258540153503, "learning_rate": 7.961709515290595e-05, "loss": 0.0269, "step": 11250 }, { "epoch": 10.008888888888889, "grad_norm": 0.44736772775650024, "learning_rate": 7.957760836583648e-05, "loss": 0.0267, "step": 11260 }, { "epoch": 10.017777777777777, "grad_norm": 0.3122999966144562, "learning_rate": 7.953809318117307e-05, "loss": 0.0306, "step": 11270 }, { "epoch": 10.026666666666667, "grad_norm": 0.35206544399261475, "learning_rate": 7.949854963685438e-05, "loss": 0.0241, "step": 11280 }, { "epoch": 10.035555555555556, "grad_norm": 0.3072154223918915, "learning_rate": 7.945897777084637e-05, "loss": 0.0243, "step": 11290 }, { "epoch": 10.044444444444444, "grad_norm": 0.33581066131591797, "learning_rate": 7.941937762114215e-05, "loss": 0.0281, "step": 11300 }, { "epoch": 10.053333333333333, "grad_norm": 0.2976914942264557, "learning_rate": 7.937974922576201e-05, "loss": 0.0314, "step": 11310 }, { "epoch": 10.062222222222223, "grad_norm": 0.3194797933101654, "learning_rate": 7.934009262275338e-05, "loss": 0.0262, "step": 11320 }, { "epoch": 10.071111111111112, "grad_norm": 0.2215932160615921, "learning_rate": 7.930040785019069e-05, "loss": 0.0266, "step": 11330 }, { "epoch": 10.08, "grad_norm": 0.32338833808898926, "learning_rate": 7.926069494617547e-05, "loss": 0.0293, "step": 11340 }, { "epoch": 10.088888888888889, "grad_norm": 0.2830520570278168, "learning_rate": 7.922095394883628e-05, "loss": 0.0272, "step": 11350 }, { "epoch": 10.097777777777777, "grad_norm": 0.33606281876564026, "learning_rate": 7.918118489632865e-05, "loss": 0.0301, "step": 11360 }, { "epoch": 10.106666666666667, "grad_norm": 0.36245089769363403, "learning_rate": 7.914138782683498e-05, "loss": 0.023, "step": 11370 }, { "epoch": 10.115555555555556, "grad_norm": 0.26947593688964844, "learning_rate": 7.910156277856462e-05, "loss": 0.0207, "step": 11380 }, { "epoch": 10.124444444444444, "grad_norm": 0.3257203698158264, "learning_rate": 7.906170978975381e-05, "loss": 0.0282, "step": 11390 }, { "epoch": 10.133333333333333, "grad_norm": 0.3752073645591736, "learning_rate": 7.902182889866554e-05, "loss": 0.0303, "step": 11400 }, { "epoch": 10.142222222222221, "grad_norm": 0.33405786752700806, "learning_rate": 7.898192014358967e-05, "loss": 0.0277, "step": 11410 }, { "epoch": 10.151111111111112, "grad_norm": 0.3923734724521637, "learning_rate": 7.894198356284276e-05, "loss": 0.0289, "step": 11420 }, { "epoch": 10.16, "grad_norm": 0.3147445023059845, "learning_rate": 7.89020191947681e-05, "loss": 0.0242, "step": 11430 }, { "epoch": 10.168888888888889, "grad_norm": 0.2973533570766449, "learning_rate": 7.886202707773565e-05, "loss": 0.0261, "step": 11440 }, { "epoch": 10.177777777777777, "grad_norm": 0.20688630640506744, "learning_rate": 7.882200725014201e-05, "loss": 0.0274, "step": 11450 }, { "epoch": 10.186666666666667, "grad_norm": 0.3064502477645874, "learning_rate": 7.878195975041043e-05, "loss": 0.0283, "step": 11460 }, { "epoch": 10.195555555555556, "grad_norm": 0.3316955268383026, "learning_rate": 7.874188461699069e-05, "loss": 0.0254, "step": 11470 }, { "epoch": 10.204444444444444, "grad_norm": 0.3041556179523468, "learning_rate": 7.870178188835906e-05, "loss": 0.027, "step": 11480 }, { "epoch": 10.213333333333333, "grad_norm": 0.32648777961730957, "learning_rate": 7.86616516030184e-05, "loss": 0.0205, "step": 11490 }, { "epoch": 10.222222222222221, "grad_norm": 0.3763282299041748, "learning_rate": 7.862149379949793e-05, "loss": 0.0244, "step": 11500 }, { "epoch": 10.231111111111112, "grad_norm": 0.3468310832977295, "learning_rate": 7.858130851635339e-05, "loss": 0.0291, "step": 11510 }, { "epoch": 10.24, "grad_norm": 0.4418081045150757, "learning_rate": 7.854109579216681e-05, "loss": 0.0285, "step": 11520 }, { "epoch": 10.248888888888889, "grad_norm": 0.3643500804901123, "learning_rate": 7.850085566554661e-05, "loss": 0.0261, "step": 11530 }, { "epoch": 10.257777777777777, "grad_norm": 0.4109536409378052, "learning_rate": 7.846058817512753e-05, "loss": 0.0305, "step": 11540 }, { "epoch": 10.266666666666667, "grad_norm": 0.32088756561279297, "learning_rate": 7.842029335957056e-05, "loss": 0.0251, "step": 11550 }, { "epoch": 10.275555555555556, "grad_norm": 0.43225717544555664, "learning_rate": 7.837997125756293e-05, "loss": 0.0256, "step": 11560 }, { "epoch": 10.284444444444444, "grad_norm": 0.38603630661964417, "learning_rate": 7.83396219078181e-05, "loss": 0.0226, "step": 11570 }, { "epoch": 10.293333333333333, "grad_norm": 0.29093268513679504, "learning_rate": 7.82992453490756e-05, "loss": 0.0243, "step": 11580 }, { "epoch": 10.302222222222222, "grad_norm": 0.3451744019985199, "learning_rate": 7.82588416201012e-05, "loss": 0.0217, "step": 11590 }, { "epoch": 10.311111111111112, "grad_norm": 0.2839917540550232, "learning_rate": 7.821841075968666e-05, "loss": 0.0235, "step": 11600 }, { "epoch": 10.32, "grad_norm": 0.33079519867897034, "learning_rate": 7.817795280664986e-05, "loss": 0.0279, "step": 11610 }, { "epoch": 10.328888888888889, "grad_norm": 0.3657057583332062, "learning_rate": 7.813746779983466e-05, "loss": 0.0228, "step": 11620 }, { "epoch": 10.337777777777777, "grad_norm": 0.3122910261154175, "learning_rate": 7.809695577811086e-05, "loss": 0.0245, "step": 11630 }, { "epoch": 10.346666666666668, "grad_norm": 0.20820283889770508, "learning_rate": 7.805641678037429e-05, "loss": 0.022, "step": 11640 }, { "epoch": 10.355555555555556, "grad_norm": 0.34736162424087524, "learning_rate": 7.801585084554658e-05, "loss": 0.0315, "step": 11650 }, { "epoch": 10.364444444444445, "grad_norm": 0.38755083084106445, "learning_rate": 7.797525801257528e-05, "loss": 0.0251, "step": 11660 }, { "epoch": 10.373333333333333, "grad_norm": 0.30341076850891113, "learning_rate": 7.793463832043374e-05, "loss": 0.0294, "step": 11670 }, { "epoch": 10.382222222222222, "grad_norm": 0.25731098651885986, "learning_rate": 7.789399180812114e-05, "loss": 0.0308, "step": 11680 }, { "epoch": 10.391111111111112, "grad_norm": 0.27498432993888855, "learning_rate": 7.785331851466233e-05, "loss": 0.0277, "step": 11690 }, { "epoch": 10.4, "grad_norm": 0.3247620463371277, "learning_rate": 7.781261847910794e-05, "loss": 0.0238, "step": 11700 }, { "epoch": 10.408888888888889, "grad_norm": 0.30719584226608276, "learning_rate": 7.777189174053429e-05, "loss": 0.0276, "step": 11710 }, { "epoch": 10.417777777777777, "grad_norm": 0.29466691613197327, "learning_rate": 7.773113833804326e-05, "loss": 0.0233, "step": 11720 }, { "epoch": 10.426666666666666, "grad_norm": 0.25773298740386963, "learning_rate": 7.769035831076239e-05, "loss": 0.0309, "step": 11730 }, { "epoch": 10.435555555555556, "grad_norm": 0.2722857594490051, "learning_rate": 7.764955169784476e-05, "loss": 0.0279, "step": 11740 }, { "epoch": 10.444444444444445, "grad_norm": 0.4068128764629364, "learning_rate": 7.760871853846899e-05, "loss": 0.0247, "step": 11750 }, { "epoch": 10.453333333333333, "grad_norm": 0.4604669213294983, "learning_rate": 7.756785887183918e-05, "loss": 0.0333, "step": 11760 }, { "epoch": 10.462222222222222, "grad_norm": 0.29569754004478455, "learning_rate": 7.752697273718487e-05, "loss": 0.0207, "step": 11770 }, { "epoch": 10.471111111111112, "grad_norm": 0.31176942586898804, "learning_rate": 7.748606017376103e-05, "loss": 0.0264, "step": 11780 }, { "epoch": 10.48, "grad_norm": 0.2906259596347809, "learning_rate": 7.744512122084798e-05, "loss": 0.0282, "step": 11790 }, { "epoch": 10.488888888888889, "grad_norm": 0.3506722152233124, "learning_rate": 7.740415591775141e-05, "loss": 0.0318, "step": 11800 }, { "epoch": 10.497777777777777, "grad_norm": 0.36894819140434265, "learning_rate": 7.736316430380228e-05, "loss": 0.0258, "step": 11810 }, { "epoch": 10.506666666666666, "grad_norm": 0.24554432928562164, "learning_rate": 7.732214641835683e-05, "loss": 0.0226, "step": 11820 }, { "epoch": 10.515555555555556, "grad_norm": 0.23827889561653137, "learning_rate": 7.728110230079649e-05, "loss": 0.0226, "step": 11830 }, { "epoch": 10.524444444444445, "grad_norm": 0.35021743178367615, "learning_rate": 7.724003199052794e-05, "loss": 0.0286, "step": 11840 }, { "epoch": 10.533333333333333, "grad_norm": 0.2993866801261902, "learning_rate": 7.719893552698292e-05, "loss": 0.0263, "step": 11850 }, { "epoch": 10.542222222222222, "grad_norm": 0.29780858755111694, "learning_rate": 7.715781294961838e-05, "loss": 0.0263, "step": 11860 }, { "epoch": 10.551111111111112, "grad_norm": 0.28403034806251526, "learning_rate": 7.711666429791626e-05, "loss": 0.0272, "step": 11870 }, { "epoch": 10.56, "grad_norm": 0.29005053639411926, "learning_rate": 7.707548961138358e-05, "loss": 0.0253, "step": 11880 }, { "epoch": 10.568888888888889, "grad_norm": 0.2845589518547058, "learning_rate": 7.703428892955232e-05, "loss": 0.0254, "step": 11890 }, { "epoch": 10.577777777777778, "grad_norm": 0.3075175881385803, "learning_rate": 7.699306229197945e-05, "loss": 0.0225, "step": 11900 }, { "epoch": 10.586666666666666, "grad_norm": 0.3107968270778656, "learning_rate": 7.695180973824685e-05, "loss": 0.0264, "step": 11910 }, { "epoch": 10.595555555555556, "grad_norm": 0.2284160852432251, "learning_rate": 7.691053130796127e-05, "loss": 0.026, "step": 11920 }, { "epoch": 10.604444444444445, "grad_norm": 0.33122581243515015, "learning_rate": 7.686922704075436e-05, "loss": 0.0265, "step": 11930 }, { "epoch": 10.613333333333333, "grad_norm": 0.3668990433216095, "learning_rate": 7.682789697628246e-05, "loss": 0.028, "step": 11940 }, { "epoch": 10.622222222222222, "grad_norm": 0.31447911262512207, "learning_rate": 7.678654115422681e-05, "loss": 0.0224, "step": 11950 }, { "epoch": 10.63111111111111, "grad_norm": 0.35301095247268677, "learning_rate": 7.674515961429327e-05, "loss": 0.0212, "step": 11960 }, { "epoch": 10.64, "grad_norm": 0.3603094220161438, "learning_rate": 7.67037523962125e-05, "loss": 0.0222, "step": 11970 }, { "epoch": 10.648888888888889, "grad_norm": 0.28335997462272644, "learning_rate": 7.66623195397397e-05, "loss": 0.0236, "step": 11980 }, { "epoch": 10.657777777777778, "grad_norm": 0.3135858476161957, "learning_rate": 7.662086108465479e-05, "loss": 0.0282, "step": 11990 }, { "epoch": 10.666666666666666, "grad_norm": 0.4036824703216553, "learning_rate": 7.657937707076217e-05, "loss": 0.0228, "step": 12000 }, { "epoch": 10.675555555555556, "grad_norm": 0.34251537919044495, "learning_rate": 7.653786753789084e-05, "loss": 0.0213, "step": 12010 }, { "epoch": 10.684444444444445, "grad_norm": 0.28866642713546753, "learning_rate": 7.64963325258943e-05, "loss": 0.0219, "step": 12020 }, { "epoch": 10.693333333333333, "grad_norm": 0.3918711841106415, "learning_rate": 7.64547720746505e-05, "loss": 0.0253, "step": 12030 }, { "epoch": 10.702222222222222, "grad_norm": 0.30129826068878174, "learning_rate": 7.641318622406183e-05, "loss": 0.0231, "step": 12040 }, { "epoch": 10.71111111111111, "grad_norm": 0.366491436958313, "learning_rate": 7.637157501405503e-05, "loss": 0.0256, "step": 12050 }, { "epoch": 10.72, "grad_norm": 0.263148695230484, "learning_rate": 7.632993848458119e-05, "loss": 0.022, "step": 12060 }, { "epoch": 10.72888888888889, "grad_norm": 0.32340118288993835, "learning_rate": 7.628827667561577e-05, "loss": 0.0218, "step": 12070 }, { "epoch": 10.737777777777778, "grad_norm": 0.27806341648101807, "learning_rate": 7.624658962715843e-05, "loss": 0.0281, "step": 12080 }, { "epoch": 10.746666666666666, "grad_norm": 0.5336754322052002, "learning_rate": 7.62048773792331e-05, "loss": 0.0258, "step": 12090 }, { "epoch": 10.755555555555556, "grad_norm": 0.38594383001327515, "learning_rate": 7.616313997188789e-05, "loss": 0.0295, "step": 12100 }, { "epoch": 10.764444444444445, "grad_norm": 0.1937389075756073, "learning_rate": 7.612137744519509e-05, "loss": 0.0234, "step": 12110 }, { "epoch": 10.773333333333333, "grad_norm": 0.32802289724349976, "learning_rate": 7.607958983925106e-05, "loss": 0.0257, "step": 12120 }, { "epoch": 10.782222222222222, "grad_norm": 0.432328999042511, "learning_rate": 7.60377771941763e-05, "loss": 0.026, "step": 12130 }, { "epoch": 10.79111111111111, "grad_norm": 0.30584481358528137, "learning_rate": 7.599593955011529e-05, "loss": 0.0246, "step": 12140 }, { "epoch": 10.8, "grad_norm": 0.326967716217041, "learning_rate": 7.595407694723656e-05, "loss": 0.0252, "step": 12150 }, { "epoch": 10.80888888888889, "grad_norm": 0.2345668375492096, "learning_rate": 7.591218942573256e-05, "loss": 0.0214, "step": 12160 }, { "epoch": 10.817777777777778, "grad_norm": 0.25147390365600586, "learning_rate": 7.58702770258197e-05, "loss": 0.0288, "step": 12170 }, { "epoch": 10.826666666666666, "grad_norm": 0.3471015691757202, "learning_rate": 7.582833978773826e-05, "loss": 0.0236, "step": 12180 }, { "epoch": 10.835555555555555, "grad_norm": 0.28587156534194946, "learning_rate": 7.578637775175239e-05, "loss": 0.0281, "step": 12190 }, { "epoch": 10.844444444444445, "grad_norm": 0.4024760127067566, "learning_rate": 7.574439095814998e-05, "loss": 0.0298, "step": 12200 }, { "epoch": 10.853333333333333, "grad_norm": 0.36477139592170715, "learning_rate": 7.570237944724277e-05, "loss": 0.0284, "step": 12210 }, { "epoch": 10.862222222222222, "grad_norm": 0.29017889499664307, "learning_rate": 7.566034325936621e-05, "loss": 0.0213, "step": 12220 }, { "epoch": 10.87111111111111, "grad_norm": 0.4490809440612793, "learning_rate": 7.56182824348794e-05, "loss": 0.0294, "step": 12230 }, { "epoch": 10.88, "grad_norm": 0.37127920985221863, "learning_rate": 7.557619701416516e-05, "loss": 0.0233, "step": 12240 }, { "epoch": 10.88888888888889, "grad_norm": 0.29528486728668213, "learning_rate": 7.553408703762987e-05, "loss": 0.0224, "step": 12250 }, { "epoch": 10.897777777777778, "grad_norm": 0.2523745000362396, "learning_rate": 7.549195254570351e-05, "loss": 0.0199, "step": 12260 }, { "epoch": 10.906666666666666, "grad_norm": 0.25951671600341797, "learning_rate": 7.54497935788396e-05, "loss": 0.0217, "step": 12270 }, { "epoch": 10.915555555555555, "grad_norm": 0.20171770453453064, "learning_rate": 7.540761017751511e-05, "loss": 0.0188, "step": 12280 }, { "epoch": 10.924444444444445, "grad_norm": 0.3630729615688324, "learning_rate": 7.536540238223061e-05, "loss": 0.0242, "step": 12290 }, { "epoch": 10.933333333333334, "grad_norm": 0.2654579281806946, "learning_rate": 7.53231702335099e-05, "loss": 0.0263, "step": 12300 }, { "epoch": 10.942222222222222, "grad_norm": 0.36841532588005066, "learning_rate": 7.528091377190028e-05, "loss": 0.0254, "step": 12310 }, { "epoch": 10.95111111111111, "grad_norm": 0.328571617603302, "learning_rate": 7.523863303797238e-05, "loss": 0.0218, "step": 12320 }, { "epoch": 10.96, "grad_norm": 0.3228000998497009, "learning_rate": 7.519632807232011e-05, "loss": 0.0292, "step": 12330 }, { "epoch": 10.96888888888889, "grad_norm": 0.2588554620742798, "learning_rate": 7.515399891556068e-05, "loss": 0.0226, "step": 12340 }, { "epoch": 10.977777777777778, "grad_norm": 0.24971751868724823, "learning_rate": 7.511164560833449e-05, "loss": 0.0232, "step": 12350 }, { "epoch": 10.986666666666666, "grad_norm": 0.29532814025878906, "learning_rate": 7.506926819130511e-05, "loss": 0.0229, "step": 12360 }, { "epoch": 10.995555555555555, "grad_norm": 0.2568766474723816, "learning_rate": 7.502686670515931e-05, "loss": 0.0236, "step": 12370 }, { "epoch": 11.004444444444445, "grad_norm": 0.3809404969215393, "learning_rate": 7.498444119060697e-05, "loss": 0.0237, "step": 12380 }, { "epoch": 11.013333333333334, "grad_norm": 0.3071889281272888, "learning_rate": 7.494199168838099e-05, "loss": 0.0253, "step": 12390 }, { "epoch": 11.022222222222222, "grad_norm": 0.30260175466537476, "learning_rate": 7.489951823923732e-05, "loss": 0.0211, "step": 12400 }, { "epoch": 11.03111111111111, "grad_norm": 0.2983306050300598, "learning_rate": 7.485702088395493e-05, "loss": 0.0235, "step": 12410 }, { "epoch": 11.04, "grad_norm": 0.26867160201072693, "learning_rate": 7.48144996633357e-05, "loss": 0.0235, "step": 12420 }, { "epoch": 11.04888888888889, "grad_norm": 0.29643845558166504, "learning_rate": 7.477195461820445e-05, "loss": 0.0236, "step": 12430 }, { "epoch": 11.057777777777778, "grad_norm": 0.23769840598106384, "learning_rate": 7.472938578940889e-05, "loss": 0.0243, "step": 12440 }, { "epoch": 11.066666666666666, "grad_norm": 0.3161536753177643, "learning_rate": 7.468679321781952e-05, "loss": 0.0237, "step": 12450 }, { "epoch": 11.075555555555555, "grad_norm": 0.2540513575077057, "learning_rate": 7.464417694432966e-05, "loss": 0.0238, "step": 12460 }, { "epoch": 11.084444444444445, "grad_norm": 0.3569740951061249, "learning_rate": 7.460153700985538e-05, "loss": 0.0315, "step": 12470 }, { "epoch": 11.093333333333334, "grad_norm": 0.2520148456096649, "learning_rate": 7.45588734553355e-05, "loss": 0.0242, "step": 12480 }, { "epoch": 11.102222222222222, "grad_norm": 0.3053121566772461, "learning_rate": 7.451618632173144e-05, "loss": 0.0243, "step": 12490 }, { "epoch": 11.11111111111111, "grad_norm": 0.2544589936733246, "learning_rate": 7.447347565002737e-05, "loss": 0.0211, "step": 12500 }, { "epoch": 11.12, "grad_norm": 0.2467353194952011, "learning_rate": 7.443074148122994e-05, "loss": 0.0198, "step": 12510 }, { "epoch": 11.12888888888889, "grad_norm": 0.2817169725894928, "learning_rate": 7.438798385636845e-05, "loss": 0.0252, "step": 12520 }, { "epoch": 11.137777777777778, "grad_norm": 0.363006591796875, "learning_rate": 7.434520281649466e-05, "loss": 0.0259, "step": 12530 }, { "epoch": 11.146666666666667, "grad_norm": 0.3284893035888672, "learning_rate": 7.430239840268287e-05, "loss": 0.0229, "step": 12540 }, { "epoch": 11.155555555555555, "grad_norm": 0.23061001300811768, "learning_rate": 7.425957065602976e-05, "loss": 0.0204, "step": 12550 }, { "epoch": 11.164444444444445, "grad_norm": 0.24731741845607758, "learning_rate": 7.421671961765447e-05, "loss": 0.0305, "step": 12560 }, { "epoch": 11.173333333333334, "grad_norm": 0.2779179513454437, "learning_rate": 7.417384532869846e-05, "loss": 0.0232, "step": 12570 }, { "epoch": 11.182222222222222, "grad_norm": 0.29773208498954773, "learning_rate": 7.413094783032551e-05, "loss": 0.0249, "step": 12580 }, { "epoch": 11.19111111111111, "grad_norm": 0.275779664516449, "learning_rate": 7.408802716372171e-05, "loss": 0.0243, "step": 12590 }, { "epoch": 11.2, "grad_norm": 0.3022674322128296, "learning_rate": 7.404508337009539e-05, "loss": 0.0243, "step": 12600 }, { "epoch": 11.20888888888889, "grad_norm": 0.2658543288707733, "learning_rate": 7.400211649067709e-05, "loss": 0.0259, "step": 12610 }, { "epoch": 11.217777777777778, "grad_norm": 0.28518038988113403, "learning_rate": 7.395912656671947e-05, "loss": 0.0201, "step": 12620 }, { "epoch": 11.226666666666667, "grad_norm": 0.3062174916267395, "learning_rate": 7.391611363949738e-05, "loss": 0.0204, "step": 12630 }, { "epoch": 11.235555555555555, "grad_norm": 0.25275179743766785, "learning_rate": 7.38730777503077e-05, "loss": 0.0265, "step": 12640 }, { "epoch": 11.244444444444444, "grad_norm": 0.3122808337211609, "learning_rate": 7.383001894046941e-05, "loss": 0.0213, "step": 12650 }, { "epoch": 11.253333333333334, "grad_norm": 0.38366371393203735, "learning_rate": 7.378693725132343e-05, "loss": 0.022, "step": 12660 }, { "epoch": 11.262222222222222, "grad_norm": 0.31082090735435486, "learning_rate": 7.374383272423268e-05, "loss": 0.0222, "step": 12670 }, { "epoch": 11.27111111111111, "grad_norm": 0.339639276266098, "learning_rate": 7.370070540058206e-05, "loss": 0.0213, "step": 12680 }, { "epoch": 11.28, "grad_norm": 0.36463218927383423, "learning_rate": 7.365755532177824e-05, "loss": 0.0219, "step": 12690 }, { "epoch": 11.28888888888889, "grad_norm": 0.3119225800037384, "learning_rate": 7.361438252924987e-05, "loss": 0.0224, "step": 12700 }, { "epoch": 11.297777777777778, "grad_norm": 0.2817429304122925, "learning_rate": 7.357118706444728e-05, "loss": 0.026, "step": 12710 }, { "epoch": 11.306666666666667, "grad_norm": 0.2584955096244812, "learning_rate": 7.352796896884269e-05, "loss": 0.0217, "step": 12720 }, { "epoch": 11.315555555555555, "grad_norm": 0.3046570420265198, "learning_rate": 7.348472828392992e-05, "loss": 0.0233, "step": 12730 }, { "epoch": 11.324444444444444, "grad_norm": 0.40423673391342163, "learning_rate": 7.344146505122458e-05, "loss": 0.0238, "step": 12740 }, { "epoch": 11.333333333333334, "grad_norm": 0.2421920895576477, "learning_rate": 7.339817931226392e-05, "loss": 0.0231, "step": 12750 }, { "epoch": 11.342222222222222, "grad_norm": 0.3225543200969696, "learning_rate": 7.335487110860673e-05, "loss": 0.0245, "step": 12760 }, { "epoch": 11.351111111111111, "grad_norm": 0.3430350124835968, "learning_rate": 7.33115404818334e-05, "loss": 0.0224, "step": 12770 }, { "epoch": 11.36, "grad_norm": 0.31562530994415283, "learning_rate": 7.32681874735459e-05, "loss": 0.0242, "step": 12780 }, { "epoch": 11.36888888888889, "grad_norm": 0.32490724325180054, "learning_rate": 7.322481212536762e-05, "loss": 0.0234, "step": 12790 }, { "epoch": 11.377777777777778, "grad_norm": 0.3731115460395813, "learning_rate": 7.318141447894344e-05, "loss": 0.028, "step": 12800 }, { "epoch": 11.386666666666667, "grad_norm": 0.2975084185600281, "learning_rate": 7.313799457593963e-05, "loss": 0.0209, "step": 12810 }, { "epoch": 11.395555555555555, "grad_norm": 0.25334909558296204, "learning_rate": 7.309455245804383e-05, "loss": 0.0236, "step": 12820 }, { "epoch": 11.404444444444444, "grad_norm": 0.2669563591480255, "learning_rate": 7.3051088166965e-05, "loss": 0.0275, "step": 12830 }, { "epoch": 11.413333333333334, "grad_norm": 0.28507596254348755, "learning_rate": 7.300760174443342e-05, "loss": 0.0277, "step": 12840 }, { "epoch": 11.422222222222222, "grad_norm": 0.2949814796447754, "learning_rate": 7.29640932322006e-05, "loss": 0.027, "step": 12850 }, { "epoch": 11.431111111111111, "grad_norm": 0.25755777955055237, "learning_rate": 7.292056267203924e-05, "loss": 0.0255, "step": 12860 }, { "epoch": 11.44, "grad_norm": 0.33741092681884766, "learning_rate": 7.287701010574321e-05, "loss": 0.0194, "step": 12870 }, { "epoch": 11.448888888888888, "grad_norm": 0.3325103521347046, "learning_rate": 7.283343557512759e-05, "loss": 0.0211, "step": 12880 }, { "epoch": 11.457777777777778, "grad_norm": 0.31777432560920715, "learning_rate": 7.27898391220284e-05, "loss": 0.0206, "step": 12890 }, { "epoch": 11.466666666666667, "grad_norm": 0.24217531085014343, "learning_rate": 7.274622078830284e-05, "loss": 0.0215, "step": 12900 }, { "epoch": 11.475555555555555, "grad_norm": 0.2873959243297577, "learning_rate": 7.270258061582907e-05, "loss": 0.0214, "step": 12910 }, { "epoch": 11.484444444444444, "grad_norm": 0.25184574723243713, "learning_rate": 7.265891864650618e-05, "loss": 0.0262, "step": 12920 }, { "epoch": 11.493333333333334, "grad_norm": 0.3795645833015442, "learning_rate": 7.261523492225425e-05, "loss": 0.0224, "step": 12930 }, { "epoch": 11.502222222222223, "grad_norm": 0.27197402715682983, "learning_rate": 7.25715294850142e-05, "loss": 0.0182, "step": 12940 }, { "epoch": 11.511111111111111, "grad_norm": 0.29623234272003174, "learning_rate": 7.252780237674783e-05, "loss": 0.0224, "step": 12950 }, { "epoch": 11.52, "grad_norm": 0.3486567735671997, "learning_rate": 7.248405363943773e-05, "loss": 0.0274, "step": 12960 }, { "epoch": 11.528888888888888, "grad_norm": 0.3263501524925232, "learning_rate": 7.244028331508727e-05, "loss": 0.0244, "step": 12970 }, { "epoch": 11.537777777777778, "grad_norm": 0.2543012797832489, "learning_rate": 7.23964914457205e-05, "loss": 0.026, "step": 12980 }, { "epoch": 11.546666666666667, "grad_norm": 0.24826563894748688, "learning_rate": 7.235267807338222e-05, "loss": 0.0195, "step": 12990 }, { "epoch": 11.555555555555555, "grad_norm": 0.19988322257995605, "learning_rate": 7.230884324013783e-05, "loss": 0.0215, "step": 13000 }, { "epoch": 11.564444444444444, "grad_norm": 0.22498029470443726, "learning_rate": 7.226498698807335e-05, "loss": 0.0229, "step": 13010 }, { "epoch": 11.573333333333334, "grad_norm": 0.34753143787384033, "learning_rate": 7.222110935929538e-05, "loss": 0.0224, "step": 13020 }, { "epoch": 11.582222222222223, "grad_norm": 0.2640341818332672, "learning_rate": 7.2177210395931e-05, "loss": 0.0271, "step": 13030 }, { "epoch": 11.591111111111111, "grad_norm": 0.3026505708694458, "learning_rate": 7.21332901401278e-05, "loss": 0.0294, "step": 13040 }, { "epoch": 11.6, "grad_norm": 0.2837521731853485, "learning_rate": 7.208934863405386e-05, "loss": 0.0269, "step": 13050 }, { "epoch": 11.608888888888888, "grad_norm": 0.31026339530944824, "learning_rate": 7.204538591989757e-05, "loss": 0.0245, "step": 13060 }, { "epoch": 11.617777777777778, "grad_norm": 0.3447277843952179, "learning_rate": 7.200140203986776e-05, "loss": 0.0237, "step": 13070 }, { "epoch": 11.626666666666667, "grad_norm": 0.3338221609592438, "learning_rate": 7.19573970361935e-05, "loss": 0.0294, "step": 13080 }, { "epoch": 11.635555555555555, "grad_norm": 0.2874343693256378, "learning_rate": 7.191337095112426e-05, "loss": 0.0219, "step": 13090 }, { "epoch": 11.644444444444444, "grad_norm": 0.31238067150115967, "learning_rate": 7.186932382692961e-05, "loss": 0.0225, "step": 13100 }, { "epoch": 11.653333333333332, "grad_norm": 0.23653018474578857, "learning_rate": 7.182525570589946e-05, "loss": 0.0257, "step": 13110 }, { "epoch": 11.662222222222223, "grad_norm": 0.329276978969574, "learning_rate": 7.178116663034376e-05, "loss": 0.0218, "step": 13120 }, { "epoch": 11.671111111111111, "grad_norm": 0.30477648973464966, "learning_rate": 7.173705664259264e-05, "loss": 0.0272, "step": 13130 }, { "epoch": 11.68, "grad_norm": 0.36361584067344666, "learning_rate": 7.169292578499632e-05, "loss": 0.0239, "step": 13140 }, { "epoch": 11.688888888888888, "grad_norm": 0.24462126195430756, "learning_rate": 7.164877409992499e-05, "loss": 0.0227, "step": 13150 }, { "epoch": 11.697777777777778, "grad_norm": 0.2306142896413803, "learning_rate": 7.160460162976891e-05, "loss": 0.0238, "step": 13160 }, { "epoch": 11.706666666666667, "grad_norm": 0.27776047587394714, "learning_rate": 7.156040841693828e-05, "loss": 0.0337, "step": 13170 }, { "epoch": 11.715555555555556, "grad_norm": 0.20803675055503845, "learning_rate": 7.151619450386315e-05, "loss": 0.0198, "step": 13180 }, { "epoch": 11.724444444444444, "grad_norm": 0.27558839321136475, "learning_rate": 7.147195993299355e-05, "loss": 0.0239, "step": 13190 }, { "epoch": 11.733333333333333, "grad_norm": 0.28122463822364807, "learning_rate": 7.142770474679925e-05, "loss": 0.0225, "step": 13200 }, { "epoch": 11.742222222222223, "grad_norm": 0.31278195977211, "learning_rate": 7.138342898776989e-05, "loss": 0.0306, "step": 13210 }, { "epoch": 11.751111111111111, "grad_norm": 0.257191926240921, "learning_rate": 7.133913269841479e-05, "loss": 0.031, "step": 13220 }, { "epoch": 11.76, "grad_norm": 0.30717402696609497, "learning_rate": 7.129481592126302e-05, "loss": 0.027, "step": 13230 }, { "epoch": 11.768888888888888, "grad_norm": 0.29339495301246643, "learning_rate": 7.125047869886333e-05, "loss": 0.0241, "step": 13240 }, { "epoch": 11.777777777777779, "grad_norm": 0.374300092458725, "learning_rate": 7.120612107378408e-05, "loss": 0.0252, "step": 13250 }, { "epoch": 11.786666666666667, "grad_norm": 0.289994478225708, "learning_rate": 7.116174308861322e-05, "loss": 0.0228, "step": 13260 }, { "epoch": 11.795555555555556, "grad_norm": 0.28407564759254456, "learning_rate": 7.111734478595826e-05, "loss": 0.0298, "step": 13270 }, { "epoch": 11.804444444444444, "grad_norm": 0.2085961252450943, "learning_rate": 7.10729262084462e-05, "loss": 0.0284, "step": 13280 }, { "epoch": 11.813333333333333, "grad_norm": 0.30718860030174255, "learning_rate": 7.102848739872352e-05, "loss": 0.0266, "step": 13290 }, { "epoch": 11.822222222222223, "grad_norm": 0.21840649843215942, "learning_rate": 7.098402839945612e-05, "loss": 0.0208, "step": 13300 }, { "epoch": 11.831111111111111, "grad_norm": 0.21813766658306122, "learning_rate": 7.093954925332928e-05, "loss": 0.0231, "step": 13310 }, { "epoch": 11.84, "grad_norm": 0.32680776715278625, "learning_rate": 7.089505000304763e-05, "loss": 0.0279, "step": 13320 }, { "epoch": 11.848888888888888, "grad_norm": 0.3649505078792572, "learning_rate": 7.085053069133509e-05, "loss": 0.0201, "step": 13330 }, { "epoch": 11.857777777777777, "grad_norm": 0.27578648924827576, "learning_rate": 7.080599136093487e-05, "loss": 0.0259, "step": 13340 }, { "epoch": 11.866666666666667, "grad_norm": 0.31153982877731323, "learning_rate": 7.076143205460934e-05, "loss": 0.0282, "step": 13350 }, { "epoch": 11.875555555555556, "grad_norm": 0.21310918033123016, "learning_rate": 7.071685281514015e-05, "loss": 0.0208, "step": 13360 }, { "epoch": 11.884444444444444, "grad_norm": 0.23538850247859955, "learning_rate": 7.0672253685328e-05, "loss": 0.0216, "step": 13370 }, { "epoch": 11.893333333333333, "grad_norm": 0.2487376630306244, "learning_rate": 7.062763470799268e-05, "loss": 0.022, "step": 13380 }, { "epoch": 11.902222222222223, "grad_norm": 0.20247872173786163, "learning_rate": 7.058299592597312e-05, "loss": 0.0269, "step": 13390 }, { "epoch": 11.911111111111111, "grad_norm": 0.30929505825042725, "learning_rate": 7.053833738212717e-05, "loss": 0.0223, "step": 13400 }, { "epoch": 11.92, "grad_norm": 0.24598610401153564, "learning_rate": 7.049365911933174e-05, "loss": 0.0203, "step": 13410 }, { "epoch": 11.928888888888888, "grad_norm": 0.30081743001937866, "learning_rate": 7.044896118048259e-05, "loss": 0.0225, "step": 13420 }, { "epoch": 11.937777777777779, "grad_norm": 0.2534627318382263, "learning_rate": 7.040424360849445e-05, "loss": 0.0254, "step": 13430 }, { "epoch": 11.946666666666667, "grad_norm": 0.39771005511283875, "learning_rate": 7.035950644630081e-05, "loss": 0.0236, "step": 13440 }, { "epoch": 11.955555555555556, "grad_norm": 0.24600432813167572, "learning_rate": 7.031474973685408e-05, "loss": 0.0239, "step": 13450 }, { "epoch": 11.964444444444444, "grad_norm": 0.2398240566253662, "learning_rate": 7.026997352312534e-05, "loss": 0.0227, "step": 13460 }, { "epoch": 11.973333333333333, "grad_norm": 0.22781044244766235, "learning_rate": 7.022517784810443e-05, "loss": 0.0224, "step": 13470 }, { "epoch": 11.982222222222223, "grad_norm": 0.38611629605293274, "learning_rate": 7.01803627547999e-05, "loss": 0.0268, "step": 13480 }, { "epoch": 11.991111111111111, "grad_norm": 0.3174099624156952, "learning_rate": 7.013552828623891e-05, "loss": 0.0241, "step": 13490 }, { "epoch": 12.0, "grad_norm": 0.4602668583393097, "learning_rate": 7.009067448546722e-05, "loss": 0.0201, "step": 13500 }, { "epoch": 12.008888888888889, "grad_norm": 0.3677900433540344, "learning_rate": 7.004580139554918e-05, "loss": 0.0209, "step": 13510 }, { "epoch": 12.017777777777777, "grad_norm": 0.3168596923351288, "learning_rate": 7.000090905956765e-05, "loss": 0.0209, "step": 13520 }, { "epoch": 12.026666666666667, "grad_norm": 0.259561687707901, "learning_rate": 6.995599752062394e-05, "loss": 0.0282, "step": 13530 }, { "epoch": 12.035555555555556, "grad_norm": 0.23353905975818634, "learning_rate": 6.991106682183783e-05, "loss": 0.0207, "step": 13540 }, { "epoch": 12.044444444444444, "grad_norm": 0.3146946430206299, "learning_rate": 6.98661170063475e-05, "loss": 0.0213, "step": 13550 }, { "epoch": 12.053333333333333, "grad_norm": 0.33124473690986633, "learning_rate": 6.982114811730944e-05, "loss": 0.0218, "step": 13560 }, { "epoch": 12.062222222222223, "grad_norm": 0.26577937602996826, "learning_rate": 6.977616019789852e-05, "loss": 0.0203, "step": 13570 }, { "epoch": 12.071111111111112, "grad_norm": 0.35745370388031006, "learning_rate": 6.973115329130782e-05, "loss": 0.0249, "step": 13580 }, { "epoch": 12.08, "grad_norm": 0.27764683961868286, "learning_rate": 6.968612744074867e-05, "loss": 0.029, "step": 13590 }, { "epoch": 12.088888888888889, "grad_norm": 0.29689449071884155, "learning_rate": 6.964108268945056e-05, "loss": 0.027, "step": 13600 }, { "epoch": 12.097777777777777, "grad_norm": 0.2417519986629486, "learning_rate": 6.959601908066122e-05, "loss": 0.0246, "step": 13610 }, { "epoch": 12.106666666666667, "grad_norm": 0.2224150002002716, "learning_rate": 6.955093665764638e-05, "loss": 0.0188, "step": 13620 }, { "epoch": 12.115555555555556, "grad_norm": 0.22972023487091064, "learning_rate": 6.950583546368991e-05, "loss": 0.0196, "step": 13630 }, { "epoch": 12.124444444444444, "grad_norm": 0.28134629130363464, "learning_rate": 6.946071554209361e-05, "loss": 0.0253, "step": 13640 }, { "epoch": 12.133333333333333, "grad_norm": 0.27405083179473877, "learning_rate": 6.941557693617738e-05, "loss": 0.0216, "step": 13650 }, { "epoch": 12.142222222222221, "grad_norm": 0.49325740337371826, "learning_rate": 6.937041968927896e-05, "loss": 0.0253, "step": 13660 }, { "epoch": 12.151111111111112, "grad_norm": 0.3883151113986969, "learning_rate": 6.932524384475403e-05, "loss": 0.0222, "step": 13670 }, { "epoch": 12.16, "grad_norm": 0.2691245973110199, "learning_rate": 6.928004944597615e-05, "loss": 0.0228, "step": 13680 }, { "epoch": 12.168888888888889, "grad_norm": 0.2971879541873932, "learning_rate": 6.923483653633662e-05, "loss": 0.0208, "step": 13690 }, { "epoch": 12.177777777777777, "grad_norm": 0.37468215823173523, "learning_rate": 6.918960515924458e-05, "loss": 0.0236, "step": 13700 }, { "epoch": 12.186666666666667, "grad_norm": 0.40784910321235657, "learning_rate": 6.914435535812685e-05, "loss": 0.0223, "step": 13710 }, { "epoch": 12.195555555555556, "grad_norm": 0.2819586396217346, "learning_rate": 6.9099087176428e-05, "loss": 0.0246, "step": 13720 }, { "epoch": 12.204444444444444, "grad_norm": 0.272192120552063, "learning_rate": 6.905380065761023e-05, "loss": 0.0184, "step": 13730 }, { "epoch": 12.213333333333333, "grad_norm": 0.28072166442871094, "learning_rate": 6.90084958451533e-05, "loss": 0.0211, "step": 13740 }, { "epoch": 12.222222222222221, "grad_norm": 0.2098255604505539, "learning_rate": 6.896317278255455e-05, "loss": 0.0207, "step": 13750 }, { "epoch": 12.231111111111112, "grad_norm": 0.3024193346500397, "learning_rate": 6.891783151332887e-05, "loss": 0.0226, "step": 13760 }, { "epoch": 12.24, "grad_norm": 0.37237659096717834, "learning_rate": 6.887247208100862e-05, "loss": 0.0226, "step": 13770 }, { "epoch": 12.248888888888889, "grad_norm": 0.3691575527191162, "learning_rate": 6.882709452914357e-05, "loss": 0.0231, "step": 13780 }, { "epoch": 12.257777777777777, "grad_norm": 0.38528627157211304, "learning_rate": 6.878169890130097e-05, "loss": 0.0234, "step": 13790 }, { "epoch": 12.266666666666667, "grad_norm": 0.3225100636482239, "learning_rate": 6.87362852410653e-05, "loss": 0.0245, "step": 13800 }, { "epoch": 12.275555555555556, "grad_norm": 0.2749731242656708, "learning_rate": 6.869085359203844e-05, "loss": 0.021, "step": 13810 }, { "epoch": 12.284444444444444, "grad_norm": 0.21672292053699493, "learning_rate": 6.864540399783953e-05, "loss": 0.0202, "step": 13820 }, { "epoch": 12.293333333333333, "grad_norm": 0.2488427758216858, "learning_rate": 6.859993650210491e-05, "loss": 0.0197, "step": 13830 }, { "epoch": 12.302222222222222, "grad_norm": 0.2708471715450287, "learning_rate": 6.855445114848816e-05, "loss": 0.0234, "step": 13840 }, { "epoch": 12.311111111111112, "grad_norm": 0.31763216853141785, "learning_rate": 6.850894798065992e-05, "loss": 0.0215, "step": 13850 }, { "epoch": 12.32, "grad_norm": 0.31963443756103516, "learning_rate": 6.846342704230803e-05, "loss": 0.0214, "step": 13860 }, { "epoch": 12.328888888888889, "grad_norm": 0.2817164659500122, "learning_rate": 6.841788837713732e-05, "loss": 0.0213, "step": 13870 }, { "epoch": 12.337777777777777, "grad_norm": 0.31978821754455566, "learning_rate": 6.837233202886968e-05, "loss": 0.026, "step": 13880 }, { "epoch": 12.346666666666668, "grad_norm": 0.23339121043682098, "learning_rate": 6.832675804124395e-05, "loss": 0.0187, "step": 13890 }, { "epoch": 12.355555555555556, "grad_norm": 0.422465056180954, "learning_rate": 6.828116645801594e-05, "loss": 0.0199, "step": 13900 }, { "epoch": 12.364444444444445, "grad_norm": 0.36834532022476196, "learning_rate": 6.823555732295832e-05, "loss": 0.0195, "step": 13910 }, { "epoch": 12.373333333333333, "grad_norm": 0.3216288685798645, "learning_rate": 6.818993067986062e-05, "loss": 0.0233, "step": 13920 }, { "epoch": 12.382222222222222, "grad_norm": 0.343928724527359, "learning_rate": 6.814428657252918e-05, "loss": 0.0214, "step": 13930 }, { "epoch": 12.391111111111112, "grad_norm": 0.2744537889957428, "learning_rate": 6.809862504478714e-05, "loss": 0.0245, "step": 13940 }, { "epoch": 12.4, "grad_norm": 0.19744327664375305, "learning_rate": 6.805294614047431e-05, "loss": 0.0192, "step": 13950 }, { "epoch": 12.408888888888889, "grad_norm": 0.22732292115688324, "learning_rate": 6.80072499034472e-05, "loss": 0.02, "step": 13960 }, { "epoch": 12.417777777777777, "grad_norm": 0.22170546650886536, "learning_rate": 6.7961536377579e-05, "loss": 0.0208, "step": 13970 }, { "epoch": 12.426666666666666, "grad_norm": 0.32497164607048035, "learning_rate": 6.791580560675945e-05, "loss": 0.0222, "step": 13980 }, { "epoch": 12.435555555555556, "grad_norm": 0.3056831657886505, "learning_rate": 6.787005763489489e-05, "loss": 0.0271, "step": 13990 }, { "epoch": 12.444444444444445, "grad_norm": 0.20780594646930695, "learning_rate": 6.782429250590813e-05, "loss": 0.0188, "step": 14000 }, { "epoch": 12.453333333333333, "grad_norm": 0.28138676285743713, "learning_rate": 6.777851026373849e-05, "loss": 0.0272, "step": 14010 }, { "epoch": 12.462222222222222, "grad_norm": 0.32628491520881653, "learning_rate": 6.77327109523417e-05, "loss": 0.0282, "step": 14020 }, { "epoch": 12.471111111111112, "grad_norm": 0.2709018588066101, "learning_rate": 6.768689461568987e-05, "loss": 0.0206, "step": 14030 }, { "epoch": 12.48, "grad_norm": 0.22903290390968323, "learning_rate": 6.764106129777149e-05, "loss": 0.0193, "step": 14040 }, { "epoch": 12.488888888888889, "grad_norm": 0.27164530754089355, "learning_rate": 6.759521104259133e-05, "loss": 0.0245, "step": 14050 }, { "epoch": 12.497777777777777, "grad_norm": 0.31813061237335205, "learning_rate": 6.75493438941704e-05, "loss": 0.0247, "step": 14060 }, { "epoch": 12.506666666666666, "grad_norm": 0.25169700384140015, "learning_rate": 6.7503459896546e-05, "loss": 