diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,6945 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 9864, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0010138710229324953, + "grad_norm": 50.731353759765625, + "learning_rate": 6.081081081081082e-07, + "loss": 1.4694069862365722, + "step": 10 + }, + { + "epoch": 0.0020277420458649905, + "grad_norm": 19.64676284790039, + "learning_rate": 1.2837837837837838e-06, + "loss": 0.2675037384033203, + "step": 20 + }, + { + "epoch": 0.003041613068797486, + "grad_norm": 14.0009126663208, + "learning_rate": 1.9594594594594595e-06, + "loss": 0.2196730136871338, + "step": 30 + }, + { + "epoch": 0.004055484091729981, + "grad_norm": 16.083576202392578, + "learning_rate": 2.6351351351351353e-06, + "loss": 0.21256353855133056, + "step": 40 + }, + { + "epoch": 0.005069355114662476, + "grad_norm": 13.417021751403809, + "learning_rate": 3.310810810810811e-06, + "loss": 0.2123706817626953, + "step": 50 + }, + { + "epoch": 0.006083226137594972, + "grad_norm": 7.97927188873291, + "learning_rate": 3.986486486486487e-06, + "loss": 0.22084062099456786, + "step": 60 + }, + { + "epoch": 0.0070970971605274664, + "grad_norm": 10.193236351013184, + "learning_rate": 4.6621621621621625e-06, + "loss": 0.19975757598876953, + "step": 70 + }, + { + "epoch": 0.008110968183459962, + "grad_norm": 10.557439804077148, + "learning_rate": 5.337837837837838e-06, + "loss": 0.22339112758636476, + "step": 80 + }, + { + "epoch": 0.009124839206392456, + "grad_norm": 8.02086067199707, + "learning_rate": 6.013513513513514e-06, + "loss": 0.2130129814147949, + "step": 90 + }, + { + "epoch": 0.010138710229324952, + "grad_norm": 6.120218276977539, + "learning_rate": 6.689189189189191e-06, + "loss": 0.19581260681152343, + "step": 100 + }, + { + "epoch": 0.011152581252257448, + "grad_norm": 8.377041816711426, + "learning_rate": 7.3648648648648655e-06, + "loss": 0.22396681308746338, + "step": 110 + }, + { + "epoch": 0.012166452275189943, + "grad_norm": 12.647221565246582, + "learning_rate": 8.040540540540541e-06, + "loss": 0.2036118507385254, + "step": 120 + }, + { + "epoch": 0.013180323298122437, + "grad_norm": 5.785091876983643, + "learning_rate": 8.716216216216217e-06, + "loss": 0.20713284015655517, + "step": 130 + }, + { + "epoch": 0.014194194321054933, + "grad_norm": 9.336621284484863, + "learning_rate": 9.391891891891893e-06, + "loss": 0.21274447441101074, + "step": 140 + }, + { + "epoch": 0.015208065343987429, + "grad_norm": 13.567805290222168, + "learning_rate": 1.0067567567567569e-05, + "loss": 0.19922657012939454, + "step": 150 + }, + { + "epoch": 0.016221936366919924, + "grad_norm": 6.87255334854126, + "learning_rate": 1.0743243243243244e-05, + "loss": 0.1818059802055359, + "step": 160 + }, + { + "epoch": 0.01723580738985242, + "grad_norm": 10.19325065612793, + "learning_rate": 1.141891891891892e-05, + "loss": 0.18367323875427247, + "step": 170 + }, + { + "epoch": 0.018249678412784912, + "grad_norm": 12.029380798339844, + "learning_rate": 1.2094594594594596e-05, + "loss": 0.1895366668701172, + "step": 180 + }, + { + "epoch": 0.019263549435717408, + "grad_norm": 12.335498809814453, + "learning_rate": 1.2770270270270272e-05, + "loss": 0.18756089210510254, + "step": 190 + }, + { + "epoch": 0.020277420458649904, + "grad_norm": 11.819610595703125, + "learning_rate": 1.3445945945945946e-05, + "loss": 0.2158430814743042, + "step": 200 + }, + { + "epoch": 0.0212912914815824, + "grad_norm": 6.693185329437256, + "learning_rate": 1.4121621621621621e-05, + "loss": 0.20627069473266602, + "step": 210 + }, + { + "epoch": 0.022305162504514895, + "grad_norm": 7.694805145263672, + "learning_rate": 1.4797297297297299e-05, + "loss": 0.1934520959854126, + "step": 220 + }, + { + "epoch": 0.02331903352744739, + "grad_norm": 7.408692359924316, + "learning_rate": 1.5472972972972976e-05, + "loss": 0.18309669494628905, + "step": 230 + }, + { + "epoch": 0.024332904550379886, + "grad_norm": 5.283211708068848, + "learning_rate": 1.614864864864865e-05, + "loss": 0.1826266050338745, + "step": 240 + }, + { + "epoch": 0.02534677557331238, + "grad_norm": 8.4596529006958, + "learning_rate": 1.6824324324324324e-05, + "loss": 0.1996974229812622, + "step": 250 + }, + { + "epoch": 0.026360646596244874, + "grad_norm": 10.048973083496094, + "learning_rate": 1.7500000000000002e-05, + "loss": 0.19887588024139405, + "step": 260 + }, + { + "epoch": 0.02737451761917737, + "grad_norm": 7.927089691162109, + "learning_rate": 1.8175675675675676e-05, + "loss": 0.19381613731384278, + "step": 270 + }, + { + "epoch": 0.028388388642109866, + "grad_norm": 7.093384742736816, + "learning_rate": 1.8851351351351353e-05, + "loss": 