0.0285, "step": 14070 }, { "epoch": 12.515555555555556, "grad_norm": 0.23328697681427002, "learning_rate": 6.745755909377151e-05, "loss": 0.0223, "step": 14080 }, { "epoch": 12.524444444444445, "grad_norm": 0.25997769832611084, "learning_rate": 6.741164152991653e-05, "loss": 0.0212, "step": 14090 }, { "epoch": 12.533333333333333, "grad_norm": 0.2833796739578247, "learning_rate": 6.736570724906673e-05, "loss": 0.0246, "step": 14100 }, { "epoch": 12.542222222222222, "grad_norm": 0.24774819612503052, "learning_rate": 6.731975629532378e-05, "loss": 0.0217, "step": 14110 }, { "epoch": 12.551111111111112, "grad_norm": 0.32720017433166504, "learning_rate": 6.727378871280541e-05, "loss": 0.0283, "step": 14120 }, { "epoch": 12.56, "grad_norm": 0.25564059615135193, "learning_rate": 6.72278045456453e-05, "loss": 0.0198, "step": 14130 }, { "epoch": 12.568888888888889, "grad_norm": 0.3881785571575165, "learning_rate": 6.718180383799307e-05, "loss": 0.0202, "step": 14140 }, { "epoch": 12.577777777777778, "grad_norm": 0.2895052433013916, "learning_rate": 6.713578663401421e-05, "loss": 0.0181, "step": 14150 }, { "epoch": 12.586666666666666, "grad_norm": 0.22810260951519012, "learning_rate": 6.708975297789001e-05, "loss": 0.0249, "step": 14160 }, { "epoch": 12.595555555555556, "grad_norm": 0.3084498345851898, "learning_rate": 6.70437029138176e-05, "loss": 0.0216, "step": 14170 }, { "epoch": 12.604444444444445, "grad_norm": 0.28718528151512146, "learning_rate": 6.699763648600986e-05, "loss": 0.0238, "step": 14180 }, { "epoch": 12.613333333333333, "grad_norm": 0.28653743863105774, "learning_rate": 6.695155373869538e-05, "loss": 0.0264, "step": 14190 }, { "epoch": 12.622222222222222, "grad_norm": 0.26975277066230774, "learning_rate": 6.690545471611841e-05, "loss": 0.0215, "step": 14200 }, { "epoch": 12.63111111111111, "grad_norm": 0.36905917525291443, "learning_rate": 6.685933946253881e-05, "loss": 0.0225, "step": 14210 }, { "epoch": 12.64, "grad_norm": 0.2783248722553253, "learning_rate": 6.681320802223205e-05, "loss": 0.0203, "step": 14220 }, { "epoch": 12.648888888888889, "grad_norm": 0.24623630940914154, "learning_rate": 6.676706043948915e-05, "loss": 0.0208, "step": 14230 }, { "epoch": 12.657777777777778, "grad_norm": 0.3119873106479645, "learning_rate": 6.672089675861658e-05, "loss": 0.0218, "step": 14240 }, { "epoch": 12.666666666666666, "grad_norm": 0.26775822043418884, "learning_rate": 6.667471702393632e-05, "loss": 0.0259, "step": 14250 }, { "epoch": 12.675555555555556, "grad_norm": 0.3162439167499542, "learning_rate": 6.662852127978572e-05, "loss": 0.0239, "step": 14260 }, { "epoch": 12.684444444444445, "grad_norm": 0.34829095005989075, "learning_rate": 6.658230957051754e-05, "loss": 0.0205, "step": 14270 }, { "epoch": 12.693333333333333, "grad_norm": 0.27948522567749023, "learning_rate": 6.653608194049984e-05, "loss": 0.0209, "step": 14280 }, { "epoch": 12.702222222222222, "grad_norm": 0.29584112763404846, "learning_rate": 6.648983843411596e-05, "loss": 0.0191, "step": 14290 }, { "epoch": 12.71111111111111, "grad_norm": 0.28777679800987244, "learning_rate": 6.644357909576455e-05, "loss": 0.0212, "step": 14300 }, { "epoch": 12.72, "grad_norm": 0.24508832395076752, "learning_rate": 6.639730396985933e-05, "loss": 0.0313, "step": 14310 }, { "epoch": 12.72888888888889, "grad_norm": 0.25309571623802185, "learning_rate": 6.63510131008293e-05, "loss": 0.0198, "step": 14320 }, { "epoch": 12.737777777777778, "grad_norm": 0.3057999312877655, "learning_rate": 6.630470653311853e-05, "loss": 0.0197, "step": 14330 }, { "epoch": 12.746666666666666, "grad_norm": 0.3807729184627533, "learning_rate": 6.625838431118616e-05, "loss": 0.0217, "step": 14340 }, { "epoch": 12.755555555555556, "grad_norm": 0.23334069550037384, "learning_rate": 6.621204647950636e-05, "loss": 0.0237, "step": 14350 }, { "epoch": 12.764444444444445, "grad_norm": 0.2573685348033905, "learning_rate": 6.616569308256828e-05, "loss": 0.0222, "step": 14360 }, { "epoch": 12.773333333333333, "grad_norm": 0.21654920279979706, "learning_rate": 6.611932416487602e-05, "loss": 0.0179, "step": 14370 }, { "epoch": 12.782222222222222, "grad_norm": 0.33958300948143005, "learning_rate": 6.607293977094861e-05, "loss": 0.0197, "step": 14380 }, { "epoch": 12.79111111111111, "grad_norm": 0.21544739603996277, "learning_rate": 6.60265399453199e-05, "loss": 0.0195, "step": 14390 }, { "epoch": 12.8, "grad_norm": 0.28914889693260193, "learning_rate": 6.598012473253858e-05, "loss": 0.0233, "step": 14400 }, { "epoch": 12.80888888888889, "grad_norm": 0.2683359980583191, "learning_rate": 6.593369417716808e-05, "loss": 0.0265, "step": 14410 }, { "epoch": 12.817777777777778, "grad_norm": 0.24401822686195374, "learning_rate": 6.588724832378658e-05, "loss": 0.027, "step": 14420 }, { "epoch": 12.826666666666666, "grad_norm": 0.2572740912437439, "learning_rate": 6.584078721698701e-05, "loss": 0.0214, "step": 14430 }, { "epoch": 12.835555555555555, "grad_norm": 0.3118238151073456, "learning_rate": 6.579431090137681e-05, "loss": 0.021, "step": 14440 }, { "epoch": 12.844444444444445, "grad_norm": 0.21864013373851776, "learning_rate": 6.574781942157815e-05, "loss": 0.0237, "step": 14450 }, { "epoch": 12.853333333333333, "grad_norm": 0.29276907444000244, "learning_rate": 6.570131282222769e-05, "loss": 0.0183, "step": 14460 }, { "epoch": 12.862222222222222, "grad_norm": 0.26520612835884094, "learning_rate": 6.565479114797663e-05, "loss": 0.0184, "step": 14470 }, { "epoch": 12.87111111111111, "grad_norm": 0.2573179006576538, "learning_rate": 6.560825444349062e-05, "loss": 0.0286, "step": 14480 }, { "epoch": 12.88, "grad_norm": 0.2972376346588135, "learning_rate": 6.556170275344977e-05, "loss": 0.0232, "step": 14490 }, { "epoch": 12.88888888888889, "grad_norm": 0.2132313847541809, "learning_rate": 6.551513612254858e-05, "loss": 0.0237, "step": 14500 }, { "epoch": 12.897777777777778, "grad_norm": 0.26872149109840393, "learning_rate": 6.546855459549586e-05, "loss": 0.0234, "step": 14510 }, { "epoch": 12.906666666666666, "grad_norm": 0.24907062947750092, "learning_rate": 6.542195821701475e-05, "loss": 0.0212, "step": 14520 }, { "epoch": 12.915555555555555, "grad_norm": 0.3541968762874603, "learning_rate": 6.537534703184261e-05, "loss": 0.0235, "step": 14530 }, { "epoch": 12.924444444444445, "grad_norm": 0.3012160360813141, "learning_rate": 6.532872108473112e-05, "loss": 0.0212, "step": 14540 }, { "epoch": 12.933333333333334, "grad_norm": 0.3640080392360687, "learning_rate": 6.5282080420446e-05, "loss": 0.0224, "step": 14550 }, { "epoch": 12.942222222222222, "grad_norm": 0.2537399232387543, "learning_rate": 6.523542508376717e-05, "loss": 0.0235, "step": 14560 }, { "epoch": 12.95111111111111, "grad_norm": 0.2425050437450409, "learning_rate": 6.518875511948863e-05, "loss": 0.024, "step": 14570 }, { "epoch": 12.96, "grad_norm": 0.24450407922267914, "learning_rate": 6.514207057241844e-05, "loss": 0.0191, "step": 14580 }, { "epoch": 12.96888888888889, "grad_norm": 0.1733064502477646, "learning_rate": 6.509537148737862e-05, "loss": 0.0176, "step": 14590 }, { "epoch": 12.977777777777778, "grad_norm": 0.2264716923236847, "learning_rate": 6.504865790920517e-05, "loss": 0.0219, "step": 14600 }, { "epoch": 12.986666666666666, "grad_norm": 0.2664300799369812, "learning_rate": 6.500192988274804e-05, "loss": 0.0183, "step": 14610 }, { "epoch": 12.995555555555555, "grad_norm": 0.2583307921886444, "learning_rate": 6.495518745287095e-05, "loss": 0.0182, "step": 14620 }, { "epoch": 13.004444444444445, "grad_norm": 0.2847904562950134, "learning_rate": 6.490843066445159e-05, "loss": 0.023, "step": 14630 }, { "epoch": 13.013333333333334, "grad_norm": 0.6640941500663757, "learning_rate": 6.486165956238132e-05, "loss": 0.0205, "step": 14640 }, { "epoch": 13.022222222222222, "grad_norm": 0.22650055587291718, "learning_rate": 6.481487419156528e-05, "loss": 0.0213, "step": 14650 }, { "epoch": 13.03111111111111, "grad_norm": 0.3861131966114044, "learning_rate": 6.476807459692235e-05, "loss": 0.0208, "step": 14660 }, { "epoch": 13.04, "grad_norm": 0.23897649347782135, "learning_rate": 6.472126082338498e-05, "loss": 0.0178, "step": 14670 }, { "epoch": 13.04888888888889, "grad_norm": 0.3555122911930084, "learning_rate": 6.467443291589932e-05, "loss": 0.0251, "step": 14680 }, { "epoch": 13.057777777777778, "grad_norm": 0.36616942286491394, "learning_rate": 6.462759091942504e-05, "loss": 0.0196, "step": 14690 }, { "epoch": 13.066666666666666, "grad_norm": 0.3321700394153595, "learning_rate": 6.458073487893533e-05, "loss": 0.023, "step": 14700 }, { "epoch": 13.075555555555555, "grad_norm": 0.31792184710502625, "learning_rate": 6.453386483941693e-05, "loss": 0.0278, "step": 14710 }, { "epoch": 13.084444444444445, "grad_norm": 0.31450894474983215, "learning_rate": 6.448698084586994e-05, "loss": 0.0215, "step": 14720 }, { "epoch": 13.093333333333334, "grad_norm": 0.26492950320243835, "learning_rate": 6.444008294330793e-05, "loss": 0.0228, "step": 14730 }, { "epoch": 13.102222222222222, "grad_norm": 0.24761633574962616, "learning_rate": 6.439317117675771e-05, "loss": 0.0176, "step": 14740 }, { "epoch": 13.11111111111111, "grad_norm": 0.2563458979129791, "learning_rate": 6.434624559125955e-05, "loss": 0.0216, "step": 14750 }, { "epoch": 13.12, "grad_norm": 0.29232168197631836, "learning_rate": 6.429930623186689e-05, "loss": 0.0177, "step": 14760 }, { "epoch": 13.12888888888889, "grad_norm": 0.3143885135650635, "learning_rate": 6.425235314364641e-05, "loss": 0.0225, "step": 14770 }, { "epoch": 13.137777777777778, "grad_norm": 0.26505231857299805, "learning_rate": 6.420538637167799e-05, "loss": 0.0254, "step": 14780 }, { "epoch": 13.146666666666667, "grad_norm": 0.4117182195186615, "learning_rate": 6.415840596105464e-05, "loss": 0.0216, "step": 14790 }, { "epoch": 13.155555555555555, "grad_norm": 0.3176908791065216, "learning_rate": 6.411141195688246e-05, "loss": 0.0212, "step": 14800 }, { "epoch": 13.164444444444445, "grad_norm": 0.3194044232368469, "learning_rate": 6.406440440428059e-05, "loss": 0.0208, "step": 14810 }, { "epoch": 13.173333333333334, "grad_norm": 0.3636726140975952, "learning_rate": 6.40173833483812e-05, "loss": 0.0174, "step": 14820 }, { "epoch": 13.182222222222222, "grad_norm": 0.21844886243343353, "learning_rate": 6.397034883432942e-05, "loss": 0.0208, "step": 14830 }, { "epoch": 13.19111111111111, "grad_norm": 0.3168864846229553, "learning_rate": 6.392330090728331e-05, "loss": 0.0204, "step": 14840 }, { "epoch": 13.2, "grad_norm": 0.3062882125377655, "learning_rate": 6.387623961241374e-05, "loss": 0.0196, "step": 14850 }, { "epoch": 13.20888888888889, "grad_norm": 0.28821104764938354, "learning_rate": 6.382916499490454e-05, "loss": 0.0209, "step": 14860 }, { "epoch": 13.217777777777778, "grad_norm": 0.2665702700614929, "learning_rate": 6.378207709995223e-05, "loss": 0.0195, "step": 14870 }, { "epoch": 13.226666666666667, "grad_norm": 0.3430591821670532, "learning_rate": 6.373497597276612e-05, "loss": 0.0229, "step": 14880 }, { "epoch": 13.235555555555555, "grad_norm": 0.2094268500804901, "learning_rate": 6.368786165856817e-05, "loss": 0.0226, "step": 14890 }, { "epoch": 13.244444444444444, "grad_norm": 0.23809145390987396, "learning_rate": 6.364073420259307e-05, "loss": 0.0152, "step": 14900 }, { "epoch": 13.253333333333334, "grad_norm": 0.32015639543533325, "learning_rate": 6.359359365008814e-05, "loss": 0.0274, "step": 14910 }, { "epoch": 13.262222222222222, "grad_norm": 0.26292017102241516, "learning_rate": 6.354644004631318e-05, "loss": 0.0223, "step": 14920 }, { "epoch": 13.27111111111111, "grad_norm": 0.24396534264087677, "learning_rate": 6.34992734365406e-05, "loss": 0.0231, "step": 14930 }, { "epoch": 13.28, "grad_norm": 0.3129543364048004, "learning_rate": 6.345209386605526e-05, "loss": 0.0212, "step": 14940 }, { "epoch": 13.28888888888889, "grad_norm": 0.26238301396369934, "learning_rate": 6.340490138015451e-05, "loss": 0.0206, "step": 14950 }, { "epoch": 13.297777777777778, "grad_norm": 0.331840842962265, "learning_rate": 6.335769602414802e-05, "loss": 0.0221, "step": 14960 }, { "epoch": 13.306666666666667, "grad_norm": 0.3021104335784912, "learning_rate": 6.33104778433579e-05, "loss": 0.0191, "step": 14970 }, { "epoch": 13.315555555555555, "grad_norm": 0.27757468819618225, "learning_rate": 6.326324688311852e-05, "loss": 0.0247, "step": 14980 }, { "epoch": 13.324444444444444, "grad_norm": 0.26093706488609314, "learning_rate": 6.321600318877651e-05, "loss": 0.0202, "step": 14990 }, { "epoch": 13.333333333333334, "grad_norm": 0.23868370056152344, "learning_rate": 6.31687468056908e-05, "loss": 0.0222, "step": 15000 }, { "epoch": 13.342222222222222, "grad_norm": 0.2196987122297287, "learning_rate": 6.312147777923244e-05, "loss": 0.0185, "step": 15010 }, { "epoch": 13.351111111111111, "grad_norm": 0.2625134587287903, "learning_rate": 6.307419615478463e-05, "loss": 0.0218, "step": 15020 }, { "epoch": 13.36, "grad_norm": 0.2303367704153061, "learning_rate": 6.302690197774266e-05, "loss": 0.0253, "step": 15030 }, { "epoch": 13.36888888888889, "grad_norm": 0.2647522985935211, "learning_rate": 6.29795952935139e-05, "loss": 0.0179, "step": 15040 }, { "epoch": 13.377777777777778, "grad_norm": 0.288402259349823, "learning_rate": 6.293227614751768e-05, "loss": 0.0271, "step": 15050 }, { "epoch": 13.386666666666667, "grad_norm": 0.28528571128845215, "learning_rate": 6.288494458518534e-05, "loss": 0.0239, "step": 15060 }, { "epoch": 13.395555555555555, "grad_norm": 0.2393537163734436, "learning_rate": 6.283760065196014e-05, "loss": 0.0193, "step": 15070 }, { "epoch": 13.404444444444444, "grad_norm": 0.343585342168808, "learning_rate": 6.279024439329716e-05, "loss": 0.0223, "step": 15080 }, { "epoch": 13.413333333333334, "grad_norm": 0.3962152302265167, "learning_rate": 6.27428758546634e-05, "loss": 0.0236, "step": 15090 }, { "epoch": 13.422222222222222, "grad_norm": 0.17810606956481934, "learning_rate": 6.269549508153758e-05, "loss": 0.0221, "step": 15100 }, { "epoch": 13.431111111111111, "grad_norm": 0.30574271082878113, "learning_rate": 6.264810211941022e-05, "loss": 0.0218, "step": 15110 }, { "epoch": 13.44, "grad_norm": 0.2580048739910126, "learning_rate": 6.260069701378347e-05, "loss": 0.0212, "step": 15120 }, { "epoch": 13.448888888888888, "grad_norm": 0.23603177070617676, "learning_rate": 6.255327981017122e-05, "loss": 0.0184, "step": 15130 }, { "epoch": 13.457777777777778, "grad_norm": 0.21261227130889893, "learning_rate": 6.250585055409893e-05, "loss": 0.0223, "step": 15140 }, { "epoch": 13.466666666666667, "grad_norm": 0.26375216245651245, "learning_rate": 6.245840929110363e-05, "loss": 0.022, "step": 15150 }, { "epoch": 13.475555555555555, "grad_norm": 0.2654314339160919, "learning_rate": 6.24109560667339e-05, "loss": 0.0237, "step": 15160 }, { "epoch": 13.484444444444444, "grad_norm": 0.3586564362049103, "learning_rate": 6.236349092654978e-05, "loss": 0.025, "step": 15170 }, { "epoch": 13.493333333333334, "grad_norm": 0.3405466377735138, "learning_rate": 6.231601391612279e-05, "loss": 0.0202, "step": 15180 }, { "epoch": 13.502222222222223, "grad_norm": 0.27525395154953003, "learning_rate": 6.226852508103577e-05, "loss": 0.0247, "step": 15190 }, { "epoch": 13.511111111111111, "grad_norm": 0.3566519021987915, "learning_rate": 6.222102446688303e-05, "loss": 0.0249, "step": 15200 }, { "epoch": 13.52, "grad_norm": 0.3729831278324127, "learning_rate": 6.217351211927008e-05, "loss": 0.0214, "step": 15210 }, { "epoch": 13.528888888888888, "grad_norm": 0.3038574159145355, "learning_rate": 6.212598808381372e-05, "loss": 0.0226, "step": 15220 }, { "epoch": 13.537777777777778, "grad_norm": 0.43098074197769165, "learning_rate": 6.207845240614205e-05, "loss": 0.0217, "step": 15230 }, { "epoch": 13.546666666666667, "grad_norm": 0.34489667415618896, "learning_rate": 6.203090513189423e-05, "loss": 0.0196, "step": 15240 }, { "epoch": 13.555555555555555, "grad_norm": 0.2544582486152649, "learning_rate": 6.198334630672064e-05, "loss": 0.0192, "step": 15250 }, { "epoch": 13.564444444444444, "grad_norm": 0.2640814185142517, "learning_rate": 6.193577597628268e-05, "loss": 0.0223, "step": 15260 }, { "epoch": 13.573333333333334, "grad_norm": 0.2900255024433136, "learning_rate": 6.188819418625291e-05, "loss": 0.0201, "step": 15270 }, { "epoch": 13.582222222222223, "grad_norm": 0.2611928880214691, "learning_rate": 6.184060098231474e-05, "loss": 0.0216, "step": 15280 }, { "epoch": 13.591111111111111, "grad_norm": 0.24003702402114868, "learning_rate": 6.179299641016268e-05, "loss": 0.025, "step": 15290 }, { "epoch": 13.6, "grad_norm": 0.2014537751674652, "learning_rate": 6.174538051550205e-05, "loss": 0.0168, "step": 15300 }, { "epoch": 13.608888888888888, "grad_norm": 0.317535400390625, "learning_rate": 6.169775334404912e-05, "loss": 0.0207, "step": 15310 }, { "epoch": 13.617777777777778, "grad_norm": 0.338121622800827, "learning_rate": 6.165011494153091e-05, "loss": 0.022, "step": 15320 }, { "epoch": 13.626666666666667, "grad_norm": 0.2002786248922348, "learning_rate": 6.16024653536853e-05, "loss": 0.0242, "step": 15330 }, { "epoch": 13.635555555555555, "grad_norm": 0.24488554894924164, "learning_rate": 6.155480462626086e-05, "loss": 0.0175, "step": 15340 }, { "epoch": 13.644444444444444, "grad_norm": 0.20322267711162567, "learning_rate": 6.150713280501683e-05, "loss": 0.0168, "step": 15350 }, { "epoch": 13.653333333333332, "grad_norm": 0.2397463470697403, "learning_rate": 6.14594499357232e-05, "loss": 0.0215, "step": 15360 }, { "epoch": 13.662222222222223, "grad_norm": 0.3479824662208557, "learning_rate": 6.141175606416047e-05, "loss": 0.0201, "step": 15370 }, { "epoch": 13.671111111111111, "grad_norm": 0.2558960020542145, "learning_rate": 6.136405123611975e-05, "loss": 0.0204, "step": 15380 }, { "epoch": 13.68, "grad_norm": 0.29274797439575195, "learning_rate": 6.131633549740264e-05, "loss": 0.0214, "step": 15390 }, { "epoch": 13.688888888888888, "grad_norm": 0.34559011459350586, "learning_rate": 6.126860889382127e-05, "loss": 0.0218, "step": 15400 }, { "epoch": 13.697777777777778, "grad_norm": 0.22137686610221863, "learning_rate": 6.122087147119814e-05, "loss": 0.0219, "step": 15410 }, { "epoch": 13.706666666666667, "grad_norm": 0.23089848458766937, "learning_rate": 6.117312327536616e-05, "loss": 0.0244, "step": 15420 }, { "epoch": 13.715555555555556, "grad_norm": 0.22621923685073853, "learning_rate": 6.112536435216862e-05, "loss": 0.0213, "step": 15430 }, { "epoch": 13.724444444444444, "grad_norm": 0.2163238227367401, "learning_rate": 6.107759474745902e-05, "loss": 0.0236, "step": 15440 }, { "epoch": 13.733333333333333, "grad_norm": 0.327754944562912, "learning_rate": 6.1029814507101205e-05, "loss": 0.0225, "step": 15450 }, { "epoch": 13.742222222222223, "grad_norm": 0.2700999677181244, "learning_rate": 6.098202367696918e-05, "loss": 0.0217, "step": 15460 }, { "epoch": 13.751111111111111, "grad_norm": 0.2824099361896515, "learning_rate": 6.0934222302947165e-05, "loss": 0.0186, "step": 15470 }, { "epoch": 13.76, "grad_norm": 0.32764312624931335, "learning_rate": 6.088641043092946e-05, "loss": 0.022, "step": 15480 }, { "epoch": 13.768888888888888, "grad_norm": 0.18012875318527222, "learning_rate": 6.083858810682045e-05, "loss": 0.0198, "step": 15490 }, { "epoch": 13.777777777777779, "grad_norm": 0.2811411917209625, "learning_rate": 6.0790755376534556e-05, "loss": 0.016, "step": 15500 }, { "epoch": 13.786666666666667, "grad_norm": 0.22871829569339752, "learning_rate": 6.07429122859962e-05, "loss": 0.0192, "step": 15510 }, { "epoch": 13.795555555555556, "grad_norm": 0.25454089045524597, "learning_rate": 6.0695058881139765e-05, "loss": 0.0236, "step": 15520 }, { "epoch": 13.804444444444444, "grad_norm": 0.22371132671833038, "learning_rate": 6.064719520790949e-05, "loss": 0.0192, "step": 15530 }, { "epoch": 13.813333333333333, "grad_norm": 0.32117897272109985, "learning_rate": 6.059932131225954e-05, "loss": 0.0207, "step": 15540 }, { "epoch": 13.822222222222223, "grad_norm": 0.25924891233444214, "learning_rate": 6.0551437240153796e-05, "loss": 0.0177, "step": 15550 }, { "epoch": 13.831111111111111, "grad_norm": 0.3604084551334381, "learning_rate": 6.050354303756603e-05, "loss": 0.0273, "step": 15560 }, { "epoch": 13.84, "grad_norm": 0.291311115026474, "learning_rate": 6.0455638750479656e-05, "loss": 0.0177, "step": 15570 }, { "epoch": 13.848888888888888, "grad_norm": 0.2340097874403, "learning_rate": 6.040772442488779e-05, "loss": 0.0197, "step": 15580 }, { "epoch": 13.857777777777777, "grad_norm": 0.22553184628486633, "learning_rate": 6.035980010679321e-05, "loss": 0.0231, "step": 15590 }, { "epoch": 13.866666666666667, "grad_norm": 0.20488514006137848, "learning_rate": 6.0311865842208226e-05, "loss": 0.017, "step": 15600 }, { "epoch": 13.875555555555556, "grad_norm": 0.22748041152954102, "learning_rate": 6.0263921677154786e-05, "loss": 0.0189, "step": 15610 }, { "epoch": 13.884444444444444, "grad_norm": 0.3203737139701843, "learning_rate": 6.021596765766426e-05, "loss": 0.0223, "step": 15620 }, { "epoch": 13.893333333333333, "grad_norm": 0.30566495656967163, "learning_rate": 6.016800382977755e-05, "loss": 0.0219, "step": 15630 }, { "epoch": 13.902222222222223, "grad_norm": 0.23780511319637299, "learning_rate": 6.0120030239544925e-05, "loss": 0.0158, "step": 15640 }, { "epoch": 13.911111111111111, "grad_norm": 0.3699679970741272, "learning_rate": 6.007204693302604e-05, "loss": 0.0263, "step": 15650 }, { "epoch": 13.92, "grad_norm": 0.255758672952652, "learning_rate": 6.002405395628988e-05, "loss": 0.0223, "step": 15660 }, { "epoch": 13.928888888888888, "grad_norm": 0.4749181270599365, "learning_rate": 5.997605135541472e-05, "loss": 0.0207, "step": 15670 }, { "epoch": 13.937777777777779, "grad_norm": 0.2637505531311035, "learning_rate": 5.992803917648808e-05, "loss": 0.0255, "step": 15680 }, { "epoch": 13.946666666666667, "grad_norm": 0.34774959087371826, "learning_rate": 5.988001746560666e-05, "loss": 0.0198, "step": 15690 }, { "epoch": 13.955555555555556, "grad_norm": 0.26633942127227783, "learning_rate": 5.983198626887632e-05, "loss": 0.0253, "step": 15700 }, { "epoch": 13.964444444444444, "grad_norm": 0.43607550859451294, "learning_rate": 5.978394563241202e-05, "loss": 0.0229, "step": 15710 }, { "epoch": 13.973333333333333, "grad_norm": 0.25149473547935486, "learning_rate": 5.973589560233781e-05, "loss": 0.0225, "step": 15720 }, { "epoch": 13.982222222222223, "grad_norm": 0.21421964466571808, "learning_rate": 5.968783622478672e-05, "loss": 0.0248, "step": 15730 }, { "epoch": 13.991111111111111, "grad_norm": 0.2781190574169159, "learning_rate": 5.963976754590077e-05, "loss": 0.0282, "step": 15740 }, { "epoch": 14.0, "grad_norm": 0.2714502513408661, "learning_rate": 5.9591689611830936e-05, "loss": 0.0183, "step": 15750 }, { "epoch": 14.008888888888889, "grad_norm": 0.24736998975276947, "learning_rate": 5.9543602468737024e-05, "loss": 0.0223, "step": 15760 }, { "epoch": 14.017777777777777, "grad_norm": 0.2709900140762329, "learning_rate": 5.949550616278775e-05, "loss": 0.0248, "step": 15770 }, { "epoch": 14.026666666666667, "grad_norm": 0.17452913522720337, "learning_rate": 5.944740074016057e-05, "loss": 0.0193, "step": 15780 }, { "epoch": 14.035555555555556, "grad_norm": 0.22913698852062225, "learning_rate": 5.939928624704173e-05, "loss": 0.0183, "step": 15790 }, { "epoch": 14.044444444444444, "grad_norm": 0.3353698253631592, "learning_rate": 5.935116272962614e-05, "loss": 0.0203, "step": 15800 }, { "epoch": 14.053333333333333, "grad_norm": 0.35660868883132935, "learning_rate": 5.9303030234117454e-05, "loss": 0.023, "step": 15810 }, { "epoch": 14.062222222222223, "grad_norm": 0.27661803364753723, "learning_rate": 5.925488880672785e-05, "loss": 0.0209, "step": 15820 }, { "epoch": 14.071111111111112, "grad_norm": 0.32649901509284973, "learning_rate": 5.920673849367814e-05, "loss": 0.0175, "step": 15830 }, { "epoch": 14.08, "grad_norm": 0.3255966007709503, "learning_rate": 5.9158579341197674e-05, "loss": 0.0203, "step": 15840 }, { "epoch": 14.088888888888889, "grad_norm": 0.24412259459495544, "learning_rate": 5.911041139552424e-05, "loss": 0.0191, "step": 15850 }, { "epoch": 14.097777777777777, "grad_norm": 0.2026434987783432, "learning_rate": 5.906223470290414e-05, "loss": 0.0208, "step": 15860 }, { "epoch": 14.106666666666667, "grad_norm": 0.36602288484573364, "learning_rate": 5.901404930959198e-05, "loss": 0.0181, "step": 15870 }, { "epoch": 14.115555555555556, "grad_norm": 0.299293577671051, "learning_rate": 5.896585526185081e-05, "loss": 0.0232, "step": 15880 }, { "epoch": 14.124444444444444, "grad_norm": 0.2795165777206421, "learning_rate": 5.8917652605951936e-05, "loss": 0.0168, "step": 15890 }, { "epoch": 14.133333333333333, "grad_norm": 0.3157389163970947, "learning_rate": 5.886944138817494e-05, "loss": 0.0201, "step": 15900 }, { "epoch": 14.142222222222221, "grad_norm": 0.2926185727119446, "learning_rate": 5.882122165480763e-05, "loss": 0.0206, "step": 15910 }, { "epoch": 14.151111111111112, "grad_norm": 0.3093443810939789, "learning_rate": 5.8772993452145986e-05, "loss": 0.0219, "step": 15920 }, { "epoch": 14.16, "grad_norm": 0.28134483098983765, "learning_rate": 5.872475682649413e-05, "loss": 0.0234, "step": 15930 }, { "epoch": 14.168888888888889, "grad_norm": 0.3043070435523987, "learning_rate": 5.867651182416425e-05, "loss": 0.0232, "step": 15940 }, { "epoch": 14.177777777777777, "grad_norm": 0.29550790786743164, "learning_rate": 5.862825849147662e-05, "loss": 0.0231, "step": 15950 }, { "epoch": 14.186666666666667, "grad_norm": 0.25083112716674805, "learning_rate": 5.857999687475943e-05, "loss": 0.02, "step": 15960 }, { "epoch": 14.195555555555556, "grad_norm": 0.29701030254364014, "learning_rate": 5.8531727020348904e-05, "loss": 0.0222, "step": 15970 }, { "epoch": 14.204444444444444, "grad_norm": 0.26247745752334595, "learning_rate": 5.848344897458915e-05, "loss": 0.0164, "step": 15980 }, { "epoch": 14.213333333333333, "grad_norm": 0.2333749532699585, "learning_rate": 5.843516278383211e-05, "loss": 0.025, "step": 15990 }, { "epoch": 14.222222222222221, "grad_norm": 0.24535126984119415, "learning_rate": 5.838686849443761e-05, "loss": 0.0268, "step": 16000 }, { "epoch": 14.231111111111112, "grad_norm": 0.282220721244812, "learning_rate": 5.8338566152773164e-05, "loss": 0.0205, "step": 16010 }, { "epoch": 14.24, "grad_norm": 0.2613103985786438, "learning_rate": 5.82902558052141e-05, "loss": 0.0256, "step": 16020 }, { "epoch": 14.248888888888889, "grad_norm": 0.23091456294059753, "learning_rate": 5.8241937498143387e-05, "loss": 0.0192, "step": 16030 }, { "epoch": 14.257777777777777, "grad_norm": 0.2623749375343323, "learning_rate": 5.819361127795167e-05, "loss": 0.0193, "step": 16040 }, { "epoch": 14.266666666666667, "grad_norm": 0.248978853225708, "learning_rate": 5.814527719103713e-05, "loss": 0.0195, "step": 16050 }, { "epoch": 14.275555555555556, "grad_norm": 0.2796999216079712, "learning_rate": 5.8096935283805563e-05, "loss": 0.0278, "step": 16060 }, { "epoch": 14.284444444444444, "grad_norm": 0.23910991847515106, "learning_rate": 5.8048585602670255e-05, "loss": 0.0223, "step": 16070 }, { "epoch": 14.293333333333333, "grad_norm": 0.23127345740795135, "learning_rate": 5.800022819405193e-05, "loss": 0.0191, "step": 16080 }, { "epoch": 14.302222222222222, "grad_norm": 0.2520054280757904, "learning_rate": 5.795186310437878e-05, "loss": 0.0191, "step": 16090 }, { "epoch": 14.311111111111112, "grad_norm": 0.2270239144563675, "learning_rate": 5.790349038008632e-05, "loss": 0.023, "step": 16100 }, { "epoch": 14.32, "grad_norm": 0.3234311044216156, "learning_rate": 5.7855110067617436e-05, "loss": 0.0218, "step": 16110 }, { "epoch": 14.328888888888889, "grad_norm": 0.2440541833639145, "learning_rate": 5.780672221342226e-05, "loss": 0.016, "step": 16120 }, { "epoch": 14.337777777777777, "grad_norm": 0.2221696525812149, "learning_rate": 5.7758326863958214e-05, "loss": 0.0233, "step": 16130 }, { "epoch": 14.346666666666668, "grad_norm": 0.21968792378902435, "learning_rate": 5.770992406568988e-05, "loss": 0.0204, "step": 16140 }, { "epoch": 14.355555555555556, "grad_norm": 0.257714182138443, "learning_rate": 5.766151386508899e-05, "loss": 0.0181, "step": 16150 }, { "epoch": 14.364444444444445, "grad_norm": 0.2939833104610443, "learning_rate": 5.761309630863439e-05, "loss": 0.0233, "step": 16160 }, { "epoch": 14.373333333333333, "grad_norm": 0.3263516426086426, "learning_rate": 5.7564671442812015e-05, "loss": 0.0226, "step": 16170 }, { "epoch": 14.382222222222222, "grad_norm": 0.23128527402877808, "learning_rate": 5.7516239314114786e-05, "loss": 0.0226, "step": 16180 }, { "epoch": 14.391111111111112, "grad_norm": 0.25536197423934937, "learning_rate": 5.746779996904259e-05, "loss": 0.0193, "step": 16190 }, { "epoch": 14.4, "grad_norm": 0.37137383222579956, "learning_rate": 5.741935345410228e-05, "loss": 0.023, "step": 16200 }, { "epoch": 14.408888888888889, "grad_norm": 0.27404525876045227, "learning_rate": 5.7370899815807566e-05, "loss": 0.0186, "step": 16210 }, { "epoch": 14.417777777777777, "grad_norm": 0.2673507630825043, "learning_rate": 5.7322439100679006e-05, "loss": 0.0221, "step": 16220 }, { "epoch": 14.426666666666666, "grad_norm": 0.2506387531757355, "learning_rate": 5.7273971355243926e-05, "loss": 0.0228, "step": 16230 }, { "epoch": 14.435555555555556, "grad_norm": 0.2256951481103897, "learning_rate": 5.7225496626036446e-05, "loss": 0.0201, "step": 16240 }, { "epoch": 14.444444444444445, "grad_norm": 0.24575386941432953, "learning_rate": 5.717701495959739e-05, "loss": 0.0192, "step": 16250 }, { "epoch": 14.453333333333333, "grad_norm": 0.3262012004852295, "learning_rate": 5.712852640247418e-05, "loss": 0.0227, "step": 16260 }, { "epoch": 14.462222222222222, "grad_norm": 0.2879999577999115, "learning_rate": 5.7080031001220914e-05, "loss": 0.0161, "step": 16270 }, { "epoch": 14.471111111111112, "grad_norm": 0.1657661646604538, "learning_rate": 5.7031528802398235e-05, "loss": 0.0187, "step": 16280 }, { "epoch": 14.48, "grad_norm": 0.372457355260849, "learning_rate": 5.698301985257336e-05, "loss": 0.0197, "step": 16290 }, { "epoch": 14.488888888888889, "grad_norm": 0.25743913650512695, "learning_rate": 5.69345041983199e-05, "loss": 0.0227, "step": 16300 }, { "epoch": 14.497777777777777, "grad_norm": 0.3178394138813019, "learning_rate": 5.6885981886217984e-05, "loss": 0.0218, "step": 16310 }, { "epoch": 14.506666666666666, "grad_norm": 0.43189308047294617, "learning_rate": 5.683745296285408e-05, "loss": 0.0196, "step": 16320 }, { "epoch": 14.515555555555556, "grad_norm": 0.3193722665309906, "learning_rate": 5.678891747482102e-05, "loss": 0.0194, "step": 16330 }, { "epoch": 14.524444444444445, "grad_norm": 0.40533554553985596, "learning_rate": 5.674037546871797e-05, "loss": 0.0213, "step": 16340 }, { "epoch": 14.533333333333333, "grad_norm": 0.3343496024608612, "learning_rate": 5.66918269911503e-05, "loss": 0.0203, "step": 16350 }, { "epoch": 14.542222222222222, "grad_norm": 0.25810468196868896, "learning_rate": 5.6643272088729646e-05, "loss": 0.0201, "step": 16360 }, { "epoch": 14.551111111111112, "grad_norm": 0.22527335584163666, "learning_rate": 5.659471080807376e-05, "loss": 0.0229, "step": 16370 }, { "epoch": 14.56, "grad_norm": 0.21295639872550964, "learning_rate": 5.654614319580657e-05, "loss": 0.0231, "step": 16380 }, { "epoch": 14.568888888888889, "grad_norm": 0.2894652783870697, "learning_rate": 5.649756929855803e-05, "loss": 0.0243, "step": 16390 }, { "epoch": 14.577777777777778, "grad_norm": 0.24444344639778137, "learning_rate": 5.644898916296417e-05, "loss": 0.0156, "step": 16400 }, { "epoch": 14.586666666666666, "grad_norm": 0.25050243735313416, "learning_rate": 5.640040283566702e-05, "loss": 0.0211, "step": 16410 }, { "epoch": 14.595555555555556, "grad_norm": 0.6181800961494446, "learning_rate": 5.6351810363314487e-05, "loss": 0.019, "step": 16420 }, { "epoch": 14.604444444444445, "grad_norm": 0.301251083612442, "learning_rate": 5.630321179256044e-05, "loss": 0.0233, "step": 16430 }, { "epoch": 14.613333333333333, "grad_norm": 0.28157222270965576, "learning_rate": 5.625460717006458e-05, "loss": 0.0197, "step": 16440 }, { "epoch": 14.622222222222222, "grad_norm": 0.45060184597969055, "learning_rate": 5.6205996542492435e-05, "loss": 0.0176, "step": 16450 }, { "epoch": 14.63111111111111, "grad_norm": 0.37789154052734375, "learning_rate": 5.615737995651528e-05, "loss": 0.0181, "step": 16460 }, { "epoch": 14.64, "grad_norm": 0.339242160320282, "learning_rate": 5.61087574588101e-05, "loss": 0.0192, "step": 16470 }, { "epoch": 14.648888888888889, "grad_norm": 0.24192824959754944, "learning_rate": 5.60601290960596e-05, "loss": 0.0199, "step": 16480 }, { "epoch": 14.657777777777778, "grad_norm": 0.24341998994350433, "learning_rate": 5.6011494914952055e-05, "loss": 0.0186, "step": 16490 }, { "epoch": 14.666666666666666, "grad_norm": 0.3235504627227783, "learning_rate": 5.59628549621814e-05, "loss": 0.0223, "step": 16500 }, { "epoch": 14.675555555555556, "grad_norm": 0.3781902492046356, "learning_rate": 5.5914209284447025e-05, "loss": 0.0242, "step": 16510 }, { "epoch": 14.684444444444445, "grad_norm": 0.31276950240135193, "learning_rate": 5.586555792845391e-05, "loss": 0.0162, "step": 16520 }, { "epoch": 14.693333333333333, "grad_norm": 0.26168525218963623, "learning_rate": 5.581690094091241e-05, "loss": 0.0213, "step": 16530 }, { "epoch": 14.702222222222222, "grad_norm": 0.28395894169807434, "learning_rate": 5.576823836853832e-05, "loss": 0.025, "step": 16540 }, { "epoch": 14.71111111111111, "grad_norm": 0.2656903564929962, "learning_rate": 5.571957025805279e-05, "loss": 0.0184, "step": 16550 }, { "epoch": 14.72, "grad_norm": 0.24859799444675446, "learning_rate": 5.56708966561823e-05, "loss": 0.0177, "step": 16560 }, { "epoch": 14.72888888888889, "grad_norm": 0.3054693043231964, "learning_rate": 5.56222176096586e-05, "loss": 0.0203, "step": 16570 }, { "epoch": 14.737777777777778, "grad_norm": 0.35054874420166016, "learning_rate": 5.5573533165218624e-05, "loss": 0.0213, "step": 16580 }, { "epoch": 14.746666666666666, "grad_norm": 0.2619744539260864, "learning_rate": 5.552484336960457e-05, "loss": 0.0194, "step": 16590 }, { "epoch": 14.755555555555556, "grad_norm": 0.20784345269203186, "learning_rate": 5.54761482695637e-05, "loss": 0.0191, "step": 16600 }, { "epoch": 14.764444444444445, "grad_norm": 0.30113518238067627, "learning_rate": 5.542744791184841e-05, "loss": 0.018, "step": 16610 }, { "epoch": 14.773333333333333, "grad_norm": 0.20943520963191986, "learning_rate": 5.537874234321613e-05, "loss": 0.0225, "step": 16620 }, { "epoch": 14.782222222222222, "grad_norm": 0.2886584401130676, "learning_rate": 5.533003161042929e-05, "loss": 0.0207, "step": 16630 }, { "epoch": 14.79111111111111, "grad_norm": 0.18316636979579926, "learning_rate": 5.528131576025527e-05, "loss": 0.0192, "step": 16640 }, { "epoch": 14.8, "grad_norm": 0.29903581738471985, "learning_rate": 5.523259483946639e-05, "loss": 0.0191, "step": 16650 }, { "epoch": 14.80888888888889, "grad_norm": 0.23260439932346344, "learning_rate": 5.518386889483983e-05, "loss": 0.0154, "step": 16660 }, { "epoch": 14.817777777777778, "grad_norm": 0.31311744451522827, "learning_rate": 5.513513797315758e-05, "loss": 0.0174, "step": 16670 }, { "epoch": 14.826666666666666, "grad_norm": 0.28364869952201843, "learning_rate": 5.5086402121206404e-05, "loss": 0.0203, "step": 16680 }, { "epoch": 14.835555555555555, "grad_norm": 0.17186027765274048, "learning_rate": 5.503766138577782e-05, "loss": 0.0218, "step": 16690 }, { "epoch": 14.844444444444445, "grad_norm": 0.2953161895275116, "learning_rate": 5.498891581366804e-05, "loss": 0.0162, "step": 16700 }, { "epoch": 14.853333333333333, "grad_norm": 0.2372186928987503, "learning_rate": 5.49401654516779e-05, "loss": 0.0189, "step": 16710 }, { "epoch": 14.862222222222222, "grad_norm": 0.1992373764514923, "learning_rate": 5.4891410346612816e-05, "loss": 0.0191, "step": 16720 }, { "epoch": 14.87111111111111, "grad_norm": 0.301650732755661, "learning_rate": 5.4842650545282805e-05, "loss": 0.0174, "step": 16730 }, { "epoch": 14.88, "grad_norm": 0.25732946395874023, "learning_rate": 5.4793886094502354e-05, "loss": 0.0178, "step": 16740 }, { "epoch": 14.88888888888889, "grad_norm": 0.29436078667640686, "learning_rate": 5.4745117041090454e-05, "loss": 0.0208, "step": 16750 }, { "epoch": 14.897777777777778, "grad_norm": 0.25080159306526184, "learning_rate": 5.469634343187048e-05, "loss": 0.0192, "step": 16760 }, { "epoch": 14.906666666666666, "grad_norm": 0.2518860399723053, "learning_rate": 5.464756531367019e-05, "loss": 0.0214, "step": 16770 }, { "epoch": 14.915555555555555, "grad_norm": 0.2241722196340561, "learning_rate": 5.459878273332166e-05, "loss": 0.0191, "step": 16780 }, { "epoch": 14.924444444444445, "grad_norm": 0.22872714698314667, "learning_rate": 5.45499957376613e-05, "loss": 0.017, "step": 16790 }, { "epoch": 14.933333333333334, "grad_norm": 0.20401208102703094, "learning_rate": 5.450120437352969e-05, "loss": 0.0235, "step": 16800 }, { "epoch": 14.942222222222222, "grad_norm": 0.35684698820114136, "learning_rate": 5.4452408687771626e-05, "loss": 0.0199, "step": 16810 }, { "epoch": 14.95111111111111, "grad_norm": 0.29822883009910583, "learning_rate": 5.440360872723611e-05, "loss": 0.017, "step": 16820 }, { "epoch": 14.96, "grad_norm": 0.2835635244846344, "learning_rate": 5.435480453877617e-05, "loss": 0.0196, "step": 16830 }, { "epoch": 14.96888888888889, "grad_norm": 0.276254802942276, "learning_rate": 5.430599616924894e-05, "loss": 0.019, "step": 16840 }, { "epoch": 14.977777777777778, "grad_norm": 0.2621566951274872, "learning_rate": 5.425718366551552e-05, "loss": 0.0178, "step": 16850 }, { "epoch": 14.986666666666666, "grad_norm": 0.21029075980186462, "learning_rate": 5.420836707444106e-05, "loss": 0.0234, "step": 16860 }, { "epoch": 14.995555555555555, "grad_norm": 0.2922075688838959, "learning_rate": 5.415954644289456e-05, "loss": 0.018, "step": 16870 }, { "epoch": 15.004444444444445, "grad_norm": 0.25814366340637207, "learning_rate": 5.4110721817748946e-05, "loss": 0.0176, "step": 16880 }, { "epoch": 15.013333333333334, "grad_norm": 0.22869639098644257, "learning_rate": 5.406189324588091e-05, "loss": 0.0186, "step": 