0.20363895893096923, + "step": 280 + }, + { + "epoch": 0.02940225966504236, + "grad_norm": 3.7583069801330566, + "learning_rate": 1.9527027027027027e-05, + "loss": 0.190981662273407, + "step": 290 + }, + { + "epoch": 0.030416130687974857, + "grad_norm": 5.6196465492248535, + "learning_rate": 1.999999514856865e-05, + "loss": 0.17681100368499755, + "step": 300 + }, + { + "epoch": 0.03143000171090735, + "grad_norm": 5.622128009796143, + "learning_rate": 1.999990890103106e-05, + "loss": 0.18691920042037963, + "step": 310 + }, + { + "epoch": 0.03244387273383985, + "grad_norm": 4.246124744415283, + "learning_rate": 1.999971484497808e-05, + "loss": 0.18674135208129883, + "step": 320 + }, + { + "epoch": 0.033457743756772344, + "grad_norm": 3.3632986545562744, + "learning_rate": 1.9999412982501806e-05, + "loss": 0.1738205671310425, + "step": 330 + }, + { + "epoch": 0.03447161477970484, + "grad_norm": 6.951775550842285, + "learning_rate": 1.9999003316856605e-05, + "loss": 0.17357407808303832, + "step": 340 + }, + { + "epoch": 0.03548548580263733, + "grad_norm": 6.772850036621094, + "learning_rate": 1.9998485852459068e-05, + "loss": 0.16955573558807374, + "step": 350 + }, + { + "epoch": 0.036499356825569824, + "grad_norm": 4.115789890289307, + "learning_rate": 1.9997860594887947e-05, + "loss": 0.18763513565063478, + "step": 360 + }, + { + "epoch": 0.03751322784850232, + "grad_norm": 10.037181854248047, + "learning_rate": 1.9997127550884116e-05, + "loss": 0.19218515157699584, + "step": 370 + }, + { + "epoch": 0.038527098871434816, + "grad_norm": 5.396065711975098, + "learning_rate": 1.9996286728350484e-05, + "loss": 0.1758205533027649, + "step": 380 + }, + { + "epoch": 0.03954096989436731, + "grad_norm": 5.241728782653809, + "learning_rate": 1.9995338136351913e-05, + "loss": 0.19470181465148925, + "step": 390 + }, + { + "epoch": 0.04055484091729981, + "grad_norm": 7.0719780921936035, + "learning_rate": 1.9994281785115127e-05, + "loss": 0.181782865524292, + "step": 400 + }, + { + "epoch": 0.0415687119402323, + "grad_norm": 3.386136293411255, + "learning_rate": 1.9993117686028603e-05, + "loss": 0.17601230144500732, + "step": 410 + }, + { + "epoch": 0.0425825829631648, + "grad_norm": 8.581480026245117, + "learning_rate": 1.9991845851642426e-05, + "loss": 0.2052220582962036, + "step": 420 + }, + { + "epoch": 0.043596453986097294, + "grad_norm": 4.776439189910889, + "learning_rate": 1.999046629566818e-05, + "loss": 0.17804757356643677, + "step": 430 + }, + { + "epoch": 0.04461032500902979, + "grad_norm": 3.676837205886841, + "learning_rate": 1.998897903297879e-05, + "loss": 0.1841128349304199, + "step": 440 + }, + { + "epoch": 0.045624196031962286, + "grad_norm": 8.512882232666016, + "learning_rate": 1.9987384079608362e-05, + "loss": 0.18296406269073487, + "step": 450 + }, + { + "epoch": 0.04663806705489478, + "grad_norm": 5.91367769241333, + "learning_rate": 1.998568145275201e-05, + "loss": 0.18812756538391112, + "step": 460 + }, + { + "epoch": 0.04765193807782728, + "grad_norm": 3.9949119091033936, + "learning_rate": 1.998387117076566e-05, + "loss": 0.18890690803527832, + "step": 470 + }, + { + "epoch": 0.04866580910075977, + "grad_norm": 6.400063991546631, + "learning_rate": 1.998195325316588e-05, + "loss": 0.18471870422363282, + "step": 480 + }, + { + "epoch": 0.04967968012369226, + "grad_norm": 2.673517942428589, + "learning_rate": 1.997992772062964e-05, + "loss": 0.20067477226257324, + "step": 490 + }, + { + "epoch": 0.05069355114662476, + "grad_norm": 5.934947967529297, + "learning_rate": 1.9977794594994103e-05, + "loss": 0.19323770999908446, + "step": 500 + }, + { + "epoch": 0.05170742216955725, + "grad_norm": 2.8921213150024414, + "learning_rate": 1.9975553899256388e-05, + "loss": 0.18293715715408326, + "step": 510 + }, + { + "epoch": 0.05272129319248975, + "grad_norm": 3.642382860183716, + "learning_rate": 1.9973205657573328e-05, + "loss": 0.1827467918395996, + "step": 520 + }, + { + "epoch": 0.053735164215422244, + "grad_norm": 6.452512741088867, + "learning_rate": 1.99707498952612e-05, + "loss": 0.172408127784729, + "step": 530 + }, + { + "epoch": 0.05474903523835474, + "grad_norm": 3.4119246006011963, + "learning_rate": 1.9968186638795445e-05, + "loss": 0.1622527837753296, + "step": 540 + }, + { + "epoch": 0.055762906261287236, + "grad_norm": 5.394989967346191, + "learning_rate": 1.9965515915810408e-05, + "loss": 0.16540141105651857, + "step": 550 + }, + { + "epoch": 0.05677677728421973, + "grad_norm": 4.032186985015869, + "learning_rate": 1.9962737755099023e-05, + "loss": 0.17890504598617554, + "step": 560 + }, + { + "epoch": 0.05779064830715223, + "grad_norm": 3.7435805797576904, + "learning_rate": 1.9959852186612492e-05, + "loss": 0.18007283210754393, + "step": 570 + }, + { + "epoch": 0.05880451933008472, + "grad_norm": 4.005575656890869, + "learning_rate": 1.995685924145999e-05, + "loss": 0.1741203784942627, + "step": 580 + }, + { + "epoch": 0.05981839035301722, + "grad_norm": 4.614043235778809, + "learning_rate": 1.9953758951908305e-05, + "loss": 0.159068500995636, + "step": 590 + }, + { + "epoch": 0.060832261375949714, + "grad_norm": 4.4694085121154785, + "learning_rate": 1.995055135138151e-05, + "loss": 0.18460067510604858, + "step": 600 + }, + { + "epoch": 0.06184613239888221, + "grad_norm": 4.2469658851623535, + "learning_rate": 1.994723647446058e-05, + "loss": 0.18630992174148558, + "step": 610 + }, + { + "epoch": 0.0628600034218147, + "grad_norm": 3.258877992630005, + "learning_rate": 1.994381435688303e-05, + "loss": 0.1806309461593628, + "step": 620 + }, + { + "epoch": 0.0638738744447472, + "grad_norm": 4.703103542327881, + "learning_rate": 1.994028503554255e-05, + "loss": 0.17196577787399292, + "step": 630 + }, + { + "epoch": 0.0648877454676797, + "grad_norm": 4.63681173324585, + "learning_rate": 1.9936648548488568e-05, + "loss": 0.18792276382446288, + "step": 640 + }, + { + "epoch": 0.06590161649061219, + "grad_norm": 5.628655433654785, + "learning_rate": 1.993290493492587e-05, + "loss": 0.17562625408172608, + "step": 650 + }, + { + "epoch": 0.06691548751354469, + "grad_norm": 6.253754615783691, + "learning_rate": 1.9929054235214168e-05, + "loss": 0.17272111177444457, + "step": 660 + }, + { + "epoch": 0.06792935853647718, + "grad_norm": 4.3602495193481445, + "learning_rate": 1.992509649086766e-05, + "loss": 0.1459844470024109, + "step": 670 + }, + { + "epoch": 0.06894322955940968, + "grad_norm": 3.917858123779297, + "learning_rate": 1.9921031744554588e-05, + "loss": 0.1930823802947998, + "step": 680 + }, + { + "epoch": 0.06995710058234217, + "grad_norm": 5.484121799468994, + "learning_rate": 1.991686004009678e-05, + "loss": 0.18086743354797363, + "step": 690 + }, + { + "epoch": 0.07097097160527466, + "grad_norm": 3.734393358230591, + "learning_rate": 1.991258142246917e-05, + "loss": 0.18379961252212523, + "step": 700 + }, + { + "epoch": 0.07198484262820716, + "grad_norm": 5.26226806640625, + "learning_rate": 1.990819593779932e-05, + "loss": 0.18340892791748048, + "step": 710 + }, + { + "epoch": 0.07299871365113965, + "grad_norm": 5.593852519989014, + "learning_rate": 1.9903703633366923e-05, + "loss": 0.17229235172271729, + "step": 720 + }, + { + "epoch": 0.07401258467407215, + "grad_norm": 7.495728015899658, + "learning_rate": 1.9899104557603285e-05, + "loss": 0.16693897247314454, + "step": 730 + }, + { + "epoch": 0.07502645569700464, + "grad_norm": 9.115511894226074, + "learning_rate": 1.9894398760090807e-05, + "loss": 0.19775344133377076, + "step": 740 + }, + { + "epoch": 0.07604032671993714, + "grad_norm": 3.1551353931427, + "learning_rate": 1.988958629156246e-05, + "loss": 0.17459380626678467, + "step": 750 + }, + { + "epoch": 0.07705419774286963, + "grad_norm": 3.1768381595611572, + "learning_rate": 1.988466720390123e-05, + "loss": 0.17835698127746583, + "step": 760 + }, + { + "epoch": 0.07806806876580213, + "grad_norm": 4.682513236999512, + "learning_rate": 1.9879641550139545e-05, + "loss": 0.1760738730430603, + "step": 770 + }, + { + "epoch": 0.07908193978873462, + "grad_norm": 4.256967544555664, + "learning_rate": 1.9874509384458726e-05, + "loss": 0.17808153629302978, + "step": 780 + }, + { + "epoch": 0.08009581081166713, + "grad_norm": 2.1426165103912354, + "learning_rate": 1.9869270762188402e-05, + "loss": 0.175534987449646, + "step": 790 + }, + { + "epoch": 0.08110968183459961, + "grad_norm": 5.081096172332764, + "learning_rate": 1.986392573980589e-05, + "loss": 0.186021888256073, + "step": 800 + }, + { + "epoch": 0.08212355285753212, + "grad_norm": 3.465602397918701, + "learning_rate": 1.9858474374935606e-05, + "loss": 0.17549388408660888, + "step": 810 + }, + { + "epoch": 0.0831374238804646, + "grad_norm": 2.4676618576049805, + "learning_rate": 1.9852916726348446e-05, + "loss": 0.1793988823890686, + "step": 820 + }, + { + "epoch": 0.0841512949033971, + "grad_norm": 2.2493669986724854, + "learning_rate": 1.9847252853961136e-05, + "loss": 0.16792131662368776, + "step": 830 + }, + { + "epoch": 0.0851651659263296, + "grad_norm": 5.110182762145996, + "learning_rate": 1.98414828188356e-05, + "loss": 0.1787337303161621, + "step": 840 + }, + { + "epoch": 0.08617903694926209, + "grad_norm": 3.9989397525787354, + "learning_rate": 1.98356066831783e-05, + "loss": 0.1730713963508606, + "step": 850 + }, + { + "epoch": 