16890 }, { "epoch": 15.022222222222222, "grad_norm": 0.26574382185935974, "learning_rate": 5.4013060774171055e-05, "loss": 0.0188, "step": 16900 }, { "epoch": 15.03111111111111, "grad_norm": 0.2556498944759369, "learning_rate": 5.39642244495036e-05, "loss": 0.0229, "step": 16910 }, { "epoch": 15.04, "grad_norm": 0.18329009413719177, "learning_rate": 5.391538431876652e-05, "loss": 0.0172, "step": 16920 }, { "epoch": 15.04888888888889, "grad_norm": 0.184352308511734, "learning_rate": 5.38665404288515e-05, "loss": 0.0164, "step": 16930 }, { "epoch": 15.057777777777778, "grad_norm": 0.2291458249092102, "learning_rate": 5.3817692826653697e-05, "loss": 0.0167, "step": 16940 }, { "epoch": 15.066666666666666, "grad_norm": 0.225351482629776, "learning_rate": 5.376884155907197e-05, "loss": 0.016, "step": 16950 }, { "epoch": 15.075555555555555, "grad_norm": 0.29363006353378296, "learning_rate": 5.3719986673008614e-05, "loss": 0.0189, "step": 16960 }, { "epoch": 15.084444444444445, "grad_norm": 0.20316341519355774, "learning_rate": 5.367112821536943e-05, "loss": 0.0149, "step": 16970 }, { "epoch": 15.093333333333334, "grad_norm": 0.2644878625869751, "learning_rate": 5.3622266233063647e-05, "loss": 0.0249, "step": 16980 }, { "epoch": 15.102222222222222, "grad_norm": 0.4142638146877289, "learning_rate": 5.3573400773003865e-05, "loss": 0.0226, "step": 16990 }, { "epoch": 15.11111111111111, "grad_norm": 0.3178151845932007, "learning_rate": 5.352453188210603e-05, "loss": 0.0167, "step": 17000 }, { "epoch": 15.12, "grad_norm": 0.23672336339950562, "learning_rate": 5.3475659607289374e-05, "loss": 0.024, "step": 17010 }, { "epoch": 15.12888888888889, "grad_norm": 0.25803354382514954, "learning_rate": 5.3426783995476415e-05, "loss": 0.0176, "step": 17020 }, { "epoch": 15.137777777777778, "grad_norm": 0.2361428290605545, "learning_rate": 5.337790509359283e-05, "loss": 0.0242, "step": 17030 }, { "epoch": 15.146666666666667, "grad_norm": 0.2522095739841461, "learning_rate": 5.332902294856746e-05, "loss": 0.0197, "step": 17040 }, { "epoch": 15.155555555555555, "grad_norm": 0.3701515197753906, "learning_rate": 5.328013760733227e-05, "loss": 0.0204, "step": 17050 }, { "epoch": 15.164444444444445, "grad_norm": 0.2582254111766815, "learning_rate": 5.3231249116822324e-05, "loss": 0.0229, "step": 17060 }, { "epoch": 15.173333333333334, "grad_norm": 0.2073426991701126, "learning_rate": 5.3182357523975655e-05, "loss": 0.0138, "step": 17070 }, { "epoch": 15.182222222222222, "grad_norm": 0.24096085131168365, "learning_rate": 5.313346287573331e-05, "loss": 0.0203, "step": 17080 }, { "epoch": 15.19111111111111, "grad_norm": 0.2670769691467285, "learning_rate": 5.308456521903925e-05, "loss": 0.0237, "step": 17090 }, { "epoch": 15.2, "grad_norm": 0.28113535046577454, "learning_rate": 5.303566460084034e-05, "loss": 0.0154, "step": 17100 }, { "epoch": 15.20888888888889, "grad_norm": 0.2864713966846466, "learning_rate": 5.2986761068086276e-05, "loss": 0.0158, "step": 17110 }, { "epoch": 15.217777777777778, "grad_norm": 0.337352991104126, "learning_rate": 5.293785466772957e-05, "loss": 0.016, "step": 17120 }, { "epoch": 15.226666666666667, "grad_norm": 0.24624383449554443, "learning_rate": 5.288894544672548e-05, "loss": 0.0169, "step": 17130 }, { "epoch": 15.235555555555555, "grad_norm": 0.32588687539100647, "learning_rate": 5.284003345203193e-05, "loss": 0.025, "step": 17140 }, { "epoch": 15.244444444444444, "grad_norm": 0.31252485513687134, "learning_rate": 5.279111873060959e-05, "loss": 0.019, "step": 17150 }, { "epoch": 15.253333333333334, "grad_norm": 0.23673707246780396, "learning_rate": 5.274220132942169e-05, "loss": 0.0199, "step": 17160 }, { "epoch": 15.262222222222222, "grad_norm": 0.20423223078250885, "learning_rate": 5.269328129543401e-05, "loss": 0.018, "step": 17170 }, { "epoch": 15.27111111111111, "grad_norm": 0.28028908371925354, "learning_rate": 5.264435867561497e-05, "loss": 0.0232, "step": 17180 }, { "epoch": 15.28, "grad_norm": 0.3139406144618988, "learning_rate": 5.259543351693531e-05, "loss": 0.0179, "step": 17190 }, { "epoch": 15.28888888888889, "grad_norm": 0.2052946537733078, "learning_rate": 5.254650586636835e-05, "loss": 0.0158, "step": 17200 }, { "epoch": 15.297777777777778, "grad_norm": 0.25753018260002136, "learning_rate": 5.249757577088971e-05, "loss": 0.0183, "step": 17210 }, { "epoch": 15.306666666666667, "grad_norm": 0.28061798214912415, "learning_rate": 5.2448643277477425e-05, "loss": 0.0159, "step": 17220 }, { "epoch": 15.315555555555555, "grad_norm": 0.28421011567115784, "learning_rate": 5.2399708433111786e-05, "loss": 0.0186, "step": 17230 }, { "epoch": 15.324444444444444, "grad_norm": 0.2976103127002716, "learning_rate": 5.235077128477536e-05, "loss": 0.0221, "step": 17240 }, { "epoch": 15.333333333333334, "grad_norm": 0.24501818418502808, "learning_rate": 5.230183187945291e-05, "loss": 0.0185, "step": 17250 }, { "epoch": 15.342222222222222, "grad_norm": 0.3082255721092224, "learning_rate": 5.2252890264131383e-05, "loss": 0.0214, "step": 17260 }, { "epoch": 15.351111111111111, "grad_norm": 0.2790962755680084, "learning_rate": 5.220394648579986e-05, "loss": 0.02, "step": 17270 }, { "epoch": 15.36, "grad_norm": 0.2588237226009369, "learning_rate": 5.2155000591449454e-05, "loss": 0.021, "step": 17280 }, { "epoch": 15.36888888888889, "grad_norm": 0.4451865553855896, "learning_rate": 5.2106052628073366e-05, "loss": 0.0184, "step": 17290 }, { "epoch": 15.377777777777778, "grad_norm": 0.253683865070343, "learning_rate": 5.2057102642666725e-05, "loss": 0.0179, "step": 17300 }, { "epoch": 15.386666666666667, "grad_norm": 0.155042365193367, "learning_rate": 5.200815068222665e-05, "loss": 0.0164, "step": 17310 }, { "epoch": 15.395555555555555, "grad_norm": 0.2535167634487152, "learning_rate": 5.1959196793752127e-05, "loss": 0.0165, "step": 17320 }, { "epoch": 15.404444444444444, "grad_norm": 0.2950916588306427, "learning_rate": 5.1910241024244e-05, "loss": 0.0225, "step": 17330 }, { "epoch": 15.413333333333334, "grad_norm": 0.3264063596725464, "learning_rate": 5.186128342070495e-05, "loss": 0.0192, "step": 17340 }, { "epoch": 15.422222222222222, "grad_norm": 0.3013168275356293, "learning_rate": 5.181232403013936e-05, "loss": 0.0244, "step": 17350 }, { "epoch": 15.431111111111111, "grad_norm": 0.2327616661787033, "learning_rate": 5.176336289955337e-05, "loss": 0.0196, "step": 17360 }, { "epoch": 15.44, "grad_norm": 0.2415391355752945, "learning_rate": 5.171440007595477e-05, "loss": 0.0194, "step": 17370 }, { "epoch": 15.448888888888888, "grad_norm": 0.30694711208343506, "learning_rate": 5.1665435606353005e-05, "loss": 0.0184, "step": 17380 }, { "epoch": 15.457777777777778, "grad_norm": 0.2909052073955536, "learning_rate": 5.1616469537759085e-05, "loss": 0.017, "step": 17390 }, { "epoch": 15.466666666666667, "grad_norm": 0.270821213722229, "learning_rate": 5.1567501917185524e-05, "loss": 0.0179, "step": 17400 }, { "epoch": 15.475555555555555, "grad_norm": 0.2200581133365631, "learning_rate": 5.151853279164637e-05, "loss": 0.0185, "step": 17410 }, { "epoch": 15.484444444444444, "grad_norm": 0.21081627905368805, "learning_rate": 5.146956220815708e-05, "loss": 0.0143, "step": 17420 }, { "epoch": 15.493333333333334, "grad_norm": 0.22132234275341034, "learning_rate": 5.142059021373457e-05, "loss": 0.0174, "step": 17430 }, { "epoch": 15.502222222222223, "grad_norm": 0.23977962136268616, "learning_rate": 5.1371616855397044e-05, "loss": 0.0266, "step": 17440 }, { "epoch": 15.511111111111111, "grad_norm": 0.298194944858551, "learning_rate": 5.132264218016404e-05, "loss": 0.017, "step": 17450 }, { "epoch": 15.52, "grad_norm": 0.1604355275630951, "learning_rate": 5.127366623505636e-05, "loss": 0.019, "step": 17460 }, { "epoch": 15.528888888888888, "grad_norm": 0.2204170674085617, "learning_rate": 5.1224689067096056e-05, "loss": 0.0215, "step": 17470 }, { "epoch": 15.537777777777778, "grad_norm": 0.21123148500919342, "learning_rate": 5.117571072330629e-05, "loss": 0.0233, "step": 17480 }, { "epoch": 15.546666666666667, "grad_norm": 0.24665620923042297, "learning_rate": 5.112673125071142e-05, "loss": 0.019, "step": 17490 }, { "epoch": 15.555555555555555, "grad_norm": 0.25338661670684814, "learning_rate": 5.107775069633683e-05, "loss": 0.0227, "step": 17500 }, { "epoch": 15.564444444444444, "grad_norm": 0.2750021517276764, "learning_rate": 5.102876910720897e-05, "loss": 0.0231, "step": 17510 }, { "epoch": 15.573333333333334, "grad_norm": 0.2833741009235382, "learning_rate": 5.097978653035531e-05, "loss": 0.0226, "step": 17520 }, { "epoch": 15.582222222222223, "grad_norm": 0.2910599410533905, "learning_rate": 5.0930803012804216e-05, "loss": 0.0189, "step": 17530 }, { "epoch": 15.591111111111111, "grad_norm": 0.36290091276168823, "learning_rate": 5.0881818601585005e-05, "loss": 0.0196, "step": 17540 }, { "epoch": 15.6, "grad_norm": 0.29941707849502563, "learning_rate": 5.0832833343727795e-05, "loss": 0.0156, "step": 17550 }, { "epoch": 15.608888888888888, "grad_norm": 0.26905664801597595, "learning_rate": 5.0783847286263575e-05, "loss": 0.0179, "step": 17560 }, { "epoch": 15.617777777777778, "grad_norm": 0.29305142164230347, "learning_rate": 5.073486047622408e-05, "loss": 0.0176, "step": 17570 }, { "epoch": 15.626666666666667, "grad_norm": 0.2897503077983856, "learning_rate": 5.068587296064174e-05, "loss": 0.0158, "step": 17580 }, { "epoch": 15.635555555555555, "grad_norm": 0.25299495458602905, "learning_rate": 5.063688478654972e-05, "loss": 0.0182, "step": 17590 }, { "epoch": 15.644444444444444, "grad_norm": 0.22290833294391632, "learning_rate": 5.058789600098177e-05, "loss": 0.0173, "step": 17600 }, { "epoch": 15.653333333333332, "grad_norm": 0.2603871524333954, "learning_rate": 5.053890665097225e-05, "loss": 0.0168, "step": 17610 }, { "epoch": 15.662222222222223, "grad_norm": 0.28640782833099365, "learning_rate": 5.048991678355602e-05, "loss": 0.0193, "step": 17620 }, { "epoch": 15.671111111111111, "grad_norm": 0.18087446689605713, "learning_rate": 5.044092644576852e-05, "loss": 0.0176, "step": 17630 }, { "epoch": 15.68, "grad_norm": 0.36618635058403015, "learning_rate": 5.0391935684645574e-05, "loss": 0.0197, "step": 17640 }, { "epoch": 15.688888888888888, "grad_norm": 0.24077554047107697, "learning_rate": 5.0342944547223415e-05, "loss": 0.0176, "step": 17650 }, { "epoch": 15.697777777777778, "grad_norm": 0.25324785709381104, "learning_rate": 5.029395308053866e-05, "loss": 0.0193, "step": 17660 }, { "epoch": 15.706666666666667, "grad_norm": 0.2250833511352539, "learning_rate": 5.0244961331628236e-05, "loss": 0.0159, "step": 17670 }, { "epoch": 15.715555555555556, "grad_norm": 0.24543286859989166, "learning_rate": 5.0195969347529345e-05, "loss": 0.0158, "step": 17680 }, { "epoch": 15.724444444444444, "grad_norm": 0.27458345890045166, "learning_rate": 5.014697717527941e-05, "loss": 0.0268, "step": 17690 }, { "epoch": 15.733333333333333, "grad_norm": 0.2468402236700058, "learning_rate": 5.0097984861916045e-05, "loss": 0.0216, "step": 17700 }, { "epoch": 15.742222222222223, "grad_norm": 0.5402854084968567, "learning_rate": 5.0048992454476964e-05, "loss": 0.018, "step": 17710 }, { "epoch": 15.751111111111111, "grad_norm": 0.2868307828903198, "learning_rate": 5e-05, "loss": 0.0211, "step": 17720 }, { "epoch": 15.76, "grad_norm": 0.4100521206855774, "learning_rate": 4.9951007545523054e-05, "loss": 0.0164, "step": 17730 }, { "epoch": 15.768888888888888, "grad_norm": 0.3034549653530121, "learning_rate": 4.990201513808396e-05, "loss": 0.0157, "step": 17740 }, { "epoch": 15.777777777777779, "grad_norm": 0.28356343507766724, "learning_rate": 4.9853022824720594e-05, "loss": 0.0155, "step": 17750 }, { "epoch": 15.786666666666667, "grad_norm": 0.19584694504737854, "learning_rate": 4.980403065247065e-05, "loss": 0.0163, "step": 17760 }, { "epoch": 15.795555555555556, "grad_norm": 0.2621423900127411, "learning_rate": 4.975503866837176e-05, "loss": 0.019, "step": 17770 }, { "epoch": 15.804444444444444, "grad_norm": 0.178773432970047, "learning_rate": 4.970604691946136e-05, "loss": 0.0174, "step": 17780 }, { "epoch": 15.813333333333333, "grad_norm": 0.28350523114204407, "learning_rate": 4.965705545277661e-05, "loss": 0.0188, "step": 17790 }, { "epoch": 15.822222222222223, "grad_norm": 0.2878253757953644, "learning_rate": 4.9608064315354444e-05, "loss": 0.0158, "step": 17800 }, { "epoch": 15.831111111111111, "grad_norm": 0.25876855850219727, "learning_rate": 4.955907355423149e-05, "loss": 0.0211, "step": 17810 }, { "epoch": 15.84, "grad_norm": 0.2669919729232788, "learning_rate": 4.951008321644398e-05, "loss": 0.0201, "step": 17820 }, { "epoch": 15.848888888888888, "grad_norm": 0.23970156908035278, "learning_rate": 4.946109334902776e-05, "loss": 0.0218, "step": 17830 }, { "epoch": 15.857777777777777, "grad_norm": 0.2323504090309143, "learning_rate": 4.941210399901823e-05, "loss": 0.0205, "step": 17840 }, { "epoch": 15.866666666666667, "grad_norm": 0.2351147085428238, "learning_rate": 4.936311521345027e-05, "loss": 0.0186, "step": 17850 }, { "epoch": 15.875555555555556, "grad_norm": 0.46192312240600586, "learning_rate": 4.9314127039358257e-05, "loss": 0.0186, "step": 17860 }, { "epoch": 15.884444444444444, "grad_norm": 0.2544322609901428, "learning_rate": 4.9265139523775925e-05, "loss": 0.0193, "step": 17870 }, { "epoch": 15.893333333333333, "grad_norm": 0.24158556759357452, "learning_rate": 4.921615271373644e-05, "loss": 0.0162, "step": 17880 }, { "epoch": 15.902222222222223, "grad_norm": 0.2273600995540619, "learning_rate": 4.9167166656272216e-05, "loss": 0.017, "step": 17890 }, { "epoch": 15.911111111111111, "grad_norm": 0.21839356422424316, "learning_rate": 4.911818139841501e-05, "loss": 0.0164, "step": 17900 }, { "epoch": 15.92, "grad_norm": 0.22493292391300201, "learning_rate": 4.9069196987195795e-05, "loss": 0.0177, "step": 17910 }, { "epoch": 15.928888888888888, "grad_norm": 0.3291530907154083, "learning_rate": 4.902021346964469e-05, "loss": 0.026, "step": 17920 }, { "epoch": 15.937777777777779, "grad_norm": 0.3203970193862915, "learning_rate": 4.8971230892791034e-05, "loss": 0.0231, "step": 17930 }, { "epoch": 15.946666666666667, "grad_norm": 0.24584686756134033, "learning_rate": 4.892224930366317e-05, "loss": 0.0209, "step": 17940 }, { "epoch": 15.955555555555556, "grad_norm": 0.22291962802410126, "learning_rate": 4.8873268749288595e-05, "loss": 0.0197, "step": 17950 }, { "epoch": 15.964444444444444, "grad_norm": 0.3370886445045471, "learning_rate": 4.8824289276693716e-05, "loss": 0.0165, "step": 17960 }, { "epoch": 15.973333333333333, "grad_norm": 0.24101021885871887, "learning_rate": 4.877531093290394e-05, "loss": 0.0164, "step": 17970 }, { "epoch": 15.982222222222223, "grad_norm": 0.28463393449783325, "learning_rate": 4.872633376494364e-05, "loss": 0.015, "step": 17980 }, { "epoch": 15.991111111111111, "grad_norm": 0.21605175733566284, "learning_rate": 4.867735781983598e-05, "loss": 0.0183, "step": 17990 }, { "epoch": 16.0, "grad_norm": 0.6170627474784851, "learning_rate": 4.862838314460298e-05, "loss": 0.0153, "step": 18000 }, { "epoch": 16.00888888888889, "grad_norm": 0.25662732124328613, "learning_rate": 4.857940978626544e-05, "loss": 0.0178, "step": 18010 }, { "epoch": 16.017777777777777, "grad_norm": 0.30776146054267883, "learning_rate": 4.853043779184292e-05, "loss": 0.0238, "step": 18020 }, { "epoch": 16.026666666666667, "grad_norm": 0.2506168484687805, "learning_rate": 4.848146720835364e-05, "loss": 0.0172, "step": 18030 }, { "epoch": 16.035555555555554, "grad_norm": 0.2631062865257263, "learning_rate": 4.8432498082814495e-05, "loss": 0.0157, "step": 18040 }, { "epoch": 16.044444444444444, "grad_norm": 0.18666742742061615, "learning_rate": 4.838353046224093e-05, "loss": 0.0167, "step": 18050 }, { "epoch": 16.053333333333335, "grad_norm": 0.14807327091693878, "learning_rate": 4.833456439364699e-05, "loss": 0.0176, "step": 18060 }, { "epoch": 16.06222222222222, "grad_norm": 0.2551247477531433, "learning_rate": 4.828559992404523e-05, "loss": 0.018, "step": 18070 }, { "epoch": 16.07111111111111, "grad_norm": 0.3797951638698578, "learning_rate": 4.823663710044663e-05, "loss": 0.0232, "step": 18080 }, { "epoch": 16.08, "grad_norm": 0.2609039843082428, "learning_rate": 4.8187675969860664e-05, "loss": 0.0162, "step": 18090 }, { "epoch": 16.08888888888889, "grad_norm": 0.19994911551475525, "learning_rate": 4.813871657929506e-05, "loss": 0.0193, "step": 18100 }, { "epoch": 16.09777777777778, "grad_norm": 0.31123244762420654, "learning_rate": 4.8089758975756e-05, "loss": 0.0178, "step": 18110 }, { "epoch": 16.106666666666666, "grad_norm": 0.3433266282081604, "learning_rate": 4.804080320624788e-05, "loss": 0.0163, "step": 18120 }, { "epoch": 16.115555555555556, "grad_norm": 0.2106611728668213, "learning_rate": 4.7991849317773366e-05, "loss": 0.0177, "step": 18130 }, { "epoch": 16.124444444444446, "grad_norm": 0.22870199382305145, "learning_rate": 4.7942897357333286e-05, "loss": 0.0204, "step": 18140 }, { "epoch": 16.133333333333333, "grad_norm": 0.3321162462234497, "learning_rate": 4.7893947371926645e-05, "loss": 0.0175, "step": 18150 }, { "epoch": 16.142222222222223, "grad_norm": 0.24220706522464752, "learning_rate": 4.784499940855055e-05, "loss": 0.0234, "step": 18160 }, { "epoch": 16.15111111111111, "grad_norm": 0.2623061537742615, "learning_rate": 4.779605351420014e-05, "loss": 0.0237, "step": 18170 }, { "epoch": 16.16, "grad_norm": 0.2407197803258896, "learning_rate": 4.7747109735868615e-05, "loss": 0.0221, "step": 18180 }, { "epoch": 16.16888888888889, "grad_norm": 0.18278692662715912, "learning_rate": 4.769816812054711e-05, "loss": 0.0181, "step": 18190 }, { "epoch": 16.177777777777777, "grad_norm": 0.22486960887908936, "learning_rate": 4.764922871522467e-05, "loss": 0.0186, "step": 18200 }, { "epoch": 16.186666666666667, "grad_norm": 0.3143538534641266, "learning_rate": 4.7600291566888226e-05, "loss": 0.017, "step": 18210 }, { "epoch": 16.195555555555554, "grad_norm": 0.3177400529384613, "learning_rate": 4.755135672252259e-05, "loss": 0.0168, "step": 18220 }, { "epoch": 16.204444444444444, "grad_norm": 0.30532997846603394, "learning_rate": 4.75024242291103e-05, "loss": 0.0153, "step": 18230 }, { "epoch": 16.213333333333335, "grad_norm": 0.26448702812194824, "learning_rate": 4.7453494133631664e-05, "loss": 0.0168, "step": 18240 }, { "epoch": 16.22222222222222, "grad_norm": 0.26241979002952576, "learning_rate": 4.74045664830647e-05, "loss": 0.0238, "step": 18250 }, { "epoch": 16.23111111111111, "grad_norm": 0.22123588621616364, "learning_rate": 4.7355641324385047e-05, "loss": 0.0195, "step": 18260 }, { "epoch": 16.24, "grad_norm": 0.19880373775959015, "learning_rate": 4.7306718704565985e-05, "loss": 0.0189, "step": 18270 }, { "epoch": 16.24888888888889, "grad_norm": 0.16623450815677643, "learning_rate": 4.725779867057833e-05, "loss": 0.0206, "step": 18280 }, { "epoch": 16.25777777777778, "grad_norm": 0.1775319129228592, "learning_rate": 4.720888126939043e-05, "loss": 0.0169, "step": 18290 }, { "epoch": 16.266666666666666, "grad_norm": 0.16278605163097382, "learning_rate": 4.715996654796808e-05, "loss": 0.0176, "step": 18300 }, { "epoch": 16.275555555555556, "grad_norm": 0.18570546805858612, "learning_rate": 4.711105455327454e-05, "loss": 0.0188, "step": 18310 }, { "epoch": 16.284444444444443, "grad_norm": 0.28981906175613403, "learning_rate": 4.706214533227045e-05, "loss": 0.0205, "step": 18320 }, { "epoch": 16.293333333333333, "grad_norm": 0.21457535028457642, "learning_rate": 4.701323893191373e-05, "loss": 0.0159, "step": 18330 }, { "epoch": 16.302222222222223, "grad_norm": 0.27357226610183716, "learning_rate": 4.696433539915968e-05, "loss": 0.018, "step": 18340 }, { "epoch": 16.31111111111111, "grad_norm": 0.3503684997558594, "learning_rate": 4.691543478096075e-05, "loss": 0.0185, "step": 18350 }, { "epoch": 16.32, "grad_norm": 0.2978048026561737, "learning_rate": 4.68665371242667e-05, "loss": 0.0144, "step": 18360 }, { "epoch": 16.32888888888889, "grad_norm": 0.17902731895446777, "learning_rate": 4.6817642476024356e-05, "loss": 0.0226, "step": 18370 }, { "epoch": 16.337777777777777, "grad_norm": 0.2693292200565338, "learning_rate": 4.6768750883177674e-05, "loss": 0.0166, "step": 18380 }, { "epoch": 16.346666666666668, "grad_norm": 0.16591981053352356, "learning_rate": 4.671986239266773e-05, "loss": 0.0151, "step": 18390 }, { "epoch": 16.355555555555554, "grad_norm": 0.19048801064491272, "learning_rate": 4.6670977051432564e-05, "loss": 0.0182, "step": 18400 }, { "epoch": 16.364444444444445, "grad_norm": 0.31399354338645935, "learning_rate": 4.66220949064072e-05, "loss": 0.025, "step": 18410 }, { "epoch": 16.373333333333335, "grad_norm": 0.20212557911872864, "learning_rate": 4.65732160045236e-05, "loss": 0.0169, "step": 18420 }, { "epoch": 16.38222222222222, "grad_norm": 0.26558154821395874, "learning_rate": 4.652434039271063e-05, "loss": 0.0196, "step": 18430 }, { "epoch": 16.391111111111112, "grad_norm": 0.15303626656532288, "learning_rate": 4.647546811789398e-05, "loss": 0.0167, "step": 18440 }, { "epoch": 16.4, "grad_norm": 0.20160678029060364, "learning_rate": 4.642659922699615e-05, "loss": 0.0169, "step": 18450 }, { "epoch": 16.40888888888889, "grad_norm": 0.14691753685474396, "learning_rate": 4.6377733766936365e-05, "loss": 0.0154, "step": 18460 }, { "epoch": 16.41777777777778, "grad_norm": 0.2581484913825989, "learning_rate": 4.632887178463057e-05, "loss": 0.0168, "step": 18470 }, { "epoch": 16.426666666666666, "grad_norm": 0.21554076671600342, "learning_rate": 4.628001332699139e-05, "loss": 0.0153, "step": 18480 }, { "epoch": 16.435555555555556, "grad_norm": 0.25632625818252563, "learning_rate": 4.623115844092803e-05, "loss": 0.0201, "step": 18490 }, { "epoch": 16.444444444444443, "grad_norm": 0.26612260937690735, "learning_rate": 4.6182307173346315e-05, "loss": 0.0155, "step": 18500 }, { "epoch": 16.453333333333333, "grad_norm": 0.23513440787792206, "learning_rate": 4.6133459571148525e-05, "loss": 0.015, "step": 18510 }, { "epoch": 16.462222222222223, "grad_norm": 0.28556710481643677, "learning_rate": 4.608461568123348e-05, "loss": 0.0181, "step": 18520 }, { "epoch": 16.47111111111111, "grad_norm": 0.2529519200325012, "learning_rate": 4.603577555049641e-05, "loss": 0.0134, "step": 18530 }, { "epoch": 16.48, "grad_norm": 0.2928880453109741, "learning_rate": 4.5986939225828964e-05, "loss": 0.0159, "step": 18540 }, { "epoch": 16.488888888888887, "grad_norm": 0.22248776257038116, "learning_rate": 4.5938106754119094e-05, "loss": 0.0187, "step": 18550 }, { "epoch": 16.497777777777777, "grad_norm": 0.2728450894355774, "learning_rate": 4.588927818225107e-05, "loss": 0.0199, "step": 18560 }, { "epoch": 16.506666666666668, "grad_norm": 0.24474753439426422, "learning_rate": 4.584045355710544e-05, "loss": 0.0156, "step": 18570 }, { "epoch": 16.515555555555554, "grad_norm": 0.23622538149356842, "learning_rate": 4.5791632925558944e-05, "loss": 0.0208, "step": 18580 }, { "epoch": 16.524444444444445, "grad_norm": 0.2696390151977539, "learning_rate": 4.574281633448448e-05, "loss": 0.0165, "step": 18590 }, { "epoch": 16.533333333333335, "grad_norm": 0.20963995158672333, "learning_rate": 4.5694003830751085e-05, "loss": 0.0124, "step": 18600 }, { "epoch": 16.54222222222222, "grad_norm": 0.2755920886993408, "learning_rate": 4.564519546122385e-05, "loss": 0.0208, "step": 18610 }, { "epoch": 16.551111111111112, "grad_norm": 0.23028627038002014, "learning_rate": 4.55963912727639e-05, "loss": 0.0175, "step": 18620 }, { "epoch": 16.56, "grad_norm": 0.27477970719337463, "learning_rate": 4.554759131222838e-05, "loss": 0.0198, "step": 18630 }, { "epoch": 16.56888888888889, "grad_norm": 0.2975420355796814, "learning_rate": 4.549879562647032e-05, "loss": 0.0188, "step": 18640 }, { "epoch": 16.57777777777778, "grad_norm": 0.20530802011489868, "learning_rate": 4.545000426233871e-05, "loss": 0.0156, "step": 18650 }, { "epoch": 16.586666666666666, "grad_norm": 0.29912373423576355, "learning_rate": 4.5401217266678343e-05, "loss": 0.0177, "step": 18660 }, { "epoch": 16.595555555555556, "grad_norm": 0.29263103008270264, "learning_rate": 4.5352434686329814e-05, "loss": 0.0189, "step": 18670 }, { "epoch": 16.604444444444443, "grad_norm": 0.27585551142692566, "learning_rate": 4.5303656568129525e-05, "loss": 0.0167, "step": 18680 }, { "epoch": 16.613333333333333, "grad_norm": 0.31042736768722534, "learning_rate": 4.525488295890954e-05, "loss": 0.0185, "step": 18690 }, { "epoch": 16.622222222222224, "grad_norm": 0.2965238690376282, "learning_rate": 4.5206113905497665e-05, "loss": 0.0197, "step": 18700 }, { "epoch": 16.63111111111111, "grad_norm": 0.2616423964500427, "learning_rate": 4.515734945471722e-05, "loss": 0.018, "step": 18710 }, { "epoch": 16.64, "grad_norm": 0.25129392743110657, "learning_rate": 4.510858965338721e-05, "loss": 0.019, "step": 18720 }, { "epoch": 16.648888888888887, "grad_norm": 0.2698293626308441, "learning_rate": 4.505983454832212e-05, "loss": 0.0207, "step": 18730 }, { "epoch": 16.657777777777778, "grad_norm": 0.2511175870895386, "learning_rate": 4.501108418633197e-05, "loss": 0.0185, "step": 18740 }, { "epoch": 16.666666666666668, "grad_norm": 0.2589429020881653, "learning_rate": 4.4962338614222186e-05, "loss": 0.0181, "step": 18750 }, { "epoch": 16.675555555555555, "grad_norm": 0.2538653016090393, "learning_rate": 4.49135978787936e-05, "loss": 0.0206, "step": 18760 }, { "epoch": 16.684444444444445, "grad_norm": 0.2975861132144928, "learning_rate": 4.4864862026842434e-05, "loss": 0.0163, "step": 18770 }, { "epoch": 16.693333333333335, "grad_norm": 0.2825506627559662, "learning_rate": 4.481613110516017e-05, "loss": 0.0184, "step": 18780 }, { "epoch": 16.702222222222222, "grad_norm": 0.2141445279121399, "learning_rate": 4.476740516053361e-05, "loss": 0.0218, "step": 18790 }, { "epoch": 16.711111111111112, "grad_norm": 0.22375744581222534, "learning_rate": 4.471868423974473e-05, "loss": 0.0146, "step": 18800 }, { "epoch": 16.72, "grad_norm": 0.28977879881858826, "learning_rate": 4.466996838957074e-05, "loss": 0.017, "step": 18810 }, { "epoch": 16.72888888888889, "grad_norm": 0.22638791799545288, "learning_rate": 4.462125765678388e-05, "loss": 0.0148, "step": 18820 }, { "epoch": 16.73777777777778, "grad_norm": 0.2737583816051483, "learning_rate": 4.4572552088151596e-05, "loss": 0.018, "step": 18830 }, { "epoch": 16.746666666666666, "grad_norm": 0.22335977852344513, "learning_rate": 4.452385173043631e-05, "loss": 0.0177, "step": 18840 }, { "epoch": 16.755555555555556, "grad_norm": 0.23904503881931305, "learning_rate": 4.447515663039543e-05, "loss": 0.0183, "step": 18850 }, { "epoch": 16.764444444444443, "grad_norm": 0.265330046415329, "learning_rate": 4.442646683478138e-05, "loss": 0.019, "step": 18860 }, { "epoch": 16.773333333333333, "grad_norm": 0.27706441283226013, "learning_rate": 4.437778239034141e-05, "loss": 0.019, "step": 18870 }, { "epoch": 16.782222222222224, "grad_norm": 0.211493581533432, "learning_rate": 4.4329103343817704e-05, "loss": 0.0156, "step": 18880 }, { "epoch": 16.79111111111111, "grad_norm": 0.29014480113983154, "learning_rate": 4.428042974194722e-05, "loss": 0.0244, "step": 18890 }, { "epoch": 16.8, "grad_norm": 0.2628314793109894, "learning_rate": 4.4231761631461685e-05, "loss": 0.0149, "step": 18900 }, { "epoch": 16.808888888888887, "grad_norm": 0.23220784962177277, "learning_rate": 4.418309905908761e-05, "loss": 0.0222, "step": 18910 }, { "epoch": 16.817777777777778, "grad_norm": 0.22730910778045654, "learning_rate": 4.4134442071546105e-05, "loss": 0.0193, "step": 18920 }, { "epoch": 16.826666666666668, "grad_norm": 0.16602881252765656, "learning_rate": 4.4085790715552986e-05, "loss": 0.0218, "step": 18930 }, { "epoch": 16.835555555555555, "grad_norm": 0.31359031796455383, "learning_rate": 4.4037145037818614e-05, "loss": 0.0172, "step": 18940 }, { "epoch": 16.844444444444445, "grad_norm": 0.3290869891643524, "learning_rate": 4.398850508504796e-05, "loss": 0.0182, "step": 18950 }, { "epoch": 16.85333333333333, "grad_norm": 0.19753104448318481, "learning_rate": 4.393987090394041e-05, "loss": 0.0155, "step": 18960 }, { "epoch": 16.862222222222222, "grad_norm": 0.23446987569332123, "learning_rate": 4.389124254118991e-05, "loss": 0.0184, "step": 18970 }, { "epoch": 16.871111111111112, "grad_norm": 0.29451844096183777, "learning_rate": 4.384262004348474e-05, "loss": 0.0154, "step": 18980 }, { "epoch": 16.88, "grad_norm": 0.22823651134967804, "learning_rate": 4.379400345750756e-05, "loss": 0.0158, "step": 18990 }, { "epoch": 16.88888888888889, "grad_norm": 0.2912134528160095, "learning_rate": 4.374539282993542e-05, "loss": 0.0137, "step": 19000 }, { "epoch": 16.897777777777776, "grad_norm": 0.17657889425754547, "learning_rate": 4.369678820743958e-05, "loss": 0.0164, "step": 19010 }, { "epoch": 16.906666666666666, "grad_norm": 0.23524267971515656, "learning_rate": 4.364818963668554e-05, "loss": 0.0218, "step": 19020 }, { "epoch": 16.915555555555557, "grad_norm": 0.19213111698627472, "learning_rate": 4.3599597164333e-05, "loss": 0.016, "step": 19030 }, { "epoch": 16.924444444444443, "grad_norm": 0.2242296189069748, "learning_rate": 4.3551010837035835e-05, "loss": 0.0186, "step": 19040 }, { "epoch": 16.933333333333334, "grad_norm": 0.28421539068222046, "learning_rate": 4.350243070144198e-05, "loss": 0.0149, "step": 19050 }, { "epoch": 16.942222222222224, "grad_norm": 0.18980461359024048, "learning_rate": 4.345385680419345e-05, "loss": 0.0187, "step": 19060 }, { "epoch": 16.95111111111111, "grad_norm": 0.19074735045433044, "learning_rate": 4.3405289191926254e-05, "loss": 0.0187, "step": 19070 }, { "epoch": 16.96, "grad_norm": 0.25161054730415344, "learning_rate": 4.335672791127036e-05, "loss": 0.0166, "step": 19080 }, { "epoch": 16.968888888888888, "grad_norm": 0.19965551793575287, "learning_rate": 4.3308173008849705e-05, "loss": 0.0176, "step": 19090 }, { "epoch": 16.977777777777778, "grad_norm": 0.23593704402446747, "learning_rate": 4.325962453128204e-05, "loss": 0.0176, "step": 19100 }, { "epoch": 16.986666666666668, "grad_norm": 0.3444717824459076, "learning_rate": 4.3211082525179e-05, "loss": 0.0141, "step": 19110 }, { "epoch": 16.995555555555555, "grad_norm": 0.2291611135005951, "learning_rate": 4.3162547037145946e-05, "loss": 0.0168, "step": 19120 }, { "epoch": 17.004444444444445, "grad_norm": 0.2870919704437256, "learning_rate": 4.311401811378204e-05, "loss": 0.0161, "step": 19130 }, { "epoch": 17.013333333333332, "grad_norm": 0.26940906047821045, "learning_rate": 4.306549580168011e-05, "loss": 0.0181, "step": 19140 }, { "epoch": 17.022222222222222, "grad_norm": 0.22763362526893616, "learning_rate": 4.301698014742666e-05, "loss": 0.0175, "step": 19150 }, { "epoch": 17.031111111111112, "grad_norm": 0.27467110753059387, "learning_rate": 4.296847119760177e-05, "loss": 0.0205, "step": 19160 }, { "epoch": 17.04, "grad_norm": 0.167091965675354, "learning_rate": 4.291996899877909e-05, "loss": 0.0172, "step": 19170 }, { "epoch": 17.04888888888889, "grad_norm": 0.19951722025871277, "learning_rate": 4.287147359752584e-05, "loss": 0.0216, "step": 19180 }, { "epoch": 17.057777777777776, "grad_norm": 0.24038371443748474, "learning_rate": 4.2822985040402617e-05, "loss": 0.0189, "step": 19190 }, { "epoch": 17.066666666666666, "grad_norm": 0.23757264018058777, "learning_rate": 4.2774503373963545e-05, "loss": 0.0175, "step": 19200 }, { "epoch": 17.075555555555557, "grad_norm": 0.22958676517009735, "learning_rate": 4.272602864475608e-05, "loss": 0.0214, "step": 19210 }, { "epoch": 17.084444444444443, "grad_norm": 0.2324765920639038, "learning_rate": 4.2677560899321026e-05, "loss": 0.0152, "step": 19220 }, { "epoch": 17.093333333333334, "grad_norm": 0.29394227266311646, "learning_rate": 4.2629100184192446e-05, "loss": 0.0163, "step": 19230 }, { "epoch": 17.102222222222224, "grad_norm": 0.24878427386283875, "learning_rate": 4.2580646545897735e-05, "loss": 0.0173, "step": 19240 }, { "epoch": 17.11111111111111, "grad_norm": 0.1672867387533188, "learning_rate": 4.253220003095743e-05, "loss": 0.0174, "step": 19250 }, { "epoch": 17.12, "grad_norm": 0.26902663707733154, "learning_rate": 4.248376068588523e-05, "loss": 0.019, "step": 19260 }, { "epoch": 17.128888888888888, "grad_norm": 0.18384207785129547, "learning_rate": 4.2435328557188e-05, "loss": 0.0162, "step": 19270 }, { "epoch": 17.137777777777778, "grad_norm": 0.22755607962608337, "learning_rate": 4.238690369136561e-05, "loss": 0.019, "step": 19280 }, { "epoch": 17.14666666666667, "grad_norm": 0.2379108965396881, "learning_rate": 4.233848613491103e-05, "loss": 0.0197, "step": 19290 }, { "epoch": 17.155555555555555, "grad_norm": 0.23577594757080078, "learning_rate": 4.229007593431013e-05, "loss": 0.0121, "step": 19300 }, { "epoch": 17.164444444444445, "grad_norm": 0.43694958090782166, "learning_rate": 4.2241673136041784e-05, "loss": 0.023, "step": 19310 }, { "epoch": 17.173333333333332, "grad_norm": 0.17972497642040253, "learning_rate": 4.2193277786577744e-05, "loss": 0.0161, "step": 19320 }, { "epoch": 17.182222222222222, "grad_norm": 0.24288970232009888, "learning_rate": 4.214488993238258e-05, "loss": 0.0158, "step": 19330 }, { "epoch": 17.191111111111113, "grad_norm": 0.20906516909599304, "learning_rate": 4.209650961991369e-05, "loss": 0.0184, "step": 19340 }, { "epoch": 17.2, "grad_norm": 0.26815879344940186, "learning_rate": 4.2048136895621224e-05, "loss": 0.0194, "step": 19350 }, { "epoch": 17.20888888888889, "grad_norm": 0.22024215757846832, "learning_rate": 4.199977180594807e-05, "loss": 0.0144, "step": 19360 }, { "epoch": 17.217777777777776, "grad_norm": 0.28244853019714355, "learning_rate": 4.195141439732975e-05, "loss": 0.0208, "step": 19370 }, { "epoch": 17.226666666666667, "grad_norm": 0.17130273580551147, "learning_rate": 4.190306471619444e-05, "loss": 0.0111, "step": 19380 }, { "epoch": 17.235555555555557, "grad_norm": 0.24296534061431885, "learning_rate": 4.185472280896288e-05, "loss": 0.0159, "step": 19390 }, { "epoch": 17.244444444444444, "grad_norm": 0.230843186378479, "learning_rate": 4.180638872204834e-05, "loss": 0.0156, "step": 19400 }, { "epoch": 17.253333333333334, "grad_norm": 0.21298737823963165, "learning_rate": 4.175806250185661e-05, "loss": 0.0169, "step": 19410 }, { "epoch": 17.26222222222222, "grad_norm": 0.2468118518590927, "learning_rate": 4.1709744194785924e-05, "loss": 0.0182, "step": 19420 }, { "epoch": 17.27111111111111, "grad_norm": 0.4959006905555725, "learning_rate": 4.166143384722686e-05, "loss": 0.0182, "step": 19430 }, { "epoch": 17.28, "grad_norm": 0.19621290266513824, "learning_rate": 4.161313150556242e-05, "loss": 0.0156, "step": 19440 }, { "epoch": 17.288888888888888, "grad_norm": 0.20804475247859955, "learning_rate": 4.1564837216167904e-05, "loss": 0.0159, "step": 19450 }, { "epoch": 17.297777777777778, "grad_norm": 0.148358553647995, "learning_rate": 4.151655102541086e-05, "loss": 0.0192, "step": 19460 }, { "epoch": 17.306666666666665, "grad_norm": 0.2910904288291931, "learning_rate": 4.14682729796511e-05, "loss": 0.0136, "step": 19470 }, { "epoch": 17.315555555555555, "grad_norm": 0.202024906873703, "learning_rate": 4.142000312524058e-05, "loss": 0.0142, "step": 19480 }, { "epoch": 17.324444444444445, "grad_norm": 0.26883137226104736, "learning_rate": 4.137174150852339e-05, "loss": 0.0141, "step": 19490 }, { "epoch": 17.333333333333332, "grad_norm": 0.22418399155139923, "learning_rate": 4.1323488175835745e-05, "loss": 0.0152, "step": 19500 }, { "epoch": 17.342222222222222, "grad_norm": 0.17172867059707642, "learning_rate": 4.127524317350586e-05, "loss": 0.013, "step": 19510 }, { "epoch": 17.351111111111113, "grad_norm": 0.3993489444255829, "learning_rate": 4.122700654785401e-05, "loss": 0.018, "step": 19520 }, { "epoch": 17.36, "grad_norm": 0.26194262504577637, "learning_rate": 4.117877834519239e-05, "loss": 0.0151, "step": 19530 }, { "epoch": 17.36888888888889, "grad_norm": 0.2558589577674866, "learning_rate": 4.1130558611825085e-05, "loss": 0.0206, "step": 19540 }, { "epoch": 17.377777777777776, "grad_norm": 0.34758901596069336, "learning_rate": 4.108234739404808e-05, "loss": 0.0179, "step": 19550 }, { "epoch": 17.386666666666667, "grad_norm": 0.1877313107252121, "learning_rate": 4.103414473814921e-05, "loss": 0.0181, "step": 19560 }, { "epoch": 17.395555555555557, "grad_norm": 0.2814737856388092, "learning_rate": 4.098595069040803e-05, "loss": 0.0147, "step": 19570 }, { "epoch": 17.404444444444444, "grad_norm": 0.33937332034111023, "learning_rate": 4.0937765297095875e-05, "loss": 0.0157, "step": 19580 }, { "epoch": 17.413333333333334, "grad_norm": 0.30933037400245667, "learning_rate": 4.088958860447576e-05, "loss": 0.0156, "step": 19590 }, { "epoch": 17.42222222222222, "grad_norm": 0.26430565118789673, "learning_rate": 4.0841420658802324e-05, "loss": 0.0158, "step": 19600 }, { "epoch": 17.43111111111111, "grad_norm": 0.2423197627067566, "learning_rate": 4.079326150632185e-05, "loss": 0.0149, "step": 19610 }, { "epoch": 17.44, "grad_norm": 0.22544726729393005, "learning_rate": 4.0745111193272164e-05, "loss": 0.0198, "step": 19620 }, { "epoch": 17.448888888888888, "grad_norm": 0.2640097141265869, "learning_rate": 4.069696976588258e-05, "loss": 0.0148, "step": 19630 }, { "epoch": 17.45777777777778, "grad_norm": 0.249282568693161, "learning_rate": 4.064883727037386e-05, "loss": 0.0154, "step": 19640 }, { "epoch": 17.466666666666665, "grad_norm": 0.2649959623813629, "learning_rate": 4.060071375295829e-05, "loss": 0.016, "step": 19650 }, { "epoch": 17.475555555555555, "grad_norm": 0.31242647767066956, "learning_rate": 4.0552599259839445e-05, "loss": 0.0183, "step": 19660 }, { "epoch": 17.484444444444446, "grad_norm": 0.19215801358222961, "learning_rate": 4.0504493837212257e-05, "loss": 0.0179, "step": 19670 }, { "epoch": 17.493333333333332, "grad_norm": 0.1915053129196167, "learning_rate": 4.045639753126299e-05, "loss": 0.0139, "step": 19680 }, { "epoch": 17.502222222222223, "grad_norm": 0.2576773464679718, "learning_rate": 4.0408310388169075e-05, "loss": 0.0156, "step": 19690 }, { "epoch": 17.511111111111113, "grad_norm": 0.20673973858356476, "learning_rate": 4.0360232454099236e-05, "loss": 0.0182, "step": 19700 }, { "epoch": 17.52, "grad_norm": 0.14759942889213562, "learning_rate": 4.03121637752133e-05, "loss": 0.018, "step": 19710 }, { "epoch": 17.52888888888889, "grad_norm": 0.21804289519786835, "learning_rate": 4.026410439766219e-05, "loss": 0.0165, "step": 19720 }, { "epoch": 17.537777777777777, "grad_norm": 0.2483036369085312, "learning_rate": 4.021605436758799e-05, "loss": 0.0163, "step": 19730 }, { "epoch": 17.546666666666667, "grad_norm": 0.18946467339992523, "learning_rate": 4.01680137311237e-05, "loss": 0.0144, "step": 19740 }, { "epoch": 17.555555555555557, "grad_norm": 0.3034687638282776, "learning_rate": 4.0119982534393364e-05, "loss": 0.0202, "step": 19750 }, { "epoch": 17.564444444444444, "grad_norm": 0.25473910570144653, "learning_rate": 4.0071960823511936e-05, "loss": 0.0151, "step": 19760 }, { "epoch": 17.573333333333334, "grad_norm": 0.3094094395637512, "learning_rate": 4.0023948644585296e-05, "loss": 0.018, "step": 19770 }, { "epoch": 17.58222222222222, "grad_norm": 0.21697048842906952, "learning_rate": 3.997594604371013e-05, "loss": 0.0158, "step": 19780 }, { "epoch": 17.59111111111111, "grad_norm": 0.21185240149497986, "learning_rate": 3.992795306697398e-05, "loss": 0.0159, "step": 19790 }, { "epoch": 17.6, "grad_norm": 0.17913506925106049, "learning_rate": 3.987996976045509e-05, "loss": 0.016, "step": 19800 }, { "epoch": 17.608888888888888, "grad_norm": 0.44323936104774475, "learning_rate": 3.9831996170222455e-05, "loss": 0.0203, "step": 19810 }, { "epoch": 17.61777777777778, "grad_norm": 0.2714117765426636, "learning_rate": 3.9784032342335736e-05, "loss": 0.0126, "step": 19820 }, { "epoch": 17.626666666666665, "grad_norm": 0.18773552775382996, "learning_rate": 3.973607832284521e-05, "loss": 0.0185, "step": 19830 }, { "epoch": 17.635555555555555, "grad_norm": 0.2256433665752411, "learning_rate": 3.968813415779179e-05, "loss": 0.0149, "step": 19840 }, { "epoch": 17.644444444444446, "grad_norm": 0.3065063953399658, "learning_rate": 3.964019989320681e-05, "loss": 0.0135, "step": 19850 }, { "epoch": 17.653333333333332, "grad_norm": 0.1348503679037094, "learning_rate": 3.959227557511222e-05, "loss": 0.0138, "step": 19860 }, { "epoch": 17.662222222222223, "grad_norm": 0.2561182379722595, "learning_rate": 3.954436124952035e-05, "loss": 0.014, "step": 19870 }, { "epoch": 17.67111111111111, "grad_norm": 0.2547704577445984, "learning_rate": 3.949645696243397e-05, "loss": 0.02, "step": 19880 }, { "epoch": 17.68, "grad_norm": 0.2929536998271942, "learning_rate": 3.944856275984621e-05, "loss": 0.015, "step": 19890 }, { "epoch": 17.68888888888889, "grad_norm": 0.27362969517707825, "learning_rate": 3.940067868774048e-05, "loss": 0.0178, "step": 19900 }, { "epoch": 17.697777777777777, "grad_norm": 0.1911071538925171, "learning_rate": 3.935280479209051e-05, "loss": 0.0169, "step": 19910 }, { "epoch": 17.706666666666667, "grad_norm": 0.19461852312088013, "learning_rate": 3.930494111886024e-05, "loss": 0.0167, "step": 19920 }, { "epoch": 17.715555555555554, "grad_norm": 0.28858500719070435, "learning_rate": 3.92570877140038e-05, "loss": 0.0168, "step": 19930 }, { "epoch": 17.724444444444444, "grad_norm": 0.2881103456020355, "learning_rate": 3.920924462346546e-05, "loss": 0.0196, "step": 19940 }, { "epoch": 17.733333333333334, "grad_norm": 0.20811952650547028, "learning_rate": 3.916141189317958e-05, "loss": 0.0145, "step": 19950 }, { "epoch": 17.74222222222222, "grad_norm": 0.24481631815433502, "learning_rate": 3.9113589569070555e-05, "loss": 0.016, "step": 19960 }, { "epoch": 17.75111111111111, "grad_norm": 0.19747447967529297, "learning_rate": 3.906577769705284e-05, "loss": 0.0168, "step": 19970 }, { "epoch": 17.76, "grad_norm": 0.183800607919693, "learning_rate": 3.9017976323030826e-05, "loss": 0.0168, "step": 19980 }, { "epoch": 17.76888888888889, "grad_norm": 0.17603620886802673, "learning_rate": 3.8970185492898806e-05, "loss": 0.0148, "step": 19990 }, { "epoch": 17.77777777777778, "grad_norm": 0.19978424906730652, "learning_rate": 3.8922405252541e-05, "loss": 0.0134, "step": 20000 }, { "epoch": 17.786666666666665, "grad_norm": 0.23738978803157806, "learning_rate": 3.88746356478314e-05, "loss": 0.0183, "step": 20010 }, { "epoch": 17.795555555555556, "grad_norm": 0.19088678061962128, "learning_rate": 3.882687672463384e-05, "loss": 0.0174, "step": 20020 }, { "epoch": 17.804444444444446, "grad_norm": 0.24212780594825745, "learning_rate": 3.877912852880187e-05, "loss": 0.0225, "step": 20030 }, { "epoch": 17.813333333333333, "grad_norm": 0.21818268299102783, "learning_rate": 3.8731391106178746e-05, "loss": 0.0217, "step": 20040 }, { "epoch": 17.822222222222223, "grad_norm": 0.19892112910747528, "learning_rate": 3.868366450259736e-05, "loss": 0.0189, "step": 20050 }, { "epoch": 17.83111111111111, "grad_norm": 0.23184219002723694, "learning_rate": 3.863594876388027e-05, "loss": 0.014, "step": 20060 }, { "epoch": 17.84, "grad_norm": 0.3602442741394043, "learning_rate": 3.858824393583955e-05, "loss": 0.0148, "step": 20070 }, { "epoch": 17.84888888888889, "grad_norm": 0.22212740778923035, "learning_rate": 3.8540550064276816e-05, "loss": 0.0164, "step": 20080 }, { "epoch": 17.857777777777777, "grad_norm": 0.22164982557296753, "learning_rate": 3.849286719498318e-05, "loss": 0.0181, "step": 20090 }, { "epoch": 17.866666666666667, "grad_norm": 0.3490807116031647, "learning_rate": 3.844519537373916e-05, "loss": 0.0147, "step": 20100 }, { "epoch": 17.875555555555554, "grad_norm": 0.21686214208602905, "learning_rate": 3.8397534646314713e-05, "loss": 0.0178, "step": 20110 }, { "epoch": 17.884444444444444, "grad_norm": 0.3138503134250641, "learning_rate": 3.83498850584691e-05, "loss": 0.0215, "step": 20120 }, { "epoch": 17.893333333333334, "grad_norm": 0.34311342239379883, "learning_rate": 3.830224665595089e-05, "loss": 0.0146, "step": 20130 }, { "epoch": 17.90222222222222, "grad_norm": 0.2814919054508209, "learning_rate": 3.8254619484497963e-05, "loss": 0.0131, "step": 20140 }, { "epoch": 17.91111111111111, "grad_norm": 0.2135837823152542, "learning_rate": 3.820700358983734e-05, "loss": 0.0181, "step": 20150 }, { "epoch": 17.92, "grad_norm": 0.19342736899852753, "learning_rate": 3.815939901768527e-05, "loss": 0.0195, "step": 20160 }, { "epoch": 17.92888888888889, "grad_norm": 0.21394550800323486, "learning_rate": 3.811180581374712e-05, "loss": 0.0151, "step": 20170 }, { "epoch": 17.93777777777778, "grad_norm": 0.2759093940258026, "learning_rate": 3.806422402371733e-05, "loss": 0.0174, "step": 20180 }, { "epoch": 17.946666666666665, "grad_norm": 0.3072455823421478, "learning_rate": 3.801665369327938e-05, "loss": 0.0173, "step": 20190 }, { "epoch": 17.955555555555556, "grad_norm": 0.19053880870342255, "learning_rate": 3.7969094868105784e-05, "loss": 0.013, "step": 20200 }, { "epoch": 17.964444444444446, "grad_norm": 0.3663347363471985, "learning_rate": 3.792154759385797e-05, "loss": 0.0191, "step": 20210 }, { "epoch": 17.973333333333333, "grad_norm": 0.17502184212207794, "learning_rate": 3.787401191618627e-05, "loss": 0.0138, "step": 20220 }, { "epoch": 17.982222222222223, "grad_norm": 0.2710953950881958, "learning_rate": 3.7826487880729936e-05, "loss": 0.0142, "step": 20230 }, { "epoch": 17.99111111111111, "grad_norm": 0.2526054382324219, "learning_rate": 3.777897553311697e-05, "loss": 0.0146, "step": 20240 }, { "epoch": 18.0, "grad_norm": 0.5866895318031311, "learning_rate": 3.773147491896423e-05, "loss": 0.0242, "step": 20250 }, { "epoch": 18.00888888888889, "grad_norm": 0.29626351594924927, "learning_rate": 3.7683986083877236e-05, "loss": 0.0151, "step": 20260 }, { "epoch": 18.017777777777777, "grad_norm": 0.27652472257614136, "learning_rate": 3.763650907345024e-05, "loss": 0.0136, "step": 20270 }, { "epoch": 18.026666666666667, "grad_norm": 0.2301950305700302, "learning_rate": 3.758904393326612e-05, "loss": 0.0186, "step": 20280 }, { "epoch": 18.035555555555554, "grad_norm": 0.3296969532966614, "learning_rate": 3.754159070889639e-05, "loss": 0.0157, "step": 20290 }, { "epoch": 18.044444444444444, "grad_norm": 0.2893105447292328, "learning_rate": 3.7494149445901086e-05, "loss": 0.0214, "step": 20300 }, { "epoch": 18.053333333333335, "grad_norm": 0.2823111116886139, "learning_rate": 3.744672018982879e-05, "loss": 0.0177, "step": 20310 }, { "epoch": 18.06222222222222, "grad_norm": 0.22126393020153046, "learning_rate": 3.7399302986216546e-05, "loss": 0.0245, "step": 20320 }, { "epoch": 18.07111111111111, "grad_norm": 0.24213217198848724, "learning_rate": 3.735189788058979e-05, "loss": 0.0146, "step": 20330 }, { "epoch": 18.08, "grad_norm": 0.32339155673980713, "learning_rate": 3.730450491846242e-05, "loss": 0.0151, "step": 20340 }, { "epoch": 18.08888888888889, "grad_norm": 0.3178067207336426, "learning_rate": 3.725712414533662e-05, "loss": 0.0159, "step": 20350 }, { "epoch": 18.09777777777778, "grad_norm": 0.27801278233528137, "learning_rate": 3.7209755606702854e-05, "loss": 0.0139, "step": 20360 }, { "epoch": 18.106666666666666, "grad_norm": 0.2628476321697235, "learning_rate": 3.716239934803989e-05, "loss": 0.0138, "step": 20370 }, { "epoch": 18.115555555555556, "grad_norm": 0.31558024883270264, "learning_rate": 3.7115055414814684e-05, "loss": 0.0176, "step": 20380 }, { "epoch": 18.124444444444446, "grad_norm": 0.37999242544174194, "learning_rate": 3.706772385248234e-05, "loss": 0.0123, "step": 20390 }, { "epoch": 18.133333333333333, "grad_norm": 0.3168180286884308, "learning_rate": 3.7020404706486125e-05, "loss": 0.0195, "step": 20400 }, { "epoch": 18.142222222222223, "grad_norm": 0.13712581992149353, "learning_rate": 3.697309802225735e-05, "loss": 0.0171, "step": 20410 }, { "epoch": 18.15111111111111, "grad_norm": 0.2009391486644745, "learning_rate": 3.692580384521538e-05, "loss": 0.0164, "step": 20420 }, { "epoch": 18.16, "grad_norm": 0.19886445999145508, "learning_rate": 3.687852222076757e-05, "loss": 0.0152, "step": 20430 }, { "epoch": 18.16888888888889, "grad_norm": 0.32020148634910583, "learning_rate": 3.683125319430919e-05, "loss": 0.0138, "step": 20440 }, { "epoch": 18.177777777777777, "grad_norm": 0.282205730676651, "learning_rate": 3.6783996811223506e-05, "loss": 0.0242, "step": 20450 }, { "epoch": 18.186666666666667, "grad_norm": 0.2888079285621643, "learning_rate": 3.673675311688151e-05, "loss": 0.0132, "step": 20460 }, { "epoch": 18.195555555555554, "grad_norm": 0.27866509556770325, "learning_rate": 3.668952215664212e-05, "loss": 0.0168, "step": 20470 }, { "epoch": 18.204444444444444, "grad_norm": 0.25685665011405945, "learning_rate": 3.664230397585199e-05, "loss": 0.0146, "step": 20480 }, { "epoch": 18.213333333333335, "grad_norm": 0.21396684646606445, "learning_rate": 3.659509861984551e-05, "loss": 0.0152, "step": 20490 }, { "epoch": 18.22222222222222, "grad_norm": 0.1760185807943344, "learning_rate": 3.654790613394474e-05, "loss": 0.0211, "step": 20500 }, { "epoch": 18.23111111111111, "grad_norm": 0.19985908269882202, "learning_rate": 3.650072656345941e-05, "loss": 0.0181, "step": 20510 }, { "epoch": 18.24, "grad_norm": 0.2069374918937683, "learning_rate": 3.645355995368683e-05, "loss": 0.0159, "step": 20520 }, { "epoch": 18.24888888888889, "grad_norm": 0.41712653636932373, "learning_rate": 3.6406406349911866e-05, "loss": 0.0193, "step": 20530 }, { "epoch": 18.25777777777778, "grad_norm": 0.23140794038772583, "learning_rate": 3.6359265797406913e-05, "loss": 0.0156, "step": 20540 }, { "epoch": 18.266666666666666, "grad_norm": 0.24597440659999847, "learning_rate": 3.631213834143184e-05, "loss": 0.0159, "step": 20550 }, { "epoch": 18.275555555555556, "grad_norm": 0.19862835109233856, "learning_rate": 3.626502402723392e-05, "loss": 0.0186, "step": 20560 }, { "epoch": 18.284444444444443, "grad_norm": 0.24405580759048462, "learning_rate": 3.621792290004778e-05, "loss": 0.0169, "step": 20570 }, { "epoch": 18.293333333333333, "grad_norm": 0.2961651086807251, "learning_rate": 3.6170835005095464e-05, "loss": 0.0168, "step": 20580 }, { "epoch": 18.302222222222223, "grad_norm": 0.23413386940956116, "learning_rate": 3.612376038758626e-05, "loss": 0.0165, "step": 20590 }, { "epoch": 18.31111111111111, "grad_norm": 0.1602816879749298, "learning_rate": 3.607669909271672e-05, "loss": 0.0133, "step": 20600 }, { "epoch": 18.32, "grad_norm": 0.1705757975578308, "learning_rate": 3.602965116567059e-05, "loss": 0.013, "step": 20610 }, { "epoch": 18.32888888888889, "grad_norm": 0.23042865097522736, "learning_rate": 3.598261665161881e-05, "loss": 0.0128, "step": 20620 }, { "epoch": 18.337777777777777, "grad_norm": 0.274274617433548, "learning_rate": 3.593559559571942e-05, "loss": 0.0146, "step": 20630 }, { "epoch": 18.346666666666668, "grad_norm": 0.17341354489326477, "learning_rate": 3.5888588043117556e-05, "loss": 0.0136, "step": 20640 }, { "epoch": 18.355555555555554, "grad_norm": 0.32644253969192505, "learning_rate": 3.584159403894536e-05, "loss": 0.0162, "step": 20650 }, { "epoch": 18.364444444444445, "grad_norm": 0.2915499210357666, "learning_rate": 3.579461362832201e-05, "loss": 0.0168, "step": 20660 }, { "epoch": 18.373333333333335, "grad_norm": 0.2093273550271988, "learning_rate": 3.57476468563536e-05, "loss": 0.018, "step": 20670 }, { "epoch": 18.38222222222222, "grad_norm": 0.31164225935935974, "learning_rate": 3.5700693768133125e-05, "loss": 0.0135, "step": 20680 }, { "epoch": 18.391111111111112, "grad_norm": 0.23874767124652863, "learning_rate": 3.565375440874046e-05, "loss": 0.0158, "step": 20690 }, { "epoch": 18.4, "grad_norm": 0.2012593299150467, "learning_rate": 3.5606828823242295e-05, "loss": 0.0136, "step": 20700 }, { "epoch": 18.40888888888889, "grad_norm": 0.2596149444580078, "learning_rate": 3.5559917056692096e-05, "loss": 0.0171, "step": 20710 }, { "epoch": 18.41777777777778, "grad_norm": 0.4005165994167328, "learning_rate": 3.5513019154130054e-05, "loss": 0.0161, "step": 20720 }, { "epoch": 18.426666666666666, "grad_norm": 0.20327319204807281, "learning_rate": 3.546613516058308e-05, "loss": 0.0172, "step": 20730 }, { "epoch": 18.435555555555556, "grad_norm": 0.259213924407959, "learning_rate": 3.541926512106466e-05, "loss": 0.0138, "step": 20740 }, { "epoch": 18.444444444444443, "grad_norm": 0.2563748061656952, "learning_rate": 3.5372409080574976e-05, "loss": 0.0128, "step": 20750 }, { "epoch": 18.453333333333333, "grad_norm": 0.2942011058330536, "learning_rate": 3.532556708410071e-05, "loss": 0.0152, "step": 20760 }, { "epoch": 18.462222222222223, "grad_norm": 0.3980710208415985, "learning_rate": 3.527873917661504e-05, "loss": 0.0148, "step": 20770 }, { "epoch": 18.47111111111111, "grad_norm": 0.23921452462673187, "learning_rate": 3.5231925403077676e-05, "loss": 0.0149, "step": 20780 }, { "epoch": 18.48, "grad_norm": 0.3240032494068146, "learning_rate": 3.518512580843473e-05, "loss": 0.0167, "step": 20790 }, { "epoch": 18.488888888888887, "grad_norm": 0.2865414321422577, "learning_rate": 3.513834043761869e-05, "loss": 0.0187, "step": 20800 }, { "epoch": 18.497777777777777, "grad_norm": 0.2557905614376068, "learning_rate": 3.509156933554842e-05, "loss": 0.0201, "step": 20810 }, { "epoch": 18.506666666666668, "grad_norm": 0.17176192998886108, "learning_rate": 3.504481254712905e-05, "loss": 0.0159, "step": 20820 }, { "epoch": 18.515555555555554, "grad_norm": 0.3760502338409424, "learning_rate": 3.4998070117251977e-05, "loss": 0.0167, "step": 20830 }, { "epoch": 18.524444444444445, "grad_norm": 0.2348751276731491, "learning_rate": 3.495134209079483e-05, "loss": 0.0207, "step": 20840 }, { "epoch": 18.533333333333335, "grad_norm": 0.2428865283727646, "learning_rate": 3.4904628512621386e-05, "loss": 0.0175, "step": 20850 }, { "epoch": 18.54222222222222, "grad_norm": 0.32928889989852905, "learning_rate": 3.485792942758157e-05, "loss": 0.0136, "step": 20860 }, { "epoch": 18.551111111111112, "grad_norm": 0.2919588088989258, "learning_rate": 3.4811244880511386e-05, "loss": 0.0133, "step": 20870 }, { "epoch": 18.56, "grad_norm": 0.14602060616016388, "learning_rate": 3.4764574916232854e-05, "loss": 0.0102, "step": 20880 }, { "epoch": 18.56888888888889, "grad_norm": 0.22426171600818634, "learning_rate": 3.471791957955402e-05, "loss": 0.0164, "step": 20890 }, { "epoch": 18.57777777777778, "grad_norm": 0.253903329372406, "learning_rate": 3.46712789152689e-05, "loss": 0.0146, "step": 20900 }, { "epoch": 18.586666666666666, "grad_norm": 0.21845407783985138, "learning_rate": 3.462465296815739e-05, "loss": 0.0145, "step": 20910 }, { "epoch": 18.595555555555556, "grad_norm": 0.2430647611618042, "learning_rate": 3.457804178298527e-05, "loss": 0.0194, "step": 20920 }, { "epoch": 18.604444444444443, "grad_norm": 0.2104225903749466, "learning_rate": 3.453144540450416e-05, "loss": 0.0147, "step": 20930 }, { "epoch": 18.613333333333333, "grad_norm": 0.23042413592338562, "learning_rate": 3.448486387745142e-05, "loss": 0.0178, "step": 20940 }, { "epoch": 18.622222222222224, "grad_norm": 0.2755383849143982, "learning_rate": 3.443829724655023e-05, "loss": 0.0209, "step": 20950 }, { "epoch": 18.63111111111111, "grad_norm": 0.1877671778202057, "learning_rate": 3.439174555650938e-05, "loss": 0.0153, "step": 20960 }, { "epoch": 18.64, "grad_norm": 0.22751620411872864, "learning_rate": 3.4345208852023394e-05, "loss": 0.0153, "step": 20970 }, { "epoch": 18.648888888888887, "grad_norm": 0.3118877708911896, "learning_rate": 3.4298687177772316e-05, "loss": 0.0165, "step": 20980 }, { "epoch": 18.657777777777778, "grad_norm": 0.3294764459133148, "learning_rate": 3.4252180578421855e-05, "loss": 0.022, "step": 20990 }, { "epoch": 18.666666666666668, "grad_norm": 0.2234039455652237, "learning_rate": 3.42056890986232e-05, "loss": 0.0148, "step": 21000 }, { "epoch": 18.675555555555555, "grad_norm": 0.33817723393440247, "learning_rate": 3.415921278301301e-05, "loss": 0.0171, "step": 21010 }, { "epoch": 18.684444444444445, "grad_norm": 0.23793213069438934, "learning_rate": 3.4112751676213414e-05, "loss": 0.0161, "step": 21020 }, { "epoch": 18.693333333333335, "grad_norm": 0.23779965937137604, "learning_rate": 3.406630582283193e-05, "loss": 0.0219, "step": 21030 }, { "epoch": 18.702222222222222, "grad_norm": 0.2641121745109558, "learning_rate": 3.4019875267461435e-05, "loss": 0.0168, "step": 21040 }, { "epoch": 18.711111111111112, "grad_norm": 0.20964914560317993, "learning_rate": 3.39734600546801e-05, "loss": 0.014, "step": 21050 }, { "epoch": 18.72, "grad_norm": 0.20249123871326447, "learning_rate": 3.392706022905139e-05, "loss": 0.0139, "step": 21060 }, { "epoch": 18.72888888888889, "grad_norm": 0.3618154525756836, "learning_rate": 3.388067583512399e-05, "loss": 0.0209, "step": 21070 }, { "epoch": 18.73777777777778, "grad_norm": 0.35834965109825134, "learning_rate": 3.3834306917431745e-05, "loss": 0.0142, "step": 21080 }, { "epoch": 18.746666666666666, "grad_norm": 0.34253937005996704, "learning_rate": 3.378795352049367e-05, "loss": 0.0185, "step": 21090 }, { "epoch": 18.755555555555556, "grad_norm": 0.12523144483566284, "learning_rate": 3.374161568881386e-05, "loss": 0.0119, "step": 21100 }, { "epoch": 18.764444444444443, "grad_norm": 0.2933434545993805, "learning_rate": 3.369529346688148e-05, "loss": 0.0211, "step": 21110 }, { "epoch": 18.773333333333333, "grad_norm": 0.33856329321861267, "learning_rate": 3.36489868991707e-05, "loss": 0.0153, "step": 21120 }, { "epoch": 18.782222222222224, "grad_norm": 0.21965175867080688, "learning_rate": 3.3602696030140676e-05, "loss": 0.0139, "step": 21130 }, { "epoch": 18.79111111111111, "grad_norm": 0.27339357137680054, "learning_rate": 3.355642090423547e-05, "loss": 0.0179, "step": 21140 }, { "epoch": 18.8, "grad_norm": 0.20692169666290283, "learning_rate": 3.3510161565884026e-05, "loss": 0.0161, "step": 21150 }, { "epoch": 18.808888888888887, "grad_norm": 0.16585177183151245, "learning_rate": 3.346391805950017e-05, "loss": 0.0204, "step": 21160 }, { "epoch": 18.817777777777778, "grad_norm": 0.27508988976478577, "learning_rate": 3.3417690429482475e-05, "loss": 0.0255, "step": 21170 }, { "epoch": 18.826666666666668, "grad_norm": 0.17188018560409546, "learning_rate": 3.33714787202143e-05, "loss": 0.0143, "step": 21180 }, { "epoch": 18.835555555555555, "grad_norm": 0.21730883419513702, "learning_rate": 3.3325282976063694e-05, "loss": 0.0152, "step": 21190 }, { "epoch": 18.844444444444445, "grad_norm": 0.29784971475601196, "learning_rate": 3.327910324138344e-05, "loss": 0.0154, "step": 21200 }, { "epoch": 18.85333333333333, "grad_norm": 0.258808434009552, "learning_rate": 3.323293956051086e-05, "loss": 0.0175, "step": 21210 }, { "epoch": 18.862222222222222, "grad_norm": 0.19778914749622345, "learning_rate": 3.318679197776795e-05, "loss": 0.0158, "step": 21220 }, { "epoch": 18.871111111111112, "grad_norm": 0.268688827753067, "learning_rate": 3.31406605374612e-05, "loss": 0.0162, "step": 21230 }, { "epoch": 18.88, "grad_norm": 0.2377559244632721, "learning_rate": 3.30945452838816e-05, "loss": 0.0138, "step": 21240 }, { "epoch": 18.88888888888889, "grad_norm": 0.24800534546375275, "learning_rate": 3.304844626130462e-05, "loss": 0.0215, "step": 21250 }, { "epoch": 18.897777777777776, "grad_norm": 0.22315804660320282, "learning_rate": 3.300236351399013e-05, "loss": 0.0163, "step": 21260 }, { "epoch": 18.906666666666666, "grad_norm": 0.24996261298656464, "learning_rate": 3.2956297086182404e-05, "loss": 0.0154, "step": 21270 }, { "epoch": 18.915555555555557, "grad_norm": 0.27651068568229675, "learning_rate": 3.291024702211001e-05, "loss": 0.0169, "step": 21280 }, { "epoch": 18.924444444444443, "grad_norm": 0.20950165390968323, "learning_rate": 3.286421336598582e-05, "loss": 0.0171, "step": 21290 }, { "epoch": 18.933333333333334, "grad_norm": 0.18638715147972107, "learning_rate": 3.281819616200693e-05, "loss": 0.0164, "step": 21300 }, { "epoch": 18.942222222222224, "grad_norm": 0.41163742542266846, "learning_rate": 3.2772195454354704e-05, "loss": 0.0197, "step": 21310 }, { "epoch": 18.95111111111111, "grad_norm": 0.2402658611536026, "learning_rate": 3.272621128719461e-05, "loss": 0.0165, "step": 21320 }, { "epoch": 18.96, "grad_norm": 0.25940030813217163, "learning_rate": 3.268024370467624e-05, "loss": 0.0163, "step": 21330 }, { "epoch": 18.968888888888888, "grad_norm": 0.18777190148830414, "learning_rate": 3.263429275093329e-05, "loss": 0.0158, "step": 21340 }, { "epoch": 18.977777777777778, "grad_norm": 0.20755887031555176, "learning_rate": 3.2588358470083465e-05, "loss": 0.0178, "step": 21350 }, { "epoch": 18.986666666666668, "grad_norm": 0.35836276412010193, "learning_rate": 3.2542440906228487e-05, "loss": 0.0194, "step": 21360 }, { "epoch": 18.995555555555555, "grad_norm": 0.26252681016921997, "learning_rate": 3.2496540103454015e-05, "loss": 0.0149, "step": 21370 }, { "epoch": 19.004444444444445, "grad_norm": 0.2612445652484894, "learning_rate": 3.2450656105829615e-05, "loss": 0.0169, "step": 21380 }, { "epoch": 19.013333333333332, "grad_norm": 0.24550341069698334, "learning_rate": 3.240478895740869e-05, "loss": 0.0141, "step": 21390 }, { "epoch": 19.022222222222222, "grad_norm": 0.33342504501342773, "learning_rate": 3.235893870222853e-05, "loss": 0.0194, "step": 21400 }, { "epoch": 19.031111111111112, "grad_norm": 0.23681075870990753, "learning_rate": 3.231310538431015e-05, "loss": 0.0143, "step": 21410 }, { "epoch": 19.04, "grad_norm": 0.2638393044471741, "learning_rate": 3.2267289047658314e-05, "loss": 0.0164, "step": 21420 }, { "epoch": 19.04888888888889, "grad_norm": 0.38853418827056885, "learning_rate": 3.222148973626152e-05, "loss": 0.0127, "step": 21430 }, { "epoch": 19.057777777777776, "grad_norm": 0.2528640031814575, "learning_rate": 3.217570749409187e-05, "loss": 0.0142, "step": 21440 }, { "epoch": 19.066666666666666, "grad_norm": 0.22114290297031403, "learning_rate": 3.2129942365105116e-05, "loss": 0.0137, "step": 21450 }, { "epoch": 19.075555555555557, "grad_norm": 0.20625513792037964, "learning_rate": 3.208419439324055e-05, "loss": 0.0144, "step": 21460 }, { "epoch": 19.084444444444443, "grad_norm": 0.1973508596420288, "learning_rate": 3.2038463622421006e-05, "loss": 0.0164, "step": 21470 }, { "epoch": 19.093333333333334, "grad_norm": 0.26147663593292236, "learning_rate": 3.199275009655282e-05, "loss": 0.0145, "step": 21480 }, { "epoch": 19.102222222222224, "grad_norm": 0.29475557804107666, "learning_rate": 3.1947053859525724e-05, "loss": 0.0171, "step": 21490 }, { "epoch": 19.11111111111111, "grad_norm": 0.22964932024478912, "learning_rate": 3.1901374955212886e-05, "loss": 0.0108, "step": 21500 }, { "epoch": 19.12, "grad_norm": 0.19219917058944702, "learning_rate": 3.1855713427470826e-05, "loss": 0.0154, "step": 21510 }, { "epoch": 19.128888888888888, "grad_norm": 0.24327412247657776, "learning_rate": 3.181006932013939e-05, "loss": 0.0179, "step": 21520 }, { "epoch": 19.137777777777778, "grad_norm": 0.2940564453601837, "learning_rate": 3.176444267704169e-05, "loss": 0.0139, "step": 21530 }, { "epoch": 19.14666666666667, "grad_norm": 0.1956951916217804, "learning_rate": 3.171883354198406e-05, "loss": 0.0162, "step": 21540 }, { "epoch": 19.155555555555555, "grad_norm": 0.20033742487430573, "learning_rate": 3.167324195875605e-05, "loss": 0.0134, "step": 21550 }, { "epoch": 19.164444444444445, "grad_norm": 0.1877865344285965, "learning_rate": 3.162766797113032e-05, "loss": 0.0179, "step": 21560 }, { "epoch": 19.173333333333332, "grad_norm": 0.20479543507099152, "learning_rate": 3.158211162286268e-05, "loss": 0.0182, "step": 21570 }, { "epoch": 19.182222222222222, "grad_norm": 0.20582018792629242, "learning_rate": 3.153657295769197e-05, "loss": 0.0143, "step": 21580 }, { "epoch": 19.191111111111113, "grad_norm": 0.20946747064590454, "learning_rate": 3.1491052019340096e-05, "loss": 0.0108, "step": 21590 }, { "epoch": 19.2, "grad_norm": 0.21208973228931427, "learning_rate": 3.1445548851511855e-05, "loss": 0.0116, "step": 21600 }, { "epoch": 19.20888888888889, "grad_norm": 0.2241937667131424, "learning_rate": 3.140006349789509e-05, "loss": 0.0161, "step": 21610 }, { "epoch": 19.217777777777776, "grad_norm": 0.23196172714233398, "learning_rate": 3.135459600216047e-05, "loss": 0.017, "step": 21620 }, { "epoch": 19.226666666666667, "grad_norm": 0.2635852098464966, "learning_rate": 3.130914640796157e-05, "loss": 0.0155, "step": 21630 }, { "epoch": 19.235555555555557, "grad_norm": 0.22888678312301636, "learning_rate": 3.126371475893471e-05, "loss": 0.0149, "step": 21640 }, { "epoch": 19.244444444444444, "grad_norm": 0.19230180978775024, "learning_rate": 3.1218301098699045e-05, "loss": 0.0176, "step": 21650 }, { "epoch": 19.253333333333334, "grad_norm": 0.31586953997612, "learning_rate": 3.1172905470856415e-05, "loss": 0.0134, "step": 21660 }, { "epoch": 19.26222222222222, "grad_norm": 0.27572736144065857, "learning_rate": 3.112752791899138e-05, "loss": 0.0135, "step": 21670 }, { "epoch": 19.27111111111111, "grad_norm": 0.23354366421699524, "learning_rate": 3.108216848667113e-05, "loss": 0.0129, "step": 21680 }, { "epoch": 19.28, "grad_norm": 0.23465348780155182, "learning_rate": 3.103682721744547e-05, "loss": 0.0168, "step": 21690 }, { "epoch": 19.288888888888888, "grad_norm": 0.22522558271884918, "learning_rate": 3.099150415484673e-05, "loss": 0.0177, "step": 21700 }, { "epoch": 19.297777777777778, "grad_norm": 0.19032734632492065, "learning_rate": 3.0946199342389775e-05, "loss": 0.0165, "step": 21710 }, { "epoch": 19.306666666666665, "grad_norm": 0.20450225472450256, "learning_rate": 3.090091282357199e-05, "loss": 0.0139, "step": 21720 }, { "epoch": 19.315555555555555, "grad_norm": 0.2969111204147339, "learning_rate": 3.085564464187316e-05, "loss": 0.0133, "step": 21730 }, { "epoch": 19.324444444444445, "grad_norm": 0.23028424382209778, "learning_rate": 3.081039484075544e-05, "loss": 0.0154, "step": 21740 }, { "epoch": 19.333333333333332, "grad_norm": 0.2666028141975403, "learning_rate": 3.07651634636634e-05, "loss": 0.0146, "step": 21750 }, { "epoch": 19.342222222222222, "grad_norm": 0.21642820537090302, "learning_rate": 3.071995055402387e-05, "loss": 0.0126, "step": 21760 }, { "epoch": 19.351111111111113, "grad_norm": 0.2581894099712372, "learning_rate": 3.067475615524597e-05, "loss": 0.0115, "step": 21770 }, { "epoch": 19.36, "grad_norm": 0.222438782453537, "learning_rate": 3.0629580310721054e-05, "loss": 0.0173, "step": 21780 }, { "epoch": 19.36888888888889, "grad_norm": 0.46006739139556885, "learning_rate": 3.058442306382264e-05, "loss": 0.017, "step": 21790 }, { "epoch": 19.377777777777776, "grad_norm": 0.3157570958137512, "learning_rate": 3.05392844579064e-05, "loss": 0.0159, "step": 21800 }, { "epoch": 19.386666666666667, "grad_norm": 0.13815897703170776, "learning_rate": 3.049416453631012e-05, "loss": 0.0136, "step": 21810 }, { "epoch": 19.395555555555557, "grad_norm": 0.27146387100219727, "learning_rate": 3.0449063342353633e-05, "loss": 0.0195, "step": 21820 }, { "epoch": 19.404444444444444, "grad_norm": 0.2210850715637207, "learning_rate": 3.0403980919338794e-05, "loss": 0.0171, "step": 21830 }, { "epoch": 19.413333333333334, "grad_norm": 0.24202103912830353, "learning_rate": 3.0358917310549446e-05, "loss": 0.015, "step": 21840 }, { "epoch": 19.42222222222222, "grad_norm": 0.20739643275737762, "learning_rate": 3.0313872559251357e-05, "loss": 0.0133, "step": 21850 }, { "epoch": 19.43111111111111, "grad_norm": 0.22456535696983337, "learning_rate": 3.02688467086922e-05, "loss": 0.0165, "step": 21860 }, { "epoch": 19.44, "grad_norm": 0.34055444598197937, "learning_rate": 3.022383980210149e-05, "loss": 0.0186, "step": 21870 }, { "epoch": 19.448888888888888, "grad_norm": 0.22631113231182098, "learning_rate": 3.017885188269055e-05, "loss": 0.0135, "step": 21880 }, { "epoch": 19.45777777777778, "grad_norm": 0.2928323447704315, "learning_rate": 3.0133882993652518e-05, "loss": 0.0184, "step": 21890 }, { "epoch": 19.466666666666665, "grad_norm": 0.19482943415641785, "learning_rate": 3.0088933178162182e-05, "loss": 0.0137, "step": 21900 }, { "epoch": 19.475555555555555, "grad_norm": 0.2033020257949829, "learning_rate": 3.0044002479376075e-05, "loss": 0.0114, "step": 21910 }, { "epoch": 19.484444444444446, "grad_norm": 0.2664470970630646, "learning_rate": 2.999909094043237e-05, "loss": 0.0143, "step": 21920 }, { "epoch": 19.493333333333332, "grad_norm": 0.29129257798194885, "learning_rate": 2.995419860445084e-05, "loss": 0.017, "step": 21930 }, { "epoch": 19.502222222222223, "grad_norm": 0.2426452785730362, "learning_rate": 2.9909325514532793e-05, "loss": 0.0161, "step": 21940 }, { "epoch": 19.511111111111113, "grad_norm": 0.21799777448177338, "learning_rate": 2.986447171376111e-05, "loss": 0.016, "step": 21950 }, { "epoch": 19.52, "grad_norm": 0.1652263104915619, "learning_rate": 2.9819637245200117e-05, "loss": 0.0192, "step": 21960 }, { "epoch": 19.52888888888889, "grad_norm": 0.19840271770954132, "learning_rate": 2.977482215189557e-05, "loss": 0.0141, "step": 21970 }, { "epoch": 19.537777777777777, "grad_norm": 0.28828850388526917, "learning_rate": 2.9730026476874673e-05, "loss": 0.0211, "step": 21980 }, { "epoch": 19.546666666666667, "grad_norm": 0.2745564579963684, "learning_rate": 2.9685250263145918e-05, "loss": 0.0185, "step": 21990 }, { "epoch": 19.555555555555557, "grad_norm": 0.17435051500797272, "learning_rate": 2.964049355369919e-05, "loss": 0.0107, "step": 22000 }, { "epoch": 19.564444444444444, "grad_norm": 0.20065705478191376, "learning_rate": 2.9595756391505576e-05, "loss": 0.0189, "step": 22010 }, { "epoch": 19.573333333333334, "grad_norm": 0.2029644250869751, "learning_rate": 2.955103881951743e-05, "loss": 0.0141, "step": 22020 }, { "epoch": 19.58222222222222, "grad_norm": 0.22097273170948029, "learning_rate": 2.9506340880668282e-05, "loss": 0.0116, "step": 22030 }, { "epoch": 19.59111111111111, "grad_norm": 0.2499135434627533, "learning_rate": 2.9461662617872844e-05, "loss": 0.0171, "step": 22040 }, { "epoch": 19.6, "grad_norm": 0.22455735504627228, "learning_rate": 2.94170040740269e-05, "loss": 0.0162, "step": 22050 }, { "epoch": 19.608888888888888, "grad_norm": 0.32814979553222656, "learning_rate": 2.9372365292007332e-05, "loss": 0.0151, "step": 22060 }, { "epoch": 19.61777777777778, "grad_norm": 0.2888753116130829, "learning_rate": 2.932774631467202e-05, "loss": 0.0156, "step": 22070 }, { "epoch": 19.626666666666665, "grad_norm": 0.23517656326293945, "learning_rate": 2.9283147184859848e-05, "loss": 0.0113, "step": 22080 }, { "epoch": 19.635555555555555, "grad_norm": 0.2535247504711151, "learning_rate": 2.9238567945390648e-05, "loss": 0.0138, "step": 22090 }, { "epoch": 19.644444444444446, "grad_norm": 0.17783501744270325, "learning_rate": 2.9194008639065153e-05, "loss": 0.0108, "step": 22100 }, { "epoch": 19.653333333333332, "grad_norm": 0.29511120915412903, "learning_rate": 2.9149469308664923e-05, "loss": 0.0134, "step": 22110 }, { "epoch": 19.662222222222223, "grad_norm": 0.24729685485363007, "learning_rate": 2.9104949996952392e-05, "loss": 0.0121, "step": 22120 }, { "epoch": 19.67111111111111, "grad_norm": 0.21832166612148285, "learning_rate": 2.9060450746670738e-05, "loss": 0.0171, "step": 22130 }, { "epoch": 19.68, "grad_norm": 0.25753307342529297, "learning_rate": 2.9015971600543902e-05, "loss": 0.0236, "step": 22140 }, { "epoch": 19.68888888888889, "grad_norm": 0.3267807364463806, "learning_rate": 2.8971512601276497e-05, "loss": 0.0189, "step": 22150 }, { "epoch": 19.697777777777777, "grad_norm": 0.20042915642261505, "learning_rate": 2.89270737915538e-05, "loss": 0.0153, "step": 22160 }, { "epoch": 19.706666666666667, "grad_norm": 0.27853965759277344, "learning_rate": 2.8882655214041743e-05, "loss": 0.0147, "step": 22170 }, { "epoch": 19.715555555555554, "grad_norm": 0.19124053418636322, "learning_rate": 2.8838256911386785e-05, "loss": 0.0203, "step": 22180 }, { "epoch": 19.724444444444444, "grad_norm": 0.20070719718933105, "learning_rate": 2.8793878926215924e-05, "loss": 0.0119, "step": 22190 }, { "epoch": 19.733333333333334, "grad_norm": 0.16448889672756195, "learning_rate": 2.874952130113668e-05, "loss": 0.0124, "step": 22200 }, { "epoch": 19.74222222222222, "grad_norm": 0.23007546365261078, "learning_rate": 2.870518407873699e-05, "loss": 0.0129, "step": 22210 }, { "epoch": 19.75111111111111, "grad_norm": 0.21961109340190887, "learning_rate": 2.8660867301585232e-05, "loss": 0.0171, "step": 22220 }, { "epoch": 19.76, "grad_norm": 0.2001187652349472, "learning_rate": 2.861657101223013e-05, "loss": 0.013, "step": 22230 }, { "epoch": 19.76888888888889, "grad_norm": 0.2680734097957611, "learning_rate": 2.8572295253200764e-05, "loss": 0.0138, "step": 22240 }, { "epoch": 19.77777777777778, "grad_norm": 0.30432477593421936, "learning_rate": 2.8528040067006457e-05, "loss": 0.0125, "step": 22250 }, { "epoch": 19.786666666666665, "grad_norm": 0.31823965907096863, "learning_rate": 2.848380549613685e-05, "loss": 0.0126, "step": 22260 }, { "epoch": 19.795555555555556, "grad_norm": 0.16031889617443085, "learning_rate": 2.8439591583061743e-05, "loss": 0.0125, "step": 22270 }, { "epoch": 19.804444444444446, "grad_norm": 0.25483110547065735, "learning_rate": 2.8395398370231095e-05, "loss": 0.0148, "step": 22280 }, { "epoch": 19.813333333333333, "grad_norm": 0.17337895929813385, "learning_rate": 2.8351225900075023e-05, "loss": 0.0168, "step": 22290 }, { "epoch": 19.822222222222223, "grad_norm": 0.31845054030418396, "learning_rate": 2.830707421500369e-05, "loss": 0.015, "step": 22300 }, { "epoch": 19.83111111111111, "grad_norm": 0.1981353610754013, "learning_rate": 2.8262943357407373e-05, "loss": 0.0116, "step": 22310 }, { "epoch": 19.84, "grad_norm": 0.22226476669311523, "learning_rate": 2.8218833369656263e-05, "loss": 0.0148, "step": 22320 }, { "epoch": 19.84888888888889, "grad_norm": 0.34562450647354126, "learning_rate": 2.8174744294100563e-05, "loss": 0.0133, "step": 22330 }, { "epoch": 19.857777777777777, "grad_norm": 0.1786537617444992, "learning_rate": 2.813067617307038e-05, "loss": 0.0141, "step": 22340 }, { "epoch": 19.866666666666667, "grad_norm": 0.29418349266052246, "learning_rate": 2.8086629048875755e-05, "loss": 0.0146, "step": 22350 }, { "epoch": 19.875555555555554, "grad_norm": 0.43073078989982605, "learning_rate": 2.8042602963806502e-05, "loss": 0.0132, "step": 22360 }, { "epoch": 19.884444444444444, "grad_norm": 0.2683127224445343, "learning_rate": 2.7998597960132268e-05, "loss": 0.0141, "step": 22370 }, { "epoch": 19.893333333333334, "grad_norm": 0.15879884362220764, "learning_rate": 2.7954614080102442e-05, "loss": 0.0128, "step": 22380 }, { "epoch": 19.90222222222222, "grad_norm": 0.1991586536169052, "learning_rate": 2.7910651365946137e-05, "loss": 0.0167, "step": 22390 }, { "epoch": 19.91111111111111, "grad_norm": 0.20023439824581146, "learning_rate": 2.7866709859872188e-05, "loss": 0.0123, "step": 22400 }, { "epoch": 19.92, "grad_norm": 0.2694770395755768, "learning_rate": 2.782278960406903e-05, "loss": 0.0119, "step": 22410 }, { "epoch": 19.92888888888889, "grad_norm": 0.23117171227931976, "learning_rate": 2.7778890640704658e-05, "loss": 0.0152, "step": 22420 }, { "epoch": 19.93777777777778, "grad_norm": 0.17973051965236664, "learning_rate": 2.7735013011926658e-05, "loss": 0.0165, "step": 22430 }, { "epoch": 19.946666666666665, "grad_norm": 0.23853901028633118, "learning_rate": 2.769115675986218e-05, "loss": 0.0157, "step": 22440 }, { "epoch": 19.955555555555556, "grad_norm": 0.22592876851558685, "learning_rate": 2.7647321926617798e-05, "loss": 0.0142, "step": 22450 }, { "epoch": 19.964444444444446, "grad_norm": 0.242103710770607, "learning_rate": 2.7603508554279512e-05, "loss": 0.0148, "step": 22460 }, { "epoch": 19.973333333333333, "grad_norm": 0.2620105743408203, "learning_rate": 2.7559716684912752e-05, "loss": 0.0149, "step": 22470 }, { "epoch": 19.982222222222223, "grad_norm": 0.2991107106208801, "learning_rate": 2.7515946360562262e-05, "loss": 0.0161, "step": 22480 }, { "epoch": 19.99111111111111, "grad_norm": 0.3743143081665039, "learning_rate": 2.7472197623252165e-05, "loss": 0.0137, "step": 22490 }, { "epoch": 20.0, "grad_norm": 0.26702767610549927, "learning_rate": 2.7428470514985805e-05, "loss": 0.0132, "step": 22500 }, { "epoch": 20.00888888888889, "grad_norm": 0.21275553107261658, "learning_rate": 2.7384765077745776e-05, "loss": 0.0129, "step": 22510 }, { "epoch": 20.017777777777777, "grad_norm": 0.1537308245897293, "learning_rate": 2.734108135349383e-05, "loss": 0.0099, "step": 22520 }, { "epoch": 20.026666666666667, "grad_norm": 