0.08719290797219459, + "grad_norm": 4.167379379272461, + "learning_rate": 1.982962451033956e-05, + "loss": 0.17176637649536133, + "step": 860 + }, + { + "epoch": 0.08820677899512708, + "grad_norm": 3.8491177558898926, + "learning_rate": 1.9823536364812875e-05, + "loss": 0.16533188819885253, + "step": 870 + }, + { + "epoch": 0.08922065001805958, + "grad_norm": 2.8235135078430176, + "learning_rate": 1.981734231223425e-05, + "loss": 0.15635806322097778, + "step": 880 + }, + { + "epoch": 0.09023452104099207, + "grad_norm": 5.2434611320495605, + "learning_rate": 1.9811042419381444e-05, + "loss": 0.17323405742645265, + "step": 890 + }, + { + "epoch": 0.09124839206392457, + "grad_norm": 5.151634693145752, + "learning_rate": 1.980463675417329e-05, + "loss": 0.17093746662139891, + "step": 900 + }, + { + "epoch": 0.09226226308685706, + "grad_norm": 5.2634406089782715, + "learning_rate": 1.9798125385668942e-05, + "loss": 0.1776280164718628, + "step": 910 + }, + { + "epoch": 0.09327613410978956, + "grad_norm": 3.442875623703003, + "learning_rate": 1.9791508384067138e-05, + "loss": 0.1658215641975403, + "step": 920 + }, + { + "epoch": 0.09429000513272205, + "grad_norm": 3.128662347793579, + "learning_rate": 1.9784785820705435e-05, + "loss": 0.17807645797729493, + "step": 930 + }, + { + "epoch": 0.09530387615565455, + "grad_norm": 5.565284252166748, + "learning_rate": 1.9777957768059456e-05, + "loss": 0.17322582006454468, + "step": 940 + }, + { + "epoch": 0.09631774717858704, + "grad_norm": 7.804746627807617, + "learning_rate": 1.977102429974209e-05, + "loss": 0.17686302661895753, + "step": 950 + }, + { + "epoch": 0.09733161820151955, + "grad_norm": 4.371828556060791, + "learning_rate": 1.9763985490502714e-05, + "loss": 0.14965968132019042, + "step": 960 + }, + { + "epoch": 0.09834548922445203, + "grad_norm": 6.301712512969971, + "learning_rate": 1.9756841416226372e-05, + "loss": 0.18070511817932128, + "step": 970 + }, + { + "epoch": 0.09935936024738452, + "grad_norm": 3.5437405109405518, + "learning_rate": 1.9749592153932974e-05, + "loss": 0.1695558547973633, + "step": 980 + }, + { + "epoch": 0.10037323127031703, + "grad_norm": 6.871677875518799, + "learning_rate": 1.9742237781776455e-05, + "loss": 0.17027852535247803, + "step": 990 + }, + { + "epoch": 0.10138710229324951, + "grad_norm": 3.822228193283081, + "learning_rate": 1.9734778379043923e-05, + "loss": 0.18501952886581421, + "step": 1000 + }, + { + "epoch": 0.10240097331618202, + "grad_norm": 3.3379194736480713, + "learning_rate": 1.9727214026154827e-05, + "loss": 0.16618814468383789, + "step": 1010 + }, + { + "epoch": 0.1034148443391145, + "grad_norm": 4.984459400177002, + "learning_rate": 1.9719544804660084e-05, + "loss": 0.18112580776214598, + "step": 1020 + }, + { + "epoch": 0.10442871536204701, + "grad_norm": 8.099133491516113, + "learning_rate": 1.9711770797241174e-05, + "loss": 0.1802734375, + "step": 1030 + }, + { + "epoch": 0.1054425863849795, + "grad_norm": 4.454706192016602, + "learning_rate": 1.9703892087709287e-05, + "loss": 0.1534941554069519, + "step": 1040 + }, + { + "epoch": 0.106456457407912, + "grad_norm": 2.3928637504577637, + "learning_rate": 1.9695908761004392e-05, + "loss": 0.16265038251876832, + "step": 1050 + }, + { + "epoch": 0.10747032843084449, + "grad_norm": 2.9390366077423096, + "learning_rate": 1.9687820903194335e-05, + "loss": 0.15163285732269288, + "step": 1060 + }, + { + "epoch": 0.10848419945377699, + "grad_norm": 3.5900163650512695, + "learning_rate": 1.9679628601473905e-05, + "loss": 0.1865153431892395, + "step": 1070 + }, + { + "epoch": 0.10949807047670948, + "grad_norm": 5.388684272766113, + "learning_rate": 1.967133194416389e-05, + "loss": 0.1708207130432129, + "step": 1080 + }, + { + "epoch": 0.11051194149964198, + "grad_norm": 4.638717174530029, + "learning_rate": 1.9662931020710138e-05, + "loss": 0.17335765361785888, + "step": 1090 + }, + { + "epoch": 0.11152581252257447, + "grad_norm": 2.324394941329956, + "learning_rate": 1.965442592168259e-05, + "loss": 0.19155402183532716, + "step": 1100 + }, + { + "epoch": 0.11253968354550696, + "grad_norm": 4.08584451675415, + "learning_rate": 1.964581673877428e-05, + "loss": 0.1603313684463501, + "step": 1110 + }, + { + "epoch": 0.11355355456843946, + "grad_norm": 1.908292293548584, + "learning_rate": 1.9637103564800384e-05, + "loss": 0.16062567234039307, + "step": 1120 + }, + { + "epoch": 0.11456742559137195, + "grad_norm": 3.332932472229004, + "learning_rate": 1.962828649369719e-05, + "loss": 0.1790619134902954, + "step": 1130 + }, + { + "epoch": 0.11558129661430445, + "grad_norm": 5.628213882446289, + "learning_rate": 1.96193656205211e-05, + "loss": 