0.2884328067302704, "learning_rate": 2.7297419384170952e-05, "loss": 0.0131, "step": 22530 }, { "epoch": 20.035555555555554, "grad_norm": 0.19169169664382935, "learning_rate": 2.725377921169717e-05, "loss": 0.0113, "step": 22540 }, { "epoch": 20.044444444444444, "grad_norm": 0.24425622820854187, "learning_rate": 2.7210160877971615e-05, "loss": 0.015, "step": 22550 }, { "epoch": 20.053333333333335, "grad_norm": 0.30562838912010193, "learning_rate": 2.716656442487244e-05, "loss": 0.0179, "step": 22560 }, { "epoch": 20.06222222222222, "grad_norm": 0.32142314314842224, "learning_rate": 2.712298989425678e-05, "loss": 0.0152, "step": 22570 }, { "epoch": 20.07111111111111, "grad_norm": 0.21070171892642975, "learning_rate": 2.7079437327960777e-05, "loss": 0.0111, "step": 22580 }, { "epoch": 20.08, "grad_norm": 0.23513653874397278, "learning_rate": 2.703590676779941e-05, "loss": 0.0138, "step": 22590 }, { "epoch": 20.08888888888889, "grad_norm": 0.23869775235652924, "learning_rate": 2.6992398255566588e-05, "loss": 0.0109, "step": 22600 }, { "epoch": 20.09777777777778, "grad_norm": 0.15430688858032227, "learning_rate": 2.6948911833035005e-05, "loss": 0.0135, "step": 22610 }, { "epoch": 20.106666666666666, "grad_norm": 0.1573844701051712, "learning_rate": 2.6905447541956186e-05, "loss": 0.0133, "step": 22620 }, { "epoch": 20.115555555555556, "grad_norm": 0.18667350709438324, "learning_rate": 2.686200542406039e-05, "loss": 0.0137, "step": 22630 }, { "epoch": 20.124444444444446, "grad_norm": 0.23556800186634064, "learning_rate": 2.6818585521056573e-05, "loss": 0.0127, "step": 22640 }, { "epoch": 20.133333333333333, "grad_norm": 0.4849052429199219, "learning_rate": 2.677518787463239e-05, "loss": 0.0121, "step": 22650 }, { "epoch": 20.142222222222223, "grad_norm": 0.5560237169265747, "learning_rate": 2.6731812526454104e-05, "loss": 0.0145, "step": 22660 }, { "epoch": 20.15111111111111, "grad_norm": 0.18349380791187286, "learning_rate": 2.6688459518166596e-05, "loss": 0.0146, "step": 22670 }, { "epoch": 20.16, "grad_norm": 0.21170426905155182, "learning_rate": 2.664512889139329e-05, "loss": 0.0136, "step": 22680 }, { "epoch": 20.16888888888889, "grad_norm": 0.22021745145320892, "learning_rate": 2.6601820687736094e-05, "loss": 0.0172, "step": 22690 }, { "epoch": 20.177777777777777, "grad_norm": 0.19825851917266846, "learning_rate": 2.6558534948775414e-05, "loss": 0.013, "step": 22700 }, { "epoch": 20.186666666666667, "grad_norm": 0.21827784180641174, "learning_rate": 2.6515271716070074e-05, "loss": 0.0164, "step": 22710 }, { "epoch": 20.195555555555554, "grad_norm": 0.200856551527977, "learning_rate": 2.6472031031157336e-05, "loss": 0.0152, "step": 22720 }, { "epoch": 20.204444444444444, "grad_norm": 0.37475189566612244, "learning_rate": 2.6428812935552728e-05, "loss": 0.0155, "step": 22730 }, { "epoch": 20.213333333333335, "grad_norm": 0.2939636707305908, "learning_rate": 2.6385617470750158e-05, "loss": 0.015, "step": 22740 }, { "epoch": 20.22222222222222, "grad_norm": 0.18094338476657867, "learning_rate": 2.634244467822175e-05, "loss": 0.0127, "step": 22750 }, { "epoch": 20.23111111111111, "grad_norm": 0.3713984489440918, "learning_rate": 2.6299294599417957e-05, "loss": 0.0153, "step": 22760 }, { "epoch": 20.24, "grad_norm": 0.27797389030456543, "learning_rate": 2.6256167275767318e-05, "loss": 0.0154, "step": 22770 }, { "epoch": 20.24888888888889, "grad_norm": 0.2653930187225342, "learning_rate": 2.6213062748676594e-05, "loss": 0.0123, "step": 22780 }, { "epoch": 20.25777777777778, "grad_norm": 0.19529958069324493, "learning_rate": 2.6169981059530614e-05, "loss": 0.0137, "step": 22790 }, { "epoch": 20.266666666666666, "grad_norm": 0.3113410174846649, "learning_rate": 2.612692224969229e-05, "loss": 0.0163, "step": 22800 }, { "epoch": 20.275555555555556, "grad_norm": 0.15319372713565826, "learning_rate": 2.6083886360502617e-05, "loss": 0.0136, "step": 22810 }, { "epoch": 20.284444444444443, "grad_norm": 0.2706764042377472, "learning_rate": 2.604087343328055e-05, "loss": 0.0147, "step": 22820 }, { "epoch": 20.293333333333333, "grad_norm": 0.20668596029281616, "learning_rate": 2.5997883509322918e-05, "loss": 0.0154, "step": 22830 }, { "epoch": 20.302222222222223, "grad_norm": 0.20580416917800903, "learning_rate": 2.5954916629904615e-05, "loss": 0.0137, "step": 22840 }, { "epoch": 20.31111111111111, "grad_norm": 0.2547217309474945, "learning_rate": 2.5911972836278297e-05, "loss": 0.0109, "step": 22850 }, { "epoch": 20.32, "grad_norm": 0.2768758237361908, "learning_rate": 2.5869052169674514e-05, "loss": 0.0156, "step": 22860 }, { "epoch": 20.32888888888889, "grad_norm": 0.14281733334064484, "learning_rate": 2.5826154671301567e-05, "loss": 0.0147, "step": 22870 }, { "epoch": 20.337777777777777, "grad_norm": 0.26929354667663574, "learning_rate": 2.5783280382345532e-05, "loss": 0.0141, "step": 22880 }, { "epoch": 20.346666666666668, "grad_norm": 0.3776448369026184, "learning_rate": 2.5740429343970235e-05, "loss": 0.0131, "step": 22890 }, { "epoch": 20.355555555555554, "grad_norm": 0.2415129542350769, "learning_rate": 2.5697601597317133e-05, "loss": 0.0133, "step": 22900 }, { "epoch": 20.364444444444445, "grad_norm": 0.2167353630065918, "learning_rate": 2.565479718350534e-05, "loss": 0.0113, "step": 22910 }, { "epoch": 20.373333333333335, "grad_norm": 0.25477883219718933, "learning_rate": 2.561201614363157e-05, "loss": 0.0113, "step": 22920 }, { "epoch": 20.38222222222222, "grad_norm": 0.22062766551971436, "learning_rate": 2.556925851877008e-05, "loss": 0.0154, "step": 22930 }, { "epoch": 20.391111111111112, "grad_norm": 0.26178184151649475, "learning_rate": 2.552652434997265e-05, "loss": 0.0125, "step": 22940 }, { "epoch": 20.4, "grad_norm": 0.24109505116939545, "learning_rate": 2.5483813678268565e-05, "loss": 0.0159, "step": 22950 }, { "epoch": 20.40888888888889, "grad_norm": 0.35881954431533813, "learning_rate": 2.5441126544664527e-05, "loss": 0.0117, "step": 22960 }, { "epoch": 20.41777777777778, "grad_norm": 0.15812359750270844, "learning_rate": 2.5398462990144612e-05, "loss": 0.0119, "step": 22970 }, { "epoch": 20.426666666666666, "grad_norm": 0.28475168347358704, "learning_rate": 2.5355823055670348e-05, "loss": 0.0145, "step": 22980 }, { "epoch": 20.435555555555556, "grad_norm": 0.2991531193256378, "learning_rate": 2.5313206782180486e-05, "loss": 0.0141, "step": 22990 }, { "epoch": 20.444444444444443, "grad_norm": 0.28869447112083435, "learning_rate": 2.5270614210591113e-05, "loss": 0.0155, "step": 23000 }, { "epoch": 20.453333333333333, "grad_norm": 0.28044167160987854, "learning_rate": 2.522804538179555e-05, "loss": 0.015, "step": 23010 }, { "epoch": 20.462222222222223, "grad_norm": 0.3042050004005432, "learning_rate": 2.518550033666429e-05, "loss": 0.0156, "step": 23020 }, { "epoch": 20.47111111111111, "grad_norm": 0.22476081550121307, "learning_rate": 2.514297911604509e-05, "loss": 0.0117, "step": 23030 }, { "epoch": 20.48, "grad_norm": 0.19497133791446686, "learning_rate": 2.5100481760762695e-05, "loss": 0.0135, "step": 23040 }, { "epoch": 20.488888888888887, "grad_norm": 0.23303844034671783, "learning_rate": 2.5058008311619035e-05, "loss": 0.0137, "step": 23050 }, { "epoch": 20.497777777777777, "grad_norm": 0.20265015959739685, "learning_rate": 2.5015558809393036e-05, "loss": 0.0141, "step": 23060 }, { "epoch": 20.506666666666668, "grad_norm": 0.24057546257972717, "learning_rate": 2.4973133294840683e-05, "loss": 0.0125, "step": 23070 }, { "epoch": 20.515555555555554, "grad_norm": 0.27091413736343384, "learning_rate": 2.49307318086949e-05, "loss": 0.011, "step": 23080 }, { "epoch": 20.524444444444445, "grad_norm": 0.14452606439590454, "learning_rate": 2.488835439166553e-05, "loss": 0.0152, "step": 23090 }, { "epoch": 20.533333333333335, "grad_norm": 0.24642245471477509, "learning_rate": 2.484600108443933e-05, "loss": 0.0175, "step": 23100 }, { "epoch": 20.54222222222222, "grad_norm": 0.21309055387973785, "learning_rate": 2.480367192767987e-05, "loss": 0.0199, "step": 23110 }, { "epoch": 20.551111111111112, "grad_norm": 0.2126639187335968, "learning_rate": 2.4761366962027614e-05, "loss": 0.012, "step": 23120 }, { "epoch": 20.56, "grad_norm": 0.41326144337654114, "learning_rate": 2.4719086228099735e-05, "loss": 0.0131, "step": 23130 }, { "epoch": 20.56888888888889, "grad_norm": 0.23714320361614227, "learning_rate": 2.467682976649013e-05, "loss": 0.0156, "step": 23140 }, { "epoch": 20.57777777777778, "grad_norm": 0.24249406158924103, "learning_rate": 2.46345976177694e-05, "loss": 0.0115, "step": 23150 }, { "epoch": 20.586666666666666, "grad_norm": 0.21767011284828186, "learning_rate": 2.4592389822484874e-05, "loss": 0.0145, "step": 23160 }, { "epoch": 20.595555555555556, "grad_norm": 0.1435684859752655, "learning_rate": 2.4550206421160415e-05, "loss": 0.0138, "step": 23170 }, { "epoch": 20.604444444444443, "grad_norm": 0.2568807303905487, "learning_rate": 2.45080474542965e-05, "loss": 0.014, "step": 23180 }, { "epoch": 20.613333333333333, "grad_norm": 0.16472914814949036, "learning_rate": 2.4465912962370142e-05, "loss": 0.0196, "step": 23190 }, { "epoch": 20.622222222222224, "grad_norm": 0.2299540638923645, "learning_rate": 2.442380298583483e-05, "loss": 0.0154, "step": 23200 }, { "epoch": 20.63111111111111, "grad_norm": 0.21231044828891754, "learning_rate": 2.4381717565120593e-05, "loss": 0.0158, "step": 23210 }, { "epoch": 20.64, "grad_norm": 0.24403047561645508, "learning_rate": 2.4339656740633794e-05, "loss": 0.0156, "step": 23220 }, { "epoch": 20.648888888888887, "grad_norm": 0.18000268936157227, "learning_rate": 2.429762055275725e-05, "loss": 0.0108, "step": 23230 }, { "epoch": 20.657777777777778, "grad_norm": 0.2029264122247696, "learning_rate": 2.425560904185003e-05, "loss": 0.0108, "step": 23240 }, { "epoch": 20.666666666666668, "grad_norm": 0.17307843267917633, "learning_rate": 2.4213622248247632e-05, "loss": 0.0125, "step": 23250 }, { "epoch": 20.675555555555555, "grad_norm": 0.41854235529899597, "learning_rate": 2.417166021226175e-05, "loss": 0.0112, "step": 23260 }, { "epoch": 20.684444444444445, "grad_norm": 0.219317227602005, "learning_rate": 2.4129722974180315e-05, "loss": 0.0128, "step": 23270 }, { "epoch": 20.693333333333335, "grad_norm": 0.2477978765964508, "learning_rate": 2.4087810574267454e-05, "loss": 0.0127, "step": 23280 }, { "epoch": 20.702222222222222, "grad_norm": 0.24142271280288696, "learning_rate": 2.404592305276344e-05, "loss": 0.0106, "step": 23290 }, { "epoch": 20.711111111111112, "grad_norm": 0.22803926467895508, "learning_rate": 2.4004060449884708e-05, "loss": 0.0103, "step": 23300 }, { "epoch": 20.72, "grad_norm": 0.17607207596302032, "learning_rate": 2.3962222805823703e-05, "loss": 0.0124, "step": 23310 }, { "epoch": 20.72888888888889, "grad_norm": 0.17830201983451843, "learning_rate": 2.392041016074894e-05, "loss": 0.0167, "step": 23320 }, { "epoch": 20.73777777777778, "grad_norm": 0.20818091928958893, "learning_rate": 2.387862255480492e-05, "loss": 0.0124, "step": 23330 }, { "epoch": 20.746666666666666, "grad_norm": 0.21356865763664246, "learning_rate": 2.3836860028112118e-05, "loss": 0.0124, "step": 23340 }, { "epoch": 20.755555555555556, "grad_norm": 0.24569261074066162, "learning_rate": 2.379512262076692e-05, "loss": 0.0113, "step": 23350 }, { "epoch": 20.764444444444443, "grad_norm": 0.219884991645813, "learning_rate": 2.375341037284159e-05, "loss": 0.018, "step": 23360 }, { "epoch": 20.773333333333333, "grad_norm": 0.3093731105327606, "learning_rate": 2.3711723324384255e-05, "loss": 0.0167, "step": 23370 }, { "epoch": 20.782222222222224, "grad_norm": 0.2167966067790985, "learning_rate": 2.3670061515418805e-05, "loss": 0.0161, "step": 23380 }, { "epoch": 20.79111111111111, "grad_norm": 0.2867511212825775, "learning_rate": 2.362842498594498e-05, "loss": 0.0113, "step": 23390 }, { "epoch": 20.8, "grad_norm": 0.20045524835586548, "learning_rate": 2.3586813775938177e-05, "loss": 0.0152, "step": 23400 }, { "epoch": 20.808888888888887, "grad_norm": 0.3194568157196045, "learning_rate": 2.3545227925349495e-05, "loss": 0.0167, "step": 23410 }, { "epoch": 20.817777777777778, "grad_norm": 0.28291866183280945, "learning_rate": 2.35036674741057e-05, "loss": 0.0165, "step": 23420 }, { "epoch": 20.826666666666668, "grad_norm": 0.18429364264011383, "learning_rate": 2.3462132462109156e-05, "loss": 0.012, "step": 23430 }, { "epoch": 20.835555555555555, "grad_norm": 0.28881892561912537, "learning_rate": 2.3420622929237855e-05, "loss": 0.0143, "step": 23440 }, { "epoch": 20.844444444444445, "grad_norm": 0.21448951959609985, "learning_rate": 2.3379138915345238e-05, "loss": 0.014, "step": 23450 }, { "epoch": 20.85333333333333, "grad_norm": 0.22215282917022705, "learning_rate": 2.3337680460260313e-05, "loss": 0.0148, "step": 23460 }, { "epoch": 20.862222222222222, "grad_norm": 0.2687714993953705, "learning_rate": 2.329624760378751e-05, "loss": 0.0157, "step": 23470 }, { "epoch": 20.871111111111112, "grad_norm": 0.25485333800315857, "learning_rate": 2.325484038570672e-05, "loss": 0.0118, "step": 23480 }, { "epoch": 20.88, "grad_norm": 0.1901761293411255, "learning_rate": 2.32134588457732e-05, "loss": 0.0138, "step": 23490 }, { "epoch": 20.88888888888889, "grad_norm": 0.18413092195987701, "learning_rate": 2.3172103023717545e-05, "loss": 0.012, "step": 23500 }, { "epoch": 20.897777777777776, "grad_norm": 0.2023649364709854, "learning_rate": 2.3130772959245662e-05, "loss": 0.0186, "step": 23510 }, { "epoch": 20.906666666666666, "grad_norm": 0.19729264080524445, "learning_rate": 2.3089468692038714e-05, "loss": 0.0135, "step": 23520 }, { "epoch": 20.915555555555557, "grad_norm": 0.3075060546398163, "learning_rate": 2.304819026175315e-05, "loss": 0.0174, "step": 23530 }, { "epoch": 20.924444444444443, "grad_norm": 0.187406986951828, "learning_rate": 2.300693770802057e-05, "loss": 0.0182, "step": 23540 }, { "epoch": 20.933333333333334, "grad_norm": 0.3038090467453003, "learning_rate": 2.2965711070447703e-05, "loss": 0.0168, "step": 23550 }, { "epoch": 20.942222222222224, "grad_norm": 0.277171790599823, "learning_rate": 2.2924510388616434e-05, "loss": 0.0131, "step": 23560 }, { "epoch": 20.95111111111111, "grad_norm": 0.18923619389533997, "learning_rate": 2.2883335702083747e-05, "loss": 0.0138, "step": 23570 }, { "epoch": 20.96, "grad_norm": 0.1735076755285263, "learning_rate": 2.2842187050381625e-05, "loss": 0.0152, "step": 23580 }, { "epoch": 20.968888888888888, "grad_norm": 0.2354089319705963, "learning_rate": 2.2801064473017085e-05, "loss": 0.0127, "step": 23590 }, { "epoch": 20.977777777777778, "grad_norm": 0.23450301587581635, "learning_rate": 2.2759968009472083e-05, "loss": 0.014, "step": 23600 }, { "epoch": 20.986666666666668, "grad_norm": 0.14596326649188995, "learning_rate": 2.2718897699203508e-05, "loss": 0.0131, "step": 23610 }, { "epoch": 20.995555555555555, "grad_norm": 0.26254507899284363, "learning_rate": 2.2677853581643183e-05, "loss": 0.012, "step": 23620 }, { "epoch": 21.004444444444445, "grad_norm": 0.2193514108657837, "learning_rate": 2.2636835696197723e-05, "loss": 0.0112, "step": 23630 }, { "epoch": 21.013333333333332, "grad_norm": 0.2274070382118225, "learning_rate": 2.2595844082248595e-05, "loss": 0.011, "step": 23640 }, { "epoch": 21.022222222222222, "grad_norm": 0.1562492400407791, "learning_rate": 2.255487877915202e-05, "loss": 0.0121, "step": 23650 }, { "epoch": 21.031111111111112, "grad_norm": 0.2649695575237274, "learning_rate": 2.2513939826238984e-05, "loss": 0.0158, "step": 23660 }, { "epoch": 21.04, "grad_norm": 0.2018701136112213, "learning_rate": 2.2473027262815137e-05, "loss": 0.0132, "step": 23670 }, { "epoch": 21.04888888888889, "grad_norm": 0.2245829552412033, "learning_rate": 2.2432141128160832e-05, "loss": 0.0147, "step": 23680 }, { "epoch": 21.057777777777776, "grad_norm": 0.1610398143529892, "learning_rate": 2.2391281461531023e-05, "loss": 0.0104, "step": 23690 }, { "epoch": 21.066666666666666, "grad_norm": 0.25806042551994324, "learning_rate": 2.2350448302155236e-05, "loss": 0.0126, "step": 23700 }, { "epoch": 21.075555555555557, "grad_norm": 0.19560568034648895, "learning_rate": 2.2309641689237614e-05, "loss": 0.0152, "step": 23710 }, { "epoch": 21.084444444444443, "grad_norm": 0.18183311820030212, "learning_rate": 2.2268861661956746e-05, "loss": 0.0123, "step": 23720 }, { "epoch": 21.093333333333334, "grad_norm": 0.48290449380874634, "learning_rate": 2.2228108259465718e-05, "loss": 0.0125, "step": 23730 }, { "epoch": 21.102222222222224, "grad_norm": 0.34821927547454834, "learning_rate": 2.2187381520892057e-05, "loss": 0.012, "step": 23740 }, { "epoch": 21.11111111111111, "grad_norm": 0.16533435881137848, "learning_rate": 2.2146681485337682e-05, "loss": 0.0127, "step": 23750 }, { "epoch": 21.12, "grad_norm": 0.23061802983283997, "learning_rate": 2.2106008191878885e-05, "loss": 0.0144, "step": 23760 }, { "epoch": 21.128888888888888, "grad_norm": 0.23318462073802948, "learning_rate": 2.206536167956627e-05, "loss": 0.0129, "step": 23770 }, { "epoch": 21.137777777777778, "grad_norm": 0.16760075092315674, "learning_rate": 2.202474198742474e-05, "loss": 0.013, "step": 23780 }, { "epoch": 21.14666666666667, "grad_norm": 0.1944071352481842, "learning_rate": 2.198414915445342e-05, "loss": 0.0167, "step": 23790 }, { "epoch": 21.155555555555555, "grad_norm": 0.23009148240089417, "learning_rate": 2.1943583219625714e-05, "loss": 0.0125, "step": 23800 }, { "epoch": 21.164444444444445, "grad_norm": 0.24028640985488892, "learning_rate": 2.1903044221889134e-05, "loss": 0.013, "step": 23810 }, { "epoch": 21.173333333333332, "grad_norm": 0.15505152940750122, "learning_rate": 2.1862532200165358e-05, "loss": 0.0096, "step": 23820 }, { "epoch": 21.182222222222222, "grad_norm": 0.19410760700702667, "learning_rate": 2.1822047193350148e-05, "loss": 0.0133, "step": 23830 }, { "epoch": 21.191111111111113, "grad_norm": 0.31208041310310364, "learning_rate": 2.1781589240313332e-05, "loss": 0.0146, "step": 23840 }, { "epoch": 21.2, "grad_norm": 0.14839830994606018, "learning_rate": 2.1741158379898823e-05, "loss": 0.0102, "step": 23850 }, { "epoch": 21.20888888888889, "grad_norm": 0.21397580206394196, "learning_rate": 2.1700754650924415e-05, "loss": 0.012, "step": 23860 }, { "epoch": 21.217777777777776, "grad_norm": 0.22075647115707397, "learning_rate": 2.166037809218193e-05, "loss": 0.0101, "step": 23870 }, { "epoch": 21.226666666666667, "grad_norm": 0.21137839555740356, "learning_rate": 2.1620028742437066e-05, "loss": 0.0123, "step": 23880 }, { "epoch": 21.235555555555557, "grad_norm": 0.22954894602298737, "learning_rate": 2.1579706640429444e-05, "loss": 0.0135, "step": 23890 }, { "epoch": 21.244444444444444, "grad_norm": 0.16762250661849976, "learning_rate": 2.1539411824872474e-05, "loss": 0.0149, "step": 23900 }, { "epoch": 21.253333333333334, "grad_norm": 0.6384797096252441, "learning_rate": 2.1499144334453398e-05, "loss": 0.0174, "step": 23910 }, { "epoch": 21.26222222222222, "grad_norm": 0.20703665912151337, "learning_rate": 2.1458904207833208e-05, "loss": 0.0124, "step": 23920 }, { "epoch": 21.27111111111111, "grad_norm": 0.17645515501499176, "learning_rate": 2.1418691483646608e-05, "loss": 0.0152, "step": 23930 }, { "epoch": 21.28, "grad_norm": 0.3158314824104309, "learning_rate": 2.137850620050206e-05, "loss": 0.0128, "step": 23940 }, { "epoch": 21.288888888888888, "grad_norm": 0.19026894867420197, "learning_rate": 2.1338348396981628e-05, "loss": 0.0107, "step": 23950 }, { "epoch": 21.297777777777778, "grad_norm": 0.19023579359054565, "learning_rate": 2.129821811164096e-05, "loss": 0.0225, "step": 23960 }, { "epoch": 21.306666666666665, "grad_norm": 0.2107129991054535, "learning_rate": 2.1258115383009325e-05, "loss": 0.0126, "step": 23970 }, { "epoch": 21.315555555555555, "grad_norm": 0.2416360229253769, "learning_rate": 2.1218040249589573e-05, "loss": 0.0137, "step": 23980 }, { "epoch": 21.324444444444445, "grad_norm": 0.5105032324790955, "learning_rate": 2.1177992749857988e-05, "loss": 0.0097, "step": 23990 }, { "epoch": 21.333333333333332, "grad_norm": 0.2327822893857956, "learning_rate": 2.113797292226437e-05, "loss": 0.0124, "step": 24000 }, { "epoch": 21.342222222222222, "grad_norm": 0.22445109486579895, "learning_rate": 2.109798080523192e-05, "loss": 0.0121, "step": 24010 }, { "epoch": 21.351111111111113, "grad_norm": 0.30702564120292664, "learning_rate": 2.1058016437157242e-05, "loss": 0.0136, "step": 24020 }, { "epoch": 21.36, "grad_norm": 0.2970861792564392, "learning_rate": 2.1018079856410328e-05, "loss": 0.0121, "step": 24030 }, { "epoch": 21.36888888888889, "grad_norm": 0.20603792369365692, "learning_rate": 2.097817110133446e-05, "loss": 0.011, "step": 24040 }, { "epoch": 21.377777777777776, "grad_norm": 0.1705019176006317, "learning_rate": 2.0938290210246203e-05, "loss": 0.0125, "step": 24050 }, { "epoch": 21.386666666666667, "grad_norm": 0.18153831362724304, "learning_rate": 2.089843722143539e-05, "loss": 0.0095, "step": 24060 }, { "epoch": 21.395555555555557, "grad_norm": 0.2282872498035431, "learning_rate": 2.0858612173165044e-05, "loss": 0.017, "step": 24070 }, { "epoch": 21.404444444444444, "grad_norm": 0.13587334752082825, "learning_rate": 2.0818815103671373e-05, "loss": 0.0098, "step": 24080 }, { "epoch": 21.413333333333334, "grad_norm": 0.24785780906677246, "learning_rate": 2.0779046051163732e-05, "loss": 0.0138, "step": 24090 }, { "epoch": 21.42222222222222, "grad_norm": 0.17177079617977142, "learning_rate": 2.073930505382455e-05, "loss": 0.0119, "step": 24100 }, { "epoch": 21.43111111111111, "grad_norm": 0.13848531246185303, "learning_rate": 2.0699592149809326e-05, "loss": 0.0137, "step": 24110 }, { "epoch": 21.44, "grad_norm": 0.26593831181526184, "learning_rate": 2.0659907377246636e-05, "loss": 0.0161, "step": 24120 }, { "epoch": 21.448888888888888, "grad_norm": 0.20615488290786743, "learning_rate": 2.062025077423798e-05, "loss": 0.015, "step": 24130 }, { "epoch": 21.45777777777778, "grad_norm": 0.19071362912654877, "learning_rate": 2.058062237885785e-05, "loss": 0.0132, "step": 24140 }, { "epoch": 21.466666666666665, "grad_norm": 0.30278465151786804, "learning_rate": 2.054102222915364e-05, "loss": 0.0117, "step": 24150 }, { "epoch": 21.475555555555555, "grad_norm": 0.22482289373874664, "learning_rate": 2.050145036314563e-05, "loss": 0.0122, "step": 24160 }, { "epoch": 21.484444444444446, "grad_norm": 0.18075379729270935, "learning_rate": 2.0461906818826955e-05, "loss": 0.0102, "step": 24170 }, { "epoch": 21.493333333333332, "grad_norm": 0.289985328912735, "learning_rate": 2.042239163416353e-05, "loss": 0.0131, "step": 24180 }, { "epoch": 21.502222222222223, "grad_norm": 0.2945682108402252, "learning_rate": 2.0382904847094074e-05, "loss": 0.01, "step": 24190 }, { "epoch": 21.511111111111113, "grad_norm": 0.1972527951002121, "learning_rate": 2.034344649553001e-05, "loss": 0.0109, "step": 24200 }, { "epoch": 21.52, "grad_norm": 0.2147742658853531, "learning_rate": 2.030401661735551e-05, "loss": 0.0147, "step": 24210 }, { "epoch": 21.52888888888889, "grad_norm": 0.3052099347114563, "learning_rate": 2.026461525042736e-05, "loss": 0.0121, "step": 24220 }, { "epoch": 21.537777777777777, "grad_norm": 0.18546012043952942, "learning_rate": 2.0225242432574985e-05, "loss": 0.0119, "step": 24230 }, { "epoch": 21.546666666666667, "grad_norm": 0.3268996775150299, "learning_rate": 2.0185898201600405e-05, "loss": 0.0163, "step": 24240 }, { "epoch": 21.555555555555557, "grad_norm": 0.1985911875963211, "learning_rate": 2.0146582595278172e-05, "loss": 0.0111, "step": 24250 }, { "epoch": 21.564444444444444, "grad_norm": 0.16856905817985535, "learning_rate": 2.010729565135544e-05, "loss": 0.0115, "step": 24260 }, { "epoch": 21.573333333333334, "grad_norm": 0.21099509298801422, "learning_rate": 2.006803740755171e-05, "loss": 0.0137, "step": 24270 }, { "epoch": 21.58222222222222, "grad_norm": 0.233989417552948, "learning_rate": 2.0028807901559026e-05, "loss": 0.0115, "step": 24280 }, { "epoch": 21.59111111111111, "grad_norm": 0.2546605169773102, "learning_rate": 1.998960717104179e-05, "loss": 0.0134, "step": 24290 }, { "epoch": 21.6, "grad_norm": 0.1926691234111786, "learning_rate": 1.9950435253636834e-05, "loss": 0.0106, "step": 24300 }, { "epoch": 21.608888888888888, "grad_norm": 0.3033055365085602, "learning_rate": 1.991129218695328e-05, "loss": 0.0104, "step": 24310 }, { "epoch": 21.61777777777778, "grad_norm": 0.23188620805740356, "learning_rate": 1.9872178008572545e-05, "loss": 0.0161, "step": 24320 }, { "epoch": 21.626666666666665, "grad_norm": 0.3323211967945099, "learning_rate": 1.9833092756048344e-05, "loss": 0.0106, "step": 24330 }, { "epoch": 21.635555555555555, "grad_norm": 0.19293831288814545, "learning_rate": 1.9794036466906572e-05, "loss": 0.0134, "step": 24340 }, { "epoch": 21.644444444444446, "grad_norm": 0.3535570502281189, "learning_rate": 1.9755009178645396e-05, "loss": 0.0165, "step": 24350 }, { "epoch": 21.653333333333332, "grad_norm": 0.3003937304019928, "learning_rate": 1.9716010928735058e-05, "loss": 0.0122, "step": 24360 }, { "epoch": 21.662222222222223, "grad_norm": 0.196956068277359, "learning_rate": 1.9677041754617977e-05, "loss": 0.0131, "step": 24370 }, { "epoch": 21.67111111111111, "grad_norm": 0.2166542261838913, "learning_rate": 1.9638101693708576e-05, "loss": 0.0154, "step": 24380 }, { "epoch": 21.68, "grad_norm": 0.2301866114139557, "learning_rate": 1.959919078339344e-05, "loss": 0.0123, "step": 24390 }, { "epoch": 21.68888888888889, "grad_norm": 0.14484506845474243, "learning_rate": 1.956030906103108e-05, "loss": 0.0171, "step": 24400 }, { "epoch": 21.697777777777777, "grad_norm": 0.12841378152370453, "learning_rate": 1.9521456563952018e-05, "loss": 0.0124, "step": 24410 }, { "epoch": 21.706666666666667, "grad_norm": 0.10309533029794693, "learning_rate": 1.9482633329458712e-05, "loss": 0.0102, "step": 24420 }, { "epoch": 21.715555555555554, "grad_norm": 0.16921944916248322, "learning_rate": 1.94438393948255e-05, "loss": 0.0149, "step": 24430 }, { "epoch": 21.724444444444444, "grad_norm": 0.23969946801662445, "learning_rate": 1.940507479729865e-05, "loss": 0.0136, "step": 24440 }, { "epoch": 21.733333333333334, "grad_norm": 0.2099115401506424, "learning_rate": 1.936633957409621e-05, "loss": 0.0124, "step": 24450 }, { "epoch": 21.74222222222222, "grad_norm": 0.2624823749065399, "learning_rate": 1.932763376240803e-05, "loss": 0.0095, "step": 24460 }, { "epoch": 21.75111111111111, "grad_norm": 0.294684499502182, "learning_rate": 1.928895739939575e-05, "loss": 0.0116, "step": 24470 }, { "epoch": 21.76, "grad_norm": 0.20736834406852722, "learning_rate": 1.9250310522192703e-05, "loss": 0.0098, "step": 24480 }, { "epoch": 21.76888888888889, "grad_norm": 0.1759539395570755, "learning_rate": 1.9211693167903945e-05, "loss": 0.0128, "step": 24490 }, { "epoch": 21.77777777777778, "grad_norm": 0.2915398180484772, "learning_rate": 1.9173105373606153e-05, "loss": 0.0159, "step": 24500 }, { "epoch": 21.786666666666665, "grad_norm": 0.2193455994129181, "learning_rate": 1.913454717634765e-05, "loss": 0.0134, "step": 24510 }, { "epoch": 21.795555555555556, "grad_norm": 0.305303156375885, "learning_rate": 1.9096018613148313e-05, "loss": 0.013, "step": 24520 }, { "epoch": 21.804444444444446, "grad_norm": 0.19961167871952057, "learning_rate": 1.905751972099963e-05, "loss": 0.0121, "step": 24530 }, { "epoch": 21.813333333333333, "grad_norm": 0.1801750212907791, "learning_rate": 1.901905053686453e-05, "loss": 0.014, "step": 24540 }, { "epoch": 21.822222222222223, "grad_norm": 0.22682802379131317, "learning_rate": 1.8980611097677453e-05, "loss": 0.0134, "step": 24550 }, { "epoch": 21.83111111111111, "grad_norm": 0.32175034284591675, "learning_rate": 1.8942201440344287e-05, "loss": 0.0118, "step": 24560 }, { "epoch": 21.84, "grad_norm": 0.19222937524318695, "learning_rate": 1.89038216017423e-05, "loss": 0.0139, "step": 24570 }, { "epoch": 21.84888888888889, "grad_norm": 0.8012768030166626, "learning_rate": 1.8865471618720154e-05, "loss": 0.0136, "step": 24580 }, { "epoch": 21.857777777777777, "grad_norm": 0.22120843827724457, "learning_rate": 1.882715152809784e-05, "loss": 0.0136, "step": 24590 }, { "epoch": 21.866666666666667, "grad_norm": 0.14398542046546936, "learning_rate": 1.878886136666665e-05, "loss": 0.0104, "step": 24600 }, { "epoch": 21.875555555555554, "grad_norm": 0.22997497022151947, "learning_rate": 1.8750601171189115e-05, "loss": 0.0099, "step": 24610 }, { "epoch": 21.884444444444444, "grad_norm": 0.21705782413482666, "learning_rate": 1.8712370978399052e-05, "loss": 0.0103, "step": 24620 }, { "epoch": 21.893333333333334, "grad_norm": 0.2263585329055786, "learning_rate": 1.8674170825001437e-05, "loss": 0.0141, "step": 24630 }, { "epoch": 21.90222222222222, "grad_norm": 0.44698306918144226, "learning_rate": 1.8636000747672394e-05, "loss": 0.0155, "step": 24640 }, { "epoch": 21.91111111111111, "grad_norm": 0.26306527853012085, "learning_rate": 1.859786078305919e-05, "loss": 0.0147, "step": 24650 }, { "epoch": 21.92, "grad_norm": 0.28592202067375183, "learning_rate": 1.8559750967780165e-05, "loss": 0.0155, "step": 24660 }, { "epoch": 21.92888888888889, "grad_norm": 0.2749854028224945, "learning_rate": 1.8521671338424763e-05, "loss": 0.0105, "step": 24670 }, { "epoch": 21.93777777777778, "grad_norm": 0.22432971000671387, "learning_rate": 1.848362193155336e-05, "loss": 0.0149, "step": 24680 }, { "epoch": 21.946666666666665, "grad_norm": 0.20985762774944305, "learning_rate": 1.8445602783697374e-05, "loss": 0.014, "step": 24690 }, { "epoch": 21.955555555555556, "grad_norm": 0.18399108946323395, "learning_rate": 1.840761393135915e-05, "loss": 0.0095, "step": 24700 }, { "epoch": 21.964444444444446, "grad_norm": 0.1822250932455063, "learning_rate": 1.836965541101199e-05, "loss": 0.0124, "step": 24710 }, { "epoch": 21.973333333333333, "grad_norm": 0.29512885212898254, "learning_rate": 1.8331727259100013e-05, "loss": 0.0153, "step": 24720 }, { "epoch": 21.982222222222223, "grad_norm": 0.20416587591171265, "learning_rate": 1.8293829512038207e-05, "loss": 0.0096, "step": 24730 }, { "epoch": 21.99111111111111, "grad_norm": 0.2098909467458725, "learning_rate": 1.825596220621238e-05, "loss": 0.0088, "step": 24740 }, { "epoch": 22.0, "grad_norm": 0.4185837507247925, "learning_rate": 1.8218125377979074e-05, "loss": 0.0127, "step": 24750 }, { "epoch": 22.00888888888889, "grad_norm": 0.1637989580631256, "learning_rate": 1.818031906366563e-05, "loss": 0.0113, "step": 24760 }, { "epoch": 22.017777777777777, "grad_norm": 0.16566620767116547, "learning_rate": 1.8142543299570046e-05, "loss": 0.01, "step": 24770 }, { "epoch": 22.026666666666667, "grad_norm": 0.14865295588970184, "learning_rate": 1.8104798121961026e-05, "loss": 0.0134, "step": 24780 }, { "epoch": 22.035555555555554, "grad_norm": 0.20500653982162476, "learning_rate": 1.8067083567077823e-05, "loss": 0.013, "step": 24790 }, { "epoch": 22.044444444444444, "grad_norm": 0.270345538854599, "learning_rate": 1.8029399671130402e-05, "loss": 0.0119, "step": 24800 }, { "epoch": 22.053333333333335, "grad_norm": 0.20533454418182373, "learning_rate": 1.7991746470299227e-05, "loss": 0.0155, "step": 24810 }, { "epoch": 22.06222222222222, "grad_norm": 0.13658125698566437, "learning_rate": 1.79541240007353e-05, "loss": 0.011, "step": 24820 }, { "epoch": 22.07111111111111, "grad_norm": 0.2703021466732025, "learning_rate": 1.7916532298560125e-05, "loss": 0.012, "step": 24830 }, { "epoch": 22.08, "grad_norm": 0.2629704475402832, "learning_rate": 1.7878971399865647e-05, "loss": 0.0139, "step": 24840 }, { "epoch": 22.08888888888889, "grad_norm": 0.23449017107486725, "learning_rate": 1.7841441340714284e-05, "loss": 0.0124, "step": 24850 }, { "epoch": 22.09777777777778, "grad_norm": 0.29910969734191895, "learning_rate": 1.7803942157138793e-05, "loss": 0.0139, "step": 24860 }, { "epoch": 22.106666666666666, "grad_norm": 0.21623140573501587, "learning_rate": 1.7766473885142314e-05, "loss": 0.015, "step": 24870 }, { "epoch": 22.115555555555556, "grad_norm": 0.17897970974445343, "learning_rate": 1.7729036560698302e-05, "loss": 0.0152, "step": 24880 }, { "epoch": 22.124444444444446, "grad_norm": 0.21474811434745789, "learning_rate": 1.7691630219750495e-05, "loss": 0.0123, "step": 24890 }, { "epoch": 22.133333333333333, "grad_norm": 0.2568870484828949, "learning_rate": 1.7654254898212896e-05, "loss": 0.0177, "step": 24900 }, { "epoch": 22.142222222222223, "grad_norm": 0.15413430333137512, "learning_rate": 1.7616910631969708e-05, "loss": 0.0128, "step": 24910 }, { "epoch": 22.15111111111111, "grad_norm": 0.3591056168079376, "learning_rate": 1.7579597456875337e-05, "loss": 0.01, "step": 24920 }, { "epoch": 22.16, "grad_norm": 0.3025816082954407, "learning_rate": 1.7542315408754306e-05, "loss": 0.0126, "step": 24930 }, { "epoch": 22.16888888888889, "grad_norm": 0.27590411901474, "learning_rate": 1.7505064523401317e-05, "loss": 0.0102, "step": 24940 }, { "epoch": 22.177777777777777, "grad_norm": 0.1515084207057953, "learning_rate": 1.746784483658109e-05, "loss": 0.0109, "step": 24950 }, { "epoch": 22.186666666666667, "grad_norm": 0.2664943039417267, "learning_rate": 1.7430656384028403e-05, "loss": 0.011, "step": 24960 }, { "epoch": 22.195555555555554, "grad_norm": 0.26772040128707886, "learning_rate": 1.739349920144807e-05, "loss": 0.0132, "step": 24970 }, { "epoch": 22.204444444444444, "grad_norm": 0.22204947471618652, "learning_rate": 1.7356373324514868e-05, "loss": 0.013, "step": 24980 }, { "epoch": 22.213333333333335, "grad_norm": 0.36471641063690186, "learning_rate": 1.7319278788873505e-05, "loss": 0.0114, "step": 24990 }, { "epoch": 22.22222222222222, "grad_norm": 0.23271530866622925, "learning_rate": 1.7282215630138628e-05, "loss": 0.0183, "step": 25000 }, { "epoch": 22.23111111111111, "grad_norm": 0.27347931265830994, "learning_rate": 1.7245183883894722e-05, "loss": 0.0121, "step": 25010 }, { "epoch": 22.24, "grad_norm": 0.3069152235984802, "learning_rate": 1.7208183585696126e-05, "loss": 0.0121, "step": 25020 }, { "epoch": 22.24888888888889, "grad_norm": 0.1865713894367218, "learning_rate": 1.7171214771067022e-05, "loss": 0.013, "step": 25030 }, { "epoch": 22.25777777777778, "grad_norm": 0.20425944030284882, "learning_rate": 1.7134277475501325e-05, "loss": 0.0117, "step": 25040 }, { "epoch": 22.266666666666666, "grad_norm": 0.33731967210769653, "learning_rate": 1.7097371734462686e-05, "loss": 0.0102, "step": 25050 }, { "epoch": 22.275555555555556, "grad_norm": 0.2216380536556244, "learning_rate": 1.706049758338448e-05, "loss": 0.0142, "step": 25060 }, { "epoch": 22.284444444444443, "grad_norm": 0.14412695169448853, "learning_rate": 1.7023655057669724e-05, "loss": 0.0129, "step": 25070 }, { "epoch": 22.293333333333333, "grad_norm": 0.1460534781217575, "learning_rate": 1.6986844192691127e-05, "loss": 0.0111, "step": 25080 }, { "epoch": 22.302222222222223, "grad_norm": 0.2590576410293579, "learning_rate": 1.6950065023790963e-05, "loss": 0.0123, "step": 25090 }, { "epoch": 22.31111111111111, "grad_norm": 0.30413100123405457, "learning_rate": 1.6913317586281047e-05, "loss": 0.0094, "step": 25100 }, { "epoch": 22.32, "grad_norm": 0.222610741853714, "learning_rate": 1.6876601915442753e-05, "loss": 0.0112, "step": 25110 }, { "epoch": 22.32888888888889, "grad_norm": 0.19976459443569183, "learning_rate": 1.683991804652699e-05, "loss": 0.0098, "step": 25120 }, { "epoch": 22.337777777777777, "grad_norm": 0.16241848468780518, "learning_rate": 1.68032660147541e-05, "loss": 0.0111, "step": 25130 }, { "epoch": 22.346666666666668, "grad_norm": 0.26170650124549866, "learning_rate": 1.6766645855313856e-05, "loss": 0.012, "step": 25140 }, { "epoch": 22.355555555555554, "grad_norm": 0.19843968749046326, "learning_rate": 1.6730057603365434e-05, "loss": 0.0138, "step": 25150 }, { "epoch": 22.364444444444445, "grad_norm": 0.23872987926006317, "learning_rate": 1.6693501294037365e-05, "loss": 0.0131, "step": 25160 }, { "epoch": 22.373333333333335, "grad_norm": 0.2167704999446869, "learning_rate": 1.665697696242756e-05, "loss": 0.0102, "step": 25170 }, { "epoch": 22.38222222222222, "grad_norm": 0.1857849359512329, "learning_rate": 1.6620484643603173e-05, "loss": 0.0168, "step": 25180 }, { "epoch": 22.391111111111112, "grad_norm": 0.341842919588089, "learning_rate": 1.658402437260066e-05, "loss": 0.0117, "step": 25190 }, { "epoch": 22.4, "grad_norm": 0.25538745522499084, "learning_rate": 1.654759618442564e-05, "loss": 0.0126, "step": 25200 }, { "epoch": 22.40888888888889, "grad_norm": 0.3135729134082794, "learning_rate": 1.6511200114053026e-05, "loss": 0.0129, "step": 25210 }, { "epoch": 22.41777777777778, "grad_norm": 0.23902660608291626, "learning_rate": 1.647483619642684e-05, "loss": 0.0142, "step": 25220 }, { "epoch": 22.426666666666666, "grad_norm": 0.23528560996055603, "learning_rate": 1.6438504466460237e-05, "loss": 0.0154, "step": 25230 }, { "epoch": 22.435555555555556, "grad_norm": 0.2989983558654785, "learning_rate": 1.6402204959035468e-05, "loss": 0.0117, "step": 25240 }, { "epoch": 22.444444444444443, "grad_norm": 0.20403462648391724, "learning_rate": 1.6365937709003843e-05, "loss": 0.0119, "step": 25250 }, { "epoch": 22.453333333333333, "grad_norm": 0.1637754738330841, "learning_rate": 1.6329702751185744e-05, "loss": 0.0147, "step": 25260 }, { "epoch": 22.462222222222223, "grad_norm": 0.2709094285964966, "learning_rate": 1.6293500120370493e-05, "loss": 0.009, "step": 25270 }, { "epoch": 22.47111111111111, "grad_norm": 0.18176227807998657, "learning_rate": 1.62573298513164e-05, "loss": 0.0102, "step": 25280 }, { "epoch": 22.48, "grad_norm": 0.1893659234046936, "learning_rate": 1.6221191978750706e-05, "loss": 0.0124, "step": 25290 }, { "epoch": 22.488888888888887, "grad_norm": 0.25200024247169495, "learning_rate": 1.6185086537369544e-05, "loss": 0.011, "step": 25300 }, { "epoch": 22.497777777777777, "grad_norm": 0.2202429473400116, "learning_rate": 1.61490135618379e-05, "loss": 0.0118, "step": 25310 }, { "epoch": 22.506666666666668, "grad_norm": 0.3391597270965576, "learning_rate": 1.611297308678961e-05, "loss": 0.0117, "step": 25320 }, { "epoch": 22.515555555555554, "grad_norm": 0.2545664608478546, "learning_rate": 1.607696514682729e-05, "loss": 0.0161, "step": 25330 }, { "epoch": 22.524444444444445, "grad_norm": 0.17045556008815765, "learning_rate": 