0.17300251722335816, + "step": 1140 + }, + { + "epoch": 0.11659516763723694, + "grad_norm": 3.3096258640289307, + "learning_rate": 1.9610341041447595e-05, + "loss": 0.16023989915847778, + "step": 1150 + }, + { + "epoch": 0.11760903866016945, + "grad_norm": 2.381916046142578, + "learning_rate": 1.9601212853770214e-05, + "loss": 0.15588629245758057, + "step": 1160 + }, + { + "epoch": 0.11862290968310193, + "grad_norm": 6.42448616027832, + "learning_rate": 1.959198115589949e-05, + "loss": 0.19157402515411376, + "step": 1170 + }, + { + "epoch": 0.11963678070603444, + "grad_norm": 2.4429774284362793, + "learning_rate": 1.9582646047361888e-05, + "loss": 0.1639960527420044, + "step": 1180 + }, + { + "epoch": 0.12065065172896693, + "grad_norm": 1.7438149452209473, + "learning_rate": 1.9573207628798747e-05, + "loss": 0.16130218505859376, + "step": 1190 + }, + { + "epoch": 0.12166452275189943, + "grad_norm": 1.7322479486465454, + "learning_rate": 1.956366600196519e-05, + "loss": 0.18079793453216553, + "step": 1200 + }, + { + "epoch": 0.12267839377483192, + "grad_norm": 1.565903902053833, + "learning_rate": 1.9554021269729003e-05, + "loss": 0.15925445556640624, + "step": 1210 + }, + { + "epoch": 0.12369226479776442, + "grad_norm": 3.488713502883911, + "learning_rate": 1.9544273536069573e-05, + "loss": 0.16834355592727662, + "step": 1220 + }, + { + "epoch": 0.12470613582069691, + "grad_norm": 5.326113700866699, + "learning_rate": 1.9534422906076714e-05, + "loss": 0.17651712894439697, + "step": 1230 + }, + { + "epoch": 0.1257200068436294, + "grad_norm": 2.050661563873291, + "learning_rate": 1.9524469485949586e-05, + "loss": 0.17451609373092652, + "step": 1240 + }, + { + "epoch": 0.1267338778665619, + "grad_norm": 4.163514614105225, + "learning_rate": 1.9514413382995498e-05, + "loss": 0.17882106304168702, + "step": 1250 + }, + { + "epoch": 0.1277477488894944, + "grad_norm": 5.887022495269775, + "learning_rate": 1.9504254705628806e-05, + "loss": 0.1733614444732666, + "step": 1260 + }, + { + "epoch": 0.12876161991242688, + "grad_norm": 2.771512269973755, + "learning_rate": 1.9493993563369688e-05, + "loss": 0.1875835657119751, + "step": 1270 + }, + { + "epoch": 0.1297754909353594, + "grad_norm": 3.7256360054016113, + "learning_rate": 1.9483630066843005e-05, + "loss": 0.16062517166137696, + "step": 1280 + }, + { + "epoch": 0.13078936195829188, + "grad_norm": 2.5515639781951904, + "learning_rate": 1.9473164327777095e-05, + "loss": 0.14801760911941528, + "step": 1290 + }, + { + "epoch": 0.13180323298122437, + "grad_norm": 4.824519157409668, + "learning_rate": 1.9462596459002565e-05, + "loss": 0.17392812967300414, + "step": 1300 + }, + { + "epoch": 0.13281710400415686, + "grad_norm": 1.721830129623413, + "learning_rate": 1.9451926574451074e-05, + "loss": 0.1571485161781311, + "step": 1310 + }, + { + "epoch": 0.13383097502708938, + "grad_norm": 2.3940589427948, + "learning_rate": 1.944115478915412e-05, + "loss": 0.167417573928833, + "step": 1320 + }, + { + "epoch": 0.13484484605002187, + "grad_norm": 3.4660027027130127, + "learning_rate": 1.9430281219241767e-05, + "loss": 0.16694204807281493, + "step": 1330 + }, + { + "epoch": 0.13585871707295435, + "grad_norm": 2.04611873626709, + "learning_rate": 1.9419305981941436e-05, + "loss": 0.17357702255249025, + "step": 1340 + }, + { + "epoch": 0.13687258809588684, + "grad_norm": 3.893183708190918, + "learning_rate": 1.94082291955766e-05, + "loss": 0.15456533432006836, + "step": 1350 + }, + { + "epoch": 0.13788645911881936, + "grad_norm": 4.6778998374938965, + "learning_rate": 1.939705097956554e-05, + "loss": 0.1633974313735962, + "step": 1360 + }, + { + "epoch": 0.13890033014175185, + "grad_norm": 3.0180904865264893, + "learning_rate": 1.9385771454420048e-05, + "loss": 0.17558639049530028, + "step": 1370 + }, + { + "epoch": 0.13991420116468434, + "grad_norm": 4.071365833282471, + "learning_rate": 1.9374390741744107e-05, + "loss": 0.16226245164871217, + "step": 1380 + }, + { + "epoch": 0.14092807218761683, + "grad_norm": 6.951503753662109, + "learning_rate": 1.9362908964232616e-05, + "loss": 0.18220155239105223, + "step": 1390 + }, + { + "epoch": 0.14194194321054932, + "grad_norm": 2.839548110961914, + "learning_rate": 1.9351326245670038e-05, + "loss": 0.16862224340438842, + "step": 1400 + }, + { + "epoch": 0.14295581423348183, + "grad_norm": 3.131087064743042, + "learning_rate": 1.933964271092908e-05, + "loss": 0.15862529277801513, + "step": 1410 + }, + { + "epoch": 0.14396968525641432, + "grad_norm": 2.072535753250122, + "learning_rate": 1.932785848596934e-05, + "loss": 0.1809432864189148, + "step": 1420 + }, + { + "epoch": 0.1449835562793468, + "grad_norm": 