1.6040989776522307e-05, "loss": 0.0146, "step": 25340 }, { "epoch": 22.533333333333335, "grad_norm": 0.23581844568252563, "learning_rate": 1.6005047010414814e-05, "loss": 0.0118, "step": 25350 }, { "epoch": 22.54222222222222, "grad_norm": 0.1785620003938675, "learning_rate": 1.5969136883013597e-05, "loss": 0.0145, "step": 25360 }, { "epoch": 22.551111111111112, "grad_norm": 0.29982078075408936, "learning_rate": 1.593325942879613e-05, "loss": 0.0128, "step": 25370 }, { "epoch": 22.56, "grad_norm": 0.3449195325374603, "learning_rate": 1.5897414682208528e-05, "loss": 0.0128, "step": 25380 }, { "epoch": 22.56888888888889, "grad_norm": 0.2508891224861145, "learning_rate": 1.5861602677665467e-05, "loss": 0.0127, "step": 25390 }, { "epoch": 22.57777777777778, "grad_norm": 0.20016051828861237, "learning_rate": 1.5825823449550258e-05, "loss": 0.0148, "step": 25400 }, { "epoch": 22.586666666666666, "grad_norm": 0.1373467594385147, "learning_rate": 1.5790077032214655e-05, "loss": 0.0122, "step": 25410 }, { "epoch": 22.595555555555556, "grad_norm": 0.2606528401374817, "learning_rate": 1.5754363459978967e-05, "loss": 0.0116, "step": 25420 }, { "epoch": 22.604444444444443, "grad_norm": 0.23778709769248962, "learning_rate": 1.5718682767131943e-05, "loss": 0.0183, "step": 25430 }, { "epoch": 22.613333333333333, "grad_norm": 0.16302943229675293, "learning_rate": 1.56830349879308e-05, "loss": 0.0131, "step": 25440 }, { "epoch": 22.622222222222224, "grad_norm": 0.24473929405212402, "learning_rate": 1.5647420156601118e-05, "loss": 0.0115, "step": 25450 }, { "epoch": 22.63111111111111, "grad_norm": 0.24428841471672058, "learning_rate": 1.5611838307336858e-05, "loss": 0.0141, "step": 25460 }, { "epoch": 22.64, "grad_norm": 0.19957050681114197, "learning_rate": 1.557628947430031e-05, "loss": 0.0108, "step": 25470 }, { "epoch": 22.648888888888887, "grad_norm": 0.25809726119041443, "learning_rate": 1.554077369162205e-05, "loss": 0.0125, "step": 25480 }, { "epoch": 22.657777777777778, "grad_norm": 0.2480190545320511, "learning_rate": 1.5505290993400977e-05, "loss": 0.0124, "step": 25490 }, { "epoch": 22.666666666666668, "grad_norm": 0.20924095809459686, "learning_rate": 1.5469841413704184e-05, "loss": 0.0122, "step": 25500 }, { "epoch": 22.675555555555555, "grad_norm": 0.2146570384502411, "learning_rate": 1.5434424986566936e-05, "loss": 0.0107, "step": 25510 }, { "epoch": 22.684444444444445, "grad_norm": 0.1939656138420105, "learning_rate": 1.5399041745992703e-05, "loss": 0.0136, "step": 25520 }, { "epoch": 22.693333333333335, "grad_norm": 0.209343820810318, "learning_rate": 1.5363691725953133e-05, "loss": 0.0114, "step": 25530 }, { "epoch": 22.702222222222222, "grad_norm": 0.24882954359054565, "learning_rate": 1.532837496038792e-05, "loss": 0.0126, "step": 25540 }, { "epoch": 22.711111111111112, "grad_norm": 0.11051591485738754, "learning_rate": 1.5293091483204842e-05, "loss": 0.009, "step": 25550 }, { "epoch": 22.72, "grad_norm": 0.14847828447818756, "learning_rate": 1.5257841328279727e-05, "loss": 0.0103, "step": 25560 }, { "epoch": 22.72888888888889, "grad_norm": 0.3424551784992218, "learning_rate": 1.5222624529456387e-05, "loss": 0.0108, "step": 25570 }, { "epoch": 22.73777777777778, "grad_norm": 0.19082848727703094, "learning_rate": 1.5187441120546665e-05, "loss": 0.0113, "step": 25580 }, { "epoch": 22.746666666666666, "grad_norm": 0.18597109615802765, "learning_rate": 1.5152291135330287e-05, "loss": 0.0118, "step": 25590 }, { "epoch": 22.755555555555556, "grad_norm": 0.2573390007019043, "learning_rate": 1.5117174607554934e-05, "loss": 0.0146, "step": 25600 }, { "epoch": 22.764444444444443, "grad_norm": 0.236512690782547, "learning_rate": 1.5082091570936091e-05, "loss": 0.011, "step": 25610 }, { "epoch": 22.773333333333333, "grad_norm": 0.19493290781974792, "learning_rate": 1.5047042059157184e-05, "loss": 0.013, "step": 25620 }, { "epoch": 22.782222222222224, "grad_norm": 0.255720853805542, "learning_rate": 1.5012026105869392e-05, "loss": 0.0169, "step": 25630 }, { "epoch": 22.79111111111111, "grad_norm": 0.24535112082958221, "learning_rate": 1.4977043744691687e-05, "loss": 0.0096, "step": 25640 }, { "epoch": 22.8, "grad_norm": 0.24752256274223328, "learning_rate": 1.4942095009210789e-05, "loss": 0.012, "step": 25650 }, { "epoch": 22.808888888888887, "grad_norm": 0.2549291253089905, "learning_rate": 1.4907179932981119e-05, "loss": 0.0114, "step": 25660 }, { "epoch": 22.817777777777778, "grad_norm": 0.224941685795784, "learning_rate": 1.4872298549524838e-05, "loss": 0.0121, "step": 25670 }, { "epoch": 22.826666666666668, "grad_norm": 0.24357996881008148, "learning_rate": 1.483745089233169e-05, "loss": 0.0112, "step": 25680 }, { "epoch": 22.835555555555555, "grad_norm": 0.24497729539871216, "learning_rate": 1.480263699485907e-05, "loss": 0.009, "step": 25690 }, { "epoch": 22.844444444444445, "grad_norm": 0.17292654514312744, "learning_rate": 1.4767856890531961e-05, "loss": 0.0126, "step": 25700 }, { "epoch": 22.85333333333333, "grad_norm": 0.2006794810295105, "learning_rate": 1.4733110612742896e-05, "loss": 0.0104, "step": 25710 }, { "epoch": 22.862222222222222, "grad_norm": 0.2150603085756302, "learning_rate": 1.4698398194851931e-05, "loss": 0.0141, "step": 25720 }, { "epoch": 22.871111111111112, "grad_norm": 0.28575772047042847, "learning_rate": 1.4663719670186603e-05, "loss": 0.0133, "step": 25730 }, { "epoch": 22.88, "grad_norm": 0.25741031765937805, "learning_rate": 1.4629075072041936e-05, "loss": 0.0116, "step": 25740 }, { "epoch": 22.88888888888889, "grad_norm": 0.22067786753177643, "learning_rate": 1.4594464433680338e-05, "loss": 0.0122, "step": 25750 }, { "epoch": 22.897777777777776, "grad_norm": 0.15600574016571045, "learning_rate": 1.4559887788331672e-05, "loss": 0.011, "step": 25760 }, { "epoch": 22.906666666666666, "grad_norm": 0.12646980583667755, "learning_rate": 1.4525345169193106e-05, "loss": 0.0123, "step": 25770 }, { "epoch": 22.915555555555557, "grad_norm": 0.20477469265460968, "learning_rate": 1.4490836609429176e-05, "loss": 0.0136, "step": 25780 }, { "epoch": 22.924444444444443, "grad_norm": 0.23756495118141174, "learning_rate": 1.4456362142171692e-05, "loss": 0.0101, "step": 25790 }, { "epoch": 22.933333333333334, "grad_norm": 0.22864443063735962, "learning_rate": 1.4421921800519738e-05, "loss": 0.0141, "step": 25800 }, { "epoch": 22.942222222222224, "grad_norm": 0.21471622586250305, "learning_rate": 1.4387515617539682e-05, "loss": 0.0107, "step": 25810 }, { "epoch": 22.95111111111111, "grad_norm": 0.20771819353103638, "learning_rate": 1.4353143626265008e-05, "loss": 0.0092, "step": 25820 }, { "epoch": 22.96, "grad_norm": 0.21719345450401306, "learning_rate": 1.4318805859696438e-05, "loss": 0.0135, "step": 25830 }, { "epoch": 22.968888888888888, "grad_norm": 0.25745370984077454, "learning_rate": 1.4284502350801793e-05, "loss": 0.0125, "step": 25840 }, { "epoch": 22.977777777777778, "grad_norm": 0.21046233177185059, "learning_rate": 1.4250233132516061e-05, "loss": 0.0116, "step": 25850 }, { "epoch": 22.986666666666668, "grad_norm": 0.19565655291080475, "learning_rate": 1.4215998237741252e-05, "loss": 0.0127, "step": 25860 }, { "epoch": 22.995555555555555, "grad_norm": 0.15429292619228363, "learning_rate": 1.4181797699346445e-05, "loss": 0.0162, "step": 25870 }, { "epoch": 23.004444444444445, "grad_norm": 0.2590973973274231, "learning_rate": 1.4147631550167723e-05, "loss": 0.0088, "step": 25880 }, { "epoch": 23.013333333333332, "grad_norm": 0.2467617690563202, "learning_rate": 1.4113499823008152e-05, "loss": 0.0146, "step": 25890 }, { "epoch": 23.022222222222222, "grad_norm": 0.2033395916223526, "learning_rate": 1.4079402550637782e-05, "loss": 0.0098, "step": 25900 }, { "epoch": 23.031111111111112, "grad_norm": 0.258000910282135, "learning_rate": 1.4045339765793564e-05, "loss": 0.0137, "step": 25910 }, { "epoch": 23.04, "grad_norm": 0.20602576434612274, "learning_rate": 1.4011311501179286e-05, "loss": 0.0118, "step": 25920 }, { "epoch": 23.04888888888889, "grad_norm": 0.19731101393699646, "learning_rate": 1.397731778946565e-05, "loss": 0.011, "step": 25930 }, { "epoch": 23.057777777777776, "grad_norm": 0.3263006806373596, "learning_rate": 1.394335866329019e-05, "loss": 0.0128, "step": 25940 }, { "epoch": 23.066666666666666, "grad_norm": 0.18633420765399933, "learning_rate": 1.3909434155257201e-05, "loss": 0.0099, "step": 25950 }, { "epoch": 23.075555555555557, "grad_norm": 0.1588805615901947, "learning_rate": 1.3875544297937764e-05, "loss": 0.0116, "step": 25960 }, { "epoch": 23.084444444444443, "grad_norm": 0.1902308464050293, "learning_rate": 1.3841689123869678e-05, "loss": 0.0098, "step": 25970 }, { "epoch": 23.093333333333334, "grad_norm": 0.24751222133636475, "learning_rate": 1.3807868665557427e-05, "loss": 0.0125, "step": 25980 }, { "epoch": 23.102222222222224, "grad_norm": 0.1255027800798416, "learning_rate": 1.3774082955472218e-05, "loss": 0.0099, "step": 25990 }, { "epoch": 23.11111111111111, "grad_norm": 0.18048429489135742, "learning_rate": 1.3740332026051844e-05, "loss": 0.0096, "step": 26000 }, { "epoch": 23.12, "grad_norm": 0.18753677606582642, "learning_rate": 1.3706615909700749e-05, "loss": 0.017, "step": 26010 }, { "epoch": 23.128888888888888, "grad_norm": 0.22587570548057556, "learning_rate": 1.3672934638789869e-05, "loss": 0.0129, "step": 26020 }, { "epoch": 23.137777777777778, "grad_norm": 0.1849626898765564, "learning_rate": 1.3639288245656796e-05, "loss": 0.0125, "step": 26030 }, { "epoch": 23.14666666666667, "grad_norm": 0.2625523507595062, "learning_rate": 1.3605676762605557e-05, "loss": 0.0114, "step": 26040 }, { "epoch": 23.155555555555555, "grad_norm": 0.12550987303256989, "learning_rate": 1.3572100221906698e-05, "loss": 0.0071, "step": 26050 }, { "epoch": 23.164444444444445, "grad_norm": 0.21835565567016602, "learning_rate": 1.3538558655797207e-05, "loss": 0.0099, "step": 26060 }, { "epoch": 23.173333333333332, "grad_norm": 0.1750466227531433, "learning_rate": 1.3505052096480464e-05, "loss": 0.0183, "step": 26070 }, { "epoch": 23.182222222222222, "grad_norm": 0.19134818017482758, "learning_rate": 1.3471580576126308e-05, "loss": 0.0123, "step": 26080 }, { "epoch": 23.191111111111113, "grad_norm": 0.2348521649837494, "learning_rate": 1.3438144126870878e-05, "loss": 0.0121, "step": 26090 }, { "epoch": 23.2, "grad_norm": 0.35233768820762634, "learning_rate": 1.3404742780816654e-05, "loss": 0.0103, "step": 26100 }, { "epoch": 23.20888888888889, "grad_norm": 0.24367643892765045, "learning_rate": 1.3371376570032418e-05, "loss": 0.0127, "step": 26110 }, { "epoch": 23.217777777777776, "grad_norm": 0.22742266952991486, "learning_rate": 1.3338045526553222e-05, "loss": 0.012, "step": 26120 }, { "epoch": 23.226666666666667, "grad_norm": 0.1071775034070015, "learning_rate": 1.3304749682380351e-05, "loss": 0.012, "step": 26130 }, { "epoch": 23.235555555555557, "grad_norm": 0.2840428650379181, "learning_rate": 1.3271489069481286e-05, "loss": 0.0109, "step": 26140 }, { "epoch": 23.244444444444444, "grad_norm": 0.21054375171661377, "learning_rate": 1.32382637197897e-05, "loss": 0.0117, "step": 26150 }, { "epoch": 23.253333333333334, "grad_norm": 0.10815328359603882, "learning_rate": 1.3205073665205375e-05, "loss": 0.0102, "step": 26160 }, { "epoch": 23.26222222222222, "grad_norm": 0.3528992235660553, "learning_rate": 1.3171918937594269e-05, "loss": 0.0073, "step": 26170 }, { "epoch": 23.27111111111111, "grad_norm": 0.26749274134635925, "learning_rate": 1.3138799568788363e-05, "loss": 0.0133, "step": 26180 }, { "epoch": 23.28, "grad_norm": 0.23333585262298584, "learning_rate": 1.310571559058571e-05, "loss": 0.0108, "step": 26190 }, { "epoch": 23.288888888888888, "grad_norm": 0.1875968724489212, "learning_rate": 1.3072667034750397e-05, "loss": 0.011, "step": 26200 }, { "epoch": 23.297777777777778, "grad_norm": 0.20837616920471191, "learning_rate": 1.3039653933012458e-05, "loss": 0.0118, "step": 26210 }, { "epoch": 23.306666666666665, "grad_norm": 0.2190576046705246, "learning_rate": 1.3006676317067978e-05, "loss": 0.0168, "step": 26220 }, { "epoch": 23.315555555555555, "grad_norm": 0.21219132840633392, "learning_rate": 1.2973734218578865e-05, "loss": 0.0107, "step": 26230 }, { "epoch": 23.324444444444445, "grad_norm": 0.20089153945446014, "learning_rate": 1.2940827669172973e-05, "loss": 0.0118, "step": 26240 }, { "epoch": 23.333333333333332, "grad_norm": 0.27749061584472656, "learning_rate": 1.2907956700444057e-05, "loss": 0.0074, "step": 26250 }, { "epoch": 23.342222222222222, "grad_norm": 0.2920725643634796, "learning_rate": 1.2875121343951668e-05, "loss": 0.0134, "step": 26260 }, { "epoch": 23.351111111111113, "grad_norm": 0.2700279653072357, "learning_rate": 1.2842321631221176e-05, "loss": 0.0123, "step": 26270 }, { "epoch": 23.36, "grad_norm": 0.14764003455638885, "learning_rate": 1.2809557593743732e-05, "loss": 0.012, "step": 26280 }, { "epoch": 23.36888888888889, "grad_norm": 0.2268606722354889, "learning_rate": 1.2776829262976243e-05, "loss": 0.0087, "step": 26290 }, { "epoch": 23.377777777777776, "grad_norm": 0.25908100605010986, "learning_rate": 1.2744136670341294e-05, "loss": 0.0109, "step": 26300 }, { "epoch": 23.386666666666667, "grad_norm": 0.17933572828769684, "learning_rate": 1.2711479847227237e-05, "loss": 0.0121, "step": 26310 }, { "epoch": 23.395555555555557, "grad_norm": 0.2852003276348114, "learning_rate": 1.2678858824988022e-05, "loss": 0.0092, "step": 26320 }, { "epoch": 23.404444444444444, "grad_norm": 0.2519080638885498, "learning_rate": 1.2646273634943196e-05, "loss": 0.0106, "step": 26330 }, { "epoch": 23.413333333333334, "grad_norm": 0.17292985320091248, "learning_rate": 1.2613724308377989e-05, "loss": 0.0093, "step": 26340 }, { "epoch": 23.42222222222222, "grad_norm": 0.31645432114601135, "learning_rate": 1.2581210876543148e-05, "loss": 0.0131, "step": 26350 }, { "epoch": 23.43111111111111, "grad_norm": 0.1876762956380844, "learning_rate": 1.2548733370654953e-05, "loss": 0.0108, "step": 26360 }, { "epoch": 23.44, "grad_norm": 0.2226199358701706, "learning_rate": 1.251629182189521e-05, "loss": 0.0096, "step": 26370 }, { "epoch": 23.448888888888888, "grad_norm": 0.15590117871761322, "learning_rate": 1.2483886261411182e-05, "loss": 0.0164, "step": 26380 }, { "epoch": 23.45777777777778, "grad_norm": 0.14505432546138763, "learning_rate": 1.2451516720315615e-05, "loss": 0.0104, "step": 26390 }, { "epoch": 23.466666666666665, "grad_norm": 0.15713393688201904, "learning_rate": 1.2419183229686638e-05, "loss": 0.0148, "step": 26400 }, { "epoch": 23.475555555555555, "grad_norm": 0.17469114065170288, "learning_rate": 1.2386885820567785e-05, "loss": 0.0109, "step": 26410 }, { "epoch": 23.484444444444446, "grad_norm": 0.14535650610923767, "learning_rate": 1.2354624523967933e-05, "loss": 0.0119, "step": 26420 }, { "epoch": 23.493333333333332, "grad_norm": 0.11864432692527771, "learning_rate": 1.2322399370861304e-05, "loss": 0.0081, "step": 26430 }, { "epoch": 23.502222222222223, "grad_norm": 0.18320442736148834, "learning_rate": 1.2290210392187407e-05, "loss": 0.0093, "step": 26440 }, { "epoch": 23.511111111111113, "grad_norm": 0.2841358482837677, "learning_rate": 1.2258057618851032e-05, "loss": 0.0091, "step": 26450 }, { "epoch": 23.52, "grad_norm": 0.1412035971879959, "learning_rate": 1.2225941081722193e-05, "loss": 0.0101, "step": 26460 }, { "epoch": 23.52888888888889, "grad_norm": 0.22607092559337616, "learning_rate": 1.219386081163611e-05, "loss": 0.0127, "step": 26470 }, { "epoch": 23.537777777777777, "grad_norm": 0.2323736548423767, "learning_rate": 1.216181683939322e-05, "loss": 0.0134, "step": 26480 }, { "epoch": 23.546666666666667, "grad_norm": 0.1941043734550476, "learning_rate": 1.2129809195759067e-05, "loss": 0.0122, "step": 26490 }, { "epoch": 23.555555555555557, "grad_norm": 0.22675015032291412, "learning_rate": 1.2097837911464344e-05, "loss": 0.0111, "step": 26500 }, { "epoch": 23.564444444444444, "grad_norm": 0.21676088869571686, "learning_rate": 1.206590301720481e-05, "loss": 0.0085, "step": 26510 }, { "epoch": 23.573333333333334, "grad_norm": 0.17867739498615265, "learning_rate": 1.2034004543641292e-05, "loss": 0.011, "step": 26520 }, { "epoch": 23.58222222222222, "grad_norm": 0.2026766687631607, "learning_rate": 1.20021425213997e-05, "loss": 0.0146, "step": 26530 }, { "epoch": 23.59111111111111, "grad_norm": 0.22342705726623535, "learning_rate": 1.1970316981070868e-05, "loss": 0.0144, "step": 26540 }, { "epoch": 23.6, "grad_norm": 0.13033637404441833, "learning_rate": 1.1938527953210638e-05, "loss": 0.0097, "step": 26550 }, { "epoch": 23.608888888888888, "grad_norm": 0.1266183853149414, "learning_rate": 1.1906775468339798e-05, "loss": 0.0142, "step": 26560 }, { "epoch": 23.61777777777778, "grad_norm": 0.20786520838737488, "learning_rate": 1.1875059556944068e-05, "loss": 0.0113, "step": 26570 }, { "epoch": 23.626666666666665, "grad_norm": 0.1927679181098938, "learning_rate": 1.1843380249474028e-05, "loss": 0.015, "step": 26580 }, { "epoch": 23.635555555555555, "grad_norm": 0.12447602301836014, "learning_rate": 1.1811737576345123e-05, "loss": 0.0108, "step": 26590 }, { "epoch": 23.644444444444446, "grad_norm": 0.4034142792224884, "learning_rate": 1.1780131567937625e-05, "loss": 0.0105, "step": 26600 }, { "epoch": 23.653333333333332, "grad_norm": 0.15835300087928772, "learning_rate": 1.1748562254596585e-05, "loss": 0.0083, "step": 26610 }, { "epoch": 23.662222222222223, "grad_norm": 0.19128014147281647, "learning_rate": 1.171702966663188e-05, "loss": 0.0154, "step": 26620 }, { "epoch": 23.67111111111111, "grad_norm": 0.263559490442276, "learning_rate": 1.1685533834318073e-05, "loss": 0.0161, "step": 26630 }, { "epoch": 23.68, "grad_norm": 0.16297659277915955, "learning_rate": 1.1654074787894441e-05, "loss": 0.012, "step": 26640 }, { "epoch": 23.68888888888889, "grad_norm": 0.2136104851961136, "learning_rate": 1.1622652557564945e-05, "loss": 0.0099, "step": 26650 }, { "epoch": 23.697777777777777, "grad_norm": 0.2072393149137497, "learning_rate": 1.1591267173498243e-05, "loss": 0.0117, "step": 26660 }, { "epoch": 23.706666666666667, "grad_norm": 0.2615163028240204, "learning_rate": 1.1559918665827563e-05, "loss": 0.0127, "step": 26670 }, { "epoch": 23.715555555555554, "grad_norm": 0.227461040019989, "learning_rate": 1.1528607064650748e-05, "loss": 0.0101, "step": 26680 }, { "epoch": 23.724444444444444, "grad_norm": 0.1530824601650238, "learning_rate": 1.1497332400030213e-05, "loss": 0.0097, "step": 26690 }, { "epoch": 23.733333333333334, "grad_norm": 0.24422793090343475, "learning_rate": 1.1466094701992886e-05, "loss": 0.0104, "step": 26700 }, { "epoch": 23.74222222222222, "grad_norm": 0.23198530077934265, "learning_rate": 1.1434894000530239e-05, "loss": 0.0113, "step": 26710 }, { "epoch": 23.75111111111111, "grad_norm": 0.18568123877048492, "learning_rate": 1.1403730325598206e-05, "loss": 0.0107, "step": 26720 }, { "epoch": 23.76, "grad_norm": 0.14810414612293243, "learning_rate": 1.1372603707117174e-05, "loss": 0.0087, "step": 26730 }, { "epoch": 23.76888888888889, "grad_norm": 0.1430690437555313, "learning_rate": 1.1341514174971907e-05, "loss": 0.0107, "step": 26740 }, { "epoch": 23.77777777777778, "grad_norm": 0.221278578042984, "learning_rate": 1.1310461759011653e-05, "loss": 0.0095, "step": 26750 }, { "epoch": 23.786666666666665, "grad_norm": 0.17368175089359283, "learning_rate": 1.1279446489049956e-05, "loss": 0.0095, "step": 26760 }, { "epoch": 23.795555555555556, "grad_norm": 0.23779454827308655, "learning_rate": 1.1248468394864713e-05, "loss": 0.0128, "step": 26770 }, { "epoch": 23.804444444444446, "grad_norm": 0.14516310393810272, "learning_rate": 1.1217527506198139e-05, "loss": 0.0082, "step": 26780 }, { "epoch": 23.813333333333333, "grad_norm": 0.27395346760749817, "learning_rate": 1.11866238527567e-05, "loss": 0.0091, "step": 26790 }, { "epoch": 23.822222222222223, "grad_norm": 0.19112397730350494, "learning_rate": 1.1155757464211158e-05, "loss": 0.0114, "step": 26800 }, { "epoch": 23.83111111111111, "grad_norm": 0.34953808784484863, "learning_rate": 1.1124928370196458e-05, "loss": 0.009, "step": 26810 }, { "epoch": 23.84, "grad_norm": 0.2410433441400528, "learning_rate": 1.1094136600311755e-05, "loss": 0.0114, "step": 26820 }, { "epoch": 23.84888888888889, "grad_norm": 0.19664309918880463, "learning_rate": 1.1063382184120357e-05, "loss": 0.0097, "step": 26830 }, { "epoch": 23.857777777777777, "grad_norm": 0.23512950539588928, "learning_rate": 1.1032665151149724e-05, "loss": 0.0106, "step": 26840 }, { "epoch": 23.866666666666667, "grad_norm": 0.20722690224647522, "learning_rate": 1.1001985530891413e-05, "loss": 0.0107, "step": 26850 }, { "epoch": 23.875555555555554, "grad_norm": 0.12087802588939667, "learning_rate": 1.0971343352801061e-05, "loss": 0.0112, "step": 26860 }, { "epoch": 23.884444444444444, "grad_norm": 0.25682783126831055, "learning_rate": 1.094073864629837e-05, "loss": 0.0105, "step": 26870 }, { "epoch": 23.893333333333334, "grad_norm": 0.2935711741447449, "learning_rate": 1.0910171440767031e-05, "loss": 0.0131, "step": 26880 }, { "epoch": 23.90222222222222, "grad_norm": 0.19921912252902985, "learning_rate": 1.0879641765554794e-05, "loss": 0.0078, "step": 26890 }, { "epoch": 23.91111111111111, "grad_norm": 0.27297598123550415, "learning_rate": 1.0849149649973305e-05, "loss": 0.0151, "step": 26900 }, { "epoch": 23.92, "grad_norm": 0.12878775596618652, "learning_rate": 1.08186951232982e-05, "loss": 0.0122, "step": 26910 }, { "epoch": 23.92888888888889, "grad_norm": 0.26666921377182007, "learning_rate": 1.0788278214768993e-05, "loss": 0.0137, "step": 26920 }, { "epoch": 23.93777777777778, "grad_norm": 0.1377442628145218, "learning_rate": 1.0757898953589091e-05, "loss": 0.0109, "step": 26930 }, { "epoch": 23.946666666666665, "grad_norm": 0.22246715426445007, "learning_rate": 1.072755736892579e-05, "loss": 0.0113, "step": 26940 }, { "epoch": 23.955555555555556, "grad_norm": 0.1507086604833603, "learning_rate": 1.0697253489910136e-05, "loss": 0.014, "step": 26950 }, { "epoch": 23.964444444444446, "grad_norm": 0.16194523870944977, "learning_rate": 1.0666987345637047e-05, "loss": 0.0079, "step": 26960 }, { "epoch": 23.973333333333333, "grad_norm": 0.14722757041454315, "learning_rate": 1.0636758965165161e-05, "loss": 0.0123, "step": 26970 }, { "epoch": 23.982222222222223, "grad_norm": 0.16170580685138702, "learning_rate": 1.0606568377516906e-05, "loss": 0.012, "step": 26980 }, { "epoch": 23.99111111111111, "grad_norm": 0.17281919717788696, "learning_rate": 1.0576415611678392e-05, "loss": 0.0098, "step": 26990 }, { "epoch": 24.0, "grad_norm": 0.3920215964317322, "learning_rate": 1.0546300696599426e-05, "loss": 0.0162, "step": 27000 }, { "epoch": 24.00888888888889, "grad_norm": 0.13438843190670013, "learning_rate": 1.0516223661193464e-05, "loss": 0.0113, "step": 27010 }, { "epoch": 24.017777777777777, "grad_norm": 0.28679358959198, "learning_rate": 1.0486184534337595e-05, "loss": 0.0135, "step": 27020 }, { "epoch": 24.026666666666667, "grad_norm": 0.1581319123506546, "learning_rate": 1.0456183344872544e-05, "loss": 0.0126, "step": 27030 }, { "epoch": 24.035555555555554, "grad_norm": 0.20108124613761902, "learning_rate": 1.0426220121602588e-05, "loss": 0.0121, "step": 27040 }, { "epoch": 24.044444444444444, "grad_norm": 0.15850886702537537, "learning_rate": 1.0396294893295521e-05, "loss": 0.012, "step": 27050 }, { "epoch": 24.053333333333335, "grad_norm": 0.14911819994449615, "learning_rate": 1.0366407688682694e-05, "loss": 0.0124, "step": 27060 }, { "epoch": 24.06222222222222, "grad_norm": 0.2176523208618164, "learning_rate": 1.0336558536458968e-05, "loss": 0.0116, "step": 27070 }, { "epoch": 24.07111111111111, "grad_norm": 0.21693246066570282, "learning_rate": 1.030674746528264e-05, "loss": 0.0102, "step": 27080 }, { "epoch": 24.08, "grad_norm": 0.24548675119876862, "learning_rate": 1.0276974503775445e-05, "loss": 0.0102, "step": 27090 }, { "epoch": 24.08888888888889, "grad_norm": 0.12987898290157318, "learning_rate": 1.024723968052254e-05, "loss": 0.0095, "step": 27100 }, { "epoch": 24.09777777777778, "grad_norm": 0.15087313950061798, "learning_rate": 1.0217543024072452e-05, "loss": 0.0127, "step": 27110 }, { "epoch": 24.106666666666666, "grad_norm": 0.18754148483276367, "learning_rate": 1.0187884562937095e-05, "loss": 0.0087, "step": 27120 }, { "epoch": 24.115555555555556, "grad_norm": 0.24597567319869995, "learning_rate": 1.0158264325591676e-05, "loss": 0.0113, "step": 27130 }, { "epoch": 24.124444444444446, "grad_norm": 0.24601148068904877, "learning_rate": 1.0128682340474721e-05, "loss": 0.0101, "step": 27140 }, { "epoch": 24.133333333333333, "grad_norm": 0.14961457252502441, "learning_rate": 1.0099138635988026e-05, "loss": 0.0121, "step": 27150 }, { "epoch": 24.142222222222223, "grad_norm": 0.29618337750434875, "learning_rate": 1.0069633240496634e-05, "loss": 0.0116, "step": 27160 }, { "epoch": 24.15111111111111, "grad_norm": 0.2536132335662842, "learning_rate": 1.0040166182328803e-05, "loss": 0.0134, "step": 27170 }, { "epoch": 24.16, "grad_norm": 0.18167299032211304, "learning_rate": 1.001073748977599e-05, "loss": 0.0091, "step": 27180 }, { "epoch": 24.16888888888889, "grad_norm": 0.1648467779159546, "learning_rate": 9.98134719109281e-06, "loss": 0.0112, "step": 27190 }, { "epoch": 24.177777777777777, "grad_norm": 0.31653502583503723, "learning_rate": 9.951995314497015e-06, "loss": 0.0098, "step": 27200 }, { "epoch": 24.186666666666667, "grad_norm": 0.37203335762023926, "learning_rate": 9.922681888169494e-06, "loss": 0.01, "step": 27210 }, { "epoch": 24.195555555555554, "grad_norm": 0.2629411518573761, "learning_rate": 9.893406940254185e-06, "loss": 0.0102, "step": 27220 }, { "epoch": 24.204444444444444, "grad_norm": 0.1742517054080963, "learning_rate": 9.8641704988581e-06, "loss": 0.0091, "step": 27230 }, { "epoch": 24.213333333333335, "grad_norm": 0.31625795364379883, "learning_rate": 9.834972592051278e-06, "loss": 0.0089, "step": 27240 }, { "epoch": 24.22222222222222, "grad_norm": 0.1848197877407074, "learning_rate": 9.805813247866752e-06, "loss": 0.0109, "step": 27250 }, { "epoch": 24.23111111111111, "grad_norm": 0.27314293384552, "learning_rate": 9.776692494300543e-06, "loss": 0.011, "step": 27260 }, { "epoch": 24.24, "grad_norm": 0.15141908824443817, "learning_rate": 9.74761035931162e-06, "loss": 0.0093, "step": 27270 }, { "epoch": 24.24888888888889, "grad_norm": 0.14292550086975098, "learning_rate": 9.718566870821865e-06, "loss": 0.0086, "step": 27280 }, { "epoch": 24.25777777777778, "grad_norm": 0.5070074796676636, "learning_rate": 9.689562056716051e-06, "loss": 0.0146, "step": 27290 }, { "epoch": 24.266666666666666, "grad_norm": 0.242275208234787, "learning_rate": 9.660595944841861e-06, "loss": 0.0102, "step": 27300 }, { "epoch": 24.275555555555556, "grad_norm": 0.2134469598531723, "learning_rate": 9.63166856300976e-06, "loss": 0.0113, "step": 27310 }, { "epoch": 24.284444444444443, "grad_norm": 0.3297756016254425, "learning_rate": 9.602779938993073e-06, "loss": 0.0111, "step": 27320 }, { "epoch": 24.293333333333333, "grad_norm": 0.24283146858215332, "learning_rate": 9.57393010052789e-06, "loss": 0.0127, "step": 27330 }, { "epoch": 24.302222222222223, "grad_norm": 0.24007068574428558, "learning_rate": 9.545119075313058e-06, "loss": 0.0153, "step": 27340 }, { "epoch": 24.31111111111111, "grad_norm": 0.17381979525089264, "learning_rate": 9.516346891010208e-06, "loss": 0.0123, "step": 27350 }, { "epoch": 24.32, "grad_norm": 0.1964290887117386, "learning_rate": 9.487613575243603e-06, "loss": 0.013, "step": 27360 }, { "epoch": 24.32888888888889, "grad_norm": 0.2363298535346985, "learning_rate": 9.458919155600243e-06, "loss": 0.0135, "step": 27370 }, { "epoch": 24.337777777777777, "grad_norm": 0.19435954093933105, "learning_rate": 9.430263659629763e-06, "loss": 0.0106, "step": 27380 }, { "epoch": 24.346666666666668, "grad_norm": 0.2324320524930954, "learning_rate": 9.401647114844442e-06, "loss": 0.0104, "step": 27390 }, { "epoch": 24.355555555555554, "grad_norm": 0.23716668784618378, "learning_rate": 9.373069548719155e-06, "loss": 0.0141, "step": 27400 }, { "epoch": 24.364444444444445, "grad_norm": 0.16137292981147766, "learning_rate": 9.344530988691341e-06, "loss": 0.0114, "step": 27410 }, { "epoch": 24.373333333333335, "grad_norm": 0.09891986846923828, "learning_rate": 9.316031462161002e-06, "loss": 0.0109, "step": 27420 }, { "epoch": 24.38222222222222, "grad_norm": 0.19035787880420685, "learning_rate": 9.287570996490652e-06, "loss": 0.0094, "step": 27430 }, { "epoch": 24.391111111111112, "grad_norm": 0.1980881541967392, "learning_rate": 9.25914961900533e-06, "loss": 0.0119, "step": 27440 }, { "epoch": 24.4, "grad_norm": 0.19054315984249115, "learning_rate": 9.230767356992532e-06, "loss": 0.0116, "step": 27450 }, { "epoch": 24.40888888888889, "grad_norm": 0.29284849762916565, "learning_rate": 9.202424237702172e-06, "loss": 0.0146, "step": 27460 }, { "epoch": 24.41777777777778, "grad_norm": 0.20855651795864105, "learning_rate": 9.174120288346599e-06, "loss": 0.0115, "step": 27470 }, { "epoch": 24.426666666666666, "grad_norm": 0.16889655590057373, "learning_rate": 9.14585553610059e-06, "loss": 0.0105, "step": 27480 }, { "epoch": 24.435555555555556, "grad_norm": 0.1800101399421692, "learning_rate": 9.117630008101241e-06, "loss": 0.013, "step": 27490 }, { "epoch": 24.444444444444443, "grad_norm": 0.18354281783103943, "learning_rate": 9.089443731448021e-06, "loss": 0.0099, "step": 27500 }, { "epoch": 24.453333333333333, "grad_norm": 0.16465869545936584, "learning_rate": 9.06129673320269e-06, "loss": 0.0108, "step": 27510 }, { "epoch": 24.462222222222223, "grad_norm": 0.33240851759910583, "learning_rate": 9.033189040389296e-06, "loss": 0.0137, "step": 27520 }, { "epoch": 24.47111111111111, "grad_norm": 0.21743819117546082, "learning_rate": 9.005120679994183e-06, "loss": 0.0126, "step": 27530 }, { "epoch": 24.48, "grad_norm": 0.19065852463245392, "learning_rate": 8.9770916789659e-06, "loss": 0.0152, "step": 27540 }, { "epoch": 24.488888888888887, "grad_norm": 0.18292690813541412, "learning_rate": 8.94910206421522e-06, "loss": 0.0161, "step": 27550 }, { "epoch": 24.497777777777777, "grad_norm": 0.2343302220106125, "learning_rate": 8.92115186261509e-06, "loss": 0.0128, "step": 27560 }, { "epoch": 24.506666666666668, "grad_norm": 0.3566910922527313, "learning_rate": 8.893241101000627e-06, "loss": 0.0146, "step": 27570 }, { "epoch": 24.515555555555554, "grad_norm": 0.20573027431964874, "learning_rate": 8.865369806169083e-06, "loss": 0.0112, "step": 27580 }, { "epoch": 24.524444444444445, "grad_norm": 0.19456163048744202, "learning_rate": 8.837538004879803e-06, "loss": 0.0137, "step": 27590 }, { "epoch": 24.533333333333335, "grad_norm": 0.17918357253074646, "learning_rate": 8.809745723854224e-06, "loss": 0.011, "step": 27600 }, { "epoch": 24.54222222222222, "grad_norm": 0.2174484133720398, "learning_rate": 8.781992989775839e-06, "loss": 0.0145, "step": 27610 }, { "epoch": 24.551111111111112, "grad_norm": 0.22625452280044556, "learning_rate": 8.75427982929018e-06, "loss": 0.012, "step": 27620 }, { "epoch": 24.56, "grad_norm": 0.19930271804332733, "learning_rate": 8.726606269004777e-06, "loss": 0.0117, "step": 27630 }, { "epoch": 24.56888888888889, "grad_norm": 0.19016757607460022, "learning_rate": 8.698972335489136e-06, "loss": 0.0107, "step": 27640 }, { "epoch": 24.57777777777778, "grad_norm": 0.13653865456581116, "learning_rate": 8.671378055274715e-06, "loss": 0.0121, "step": 27650 }, { "epoch": 24.586666666666666, "grad_norm": 0.2945425510406494, "learning_rate": 8.643823454854915e-06, "loss": 0.0097, "step": 27660 }, { "epoch": 24.595555555555556, "grad_norm": 0.22562111914157867, "learning_rate": 8.61630856068502e-06, "loss": 0.0107, "step": 27670 }, { "epoch": 24.604444444444443, "grad_norm": 0.420655757188797, "learning_rate": 8.588833399182212e-06, "loss": 0.0126, "step": 27680 }, { "epoch": 24.613333333333333, "grad_norm": 0.17704655230045319, "learning_rate": 8.561397996725506e-06, "loss": 0.0108, "step": 27690 }, { "epoch": 24.622222222222224, "grad_norm": 0.36299124360084534, "learning_rate": 8.534002379655753e-06, "loss": 0.0091, "step": 27700 }, { "epoch": 24.63111111111111, "grad_norm": 0.14988091588020325, "learning_rate": 8.506646574275628e-06, "loss": 0.0087, "step": 27710 }, { "epoch": 24.64, "grad_norm": 0.2442554235458374, "learning_rate": 8.479330606849544e-06, "loss": 0.0118, "step": 27720 }, { "epoch": 24.648888888888887, "grad_norm": 0.22386707365512848, "learning_rate": 8.452054503603695e-06, "loss": 0.01, "step": 27730 }, { "epoch": 24.657777777777778, "grad_norm": 0.25647488236427307, "learning_rate": 8.424818290725984e-06, "loss": 0.0139, "step": 27740 }, { "epoch": 24.666666666666668, "grad_norm": 0.1954212188720703, "learning_rate": 8.397621994366e-06, "loss": 0.0099, "step": 27750 }, { "epoch": 24.675555555555555, "grad_norm": 0.2548512816429138, "learning_rate": 8.370465640635079e-06, "loss": 0.011, "step": 27760 }, { "epoch": 24.684444444444445, "grad_norm": 0.11220848560333252, "learning_rate": 8.343349255606115e-06, "loss": 0.0102, "step": 27770 }, { "epoch": 24.693333333333335, "grad_norm": 0.21980035305023193, "learning_rate": 8.316272865313679e-06, "loss": 0.0103, "step": 27780 }, { "epoch": 24.702222222222222, "grad_norm": 0.3126482367515564, "learning_rate": 8.289236495753932e-06, "loss": 0.0073, "step": 27790 }, { "epoch": 24.711111111111112, "grad_norm": 0.18589533865451813, "learning_rate": 8.262240172884628e-06, "loss": 0.0127, "step": 27800 }, { "epoch": 24.72, "grad_norm": 0.1922875940799713, "learning_rate": 8.235283922625048e-06, "loss": 0.0118, "step": 27810 }, { "epoch": 24.72888888888889, "grad_norm": 0.18335191905498505, "learning_rate": 8.208367770856007e-06, "loss": 0.0116, "step": 27820 }, { "epoch": 24.73777777777778, "grad_norm": 0.1449785679578781, "learning_rate": 8.181491743419823e-06, "loss": 0.0121, "step": 27830 }, { "epoch": 24.746666666666666, "grad_norm": 0.18011446297168732, "learning_rate": 8.154655866120286e-06, "loss": 0.0139, "step": 27840 }, { "epoch": 24.755555555555556, "grad_norm": 0.23604266345500946, "learning_rate": 8.127860164722656e-06, "loss": 0.0135, "step": 27850 }, { "epoch": 24.764444444444443, "grad_norm": 0.2653646767139435, "learning_rate": 8.101104664953597e-06, "loss": 0.0126, "step": 27860 }, { "epoch": 24.773333333333333, "grad_norm": 0.3328014314174652, "learning_rate": 8.074389392501192e-06, "loss": 0.0107, "step": 27870 }, { "epoch": 24.782222222222224, "grad_norm": 0.22810791432857513, "learning_rate": 8.047714373014864e-06, "loss": 0.0086, "step": 27880 }, { "epoch": 24.79111111111111, "grad_norm": 0.3385538160800934, "learning_rate": 8.02107963210545e-06, "loss": 0.013, "step": 27890 }, { "epoch": 24.8, "grad_norm": 0.17570675909519196, "learning_rate": 7.994485195345074e-06, "loss": 0.0095, "step": 27900 }, { "epoch": 24.808888888888887, "grad_norm": 0.13836295902729034, "learning_rate": 7.967931088267178e-06, "loss": 0.0099, "step": 27910 }, { "epoch": 24.817777777777778, "grad_norm": 0.11109130084514618, "learning_rate": 7.94141733636647e-06, "loss": 0.0092, "step": 27920 }, { "epoch": 24.826666666666668, "grad_norm": 0.2868214547634125, "learning_rate": 7.914943965098908e-06, "loss": 0.0084, "step": 27930 }, { "epoch": 24.835555555555555, "grad_norm": 0.20623454451560974, "learning_rate": 7.888510999881731e-06, "loss": 0.0099, "step": 27940 }, { "epoch": 24.844444444444445, "grad_norm": 0.172388955950737, "learning_rate": 7.862118466093321e-06, "loss": 0.0096, "step": 27950 }, { "epoch": 24.85333333333333, "grad_norm": 0.32102781534194946, "learning_rate": 7.835766389073279e-06, "loss": 0.0104, "step": 27960 }, { "epoch": 24.862222222222222, "grad_norm": 0.2310311645269394, "learning_rate": 7.809454794122344e-06, "loss": 0.0101, "step": 27970 }, { "epoch": 24.871111111111112, "grad_norm": 0.2271656095981598, "learning_rate": 7.783183706502406e-06, "loss": 0.0145, "step": 27980 }, { "epoch": 24.88, "grad_norm": 0.287925124168396, "learning_rate": 7.756953151436442e-06, "loss": 0.0113, "step": 27990 }, { "epoch": 24.88888888888889, "grad_norm": 0.344507098197937, "learning_rate": 7.730763154108528e-06, "loss": 0.0118, "step": 28000 }, { "epoch": 24.897777777777776, "grad_norm": 0.17223963141441345, "learning_rate": 7.704613739663796e-06, "loss": 0.0099, "step": 28010 }, { "epoch": 24.906666666666666, "grad_norm": 0.12462976574897766, "learning_rate": 7.678504933208414e-06, "loss": 0.012, "step": 28020 }, { "epoch": 24.915555555555557, "grad_norm": 0.11083123832941055, "learning_rate": 7.652436759809579e-06, "loss": 0.0093, "step": 28030 }, { "epoch": 24.924444444444443, "grad_norm": 0.22653871774673462, "learning_rate": 7.626409244495447e-06, "loss": 0.0154, "step": 28040 }, { "epoch": 24.933333333333334, "grad_norm": 0.24361586570739746, "learning_rate": 7.600422412255154e-06, "loss": 0.0126, "step": 28050 }, { "epoch": 24.942222222222224, "grad_norm": 0.2423054426908493, "learning_rate": 7.5744762880387785e-06, "loss": 0.0125, "step": 28060 }, { "epoch": 24.95111111111111, "grad_norm": 0.19851109385490417, "learning_rate": 7.548570896757306e-06, "loss": 0.0125, "step": 28070 }, { "epoch": 24.96, "grad_norm": 0.12606097757816315, "learning_rate": 7.52270626328262e-06, "loss": 0.0101, "step": 28080 }, { "epoch": 24.968888888888888, "grad_norm": 0.2708829939365387, "learning_rate": 7.49688241244747e-06, "loss": 0.0122, "step": 28090 }, { "epoch": 24.977777777777778, "grad_norm": 0.20652985572814941, "learning_rate": 7.471099369045448e-06, "loss": 0.012, "step": 28100 }, { "epoch": 24.986666666666668, "grad_norm": 0.27860113978385925, "learning_rate": 7.4453571578309624e-06, "loss": 0.0083, "step": 28110 }, { "epoch": 24.995555555555555, "grad_norm": 0.31638577580451965, "learning_rate": 7.419655803519243e-06, "loss": 0.009, "step": 28120 }, { "epoch": 25.004444444444445, "grad_norm": 0.2011132538318634, "learning_rate": 7.393995330786269e-06, "loss": 0.0122, "step": 28130 }, { "epoch": 25.013333333333332, "grad_norm": 0.162990003824234, "learning_rate": 7.36837576426877e-06, "loss": 0.0181, "step": 28140 }, { "epoch": 25.022222222222222, "grad_norm": 0.32856887578964233, "learning_rate": 7.34279712856421e-06, "loss": 