3.1045384407043457, + "learning_rate": 1.9315973697835966e-05, + "loss": 0.1734461784362793, + "step": 1430 + }, + { + "epoch": 0.1459974273022793, + "grad_norm": 2.9639837741851807, + "learning_rate": 1.930398847465825e-05, + "loss": 0.1647495746612549, + "step": 1440 + }, + { + "epoch": 0.14701129832521181, + "grad_norm": 2.7314109802246094, + "learning_rate": 1.9291902945648283e-05, + "loss": 0.1540939688682556, + "step": 1450 + }, + { + "epoch": 0.1480251693481443, + "grad_norm": 2.9838151931762695, + "learning_rate": 1.9279717241099554e-05, + "loss": 0.17959966659545898, + "step": 1460 + }, + { + "epoch": 0.1490390403710768, + "grad_norm": 1.6168804168701172, + "learning_rate": 1.9267431492385524e-05, + "loss": 0.16940973997116088, + "step": 1470 + }, + { + "epoch": 0.15005291139400928, + "grad_norm": 3.4817473888397217, + "learning_rate": 1.925504583195823e-05, + "loss": 0.16622445583343506, + "step": 1480 + }, + { + "epoch": 0.1510667824169418, + "grad_norm": 2.3672518730163574, + "learning_rate": 1.924256039334686e-05, + "loss": 0.1685317039489746, + "step": 1490 + }, + { + "epoch": 0.15208065343987429, + "grad_norm": 3.014786720275879, + "learning_rate": 1.9229975311156305e-05, + "loss": 0.18337373733520507, + "step": 1500 + }, + { + "epoch": 0.15309452446280677, + "grad_norm": 2.320131540298462, + "learning_rate": 1.9217290721065693e-05, + "loss": 0.17623775005340575, + "step": 1510 + }, + { + "epoch": 0.15410839548573926, + "grad_norm": 4.603436470031738, + "learning_rate": 1.9204506759826957e-05, + "loss": 0.16369402408599854, + "step": 1520 + }, + { + "epoch": 0.15512226650867175, + "grad_norm": 1.7143694162368774, + "learning_rate": 1.9191623565263342e-05, + "loss": 0.16468876600265503, + "step": 1530 + }, + { + "epoch": 0.15613613753160427, + "grad_norm": 4.20829963684082, + "learning_rate": 1.9178641276267922e-05, + "loss": 0.16598418951034546, + "step": 1540 + }, + { + "epoch": 0.15715000855453676, + "grad_norm": 2.5098955631256104, + "learning_rate": 1.91655600328021e-05, + "loss": 0.16512451171875, + "step": 1550 + }, + { + "epoch": 0.15816387957746925, + "grad_norm": 6.5186238288879395, + "learning_rate": 1.91523799758941e-05, + "loss": 0.15721583366394043, + "step": 1560 + }, + { + "epoch": 0.15917775060040174, + "grad_norm": 5.830102443695068, + "learning_rate": 1.9139101247637465e-05, + "loss": 0.16404998302459717, + "step": 1570 + }, + { + "epoch": 0.16019162162333425, + "grad_norm": 5.104887962341309, + "learning_rate": 1.912572399118949e-05, + "loss": 0.1590874433517456, + "step": 1580 + }, + { + "epoch": 0.16120549264626674, + "grad_norm": 2.4739747047424316, + "learning_rate": 1.9112248350769704e-05, + "loss": 0.17960269451141359, + "step": 1590 + }, + { + "epoch": 0.16221936366919923, + "grad_norm": 3.1461572647094727, + "learning_rate": 1.909867447165831e-05, + "loss": 0.1610199213027954, + "step": 1600 + }, + { + "epoch": 0.16323323469213172, + "grad_norm": 1.8623886108398438, + "learning_rate": 1.908500250019462e-05, + "loss": 0.16447696685791016, + "step": 1610 + }, + { + "epoch": 0.16424710571506423, + "grad_norm": 4.690032482147217, + "learning_rate": 1.9071232583775465e-05, + "loss": 0.16240527629852294, + "step": 1620 + }, + { + "epoch": 0.16526097673799672, + "grad_norm": 3.0182266235351562, + "learning_rate": 1.9057364870853628e-05, + "loss": 0.14358000755310057, + "step": 1630 + }, + { + "epoch": 0.1662748477609292, + "grad_norm": 3.9813008308410645, + "learning_rate": 1.904339951093622e-05, + "loss": 0.16875321865081788, + "step": 1640 + }, + { + "epoch": 0.1672887187838617, + "grad_norm": 1.9259511232376099, + "learning_rate": 1.902933665458309e-05, + "loss": 0.1648932933807373, + "step": 1650 + }, + { + "epoch": 0.1683025898067942, + "grad_norm": 3.811471700668335, + "learning_rate": 1.901517645340518e-05, + "loss": 0.1772339701652527, + "step": 1660 + }, + { + "epoch": 0.1693164608297267, + "grad_norm": 4.403039932250977, + "learning_rate": 1.900091906006291e-05, + "loss": 0.17537424564361573, + "step": 1670 + }, + { + "epoch": 0.1703303318526592, + "grad_norm": 1.8989040851593018, + "learning_rate": 1.8986564628264527e-05, + "loss": 0.17066320180892944, + "step": 1680 + }, + { + "epoch": 0.17134420287559168, + "grad_norm": 2.248544931411743, + "learning_rate": 1.8972113312764432e-05, + "loss": 0.15494236946105958, + "step": 1690 + }, + { + "epoch": 0.17235807389852417, + "grad_norm": 2.4127614498138428, + "learning_rate": 1.895756526936153e-05, + "loss": 0.16577097177505493, + "step": 1700 + }, + { + "epoch": 0.1733719449214567, + "grad_norm": 2.567159414291382, + "learning_rate": 1.8942920654897553e-05, + "loss": 0.16686248779296875, + "step": 1710 + }, + { + "epoch": 0.17438581594438918, + "grad_norm": 1.5967146158218384, + "learning_rate": 1.8928179627255347e-05, + "loss": 0.17398217916488648, + "step": 1720 + }, + { + "epoch": 0.17539968696732167, + "grad_norm": 3.3272957801818848, + "learning_rate": 1.891334234535719e-05, + "loss": 0.16821205615997314, + "step": 1730 + }, + { + "epoch": 0.17641355799025416, + "grad_norm": 4.121727466583252, + "learning_rate": 1.8898408969163078e-05, + "loss": 0.15749123096466064, + "step": 1740 + }, + { + "epoch": 0.17742742901318667, + "grad_norm": 2.7949209213256836, + "learning_rate": 1.8883379659668987e-05, + "loss": 0.17413833141326904, + "step": 1750 + }, + { + "epoch": 0.17844130003611916, + "grad_norm": 4.459453582763672, + "learning_rate": 1.8868254578905147e-05, + "loss": 0.1747552514076233, + "step": 1760 + }, + { + "epoch": 0.17945517105905165, + "grad_norm": 3.7542471885681152, + "learning_rate": 1.8853033889934293e-05, + "loss": 0.1445354700088501, + "step": 1770 + }, + { + "epoch": 0.18046904208198414, + "grad_norm": 1.942527413368225, + "learning_rate": 1.883771775684991e-05, + "loss": 0.14665530920028685, + "step": 1780 + }, + { + "epoch": 0.18148291310491665, + "grad_norm": 5.04036808013916, + "learning_rate": 1.882230634477446e-05, + "loss": 0.16397883892059326, + "step": 1790 + }, + { + "epoch": 0.18249678412784914, + "grad_norm": 5.832187652587891, + "learning_rate": 1.8806799819857598e-05, + "loss": 0.1804431200027466, + "step": 1800 + }, + { + "epoch": 0.18351065515078163, + "grad_norm": 2.0545175075531006, + "learning_rate": 1.8791198349274393e-05, + "loss": 0.15567407608032227, + "step": 1810 + }, + { + "epoch": 0.18452452617371412, + "grad_norm": 2.901912212371826, + "learning_rate": 1.8775502101223516e-05, + "loss": 0.17212533950805664, + "step": 1820 + }, + { + "epoch": 0.1855383971966466, + "grad_norm": 3.6515910625457764, + "learning_rate": 1.8759711244925424e-05, + "loss": 0.1622205138206482, + "step": 1830 + }, + { + "epoch": 0.18655226821957913, + "grad_norm": 2.1737754344940186, + "learning_rate": 1.874382595062055e-05, + "loss": 0.1439922571182251, + "step": 1840 + }, + { + "epoch": 0.18756613924251161, + "grad_norm": 1.4249337911605835, + "learning_rate": 1.8727846389567457e-05, + "loss": 0.15875096321105958, + "step": 1850 + }, + { + "epoch": 0.1885800102654441, + "grad_norm": 2.9065423011779785, + "learning_rate": 1.871177273404098e-05, + "loss": 0.16973226070404052, + "step": 1860 + }, + { + "epoch": 0.1895938812883766, + "grad_norm": 3.1681935787200928, + "learning_rate": 1.8695605157330398e-05, + "loss": 0.14407424926757811, + "step": 1870 + }, + { + "epoch": 0.1906077523113091, + "grad_norm": 2.3213822841644287, + "learning_rate": 1.8679343833737546e-05, + "loss": 0.1699671745300293, + "step": 1880 + }, + { + "epoch": 0.1916216233342416, + "grad_norm": 2.872251033782959, + "learning_rate": 1.8662988938574926e-05, + "loss": 0.1690751314163208, + "step": 1890 + }, + { + "epoch": 0.1926354943571741, + "grad_norm": 2.3576741218566895, + "learning_rate": 1.8646540648163853e-05, + "loss": 0.1593594431877136, + "step": 1900 + }, + { + "epoch": 0.19364936538010658, + "grad_norm": 2.7938263416290283, + "learning_rate": 1.862999913983251e-05, + "loss": 0.1736038565635681, + "step": 1910 + }, + { + "epoch": 0.1946632364030391, + "grad_norm": 3.896066665649414, + "learning_rate": 1.8613364591914073e-05, + "loss": 0.1546809434890747, + "step": 1920 + }, + { + "epoch": 0.19567710742597158, + "grad_norm": 3.4670207500457764, + "learning_rate": 1.8596637183744762e-05, + "loss": 0.15757381916046143, + "step": 1930 + }, + { + "epoch": 0.19669097844890407, + "grad_norm": 5.196355819702148, + "learning_rate": 1.857981709566193e-05, + "loss": 0.17125521898269652, + "step": 1940 + }, + { + "epoch": 0.19770484947183656, + "grad_norm": 3.148313522338867, + "learning_rate": 1.85629045090021e-05, + "loss": 0.15868176221847535, + "step": 1950 + }, + { + "epoch": 0.19871872049476905, + "grad_norm": 2.0824480056762695, + "learning_rate": 1.854589960609902e-05, + "loss": 0.14017997980117797, + "step": 1960 + }, + { + "epoch": 0.19973259151770156, + "grad_norm": 2.810434579849243, + "learning_rate": 1.8528802570281686e-05, + "loss": 0.1733710765838623, + "step": 1970 + }, + { + "epoch": 0.20074646254063405, + "grad_norm": 5.520371437072754, + "learning_rate": 1.851161358587239e-05, + "loss": 0.158262300491333, + "step": 1980 + }, + { + "epoch": 0.20176033356356654, + "grad_norm": 4.164485931396484, + "learning_rate": 1.8494332838184707e-05, + "loss": 0.17306861877441407, + "step": 1990 + }, + { + "epoch": 0.20277420458649903, + "grad_norm": 