0.0078, "step": 28150 }, { "epoch": 25.031111111111112, "grad_norm": 0.22615575790405273, "learning_rate": 7.31725944823074e-06, "loss": 0.0083, "step": 28160 }, { "epoch": 25.04, "grad_norm": 0.2773122489452362, "learning_rate": 7.291762747787223e-06, "loss": 0.0146, "step": 28170 }, { "epoch": 25.04888888888889, "grad_norm": 0.19024738669395447, "learning_rate": 7.2663070517131506e-06, "loss": 0.0098, "step": 28180 }, { "epoch": 25.057777777777776, "grad_norm": 0.11498946696519852, "learning_rate": 7.2408923844486334e-06, "loss": 0.0126, "step": 28190 }, { "epoch": 25.066666666666666, "grad_norm": 0.23978331685066223, "learning_rate": 7.215518770394403e-06, "loss": 0.0102, "step": 28200 }, { "epoch": 25.075555555555557, "grad_norm": 0.2571304142475128, "learning_rate": 7.190186233911805e-06, "loss": 0.0132, "step": 28210 }, { "epoch": 25.084444444444443, "grad_norm": 0.21632589399814606, "learning_rate": 7.164894799322714e-06, "loss": 0.0105, "step": 28220 }, { "epoch": 25.093333333333334, "grad_norm": 0.13900502026081085, "learning_rate": 7.139644490909536e-06, "loss": 0.0105, "step": 28230 }, { "epoch": 25.102222222222224, "grad_norm": 0.1731635481119156, "learning_rate": 7.114435332915221e-06, "loss": 0.0115, "step": 28240 }, { "epoch": 25.11111111111111, "grad_norm": 0.19539745151996613, "learning_rate": 7.089267349543172e-06, "loss": 0.0112, "step": 28250 }, { "epoch": 25.12, "grad_norm": 0.36959269642829895, "learning_rate": 7.064140564957306e-06, "loss": 0.0111, "step": 28260 }, { "epoch": 25.128888888888888, "grad_norm": 0.18008632957935333, "learning_rate": 7.039055003281952e-06, "loss": 0.0101, "step": 28270 }, { "epoch": 25.137777777777778, "grad_norm": 0.20967233180999756, "learning_rate": 7.0140106886018854e-06, "loss": 0.0174, "step": 28280 }, { "epoch": 25.14666666666667, "grad_norm": 0.3187423348426819, "learning_rate": 6.989007644962226e-06, "loss": 0.0105, "step": 28290 }, { "epoch": 25.155555555555555, "grad_norm": 0.36458608508110046, "learning_rate": 6.9640458963685375e-06, "loss": 0.0113, "step": 28300 }, { "epoch": 25.164444444444445, "grad_norm": 0.21886521577835083, "learning_rate": 6.9391254667867e-06, "loss": 0.0124, "step": 28310 }, { "epoch": 25.173333333333332, "grad_norm": 0.18324464559555054, "learning_rate": 6.914246380142936e-06, "loss": 0.0136, "step": 28320 }, { "epoch": 25.182222222222222, "grad_norm": 0.13113896548748016, "learning_rate": 6.889408660323754e-06, "loss": 0.0116, "step": 28330 }, { "epoch": 25.191111111111113, "grad_norm": 0.18746121227741241, "learning_rate": 6.864612331175957e-06, "loss": 0.0106, "step": 28340 }, { "epoch": 25.2, "grad_norm": 0.09454340487718582, "learning_rate": 6.839857416506629e-06, "loss": 0.0087, "step": 28350 }, { "epoch": 25.20888888888889, "grad_norm": 0.12882526218891144, "learning_rate": 6.8151439400830684e-06, "loss": 0.0129, "step": 28360 }, { "epoch": 25.217777777777776, "grad_norm": 0.24415692687034607, "learning_rate": 6.7904719256327964e-06, "loss": 0.0132, "step": 28370 }, { "epoch": 25.226666666666667, "grad_norm": 0.1766444891691208, "learning_rate": 6.765841396843514e-06, "loss": 0.0112, "step": 28380 }, { "epoch": 25.235555555555557, "grad_norm": 0.17661389708518982, "learning_rate": 6.741252377363111e-06, "loss": 0.0103, "step": 28390 }, { "epoch": 25.244444444444444, "grad_norm": 0.29812076687812805, "learning_rate": 6.716704890799608e-06, "loss": 0.0091, "step": 28400 }, { "epoch": 25.253333333333334, "grad_norm": 0.1969749629497528, "learning_rate": 6.6921989607211635e-06, "loss": 0.0095, "step": 28410 }, { "epoch": 25.26222222222222, "grad_norm": 0.0982021763920784, "learning_rate": 6.667734610656029e-06, "loss": 0.0105, "step": 28420 }, { "epoch": 25.27111111111111, "grad_norm": 0.22084352374076843, "learning_rate": 6.643311864092522e-06, "loss": 0.0114, "step": 28430 }, { "epoch": 25.28, "grad_norm": 0.2834000885486603, "learning_rate": 6.618930744479051e-06, "loss": 0.0107, "step": 28440 }, { "epoch": 25.288888888888888, "grad_norm": 0.2807549238204956, "learning_rate": 6.594591275224033e-06, "loss": 0.0109, "step": 28450 }, { "epoch": 25.297777777777778, "grad_norm": 0.1723727434873581, "learning_rate": 6.5702934796959e-06, "loss": 0.0122, "step": 28460 }, { "epoch": 25.306666666666665, "grad_norm": 0.2082793414592743, "learning_rate": 6.546037381223069e-06, "loss": 0.0154, "step": 28470 }, { "epoch": 25.315555555555555, "grad_norm": 0.2532082498073578, "learning_rate": 6.521823003093936e-06, "loss": 0.0102, "step": 28480 }, { "epoch": 25.324444444444445, "grad_norm": 0.113397017121315, "learning_rate": 6.497650368556823e-06, "loss": 0.0108, "step": 28490 }, { "epoch": 25.333333333333332, "grad_norm": 0.24627470970153809, "learning_rate": 6.4735195008199964e-06, "loss": 0.01, "step": 28500 }, { "epoch": 25.342222222222222, "grad_norm": 0.3004036247730255, "learning_rate": 6.4494304230516e-06, "loss": 0.011, "step": 28510 }, { "epoch": 25.351111111111113, "grad_norm": 0.19102109968662262, "learning_rate": 6.425383158379666e-06, "loss": 0.0095, "step": 28520 }, { "epoch": 25.36, "grad_norm": 0.17988403141498566, "learning_rate": 6.4013777298920886e-06, "loss": 0.0116, "step": 28530 }, { "epoch": 25.36888888888889, "grad_norm": 0.3283064663410187, "learning_rate": 6.377414160636586e-06, "loss": 0.0073, "step": 28540 }, { "epoch": 25.377777777777776, "grad_norm": 0.3047049045562744, "learning_rate": 6.353492473620687e-06, "loss": 0.0131, "step": 28550 }, { "epoch": 25.386666666666667, "grad_norm": 0.27931687235832214, "learning_rate": 6.329612691811704e-06, "loss": 0.0113, "step": 28560 }, { "epoch": 25.395555555555557, "grad_norm": 0.31503745913505554, "learning_rate": 6.305774838136719e-06, "loss": 0.0124, "step": 28570 }, { "epoch": 25.404444444444444, "grad_norm": 0.10827012360095978, "learning_rate": 6.281978935482574e-06, "loss": 0.0088, "step": 28580 }, { "epoch": 25.413333333333334, "grad_norm": 0.3671457767486572, "learning_rate": 6.258225006695833e-06, "loss": 0.0093, "step": 28590 }, { "epoch": 25.42222222222222, "grad_norm": 0.18268679082393646, "learning_rate": 6.2345130745827205e-06, "loss": 0.014, "step": 28600 }, { "epoch": 25.43111111111111, "grad_norm": 0.26559895277023315, "learning_rate": 6.210843161909169e-06, "loss": 0.0112, "step": 28610 }, { "epoch": 25.44, "grad_norm": 0.2442936897277832, "learning_rate": 6.187215291400794e-06, "loss": 0.008, "step": 28620 }, { "epoch": 25.448888888888888, "grad_norm": 0.21072585880756378, "learning_rate": 6.1636294857428e-06, "loss": 0.0105, "step": 28630 }, { "epoch": 25.45777777777778, "grad_norm": 0.21275325119495392, "learning_rate": 6.140085767580034e-06, "loss": 0.0069, "step": 28640 }, { "epoch": 25.466666666666665, "grad_norm": 0.13684211671352386, "learning_rate": 6.1165841595169136e-06, "loss": 0.0133, "step": 28650 }, { "epoch": 25.475555555555555, "grad_norm": 0.2662634551525116, "learning_rate": 6.093124684117441e-06, "loss": 0.0086, "step": 28660 }, { "epoch": 25.484444444444446, "grad_norm": 0.23286332190036774, "learning_rate": 6.069707363905169e-06, "loss": 0.0092, "step": 28670 }, { "epoch": 25.493333333333332, "grad_norm": 0.18972676992416382, "learning_rate": 6.046332221363177e-06, "loss": 0.0113, "step": 28680 }, { "epoch": 25.502222222222223, "grad_norm": 0.15671610832214355, "learning_rate": 6.022999278934044e-06, "loss": 0.0115, "step": 28690 }, { "epoch": 25.511111111111113, "grad_norm": 0.16633763909339905, "learning_rate": 5.999708559019801e-06, "loss": 0.0085, "step": 28700 }, { "epoch": 25.52, "grad_norm": 0.1961885690689087, "learning_rate": 5.976460083982011e-06, "loss": 0.01, "step": 28710 }, { "epoch": 25.52888888888889, "grad_norm": 0.09850001335144043, "learning_rate": 5.953253876141618e-06, "loss": 0.0101, "step": 28720 }, { "epoch": 25.537777777777777, "grad_norm": 0.18200142681598663, "learning_rate": 5.930089957779011e-06, "loss": 0.0121, "step": 28730 }, { "epoch": 25.546666666666667, "grad_norm": 0.23514842987060547, "learning_rate": 5.906968351133979e-06, "loss": 0.0096, "step": 28740 }, { "epoch": 25.555555555555557, "grad_norm": 0.18190932273864746, "learning_rate": 5.883889078405652e-06, "loss": 0.0105, "step": 28750 }, { "epoch": 25.564444444444444, "grad_norm": 0.18755628168582916, "learning_rate": 5.860852161752583e-06, "loss": 0.0095, "step": 28760 }, { "epoch": 25.573333333333334, "grad_norm": 0.219444140791893, "learning_rate": 5.837857623292592e-06, "loss": 0.0095, "step": 28770 }, { "epoch": 25.58222222222222, "grad_norm": 0.2414090931415558, "learning_rate": 5.8149054851028504e-06, "loss": 0.0115, "step": 28780 }, { "epoch": 25.59111111111111, "grad_norm": 0.16075499355793, "learning_rate": 5.791995769219799e-06, "loss": 0.0091, "step": 28790 }, { "epoch": 25.6, "grad_norm": 0.23128297924995422, "learning_rate": 5.769128497639165e-06, "loss": 0.0098, "step": 28800 }, { "epoch": 25.608888888888888, "grad_norm": 0.1866963505744934, "learning_rate": 5.746303692315913e-06, "loss": 0.01, "step": 28810 }, { "epoch": 25.61777777777778, "grad_norm": 0.15788811445236206, "learning_rate": 5.723521375164248e-06, "loss": 0.011, "step": 28820 }, { "epoch": 25.626666666666665, "grad_norm": 0.2194797843694687, "learning_rate": 5.7007815680575675e-06, "loss": 0.0083, "step": 28830 }, { "epoch": 25.635555555555555, "grad_norm": 0.1315634846687317, "learning_rate": 5.678084292828456e-06, "loss": 0.0139, "step": 28840 }, { "epoch": 25.644444444444446, "grad_norm": 0.20220068097114563, "learning_rate": 5.655429571268683e-06, "loss": 0.0089, "step": 28850 }, { "epoch": 25.653333333333332, "grad_norm": 0.16045556962490082, "learning_rate": 5.63281742512915e-06, "loss": 0.0099, "step": 28860 }, { "epoch": 25.662222222222223, "grad_norm": 0.19832739233970642, "learning_rate": 5.610247876119862e-06, "loss": 0.0083, "step": 28870 }, { "epoch": 25.67111111111111, "grad_norm": 0.1947653591632843, "learning_rate": 5.587720945909963e-06, "loss": 0.013, "step": 28880 }, { "epoch": 25.68, "grad_norm": 0.2664823830127716, "learning_rate": 5.5652366561276314e-06, "loss": 0.0118, "step": 28890 }, { "epoch": 25.68888888888889, "grad_norm": 0.1733541488647461, "learning_rate": 5.542795028360176e-06, "loss": 0.0089, "step": 28900 }, { "epoch": 25.697777777777777, "grad_norm": 0.2540658116340637, "learning_rate": 5.520396084153872e-06, "loss": 0.0097, "step": 28910 }, { "epoch": 25.706666666666667, "grad_norm": 0.19842661917209625, "learning_rate": 5.498039845014058e-06, "loss": 0.0129, "step": 28920 }, { "epoch": 25.715555555555554, "grad_norm": 0.23427170515060425, "learning_rate": 5.47572633240504e-06, "loss": 0.0109, "step": 28930 }, { "epoch": 25.724444444444444, "grad_norm": 0.15836362540721893, "learning_rate": 5.453455567750148e-06, "loss": 0.0098, "step": 28940 }, { "epoch": 25.733333333333334, "grad_norm": 0.18677003681659698, "learning_rate": 5.43122757243163e-06, "loss": 0.0095, "step": 28950 }, { "epoch": 25.74222222222222, "grad_norm": 0.2338446080684662, "learning_rate": 5.4090423677906894e-06, "loss": 0.0139, "step": 28960 }, { "epoch": 25.75111111111111, "grad_norm": 0.10215015709400177, "learning_rate": 5.38689997512743e-06, "loss": 0.0124, "step": 28970 }, { "epoch": 25.76, "grad_norm": 0.2614404857158661, "learning_rate": 5.364800415700866e-06, "loss": 0.0088, "step": 28980 }, { "epoch": 25.76888888888889, "grad_norm": 0.22617127001285553, "learning_rate": 5.342743710728887e-06, "loss": 0.0103, "step": 28990 }, { "epoch": 25.77777777777778, "grad_norm": 0.25029268860816956, "learning_rate": 5.320729881388248e-06, "loss": 0.0144, "step": 29000 }, { "epoch": 25.786666666666665, "grad_norm": 0.2914581298828125, "learning_rate": 5.298758948814497e-06, "loss": 0.0125, "step": 29010 }, { "epoch": 25.795555555555556, "grad_norm": 0.19717895984649658, "learning_rate": 5.276830934102022e-06, "loss": 0.0119, "step": 29020 }, { "epoch": 25.804444444444446, "grad_norm": 0.2734992504119873, "learning_rate": 5.254945858304039e-06, "loss": 0.0092, "step": 29030 }, { "epoch": 25.813333333333333, "grad_norm": 0.20119789242744446, "learning_rate": 5.233103742432488e-06, "loss": 0.0096, "step": 29040 }, { "epoch": 25.822222222222223, "grad_norm": 0.16952595114707947, "learning_rate": 5.211304607458084e-06, "loss": 0.0078, "step": 29050 }, { "epoch": 25.83111111111111, "grad_norm": 0.24809032678604126, "learning_rate": 5.1895484743102795e-06, "loss": 0.0095, "step": 29060 }, { "epoch": 25.84, "grad_norm": 0.2547890245914459, "learning_rate": 5.167835363877216e-06, "loss": 0.0109, "step": 29070 }, { "epoch": 25.84888888888889, "grad_norm": 0.19287632405757904, "learning_rate": 5.146165297005773e-06, "loss": 0.0104, "step": 29080 }, { "epoch": 25.857777777777777, "grad_norm": 0.27324116230010986, "learning_rate": 5.124538294501463e-06, "loss": 0.013, "step": 29090 }, { "epoch": 25.866666666666667, "grad_norm": 0.23309840261936188, "learning_rate": 5.102954377128488e-06, "loss": 0.0123, "step": 29100 }, { "epoch": 25.875555555555554, "grad_norm": 0.2634255588054657, "learning_rate": 5.0814135656096275e-06, "loss": 0.0098, "step": 29110 }, { "epoch": 25.884444444444444, "grad_norm": 0.15952865779399872, "learning_rate": 5.059915880626337e-06, "loss": 0.0082, "step": 29120 }, { "epoch": 25.893333333333334, "grad_norm": 0.15620408952236176, "learning_rate": 5.03846134281864e-06, "loss": 0.0078, "step": 29130 }, { "epoch": 25.90222222222222, "grad_norm": 0.3474004566669464, "learning_rate": 5.017049972785126e-06, "loss": 0.0087, "step": 29140 }, { "epoch": 25.91111111111111, "grad_norm": 0.13829253613948822, "learning_rate": 4.995681791082946e-06, "loss": 0.0066, "step": 29150 }, { "epoch": 25.92, "grad_norm": 0.2024381011724472, "learning_rate": 4.9743568182277875e-06, "loss": 0.0126, "step": 29160 }, { "epoch": 25.92888888888889, "grad_norm": 0.3202752470970154, "learning_rate": 4.953075074693858e-06, "loss": 0.0133, "step": 29170 }, { "epoch": 25.93777777777778, "grad_norm": 0.16790349781513214, "learning_rate": 4.931836580913856e-06, "loss": 0.0125, "step": 29180 }, { "epoch": 25.946666666666665, "grad_norm": 0.2960953414440155, "learning_rate": 4.910641357278944e-06, "loss": 0.0116, "step": 29190 }, { "epoch": 25.955555555555556, "grad_norm": 0.3693854808807373, "learning_rate": 4.88948942413876e-06, "loss": 0.0121, "step": 29200 }, { "epoch": 25.964444444444446, "grad_norm": 0.19437526166439056, "learning_rate": 4.86838080180137e-06, "loss": 0.0095, "step": 29210 }, { "epoch": 25.973333333333333, "grad_norm": 0.4909128248691559, "learning_rate": 4.847315510533246e-06, "loss": 0.0087, "step": 29220 }, { "epoch": 25.982222222222223, "grad_norm": 0.20847700536251068, "learning_rate": 4.8262935705592815e-06, "loss": 0.008, "step": 29230 }, { "epoch": 25.99111111111111, "grad_norm": 0.24041339755058289, "learning_rate": 4.8053150020627215e-06, "loss": 0.0081, "step": 29240 }, { "epoch": 26.0, "grad_norm": 0.41045695543289185, "learning_rate": 4.784379825185187e-06, "loss": 0.0121, "step": 29250 }, { "epoch": 26.00888888888889, "grad_norm": 0.17301714420318604, "learning_rate": 4.763488060026644e-06, "loss": 0.0106, "step": 29260 }, { "epoch": 26.017777777777777, "grad_norm": 0.19893324375152588, "learning_rate": 4.742639726645365e-06, "loss": 0.0089, "step": 29270 }, { "epoch": 26.026666666666667, "grad_norm": 0.16675794124603271, "learning_rate": 4.721834845057926e-06, "loss": 0.0132, "step": 29280 }, { "epoch": 26.035555555555554, "grad_norm": 0.2961096167564392, "learning_rate": 4.7010734352391895e-06, "loss": 0.0097, "step": 29290 }, { "epoch": 26.044444444444444, "grad_norm": 0.13914480805397034, "learning_rate": 4.6803555171222705e-06, "loss": 0.0076, "step": 29300 }, { "epoch": 26.053333333333335, "grad_norm": 0.2767500579357147, "learning_rate": 4.6596811105985515e-06, "loss": 0.0131, "step": 29310 }, { "epoch": 26.06222222222222, "grad_norm": 0.18252623081207275, "learning_rate": 4.639050235517606e-06, "loss": 0.0102, "step": 29320 }, { "epoch": 26.07111111111111, "grad_norm": 0.12992532551288605, "learning_rate": 4.618462911687232e-06, "loss": 0.0079, "step": 29330 }, { "epoch": 26.08, "grad_norm": 0.16229532659053802, "learning_rate": 4.5979191588734094e-06, "loss": 0.0095, "step": 29340 }, { "epoch": 26.08888888888889, "grad_norm": 0.14758215844631195, "learning_rate": 4.577418996800298e-06, "loss": 0.0115, "step": 29350 }, { "epoch": 26.09777777777778, "grad_norm": 0.24191738665103912, "learning_rate": 4.556962445150187e-06, "loss": 0.0077, "step": 29360 }, { "epoch": 26.106666666666666, "grad_norm": 0.30406999588012695, "learning_rate": 4.5365495235635106e-06, "loss": 0.0077, "step": 29370 }, { "epoch": 26.115555555555556, "grad_norm": 0.18087391555309296, "learning_rate": 4.516180251638796e-06, "loss": 0.0093, "step": 29380 }, { "epoch": 26.124444444444446, "grad_norm": 0.1972109079360962, "learning_rate": 4.495854648932662e-06, "loss": 0.0113, "step": 29390 }, { "epoch": 26.133333333333333, "grad_norm": 0.19380035996437073, "learning_rate": 4.475572734959843e-06, "loss": 0.0104, "step": 29400 }, { "epoch": 26.142222222222223, "grad_norm": 0.20563077926635742, "learning_rate": 4.455334529193083e-06, "loss": 0.0089, "step": 29410 }, { "epoch": 26.15111111111111, "grad_norm": 0.1850234866142273, "learning_rate": 4.435140051063158e-06, "loss": 0.0082, "step": 29420 }, { "epoch": 26.16, "grad_norm": 0.1187540739774704, "learning_rate": 4.4149893199588785e-06, "loss": 0.0069, "step": 29430 }, { "epoch": 26.16888888888889, "grad_norm": 0.17638258635997772, "learning_rate": 4.394882355227065e-06, "loss": 0.0089, "step": 29440 }, { "epoch": 26.177777777777777, "grad_norm": 0.2689094841480255, "learning_rate": 4.374819176172501e-06, "loss": 0.0128, "step": 29450 }, { "epoch": 26.186666666666667, "grad_norm": 0.16289061307907104, "learning_rate": 4.354799802057935e-06, "loss": 0.0082, "step": 29460 }, { "epoch": 26.195555555555554, "grad_norm": 0.30073082447052, "learning_rate": 4.334824252104053e-06, "loss": 0.012, "step": 29470 }, { "epoch": 26.204444444444444, "grad_norm": 0.17679859697818756, "learning_rate": 4.3148925454894626e-06, "loss": 0.013, "step": 29480 }, { "epoch": 26.213333333333335, "grad_norm": 0.28293320536613464, "learning_rate": 4.2950047013507025e-06, "loss": 0.0065, "step": 29490 }, { "epoch": 26.22222222222222, "grad_norm": 0.2327987104654312, "learning_rate": 4.275160738782169e-06, "loss": 0.0094, "step": 29500 }, { "epoch": 26.23111111111111, "grad_norm": 0.15678194165229797, "learning_rate": 4.255360676836151e-06, "loss": 0.0102, "step": 29510 }, { "epoch": 26.24, "grad_norm": 0.1919286996126175, "learning_rate": 4.235604534522753e-06, "loss": 0.0098, "step": 29520 }, { "epoch": 26.24888888888889, "grad_norm": 0.13881464302539825, "learning_rate": 4.215892330809956e-06, "loss": 0.0086, "step": 29530 }, { "epoch": 26.25777777777778, "grad_norm": 0.27687105536460876, "learning_rate": 4.196224084623529e-06, "loss": 0.0081, "step": 29540 }, { "epoch": 26.266666666666666, "grad_norm": 0.19503061473369598, "learning_rate": 4.1765998148470435e-06, "loss": 0.0072, "step": 29550 }, { "epoch": 26.275555555555556, "grad_norm": 0.14760535955429077, "learning_rate": 4.157019540321844e-06, "loss": 0.0075, "step": 29560 }, { "epoch": 26.284444444444443, "grad_norm": 0.18404124677181244, "learning_rate": 4.137483279847038e-06, "loss": 0.0119, "step": 29570 }, { "epoch": 26.293333333333333, "grad_norm": 0.1717691868543625, "learning_rate": 4.1179910521794875e-06, "loss": 0.0112, "step": 29580 }, { "epoch": 26.302222222222223, "grad_norm": 0.23759566247463226, "learning_rate": 4.098542876033768e-06, "loss": 0.0101, "step": 29590 }, { "epoch": 26.31111111111111, "grad_norm": 0.16107891499996185, "learning_rate": 4.079138770082147e-06, "loss": 0.0073, "step": 29600 }, { "epoch": 26.32, "grad_norm": 0.19448989629745483, "learning_rate": 4.0597787529546075e-06, "loss": 0.0145, "step": 29610 }, { "epoch": 26.32888888888889, "grad_norm": 0.18879520893096924, "learning_rate": 4.040462843238774e-06, "loss": 0.0128, "step": 29620 }, { "epoch": 26.337777777777777, "grad_norm": 0.12883466482162476, "learning_rate": 4.0211910594799505e-06, "loss": 0.0091, "step": 29630 }, { "epoch": 26.346666666666668, "grad_norm": 0.17615848779678345, "learning_rate": 4.001963420181054e-06, "loss": 0.0112, "step": 29640 }, { "epoch": 26.355555555555554, "grad_norm": 0.18308480083942413, "learning_rate": 3.982779943802622e-06, "loss": 0.0097, "step": 29650 }, { "epoch": 26.364444444444445, "grad_norm": 0.25045791268348694, "learning_rate": 3.963640648762812e-06, "loss": 0.0125, "step": 29660 }, { "epoch": 26.373333333333335, "grad_norm": 0.2209143340587616, "learning_rate": 3.94454555343734e-06, "loss": 0.009, "step": 29670 }, { "epoch": 26.38222222222222, "grad_norm": 0.12197363376617432, "learning_rate": 3.925494676159491e-06, "loss": 0.0078, "step": 29680 }, { "epoch": 26.391111111111112, "grad_norm": 0.15186116099357605, "learning_rate": 3.906488035220096e-06, "loss": 0.0108, "step": 29690 }, { "epoch": 26.4, "grad_norm": 0.20020464062690735, "learning_rate": 3.887525648867524e-06, "loss": 0.0088, "step": 29700 }, { "epoch": 26.40888888888889, "grad_norm": 0.14076752960681915, "learning_rate": 3.8686075353076225e-06, "loss": 0.0102, "step": 29710 }, { "epoch": 26.41777777777778, "grad_norm": 0.1956368386745453, "learning_rate": 3.8497337127037955e-06, "loss": 0.0112, "step": 29720 }, { "epoch": 26.426666666666666, "grad_norm": 0.30512309074401855, "learning_rate": 3.83090419917686e-06, "loss": 0.0103, "step": 29730 }, { "epoch": 26.435555555555556, "grad_norm": 0.19192807376384735, "learning_rate": 3.8121190128051042e-06, "loss": 0.0083, "step": 29740 }, { "epoch": 26.444444444444443, "grad_norm": 0.21948623657226562, "learning_rate": 3.7933781716242956e-06, "loss": 0.0085, "step": 29750 }, { "epoch": 26.453333333333333, "grad_norm": 0.12957169115543365, "learning_rate": 3.774681693627591e-06, "loss": 0.0089, "step": 29760 }, { "epoch": 26.462222222222223, "grad_norm": 0.22943875193595886, "learning_rate": 3.756029596765559e-06, "loss": 0.0112, "step": 29770 }, { "epoch": 26.47111111111111, "grad_norm": 0.22992002964019775, "learning_rate": 3.7374218989461583e-06, "loss": 0.018, "step": 29780 }, { "epoch": 26.48, "grad_norm": 0.13335996866226196, "learning_rate": 3.7188586180347263e-06, "loss": 0.0081, "step": 29790 }, { "epoch": 26.488888888888887, "grad_norm": 0.38554805517196655, "learning_rate": 3.7003397718539557e-06, "loss": 0.0086, "step": 29800 }, { "epoch": 26.497777777777777, "grad_norm": 0.23062525689601898, "learning_rate": 3.681865378183874e-06, "loss": 0.0109, "step": 29810 }, { "epoch": 26.506666666666668, "grad_norm": 0.2540562152862549, "learning_rate": 3.6634354547618265e-06, "loss": 0.0135, "step": 29820 }, { "epoch": 26.515555555555554, "grad_norm": 0.1107436865568161, "learning_rate": 3.6450500192824534e-06, "loss": 0.0113, "step": 29830 }, { "epoch": 26.524444444444445, "grad_norm": 0.26144739985466003, "learning_rate": 3.626709089397712e-06, "loss": 0.0112, "step": 29840 }, { "epoch": 26.533333333333335, "grad_norm": 0.14512908458709717, "learning_rate": 3.608412682716805e-06, "loss": 0.0095, "step": 29850 }, { "epoch": 26.54222222222222, "grad_norm": 0.1897268146276474, "learning_rate": 3.5901608168061874e-06, "loss": 0.0089, "step": 29860 }, { "epoch": 26.551111111111112, "grad_norm": 0.16545219719409943, "learning_rate": 3.5719535091895686e-06, "loss": 0.0065, "step": 29870 }, { "epoch": 26.56, "grad_norm": 0.23234258592128754, "learning_rate": 3.5537907773478486e-06, "loss": 0.0109, "step": 29880 }, { "epoch": 26.56888888888889, "grad_norm": 0.15689398348331451, "learning_rate": 3.535672638719162e-06, "loss": 0.0081, "step": 29890 }, { "epoch": 26.57777777777778, "grad_norm": 0.20708337426185608, "learning_rate": 3.517599110698816e-06, "loss": 0.0067, "step": 29900 }, { "epoch": 26.586666666666666, "grad_norm": 0.20492884516716003, "learning_rate": 3.4995702106392804e-06, "loss": 0.0087, "step": 29910 }, { "epoch": 26.595555555555556, "grad_norm": 0.13173681497573853, "learning_rate": 3.4815859558501808e-06, "loss": 0.0085, "step": 29920 }, { "epoch": 26.604444444444443, "grad_norm": 0.21655213832855225, "learning_rate": 3.463646363598283e-06, "loss": 0.008, "step": 29930 }, { "epoch": 26.613333333333333, "grad_norm": 0.4419148862361908, "learning_rate": 3.4457514511074706e-06, "loss": 0.0132, "step": 29940 }, { "epoch": 26.622222222222224, "grad_norm": 0.39601653814315796, "learning_rate": 3.427901235558728e-06, "loss": 0.0076, "step": 29950 }, { "epoch": 26.63111111111111, "grad_norm": 0.17355261743068695, "learning_rate": 3.4100957340901296e-06, "loss": 0.0095, "step": 29960 }, { "epoch": 26.64, "grad_norm": 0.082555390894413, "learning_rate": 3.3923349637968005e-06, "loss": 0.011, "step": 29970 }, { "epoch": 26.648888888888887, "grad_norm": 0.25946661829948425, "learning_rate": 3.3746189417309613e-06, "loss": 0.0101, "step": 29980 }, { "epoch": 26.657777777777778, "grad_norm": 0.2698860466480255, "learning_rate": 3.356947684901829e-06, "loss": 0.0085, "step": 29990 }, { "epoch": 26.666666666666668, "grad_norm": 0.2751452326774597, "learning_rate": 3.3393212102756643e-06, "loss": 0.0082, "step": 30000 }, { "epoch": 26.675555555555555, "grad_norm": 0.2115047574043274, "learning_rate": 3.321739534775714e-06, "loss": 0.0121, "step": 30010 }, { "epoch": 26.684444444444445, "grad_norm": 0.19484837353229523, "learning_rate": 3.304202675282225e-06, "loss": 0.0104, "step": 30020 }, { "epoch": 26.693333333333335, "grad_norm": 0.20574335753917694, "learning_rate": 3.2867106486324294e-06, "loss": 0.0082, "step": 30030 }, { "epoch": 26.702222222222222, "grad_norm": 0.25354135036468506, "learning_rate": 3.2692634716204816e-06, "loss": 0.0091, "step": 30040 }, { "epoch": 26.711111111111112, "grad_norm": 0.20808719098567963, "learning_rate": 3.251861160997505e-06, "loss": 0.0088, "step": 30050 }, { "epoch": 26.72, "grad_norm": 0.2551460862159729, "learning_rate": 3.2345037334715166e-06, "loss": 0.0108, "step": 30060 }, { "epoch": 26.72888888888889, "grad_norm": 0.12142115086317062, "learning_rate": 3.217191205707493e-06, "loss": 0.0097, "step": 30070 }, { "epoch": 26.73777777777778, "grad_norm": 0.2568332254886627, "learning_rate": 3.1999235943272477e-06, "loss": 0.0098, "step": 30080 }, { "epoch": 26.746666666666666, "grad_norm": 0.17266830801963806, "learning_rate": 3.1827009159094924e-06, "loss": 0.0139, "step": 30090 }, { "epoch": 26.755555555555556, "grad_norm": 0.24171820282936096, "learning_rate": 3.165523186989805e-06, "loss": 0.0128, "step": 30100 }, { "epoch": 26.764444444444443, "grad_norm": 0.1947808414697647, "learning_rate": 3.1483904240605876e-06, "loss": 0.0096, "step": 30110 }, { "epoch": 26.773333333333333, "grad_norm": 0.14743299782276154, "learning_rate": 3.131302643571099e-06, "loss": 0.0078, "step": 30120 }, { "epoch": 26.782222222222224, "grad_norm": 0.19477613270282745, "learning_rate": 3.1142598619273935e-06, "loss": 0.0095, "step": 30130 }, { "epoch": 26.79111111111111, "grad_norm": 0.1976245939731598, "learning_rate": 3.097262095492304e-06, "loss": 0.0104, "step": 30140 }, { "epoch": 26.8, "grad_norm": 0.1534714698791504, "learning_rate": 3.080309360585465e-06, "loss": 0.0115, "step": 30150 }, { "epoch": 26.808888888888887, "grad_norm": 0.15267989039421082, "learning_rate": 3.0634016734832905e-06, "loss": 0.0118, "step": 30160 }, { "epoch": 26.817777777777778, "grad_norm": 0.1511591225862503, "learning_rate": 3.0465390504189115e-06, "loss": 0.0122, "step": 30170 }, { "epoch": 26.826666666666668, "grad_norm": 0.16008704900741577, "learning_rate": 3.029721507582217e-06, "loss": 0.0099, "step": 30180 }, { "epoch": 26.835555555555555, "grad_norm": 0.14976473152637482, "learning_rate": 3.0129490611197963e-06, "loss": 0.0143, "step": 30190 }, { "epoch": 26.844444444444445, "grad_norm": 0.23956532776355743, "learning_rate": 2.996221727134946e-06, "loss": 0.0089, "step": 30200 }, { "epoch": 26.85333333333333, "grad_norm": 0.1108974814414978, "learning_rate": 2.979539521687669e-06, "loss": 0.0064, "step": 30210 }, { "epoch": 26.862222222222222, "grad_norm": 0.18817508220672607, "learning_rate": 2.9629024607946254e-06, "loss": 0.0091, "step": 30220 }, { "epoch": 26.871111111111112, "grad_norm": 0.3537411391735077, "learning_rate": 2.9463105604291275e-06, "loss": 0.0106, "step": 30230 }, { "epoch": 26.88, "grad_norm": 0.1439163237810135, "learning_rate": 2.9297638365211154e-06, "loss": 0.0122, "step": 30240 }, { "epoch": 26.88888888888889, "grad_norm": 0.1686868965625763, "learning_rate": 2.9132623049571927e-06, "loss": 0.0144, "step": 30250 }, { "epoch": 26.897777777777776, "grad_norm": 0.2325378954410553, "learning_rate": 2.8968059815805525e-06, "loss": 0.0145, "step": 30260 }, { "epoch": 26.906666666666666, "grad_norm": 0.13714931905269623, "learning_rate": 2.8803948821909787e-06, "loss": 0.0124, "step": 30270 }, { "epoch": 26.915555555555557, "grad_norm": 0.24520030617713928, "learning_rate": 2.8640290225448396e-06, "loss": 0.0101, "step": 30280 }, { "epoch": 26.924444444444443, "grad_norm": 0.24530260264873505, "learning_rate": 2.8477084183550717e-06, "loss": 0.0083, "step": 30290 }, { "epoch": 26.933333333333334, "grad_norm": 0.27369439601898193, "learning_rate": 2.8314330852911686e-06, "loss": 0.0093, "step": 30300 }, { "epoch": 26.942222222222224, "grad_norm": 0.15689225494861603, "learning_rate": 2.8152030389791417e-06, "loss": 0.011, "step": 30310 }, { "epoch": 26.95111111111111, "grad_norm": 0.25865888595581055, "learning_rate": 2.799018295001532e-06, "loss": 0.0123, "step": 30320 }, { "epoch": 26.96, "grad_norm": 0.2249138206243515, "learning_rate": 2.7828788688973874e-06, "loss": 0.0121, "step": 30330 }, { "epoch": 26.968888888888888, "grad_norm": 0.10920068621635437, "learning_rate": 2.76678477616224e-06, "loss": 0.0086, "step": 30340 }, { "epoch": 26.977777777777778, "grad_norm": 0.12584613263607025, "learning_rate": 2.750736032248097e-06, "loss": 0.0122, "step": 30350 }, { "epoch": 26.986666666666668, "grad_norm": 0.2109275609254837, "learning_rate": 2.7347326525634375e-06, "loss": 0.0099, "step": 30360 }, { "epoch": 26.995555555555555, "grad_norm": 0.16717959940433502, "learning_rate": 2.7187746524731717e-06, "loss": 0.0097, "step": 30370 }, { "epoch": 27.004444444444445, "grad_norm": 0.2183573842048645, "learning_rate": 2.702862047298643e-06, "loss": 0.009, "step": 30380 }, { "epoch": 27.013333333333332, "grad_norm": 0.22917835414409637, "learning_rate": 2.6869948523176312e-06, "loss": 0.0119, "step": 30390 }, { "epoch": 27.022222222222222, "grad_norm": 0.22166982293128967, "learning_rate": 2.6711730827642832e-06, "loss": 0.0116, "step": 30400 }, { "epoch": 27.031111111111112, "grad_norm": 0.10609235614538193, "learning_rate": 2.6553967538291593e-06, "loss": 0.0062, "step": 30410 }, { "epoch": 27.04, "grad_norm": 0.3267955482006073, "learning_rate": 2.6396658806591867e-06, "loss": 0.0109, "step": 30420 }, { "epoch": 27.04888888888889, "grad_norm": 0.301564484834671, "learning_rate": 2.6239804783576295e-06, "loss": 0.0084, "step": 30430 }, { "epoch": 27.057777777777776, "grad_norm": 0.2127067893743515, "learning_rate": 2.608340561984146e-06, "loss": 0.0088, "step": 30440 }, { "epoch": 27.066666666666666, "grad_norm": 0.13892556726932526, "learning_rate": 2.5927461465546533e-06, "loss": 0.0077, "step": 30450 }, { "epoch": 27.075555555555557, "grad_norm": 0.3601135313510895, "learning_rate": 2.5771972470414406e-06, "loss": 0.0123, "step": 30460 }, { "epoch": 27.084444444444443, "grad_norm": 0.1473352313041687, "learning_rate": 2.5616938783730617e-06, "loss": 0.0097, "step": 30470 }, { "epoch": 27.093333333333334, "grad_norm": 0.1843620240688324, "learning_rate": 2.5462360554343868e-06, "loss": 0.0136, "step": 30480 }, { "epoch": 27.102222222222224, "grad_norm": 0.23528851568698883, "learning_rate": 2.5308237930665334e-06, "loss": 0.0092, "step": 30490 }, { "epoch": 27.11111111111111, "grad_norm": 0.293223112821579, "learning_rate": 2.515457106066882e-06, "loss": 0.009, "step": 30500 }, { "epoch": 27.12, "grad_norm": 0.17130231857299805, "learning_rate": 2.500136009189058e-06, "loss": 0.0112, "step": 30510 }, { "epoch": 27.128888888888888, "grad_norm": 0.3127049505710602, "learning_rate": 2.4848605171429063e-06, "loss": 0.0148, "step": 30520 }, { "epoch": 27.137777777777778, "grad_norm": 0.24439440667629242, "learning_rate": 2.4696306445945115e-06, "loss": 0.0078, "step": 30530 }, { "epoch": 27.14666666666667, "grad_norm": 0.22967371344566345, "learning_rate": 2.4544464061661444e-06, "loss": 0.0082, "step": 30540 }, { "epoch": 27.155555555555555, "grad_norm": 0.11177977174520493, "learning_rate": 2.4393078164362428e-06, "loss": 0.0088, "step": 30550 }, { "epoch": 27.164444444444445, "grad_norm": 0.12512104213237762, "learning_rate": 2.42421488993943e-06, "loss": 0.0086, "step": 30560 }, { "epoch": 27.173333333333332, "grad_norm": 0.15185262262821198, "learning_rate": 2.4091676411665152e-06, "loss": 0.0074, "step": 30570 }, { "epoch": 27.182222222222222, "grad_norm": 0.4029417037963867, "learning_rate": 2.3941660845644196e-06, "loss": 0.018, "step": 30580 }, { "epoch": 27.191111111111113, "grad_norm": 0.16215172410011292, "learning_rate": 2.3792102345362054e-06, "loss": 0.0081, "step": 30590 }, { "epoch": 27.2, "grad_norm": 0.3134725093841553, "learning_rate": 2.3643001054410473e-06, "loss": 0.0087, "step": 30600 }, { "epoch": 27.20888888888889, "grad_norm": 0.17744655907154083, "learning_rate": 2.349435711594217e-06, "loss": 0.0071, "step": 30610 }, { "epoch": 27.217777777777776, "grad_norm": 0.2551135718822479, "learning_rate": 2.3346170672671087e-06, "loss": 0.0078, "step": 30620 }, { "epoch": 27.226666666666667, "grad_norm": 0.235904261469841, "learning_rate": 2.3198441866871536e-06, "loss": 0.0089, "step": 30630 }, { "epoch": 27.235555555555557, "grad_norm": 0.19147591292858124, "learning_rate": 2.3051170840378665e-06, "loss": 0.0132, "step": 30640 }, { "epoch": 27.244444444444444, "grad_norm": 0.22679422795772552, "learning_rate": 2.2904357734588045e-06, "loss": 0.0107, "step": 30650 }, { "epoch": 27.253333333333334, "grad_norm": 0.12505917251110077, "learning_rate": 2.2758002690455525e-06, "loss": 0.009, "step": 30660 }, { "epoch": 27.26222222222222, "grad_norm": 0.1669897586107254, "learning_rate": 2.2612105848497323e-06, "loss": 0.0119, "step": 30670 }, { "epoch": 27.27111111111111, "grad_norm": 0.17728103697299957, "learning_rate": 2.2466667348789604e-06, "loss": 0.0101, "step": 30680 }, { "epoch": 27.28, "grad_norm": 0.2973971664905548, "learning_rate": 2.232168733096851e-06, "loss": 0.0087, "step": 30690 }, { "epoch": 27.288888888888888, "grad_norm": 0.14075423777103424, "learning_rate": 2.217716593422997e-06, "loss": 0.0133, "step": 30700 }, { "epoch": 27.297777777777778, "grad_norm": 0.1782471090555191, "learning_rate": 2.2033103297329717e-06, "loss": 0.01, "step": 30710 }, { "epoch": 27.306666666666665, "grad_norm": 0.18774287402629852, "learning_rate": 2.1889499558582883e-06, "loss": 0.009, "step": 30720 }, { "epoch": 27.315555555555555, "grad_norm": 0.3881121575832367, "learning_rate": 2.1746354855864125e-06, "loss": 0.0123, "step": 30730 }, { "epoch": 27.324444444444445, "grad_norm": 0.2973335087299347, "learning_rate": 2.160366932660718e-06, "loss": 0.0074, "step": 30740 }, { "epoch": 27.333333333333332, "grad_norm": 0.20066313445568085, "learning_rate": 2.146144310780518e-06, "loss": 0.008, "step": 30750 }, { "epoch": 27.342222222222222, "grad_norm": 0.16349542140960693, "learning_rate": 2.131967633601006e-06, "loss": 0.0089, "step": 30760 }, { "epoch": 27.351111111111113, "grad_norm": 0.26540857553482056, "learning_rate": 2.117836914733284e-06, "loss": 0.0107, "step": 30770 }, { "epoch": 27.36, "grad_norm": 0.10120795667171478, "learning_rate": 2.103752167744305e-06, "loss": 0.008, "step": 30780 }, { "epoch": 27.36888888888889, "grad_norm": 0.17551550269126892, "learning_rate": 2.0897134061569067e-06, "loss": 0.0085, "step": 30790 }, { "epoch": 27.377777777777776, "grad_norm": 0.2564518451690674, "learning_rate": 2.0757206434497634e-06, "loss": 0.0124, "step": 30800 }, { "epoch": 27.386666666666667, "grad_norm": 0.15598450601100922, "learning_rate": 2.061773893057395e-06, "loss": 0.0062, "step": 30810 }, { "epoch": 27.395555555555557, "grad_norm": 0.15808437764644623, "learning_rate": 2.0478731683701347e-06, "loss": 0.0128, "step": 30820 }, { "epoch": 27.404444444444444, "grad_norm": 0.23481079936027527, "learning_rate": 2.0340184827341292e-06, "loss": 0.01, "step": 30830 }, { "epoch": 27.413333333333334, "grad_norm": 0.2710295617580414, "learning_rate": 2.0202098494513155e-06, "loss": 0.0089, "step": 30840 }, { "epoch": 27.42222222222222, "grad_norm": 0.16952605545520782, "learning_rate": 2.00644728177945e-06, "loss": 0.0116, "step": 30850 }, { "epoch": 27.43111111111111, "grad_norm": 0.2752940058708191, "learning_rate": 1.9927307929320127e-06, "loss": 0.0097, "step": 30860 }, { "epoch": 27.44, "grad_norm": 0.29089024662971497, "learning_rate": 1.979060396078275e-06, "loss": 0.0124, "step": 30870 }, { "epoch": 27.448888888888888, "grad_norm": 0.1972399801015854, "learning_rate": 1.9654361043432325e-06, "loss": 0.01, "step": 30880 }, { "epoch": 27.45777777777778, "grad_norm": 0.23229151964187622, "learning_rate": 1.9518579308076433e-06, "loss": 0.0081, "step": 30890 }, { "epoch": 27.466666666666665, "grad_norm": 0.13198032975196838, "learning_rate": 1.9383258885079746e-06, "loss": 0.0084, "step": 30900 }, { "epoch": 27.475555555555555, "grad_norm": 0.2732986509799957, "learning_rate": 1.924839990436389e-06, "loss": 0.0119, "step": 30910 }, { "epoch": 27.484444444444446, "grad_norm": 0.24935726821422577, "learning_rate": 1.911400249540768e-06, "loss": 0.0066, "step": 30920 }, { "epoch": 27.493333333333332, "grad_norm": 0.3242601752281189, "learning_rate": 1.8980066787246508e-06, "loss": 0.0121, "step": 30930 }, { "epoch": 27.502222222222223, "grad_norm": 0.24632306396961212, "learning_rate": 1.8846592908472848e-06, "loss": 0.0079, "step": 30940 }, { "epoch": 27.511111111111113, "grad_norm": 0.26659056544303894, "learning_rate": 1.8713580987235468e-06, "loss": 0.0108, "step": 30950 }, { "epoch": 27.52, "grad_norm": 0.22602544724941254, "learning_rate": 1.8581031151239768e-06, "loss": 0.0122, "step": 30960 }, { "epoch": 27.52888888888889, "grad_norm": 0.2393789142370224, "learning_rate": 1.8448943527747287e-06, "loss": 0.0115, "step": 30970 }, { "epoch": 27.537777777777777, "grad_norm": 0.2083200067281723, "learning_rate": 1.831731824357613e-06, "loss": 0.009, "step": 30980 }, { "epoch": 27.546666666666667, "grad_norm": 0.15634146332740784, "learning_rate": 1.818615542510016e-06, "loss": 0.0081, "step": 30990 }, { "epoch": 27.555555555555557, "grad_norm": 0.2749607264995575, "learning_rate": 1.8055455198249527e-06, "loss": 0.0064, "step": 31000 }, { "epoch": 27.564444444444444, "grad_norm": 0.16737869381904602, "learning_rate": 1.7925217688509966e-06, "loss": 0.0125, "step": 31010 }, { "epoch": 27.573333333333334, "grad_norm": 0.3974338471889496, "learning_rate": 1.7795443020923175e-06, "loss": 0.0108, "step": 31020 }, { "epoch": 27.58222222222222, "grad_norm": 0.18868964910507202, "learning_rate": 1.7666131320086376e-06, "loss": 0.0084, "step": 31030 }, { "epoch": 27.59111111111111, "grad_norm": 0.1756001114845276, "learning_rate": 1.7537282710152313e-06, "loss": 0.0138, "step": 31040 }, { "epoch": 27.6, "grad_norm": 0.1418607383966446, "learning_rate": 1.7408897314829143e-06, "loss": 0.0084, "step": 31050 }, { "epoch": 27.608888888888888, "grad_norm": 0.15227912366390228, "learning_rate": 1.7280975257380205e-06, "loss": 0.0083, "step": 31060 }, { "epoch": 27.61777777777778, "grad_norm": 0.2173752784729004, "learning_rate": 1.715351666062409e-06, "loss": 0.0098, "step": 31070 }, { "epoch": 27.626666666666665, "grad_norm": 0.22317318618297577, "learning_rate": 1.7026521646934302e-06, "loss": 0.0121, "step": 31080 }, { "epoch": 27.635555555555555, "grad_norm": 0.2544522285461426, "learning_rate": 1.6899990338239357e-06, "loss": 0.0097, "step": 31090 }, { "epoch": 27.644444444444446, "grad_norm": 0.22699777781963348, "learning_rate": 1.6773922856022583e-06, "loss": 0.0094, "step": 31100 }, { "epoch": 27.653333333333332, "grad_norm": 0.09375015646219254, "learning_rate": 1.664831932132177e-06, "loss": 0.0091, "step": 31110 }, { "epoch": 27.662222222222223, "grad_norm": 0.16013450920581818, "learning_rate": 1.6523179854729565e-06, "loss": 0.0073, "step": 31120 }, { "epoch": 27.67111111111111, "grad_norm": 0.1578095406293869, "learning_rate": 1.639850457639297e-06, "loss": 0.0116, "step": 31130 }, { "epoch": 27.68, "grad_norm": 0.220232754945755, "learning_rate": 1.627429360601318e-06, "loss": 0.0089, "step": 31140 }, { "epoch": 27.68888888888889, "grad_norm": 0.19616135954856873, "learning_rate": 1.6150547062845745e-06, "loss": 0.0123, "step": 31150 }, { "epoch": 27.697777777777777, "grad_norm": 0.2065359354019165, "learning_rate": 1.6027265065700237e-06, "loss": 0.0078, "step": 31160 }, { "epoch": 27.706666666666667, "grad_norm": 0.24581925570964813, "learning_rate": 1.590444773294031e-06, "loss": 0.0067, "step": 31170 }, { "epoch": 27.715555555555554, "grad_norm": 0.14217054843902588, "learning_rate": 1.5782095182483359e-06, "loss": 0.0102, "step": 31180 }, { "epoch": 27.724444444444444, "grad_norm": 0.17184466123580933, "learning_rate": 1.5660207531800696e-06, "loss": 0.0083, "step": 31190 }, { "epoch": 27.733333333333334, "grad_norm": 0.17983022332191467, "learning_rate": 1.5538784897917103e-06, "loss": 0.0104, "step": 31200 }, { "epoch": 27.74222222222222, "grad_norm": 0.14487634599208832, "learning_rate": 1.54178273974111e-06, "loss": 0.0085, "step": 31210 }, { "epoch": 27.75111111111111, "grad_norm": 0.1693265289068222, "learning_rate": 1.529733514641446e-06, "loss": 0.0082, "step": 31220 }, { "epoch": 27.76, "grad_norm": 0.3571673333644867, "learning_rate": 1.5177308260612366e-06, "loss": 0.0124, "step": 31230 }, { "epoch": 27.76888888888889, "grad_norm": 0.24940405786037445, "learning_rate": 1.5057746855243138e-06, "loss": 0.0083, "step": 31240 }, { "epoch": 27.77777777777778, "grad_norm": 0.16307316720485687, "learning_rate": 1.4938651045098173e-06, "loss": 0.0121, "step": 31250 }, { "epoch": 27.786666666666665, "grad_norm": 0.4804389774799347, "learning_rate": 1.4820020944522063e-06, "loss": 0.0073, "step": 31260 }, { "epoch": 27.795555555555556, "grad_norm": 0.1704680323600769, "learning_rate": 1.4701856667411917e-06, "loss": 0.0091, "step": 31270 }, { "epoch": 27.804444444444446, "grad_norm": 0.15881741046905518, "learning_rate": 1.4584158327217878e-06, "loss": 0.0099, "step": 31280 }, { "epoch": 27.813333333333333, "grad_norm": 0.17242994904518127, "learning_rate": 1.4466926036942496e-06, "loss": 0.0114, "step": 31290 }, { "epoch": 27.822222222222223, "grad_norm": 0.23058579862117767, "learning_rate": 1.4350159909141237e-06, "loss": 0.0071, "step": 31300 }, { "epoch": 27.83111111111111, "grad_norm": 0.2396666258573532, "learning_rate": 1.4233860055921645e-06, "loss": 0.0084, "step": 31310 }, { "epoch": 27.84, "grad_norm": 0.15728679299354553, "learning_rate": 1.4118026588943845e-06, "loss": 0.0101, "step": 31320 }, { "epoch": 27.84888888888889, "grad_norm": 0.15858641266822815, "learning_rate": 1.4002659619419934e-06, "loss": 0.0088, "step": 31330 }, { "epoch": 27.857777777777777, "grad_norm": 0.2095939666032791, "learning_rate": 1.3887759258114253e-06, "loss": 0.01, "step": 31340 }, { "epoch": 27.866666666666667, "grad_norm": 0.1215672492980957, "learning_rate": 1.3773325615343223e-06, "loss": 0.0057, "step": 31350 }, { "epoch": 27.875555555555554, "grad_norm": 0.20188657939434052, "learning_rate": 1.365935880097513e-06, "loss": 0.0132, "step": 31360 }, { "epoch": 27.884444444444444, "grad_norm": 0.21797586977481842, "learning_rate": 1.3545858924430054e-06, "loss": 0.0097, "step": 31370 }, { "epoch": 27.893333333333334, "grad_norm": 0.19323039054870605, "learning_rate": 1.3432826094679495e-06, "loss": 0.0109, "step": 31380 }, { "epoch": 27.90222222222222, "grad_norm": 0.22070227563381195, "learning_rate": 1.3320260420247032e-06, "loss": 0.007, "step": 31390 }, { "epoch": 27.91111111111111, "grad_norm": 0.10443749278783798, "learning_rate": 1.3208162009207325e-06, "loss": 0.0059, "step": 31400 }, { "epoch": 27.92, "grad_norm": 0.18210233747959137, "learning_rate": 1.3096530969186615e-06, "loss": 0.0095, "step": 31410 }, { "epoch": 27.92888888888889, "grad_norm": 0.1888139247894287, "learning_rate": 1.2985367407362391e-06, "loss": 0.0105, "step": 31420 }, { "epoch": 27.93777777777778, "grad_norm": 0.17596806585788727, "learning_rate": 1.287467143046317e-06, "loss": 0.008, "step": 31430 }, { "epoch": 27.946666666666665, "grad_norm": 0.3397659957408905, "learning_rate": 1.2764443144768823e-06, "loss": 0.0105, "step": 31440 }, { "epoch": 27.955555555555556, "grad_norm": 0.24012258648872375, "learning_rate": 1.2654682656109917e-06, "loss": 0.0125, "step": 31450 }, { "epoch": 27.964444444444446, "grad_norm": 0.22276870906352997, "learning_rate": 1.254539006986799e-06, "loss": 0.0076, "step": 31460 }, { "epoch": 27.973333333333333, "grad_norm": 0.28045427799224854, "learning_rate": 1.2436565490975383e-06, "loss": 0.0122, "step": 31470 }, { "epoch": 27.982222222222223, "grad_norm": 0.31102776527404785, "learning_rate": 1.2328209023915017e-06, "loss": 0.0084, "step": 31480 }, { "epoch": 27.99111111111111, "grad_norm": 0.30071496963500977, "learning_rate": 1.2220320772720396e-06, "loss": 0.0117, "step": 31490 }, { "epoch": 28.0, "grad_norm": 0.22985626757144928, "learning_rate": 1.2112900840975606e-06, "loss": 0.0074, "step": 31500 }, { "epoch": 28.00888888888889, "grad_norm": 0.18214786052703857, "learning_rate": 1.2005949331814926e-06, "loss": 0.0103, "step": 31510 }, { "epoch": 28.017777777777777, "grad_norm": 0.26299503445625305, "learning_rate": 1.1899466347922939e-06, "loss": 0.0127, "step": 31520 }, { "epoch": 28.026666666666667, "grad_norm": 0.18096651136875153, "learning_rate": 1.1793451991534587e-06, "loss": 0.0082, "step": 31530 }, { "epoch": 28.035555555555554, "grad_norm": 0.15143728256225586, "learning_rate": 1.168790636443462e-06, "loss": 0.0099, "step": 31540 }, { "epoch": 28.044444444444444, "grad_norm": 0.16967567801475525, "learning_rate": 1.1582829567957864e-06, "loss": 0.0118, "step": 31550 }, { "epoch": 28.053333333333335, "grad_norm": 0.10059212893247604, "learning_rate": 1.1478221702989067e-06, "loss": 0.0088, "step": 31560 }, { "epoch": 28.06222222222222, "grad_norm": 0.162766233086586, "learning_rate": 1.1374082869962665e-06, "loss": 0.0083, "step": 31570 }, { "epoch": 28.07111111111111, "grad_norm": 0.2689739465713501, "learning_rate": 1.1270413168862848e-06, "loss": 0.0087, "step": 31580 }, { "epoch": 28.08, "grad_norm": 0.15854036808013916, "learning_rate": 1.116721269922333e-06, "loss": 0.01, "step": 31590 }, { "epoch": 28.08888888888889, "grad_norm": 0.2793892025947571, "learning_rate": 1.1064481560127348e-06, "loss": 0.018, "step": 31600 }, { "epoch": 28.09777777777778, "grad_norm": 0.14380942285060883, "learning_rate": 1.096221985020751e-06, "loss": 0.0082, "step": 31610 }, { "epoch": 28.106666666666666, "grad_norm": 0.16460131108760834, "learning_rate": 1.0860427667645779e-06, "loss": 0.0098, "step": 31620 }, { "epoch": 28.115555555555556, "grad_norm": 0.15899720788002014, "learning_rate": 1.0759105110173306e-06, "loss": 0.0092, "step": 31630 }, { "epoch": 28.124444444444446, "grad_norm": 0.24944256246089935, "learning_rate": 1.065825227507028e-06, "loss": 0.0084, "step": 31640 }, { "epoch": 28.133333333333333, "grad_norm": 0.1176932230591774, "learning_rate": 1.0557869259165964e-06, "loss": 0.011, "step": 31650 }, { "epoch": 28.142222222222223, "grad_norm": 0.1662842333316803, "learning_rate": 1.0457956158838544e-06, "loss": 0.0079, "step": 31660 }, { "epoch": 28.15111111111111, "grad_norm": 0.16227634251117706, "learning_rate": 1.0358513070015064e-06, "loss": 0.0114, "step": 31670 }, { "epoch": 28.16, "grad_norm": 0.1924034208059311, "learning_rate": 1.0259540088171314e-06, "loss": 0.0138, "step": 31680 }, { "epoch": 28.16888888888889, "grad_norm": 0.06445229053497314, "learning_rate": 1.0161037308331567e-06, "loss": 0.0115, "step": 31690 }, { "epoch": 28.177777777777777, "grad_norm": 0.14920978248119354, "learning_rate": 1.0063004825068834e-06, "loss": 0.0081, "step": 31700 }, { "epoch": 28.186666666666667, "grad_norm": 0.30502358078956604, "learning_rate": 9.965442732504603e-07, "loss": 0.0117, "step": 31710 }, { "epoch": 28.195555555555554, "grad_norm": 0.24302737414836884, "learning_rate": 9.868351124308616e-07, "loss": 0.0079, "step": 31720 }, { "epoch": 28.204444444444444, "grad_norm": 0.16846992075443268, "learning_rate": 9.771730093699029e-07, "loss": 0.0101, "step": 31730 }, { "epoch": 28.213333333333335, "grad_norm": 0.25354519486427307, "learning_rate": 9.67557973344202e-07, "loss": 0.0105, "step": 31740 }, { "epoch": 28.22222222222222, "grad_norm": 0.16642460227012634, "learning_rate": 9.579900135851971e-07, "loss": 0.0079, "step": 31750 }, { "epoch": 28.23111111111111, "grad_norm": 0.13392789661884308, "learning_rate": 9.484691392791456e-07, "loss": 0.0075, "step": 31760 }, { "epoch": 28.24, "grad_norm": 0.15038487315177917, "learning_rate": 9.389953595670687e-07, "loss": 0.0095, "step": 31770 }, { "epoch": 28.24888888888889, "grad_norm": 0.21281783282756805, "learning_rate": 9.295686835447903e-07, "loss": 0.0106, "step": 31780 }, { "epoch": 28.25777777777778, "grad_norm": 0.1864618957042694, "learning_rate": 9.201891202628932e-07, "loss": 0.0122, "step": 31790 }, { "epoch": 28.266666666666666, "grad_norm": 0.2209198772907257, "learning_rate": 9.108566787267458e-07, "loss": 0.0079, "step": 31800 }, { "epoch": 28.275555555555556, "grad_norm": 0.12238050252199173, "learning_rate": 9.015713678964699e-07, "loss": 0.01, "step": 31810 }, { "epoch": 28.284444444444443, "grad_norm": 0.22963927686214447, "learning_rate": 8.923331966869341e-07, "loss": 0.0105, "step": 31820 }, { "epoch": 28.293333333333333, "grad_norm": 0.15653561055660248, "learning_rate": 8.831421739677437e-07, "loss": 0.0117, "step": 31830 }, { "epoch": 28.302222222222223, "grad_norm": 0.3193150460720062, "learning_rate": 8.739983085632342e-07, "loss": 0.0077, "step": 31840 }, { "epoch": 28.31111111111111, "grad_norm": 0.2232159674167633, "learning_rate": 8.649016092524831e-07, "loss": 0.0143, "step": 31850 }, { "epoch": 28.32, "grad_norm": 0.2909972071647644, "learning_rate": 8.558520847692708e-07, "loss": 0.0121, "step": 31860 }, { "epoch": 28.32888888888889, "grad_norm": 0.1765635460615158, "learning_rate": 8.468497438020806e-07, "loss": 0.0101, "step": 31870 }, { "epoch": 28.337777777777777, "grad_norm": 0.19128349423408508, "learning_rate": 8.378945949941042e-07, "loss": 0.0122, "step": 31880 }, { "epoch": 28.346666666666668, "grad_norm": 0.23191963136196136, "learning_rate": 8.289866469432194e-07, "loss": 0.0082, "step": 31890 }, { "epoch": 28.355555555555554, "grad_norm": 0.10823400318622589, "learning_rate": 8.201259082019852e-07, "loss": 0.0088, "step": 31900 }, { "epoch": 28.364444444444445, "grad_norm": 0.2372353971004486, "learning_rate": 8.113123872776407e-07, "loss": 0.0141, "step": 31910 }, { "epoch": 28.373333333333335, "grad_norm": 0.15789282321929932, "learning_rate": 8.025460926320838e-07, "loss": 0.0088, "step": 31920 }, { "epoch": 28.38222222222222, "grad_norm": 0.16445067524909973, "learning_rate": 7.938270326818653e-07, "loss": 0.0091, "step": 31930 }, { "epoch": 28.391111111111112, "grad_norm": 0.23763802647590637, "learning_rate": 7.851552157982056e-07, "loss": 0.0108, "step": 31940 }, { "epoch": 28.4, "grad_norm": 0.30863848328590393, "learning_rate": 7.765306503069502e-07, "loss": 0.0138, "step": 31950 }, { "epoch": 28.40888888888889, "grad_norm": 0.1374516785144806, "learning_rate": 7.679533444885867e-07, "loss": 0.0099, "step": 31960 }, { "epoch": 28.41777777777778, "grad_norm": 0.19450043141841888, "learning_rate": 7.594233065782219e-07, "loss": 0.0081, "step": 31970 }, { "epoch": 28.426666666666666, "grad_norm": 0.1754230558872223, "learning_rate": 7.509405447655771e-07, "loss": 0.0099, "step": 31980 }, { "epoch": 28.435555555555556, "grad_norm": 0.10845211148262024, "learning_rate": 7.425050671949985e-07, "loss": 0.0092, "step": 31990 }, { "epoch": 28.444444444444443, "grad_norm": 0.23176629841327667, "learning_rate": 7.341168819654243e-07, "loss": 0.0108, "step": 32000 }, { "epoch": 28.453333333333333, "grad_norm": 0.19052189588546753, "learning_rate": 7.257759971303901e-07, "loss": 0.0135, "step": 32010 }, { "epoch": 28.462222222222223, "grad_norm": 0.19358323514461517, "learning_rate": 7.174824206980069e-07, "loss": 0.0088, "step": 32020 }, { "epoch": 28.47111111111111, "grad_norm": 0.10826950520277023, "learning_rate": 7.092361606309827e-07, "loss": 0.0118, "step": 32030 }, { "epoch": 28.48, "grad_norm": 0.1877441257238388, "learning_rate": 7.010372248465901e-07, "loss": 0.0098, "step": 32040 }, { "epoch": 28.488888888888887, "grad_norm": 0.17118233442306519, "learning_rate": 6.928856212166601e-07, "loss": 0.0074, "step": 32050 }, { "epoch": 28.497777777777777, "grad_norm": 0.10763713717460632, "learning_rate": 6.847813575675821e-07, "loss": 0.0076, "step": 32060 }, { "epoch": 28.506666666666668, "grad_norm": 0.26426175236701965, "learning_rate": 6.767244416802987e-07, "loss": 0.0086, "step": 32070 }, { "epoch": 28.515555555555554, "grad_norm": 0.2066299170255661, "learning_rate": 6.687148812902943e-07, "loss": 0.0088, "step": 32080 }, { "epoch": 28.524444444444445, "grad_norm": 0.1911580115556717, "learning_rate": 6.607526840875844e-07, "loss": 0.0087, "step": 32090 }, { "epoch": 28.533333333333335, "grad_norm": 0.1857062429189682, "learning_rate": 6.528378577167038e-07, "loss": 0.0097, "step": 32100 }, { "epoch": 28.54222222222222, "grad_norm": 0.08462514728307724, "learning_rate": 6.449704097767073e-07, "loss": 0.0102, "step": 32110 }, { "epoch": 28.551111111111112, "grad_norm": 0.2421056032180786, "learning_rate": 6.371503478211749e-07, "loss": 0.0092, "step": 32120 }, { "epoch": 28.56, "grad_norm": 0.08873729407787323, "learning_rate": 6.293776793581896e-07, "loss": 0.0124, "step": 32130 }, { "epoch": 28.56888888888889, "grad_norm": 0.17841722071170807, "learning_rate": 6.216524118503153e-07, "loss": 0.0114, "step": 32140 }, { "epoch": 28.57777777777778, "grad_norm": 0.17117033898830414, "learning_rate": 6.13974552714619e-07, "loss": 0.0112, "step": 32150 }, { "epoch": 28.586666666666666, "grad_norm": 0.12044912576675415, "learning_rate": 6.063441093226374e-07, "loss": 0.0079, "step": 32160 }, { "epoch": 28.595555555555556, "grad_norm": 0.3052404820919037, "learning_rate": 5.987610890004159e-07, "loss": 0.0095, "step": 32170 }, { "epoch": 28.604444444444443, "grad_norm": 0.1731911152601242, "learning_rate": 5.91225499028425e-07, "loss": 0.01, "step": 32180 }, { "epoch": 28.613333333333333, "grad_norm": 0.10701994597911835, "learning_rate": 5.837373466416329e-07, "loss": 0.0102, "step": 32190 }, { "epoch": 28.622222222222224, "grad_norm": 0.212488055229187, "learning_rate": 5.762966390294388e-07, "loss": 0.0116, "step": 32200 }, { "epoch": 28.63111111111111, "grad_norm": 0.17302750051021576, "learning_rate": 5.689033833357005e-07, "loss": 0.0088, "step": 32210 }, { "epoch": 28.64, "grad_norm": 0.15671156346797943, "learning_rate": 5.615575866587174e-07, "loss": 0.0089, "step": 32220 }, { "epoch": 28.648888888888887, "grad_norm": 0.18318812549114227, "learning_rate": 5.542592560512261e-07, "loss": 0.0154, "step": 32230 }, { "epoch": 28.657777777777778, "grad_norm": 0.12446116656064987, "learning_rate": 5.470083985203822e-07, "loss": 0.012, "step": 32240 }, { "epoch": 28.666666666666668, "grad_norm": 0.11251172423362732, "learning_rate": 5.398050210277616e-07, "loss": 0.0122, "step": 32250 }, { "epoch": 28.675555555555555, "grad_norm": 0.10542918741703033, "learning_rate": 5.326491304893655e-07, "loss": 0.0078, "step": 32260 }, { "epoch": 28.684444444444445, "grad_norm": 0.16090279817581177, "learning_rate": 5.255407337755979e-07, "loss": 0.0083, "step": 32270 }, { "epoch": 28.693333333333335, "grad_norm": 0.12608112394809723, "learning_rate": 5.184798377112554e-07, "loss": 0.0074, "step": 32280 }, { "epoch": 28.702222222222222, "grad_norm": 0.19230405986309052, "learning_rate": 5.114664490755428e-07, "loss": 0.0072, "step": 32290 }, { "epoch": 28.711111111111112, "grad_norm": 0.16776396334171295, "learning_rate": 5.045005746020404e-07, "loss": 0.0091, "step": 32300 }, { "epoch": 28.72, "grad_norm": 0.12804031372070312, "learning_rate": 4.975822209787206e-07, "loss": 0.0085, "step": 32310 }, { "epoch": 28.72888888888889, "grad_norm": 0.22880469262599945, "learning_rate": 4.907113948479259e-07, "loss": 0.0103, "step": 32320 }, { "epoch": 28.73777777777778, "grad_norm": 0.1478653848171234, "learning_rate": 4.838881028063569e-07, "loss": 0.0107, "step": 32330 }, { "epoch": 28.746666666666666, "grad_norm": 0.29288479685783386, "learning_rate": 4.771123514050957e-07, "loss": 0.0096, "step": 32340 }, { "epoch": 28.755555555555556, "grad_norm": 0.29903727769851685, "learning_rate": 4.703841471495718e-07, "loss": 0.0092, "step": 32350 }, { "epoch": 28.764444444444443, "grad_norm": 0.15129528939723969, "learning_rate": 4.6370349649956233e-07, "loss": 0.0105, "step": 32360 }, { "epoch": 28.773333333333333, "grad_norm": 0.14203426241874695, "learning_rate": 4.5707040586919217e-07, "loss": 0.0083, "step": 32370 }, { "epoch": 28.782222222222224, "grad_norm": 0.12431854754686356, "learning_rate": 4.504848816269114e-07, "loss": 0.0111, "step": 32380 }, { "epoch": 28.79111111111111, "grad_norm": 0.17384375631809235, "learning_rate": 4.4394693009552343e-07, "loss": 0.0077, "step": 32390 }, { "epoch": 28.8, "grad_norm": 0.16604994237422943, "learning_rate": 4.3745655755214055e-07, "loss": 0.0109, "step": 32400 }, { "epoch": 28.808888888888887, "grad_norm": 0.3075218200683594, "learning_rate": 4.310137702282002e-07, "loss": 0.0101, "step": 32410 }, { "epoch": 28.817777777777778, "grad_norm": 0.23694370687007904, "learning_rate": 4.2461857430944884e-07, "loss": 0.0092, "step": 32420 }, { "epoch": 28.826666666666668, "grad_norm": 0.2289634793996811, "learning_rate": 4.18270975935936e-07, "loss": 0.014, "step": 32430 }, { "epoch": 28.835555555555555, "grad_norm": 0.299882173538208, "learning_rate": 4.1197098120203114e-07, "loss": 0.0114, "step": 32440 }, { "epoch": 28.844444444444445, "grad_norm": 0.1699685901403427, "learning_rate": 4.0571859615637917e-07, "loss": 0.0123, "step": 32450 }, { "epoch": 28.85333333333333, "grad_norm": 0.24647687375545502, "learning_rate": 3.995138268019283e-07, "loss": 0.0123, "step": 32460 }, { "epoch": 28.862222222222222, "grad_norm": 0.13576290011405945, "learning_rate": 3.93356679095902e-07, "loss": 0.0086, "step": 32470 }, { "epoch": 28.871111111111112, "grad_norm": 0.2218310832977295, "learning_rate": 3.8724715894980503e-07, "loss": 0.012, "step": 32480 }, { "epoch": 28.88, "grad_norm": 0.29350435733795166, "learning_rate": 3.811852722294229e-07, "loss": 0.0108, "step": 32490 }, { "epoch": 28.88888888888889, "grad_norm": 0.1997101753950119, "learning_rate": 3.7517102475478905e-07, "loss": 0.0105, "step": 32500 }, { "epoch": 28.897777777777776, "grad_norm": 0.09988665580749512, "learning_rate": 3.692044223002178e-07, "loss": 0.0092, "step": 32510 }, { "epoch": 28.906666666666666, "grad_norm": 0.13588592410087585, "learning_rate": 3.632854705942601e-07, "loss": 0.012, "step": 32520 }, { "epoch": 28.915555555555557, "grad_norm": 0.20433653891086578, "learning_rate": 3.5741417531973686e-07, "loss": 0.0078, "step": 32530 }, { "epoch": 28.924444444444443, "grad_norm": 0.21709153056144714, "learning_rate": 3.515905421137056e-07, "loss": 0.01, "step": 32540 }, { "epoch": 28.933333333333334, "grad_norm": 0.21974101662635803, "learning_rate": 3.4581457656745477e-07, "loss": 0.0098, "step": 32550 }, { "epoch": 28.942222222222224, "grad_norm": 0.2676488757133484, "learning_rate": 3.400862842265207e-07, "loss": 0.0092, "step": 32560 }, { "epoch": 28.95111111111111, "grad_norm": 0.11616649478673935, "learning_rate": 3.344056705906651e-07, "loss": 0.0172, "step": 32570 }, { "epoch": 28.96, "grad_norm": 0.14897799491882324, "learning_rate": 3.2877274111386414e-07, "loss": 0.0096, "step": 32580 }, { "epoch": 28.968888888888888, "grad_norm": 0.13196788728237152, "learning_rate": 3.231875012043195e-07, "loss": 0.0077, "step": 32590 }, { "epoch": 28.977777777777778, "grad_norm": 0.21890628337860107, "learning_rate": 3.176499562244528e-07, "loss": 0.0144, "step": 32600 }, { "epoch": 28.986666666666668, "grad_norm": 0.25631648302078247, "learning_rate": 3.1216011149087787e-07, "loss": 0.01, "step": 32610 }, { "epoch": 28.995555555555555, "grad_norm": 0.22263988852500916, "learning_rate": 3.067179722744173e-07, "loss": 0.0108, "step": 32620 }, { "epoch": 29.004444444444445, "grad_norm": 0.16103336215019226, "learning_rate": 3.013235438000972e-07, "loss": 0.0085, "step": 32630 }, { "epoch": 29.013333333333332, "grad_norm": 0.16397440433502197, "learning_rate": 2.959768312471356e-07, "loss": 0.0115, "step": 32640 }, { "epoch": 29.022222222222222, "grad_norm": 0.15609478950500488, "learning_rate": 2.906778397489318e-07, "loss": 0.0078, "step": 32650 }, { "epoch": 29.031111111111112, "grad_norm": 0.1481102705001831, "learning_rate": 2.854265743930662e-07, "loss": 0.0079, "step": 32660 }, { "epoch": 29.04, "grad_norm": 0.2087840884923935, "learning_rate": 2.802230402213113e-07, "loss": 0.0073, "step": 32670 }, { "epoch": 29.04888888888889, "grad_norm": 0.19709746539592743, "learning_rate": 2.7506724222959854e-07, "loss": 0.0107, "step": 32680 }, { "epoch": 29.057777777777776, "grad_norm": 0.22103577852249146, "learning_rate": 2.69959185368035e-07, "loss": 0.0095, "step": 32690 }, { "epoch": 29.066666666666666, "grad_norm": 0.0999704971909523, "learning_rate": 2.648988745408865e-07, "loss": 0.0086, "step": 32700 }, { "epoch": 29.075555555555557, "grad_norm": 0.13890257477760315, "learning_rate": 2.5988631460657777e-07, "loss": 0.008, "step": 32710 }, { "epoch": 29.084444444444443, "grad_norm": 0.232705220580101, "learning_rate": 2.5492151037769806e-07, "loss": 0.0069, "step": 32720 }, { "epoch": 29.093333333333334, "grad_norm": 0.23587669432163239, "learning_rate": 2.500044666209678e-07, "loss": 0.0095, "step": 32730 }, { "epoch": 29.102222222222224, "grad_norm": 0.37859249114990234, "learning_rate": 2.4513518805727165e-07, "loss": 0.0107, "step": 32740 }, { "epoch": 29.11111111111111, "grad_norm": 0.22432667016983032, "learning_rate": 2.403136793616201e-07, "loss": 0.0088, "step": 32750 }, { "epoch": 29.12, "grad_norm": 0.391062468290329, "learning_rate": 2.355399451631657e-07, "loss": 0.0085, "step": 32760 }, { "epoch": 29.128888888888888, "grad_norm": 0.26831111311912537, "learning_rate": 2.3081399004519777e-07, "loss": 0.0113, "step": 32770 }, { "epoch": 29.137777777777778, "grad_norm": 0.19532768428325653, "learning_rate": 2.261358185451201e-07, "loss": 0.0095, "step": 32780 }, { "epoch": 29.14666666666667, "grad_norm": 0.2136951982975006, "learning_rate": 2.2150543515446764e-07, "loss": 0.0147, "step": 32790 }, { "epoch": 29.155555555555555, "grad_norm": 0.15327243506908417, "learning_rate": 2.1692284431888975e-07, "loss": 0.0096, "step": 32800 }, { "epoch": 29.164444444444445, "grad_norm": 0.20629778504371643, "learning_rate": 2.1238805043816147e-07, "loss": 0.0111, "step": 32810 }, { "epoch": 29.173333333333332, "grad_norm": 0.155885249376297, "learning_rate": 2.0790105786615553e-07, "loss": 0.0127, "step": 32820 }, { "epoch": 29.182222222222222, "grad_norm": 0.08854411542415619, "learning_rate": 2.0346187091084823e-07, "loss": 0.0074, "step": 32830 }, { "epoch": 29.191111111111113, "grad_norm": 0.10432980954647064, "learning_rate": 1.9907049383431907e-07, "loss": 0.009, "step": 32840 }, { "epoch": 29.2, "grad_norm": 0.12530925869941711, "learning_rate": 1.9472693085276218e-07, "loss": 0.009, "step": 32850 }, { "epoch": 29.20888888888889, "grad_norm": 0.17657814919948578, "learning_rate": 1.904311861364416e-07, "loss": 0.0087, "step": 32860 }, { "epoch": 29.217777777777776, "grad_norm": 0.32784801721572876, "learning_rate": 1.8618326380972495e-07, "loss": 0.0076, "step": 32870 }, { "epoch": 29.226666666666667, "grad_norm": 0.15716655552387238, "learning_rate": 1.8198316795106084e-07, "loss": 0.0072, "step": 32880 }, { "epoch": 29.235555555555557, "grad_norm": 0.39083871245384216, "learning_rate": 1.778309025929792e-07, "loss": 0.0137, "step": 32890 }, { "epoch": 29.244444444444444, "grad_norm": 0.1643756926059723, "learning_rate": 1.73726471722091e-07, "loss": 0.0102, "step": 32900 }, { "epoch": 29.253333333333334, "grad_norm": 0.2601345479488373, "learning_rate": 1.6966987927908295e-07, "loss": 0.0088, "step": 32910 }, { "epoch": 29.26222222222222, "grad_norm": 0.1301746964454651, "learning_rate": 1.6566112915869515e-07, "loss": 0.0097, "step": 32920 }, { "epoch": 29.27111111111111, "grad_norm": 0.10023679584264755, "learning_rate": 1.617002252097488e-07, "loss": 0.0079, "step": 32930 }, { "epoch": 29.28, "grad_norm": 0.1454019993543625, "learning_rate": 1.577871712351353e-07, "loss": 0.0113, "step": 32940 }, { "epoch": 29.288888888888888, "grad_norm": 0.13594675064086914, "learning_rate": 1.539219709917883e-07, "loss": 0.0092, "step": 32950 }, { "epoch": 29.297777777777778, "grad_norm": 0.29065412282943726, "learning_rate": 1.5010462819070593e-07, "loss": 0.0092, "step": 32960 }, { "epoch": 29.306666666666665, "grad_norm": 0.13389593362808228, "learning_rate": 1.4633514649693424e-07, "loss": 0.0077, "step": 32970 }, { "epoch": 29.315555555555555, "grad_norm": 0.24671180546283722, "learning_rate": 1.426135295295672e-07, "loss": 0.0111, "step": 32980 }, { "epoch": 29.324444444444445, "grad_norm": 0.22125352919101715, "learning_rate": 1.3893978086175763e-07, "loss": 0.0089, "step": 32990 }, { "epoch": 29.333333333333332, "grad_norm": 0.2642894983291626, "learning_rate": 1.3531390402067302e-07, "loss": 0.0107, "step": 33000 }, { "epoch": 29.342222222222222, "grad_norm": 0.2079657018184662, "learning_rate": 1.3173590248753975e-07, "loss": 0.0071, "step": 33010 }, { "epoch": 29.351111111111113, "grad_norm": 0.1766406148672104, "learning_rate": 1.2820577969762104e-07, "loss": 0.0097, "step": 33020 }, { "epoch": 29.36, "grad_norm": 0.17377027869224548, "learning_rate": 1.247235390401891e-07, "loss": 0.0101, "step": 33030 }, { "epoch": 29.36888888888889, "grad_norm": 0.1494055986404419, "learning_rate": 1.2128918385856945e-07, "loss": 0.0064, "step": 33040 }, { "epoch": 29.377777777777776, "grad_norm": 0.2752269506454468, "learning_rate": 1.1790271745009684e-07, "loss": 0.0078, "step": 33050 }, { "epoch": 29.386666666666667, "grad_norm": 0.21817141771316528, "learning_rate": 1.1456414306613706e-07, "loss": 0.0067, "step": 33060 }, { "epoch": 29.395555555555557, "grad_norm": 0.1819126456975937, "learning_rate": 1.1127346391206494e-07, "loss": 0.0111, "step": 33070 }, { "epoch": 29.404444444444444, "grad_norm": 0.15412050485610962, "learning_rate": 1.0803068314728104e-07, "loss": 0.0102, "step": 33080 }, { "epoch": 29.413333333333334, "grad_norm": 0.22028055787086487, "learning_rate": 1.0483580388518932e-07, "loss": 0.0076, "step": 33090 }, { "epoch": 29.42222222222222, "grad_norm": 0.11501431465148926, "learning_rate": 1.0168882919320833e-07, "loss": 0.0108, "step": 33100 }, { "epoch": 29.43111111111111, "grad_norm": 0.22785663604736328, "learning_rate": 9.85897620927656e-08, "loss": 0.0105, "step": 33110 }, { "epoch": 29.44, "grad_norm": 0.15820471942424774, "learning_rate": 9.553860555929217e-08, "loss": 0.0117, "step": 33120 }, { "epoch": 29.448888888888888, "grad_norm": 0.1857900172472, "learning_rate": 9.253536252220585e-08, "loss": 0.0104, "step": 33130 }, { "epoch": 29.45777777777778, "grad_norm": 0.213301420211792, "learning_rate": 8.958003586494456e-08, "loss": 0.0117, "step": 33140 }, { "epoch": 29.466666666666665, "grad_norm": 0.2224401831626892, "learning_rate": 8.667262842492751e-08, "loss": 0.008, "step": 33150 }, { "epoch": 29.475555555555555, "grad_norm": 0.12427867203950882, "learning_rate": 8.381314299356624e-08, "loss": 0.0083, "step": 33160 }, { "epoch": 29.484444444444446, "grad_norm": 0.26772254705429077, "learning_rate": 8.10015823162702e-08, "loss": 0.0101, "step": 33170 }, { "epoch": 29.493333333333332, "grad_norm": 0.24023449420928955, "learning_rate": 7.823794909242455e-08, "loss": 0.0079, "step": 33180 }, { "epoch": 29.502222222222223, "grad_norm": 0.23810195922851562, "learning_rate": 7.552224597541235e-08, "loss": 0.0107, "step": 33190 }, { "epoch": 29.511111111111113, "grad_norm": 0.1973785161972046, "learning_rate": 7.28544755725924e-08, "loss": 0.0068, "step": 33200 }, { "epoch": 29.52, "grad_norm": 0.3858211040496826, "learning_rate": 7.023464044529915e-08, "loss": 0.0087, "step": 33210 }, { "epoch": 29.52888888888889, "grad_norm": 0.1382966935634613, "learning_rate": 6.76627431088539e-08, "loss": 0.0073, "step": 33220 }, { "epoch": 29.537777777777777, "grad_norm": 0.18616867065429688, "learning_rate": 6.513878603253697e-08, "loss": 0.015, "step": 33230 }, { "epoch": 29.546666666666667, "grad_norm": 0.08629405498504639, "learning_rate": 6.266277163961554e-08, "loss": 0.0107, "step": 33240 }, { "epoch": 29.555555555555557, "grad_norm": 0.2118198126554489, "learning_rate": 6.023470230732686e-08, "loss": 0.0091, "step": 33250 }, { "epoch": 29.564444444444444, "grad_norm": 0.20168910920619965, "learning_rate": 5.78545803668673e-08, "loss": 0.0096, "step": 33260 }, { "epoch": 29.573333333333334, "grad_norm": 0.2120639830827713, "learning_rate": 5.552240810340337e-08, "loss": 0.0082, "step": 33270 }, { "epoch": 29.58222222222222, "grad_norm": 0.16743534803390503, "learning_rate": 5.3238187756060596e-08, "loss": 0.0084, "step": 33280 }, { "epoch": 29.59111111111111, "grad_norm": 0.1453457474708557, "learning_rate": 5.10019215179347e-08, "loss": 0.0069, "step": 33290 }, { "epoch": 29.6, "grad_norm": 0.09087727963924408, "learning_rate": 4.881361153606934e-08, "loss": 0.0099, "step": 33300 }, { "epoch": 29.608888888888888, "grad_norm": 0.15936507284641266, "learning_rate": 4.6673259911472756e-08, "loss": 0.0081, "step": 33310 }, { "epoch": 29.61777777777778, "grad_norm": 0.13715404272079468, "learning_rate": 4.4580868699112267e-08, "loss": 0.0112, "step": 33320 }, { "epoch": 29.626666666666665, "grad_norm": 0.20655792951583862, "learning_rate": 4.253643990789202e-08, "loss": 0.0103, "step": 33330 }, { "epoch": 29.635555555555555, "grad_norm": 0.11239149421453476, "learning_rate": 4.0539975500680784e-08, "loss": 0.0069, "step": 33340 }, { "epoch": 29.644444444444446, "grad_norm": 0.11181444674730301, "learning_rate": 3.859147739429525e-08, "loss": 0.0099, "step": 33350 }, { "epoch": 29.653333333333332, "grad_norm": 0.11988502740859985, "learning_rate": 3.669094745950008e-08, "loss": 0.0066, "step": 33360 }, { "epoch": 29.662222222222223, "grad_norm": 0.1101899966597557, "learning_rate": 3.4838387520996776e-08, "loss": 0.0061, "step": 33370 }, { "epoch": 29.67111111111111, "grad_norm": 0.2075951099395752, "learning_rate": 3.3033799357434783e-08, "loss": 0.0078, "step": 33380 }, { "epoch": 29.68, "grad_norm": 0.13115637004375458, "learning_rate": 3.12771847014115e-08, "loss": 0.0063, "step": 33390 }, { "epoch": 29.68888888888889, "grad_norm": 0.1705673485994339, "learning_rate": 2.9568545239455624e-08, "loss": 0.0109, "step": 33400 }, { "epoch": 29.697777777777777, "grad_norm": 0.1743076741695404, "learning_rate": 2.790788261204935e-08, "loss": 0.0079, "step": 33410 }, { "epoch": 29.706666666666667, "grad_norm": 0.17617197334766388, "learning_rate": 2.629519841359507e-08, "loss": 0.0088, "step": 33420 }, { "epoch": 29.715555555555554, "grad_norm": 0.2556934058666229, "learning_rate": 2.473049419243756e-08, "loss": 0.0073, "step": 33430 }, { "epoch": 29.724444444444444, "grad_norm": 0.12542176246643066, "learning_rate": 2.3213771450858457e-08, "loss": 0.0092, "step": 33440 }, { "epoch": 29.733333333333334, "grad_norm": 0.259750097990036, "learning_rate": 2.174503164507069e-08, "loss": 0.0097, "step": 33450 }, { "epoch": 29.74222222222222, "grad_norm": 0.18822817504405975, "learning_rate": 2.0324276185218484e-08, "loss": 0.0083, "step": 33460 }, { "epoch": 29.75111111111111, "grad_norm": 0.20431265234947205, "learning_rate": 1.8951506435371803e-08, "loss": 0.0115, "step": 33470 }, { "epoch": 29.76, "grad_norm": 0.2028190791606903, "learning_rate": 1.7626723713531913e-08, "loss": 0.0107, "step": 33480 }, { "epoch": 29.76888888888889, "grad_norm": 0.22269631922245026, "learning_rate": 1.6349929291636923e-08, "loss": 0.0114, "step": 33490 }, { "epoch": 29.77777777777778, "grad_norm": 0.17523612082004547, "learning_rate": 1.5121124395539588e-08, "loss": 0.0091, "step": 33500 }, { "epoch": 29.786666666666665, "grad_norm": 0.1718907505273819, "learning_rate": 1.3940310205012853e-08, "loss": 0.0095, "step": 33510 }, { "epoch": 29.795555555555556, "grad_norm": 0.1118452325463295, "learning_rate": 1.280748785377206e-08, "loss": 0.0083, "step": 33520 }, { "epoch": 29.804444444444446, "grad_norm": 0.1932208240032196, "learning_rate": 1.1722658429441646e-08, "loss": 0.0108, "step": 33530 }, { "epoch": 29.813333333333333, "grad_norm": 0.15782316029071808, "learning_rate": 1.068582297356624e-08, "loss": 0.0076, "step": 33540 }, { "epoch": 29.822222222222223, "grad_norm": 0.17499761283397675, "learning_rate": 9.696982481621764e-09, "loss": 0.0072, "step": 33550 }, { "epoch": 29.83111111111111, "grad_norm": 0.2647162675857544, "learning_rate": 8.756137902998785e-09, "loss": 0.0088, "step": 33560 }, { "epoch": 29.84, "grad_norm": 0.20763204991817474, "learning_rate": 7.863290141002511e-09, "loss": 0.0083, "step": 33570 }, { "epoch": 29.84888888888889, "grad_norm": 0.2082856446504593, "learning_rate": 7.018440052863895e-09, "loss": 0.0108, "step": 33580 }, { "epoch": 29.857777777777777, "grad_norm": 0.19068756699562073, "learning_rate": 6.221588449728532e-09, "loss": 0.0096, "step": 33590 }, { "epoch": 29.866666666666667, "grad_norm": 0.15110057592391968, "learning_rate": 5.4727360966511095e-09, "loss": 0.0115, "step": 33600 }, { "epoch": 29.875555555555554, "grad_norm": 0.24273289740085602, "learning_rate": 4.771883712612057e-09, "loss": 0.0079, "step": 33610 }, { "epoch": 29.884444444444444, "grad_norm": 0.2139710783958435, "learning_rate": 4.119031970500897e-09, "loss": 0.0093, "step": 33620 }, { "epoch": 29.893333333333334, "grad_norm": 0.23209433257579803, "learning_rate": 3.5141814971273444e-09, "loss": 0.0084, "step": 33630 }, { "epoch": 29.90222222222222, "grad_norm": 0.23618562519550323, "learning_rate": 2.9573328732046544e-09, "loss": 0.0108, "step": 33640 }, { "epoch": 29.91111111111111, "grad_norm": 0.08592848479747772, "learning_rate": 2.448486633371827e-09, "loss": 0.0094, "step": 33650 }, { "epoch": 29.92, "grad_norm": 0.12294313311576843, "learning_rate": 1.987643266171402e-09, "loss": 0.0096, "step": 33660 }, { "epoch": 29.92888888888889, "grad_norm": 0.12456831336021423, "learning_rate": 1.5748032140661118e-09, "loss": 0.0077, "step": 33670 }, { "epoch": 29.93777777777778, "grad_norm": 0.19911231100559235, "learning_rate": 1.209966873416679e-09, "loss": 0.0093, "step": 33680 }, { "epoch": 29.946666666666665, "grad_norm": 0.1933402270078659, "learning_rate": 8.931345945095704e-10, "loss": 0.0103, "step": 33690 }, { "epoch": 29.955555555555556, "grad_norm": 0.48443588614463806, "learning_rate": 6.243066815347919e-10, "loss": 0.0103, "step": 33700 }, { "epoch": 29.964444444444446, "grad_norm": 0.16809140145778656, "learning_rate": 4.0348339259699277e-10, "loss": 0.0097, "step": 33710 }, { "epoch": 29.973333333333333, "grad_norm": 0.18648651242256165, "learning_rate": 2.3066493970991253e-10, "loss": 0.0078, "step": 33720 }, { "epoch": 29.982222222222223, "grad_norm": 0.1809154748916626, "learning_rate": 1.058514887963824e-10, "loss": 0.0113, "step": 33730 }, { "epoch": 29.99111111111111, "grad_norm": 0.17592336237430573, "learning_rate": 2.9043159688324497e-11, "loss": 0.0082, "step": 33740 }, { "epoch": 30.0, "grad_norm": 0.15305785834789276, "learning_rate": 2.4002613785434337e-13, "loss": 0.0095, "step": 33750 }, { "epoch": 30.0, "step": 33750, "total_flos": 0.0, "train_loss": 0.024522818388983056, "train_runtime": 28393.2831, "train_samples_per_second": 32.073, "train_steps_per_second": 1.189 } ], "logging_steps": 10, "max_steps": 33750, "num_input_tokens_seen": 0, "num_train_epochs": 30, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 27, "trial_name": null, "trial_params": null }