2.2761194705963135, + "learning_rate": 1.847696051352151e-05, + "loss": 0.15668495893478393, + "step": 2000 + }, + { + "epoch": 0.20378807560943155, + "grad_norm": 2.297306776046753, + "learning_rate": 1.8459496799172957e-05, + "loss": 0.15739670991897584, + "step": 2010 + }, + { + "epoch": 0.20480194663236403, + "grad_norm": 7.117006301879883, + "learning_rate": 1.8441941883414475e-05, + "loss": 0.17943084239959717, + "step": 2020 + }, + { + "epoch": 0.20581581765529652, + "grad_norm": 3.58139705657959, + "learning_rate": 1.8424295955504727e-05, + "loss": 0.15485292673110962, + "step": 2030 + }, + { + "epoch": 0.206829688678229, + "grad_norm": 2.7769317626953125, + "learning_rate": 1.840655920568358e-05, + "loss": 0.1753205418586731, + "step": 2040 + }, + { + "epoch": 0.20784355970116153, + "grad_norm": 3.5762388706207275, + "learning_rate": 1.838873182517004e-05, + "loss": 0.16791033744812012, + "step": 2050 + }, + { + "epoch": 0.20885743072409402, + "grad_norm": 2.852513551712036, + "learning_rate": 1.83708140061602e-05, + "loss": 0.16897099018096923, + "step": 2060 + }, + { + "epoch": 0.2098713017470265, + "grad_norm": 5.638570308685303, + "learning_rate": 1.835280594182517e-05, + "loss": 0.1597752571105957, + "step": 2070 + }, + { + "epoch": 0.210885172769959, + "grad_norm": 2.3849005699157715, + "learning_rate": 1.833470782630898e-05, + "loss": 0.16221199035644532, + "step": 2080 + }, + { + "epoch": 0.21189904379289148, + "grad_norm": 4.6454548835754395, + "learning_rate": 1.831651985472651e-05, + "loss": 0.16262226104736327, + "step": 2090 + }, + { + "epoch": 0.212912914815824, + "grad_norm": 5.632201194763184, + "learning_rate": 1.8298242223161362e-05, + "loss": 0.17465677261352539, + "step": 2100 + }, + { + "epoch": 0.2139267858387565, + "grad_norm": 2.3495728969573975, + "learning_rate": 1.827987512866377e-05, + "loss": 0.1586902379989624, + "step": 2110 + }, + { + "epoch": 0.21494065686168898, + "grad_norm": 4.239703178405762, + "learning_rate": 1.826141876924845e-05, + "loss": 0.15720505714416505, + "step": 2120 + }, + { + "epoch": 0.21595452788462147, + "grad_norm": 6.479139804840088, + "learning_rate": 1.8242873343892494e-05, + "loss": 0.1803435802459717, + "step": 2130 + }, + { + "epoch": 0.21696839890755398, + "grad_norm": 2.918549060821533, + "learning_rate": 1.8224239052533194e-05, + "loss": 0.16124389171600342, + "step": 2140 + }, + { + "epoch": 0.21798226993048647, + "grad_norm": 5.072047233581543, + "learning_rate": 1.820551609606591e-05, + "loss": 0.16751630306243898, + "step": 2150 + }, + { + "epoch": 0.21899614095341896, + "grad_norm": 2.5108635425567627, + "learning_rate": 1.81867046763419e-05, + "loss": 0.17503705024719238, + "step": 2160 + }, + { + "epoch": 0.22001001197635145, + "grad_norm": 2.878690481185913, + "learning_rate": 1.816780499616613e-05, + "loss": 0.17392992973327637, + "step": 2170 + }, + { + "epoch": 0.22102388299928397, + "grad_norm": 3.171135902404785, + "learning_rate": 1.81488172592951e-05, + "loss": 0.15935876369476318, + "step": 2180 + }, + { + "epoch": 0.22203775402221645, + "grad_norm": 2.0318498611450195, + "learning_rate": 1.812974167043465e-05, + "loss": 0.15613303184509278, + "step": 2190 + }, + { + "epoch": 0.22305162504514894, + "grad_norm": 2.1441421508789062, + "learning_rate": 1.8110578435237748e-05, + "loss": 0.1641915559768677, + "step": 2200 + }, + { + "epoch": 0.22406549606808143, + "grad_norm": 3.2369394302368164, + "learning_rate": 1.8091327760302267e-05, + "loss": 0.16606438159942627, + "step": 2210 + }, + { + "epoch": 0.22507936709101392, + "grad_norm": 9.453725814819336, + "learning_rate": 1.8071989853168765e-05, + "loss": 0.19781405925750734, + "step": 2220 + }, + { + "epoch": 0.22609323811394644, + "grad_norm": 1.3892203569412231, + "learning_rate": 1.8052564922318247e-05, + "loss": 0.16559085845947266, + "step": 2230 + }, + { + "epoch": 0.22710710913687893, + "grad_norm": 6.098942279815674, + "learning_rate": 1.803305317716992e-05, + "loss": 0.16889764070510865, + "step": 2240 + }, + { + "epoch": 0.22812098015981142, + "grad_norm": 2.9398958683013916, + "learning_rate": 1.8013454828078925e-05, + "loss": 0.15590085983276367, + "step": 2250 + }, + { + "epoch": 0.2291348511827439, + "grad_norm": 2.4985125064849854, + "learning_rate": 1.7993770086334082e-05, + "loss": 0.15495698451995848, + "step": 2260 + }, + { + "epoch": 0.23014872220567642, + "grad_norm": 2.469414710998535, + "learning_rate": 1.7973999164155597e-05, + "loss": 0.14815448522567748, + "step": 2270 + }, + { + "epoch": 0.2311625932286089, + "grad_norm": 3.4401633739471436, + "learning_rate": 1.795414227469279e-05, + "loss": 0.14610652923583983, + "step": 2280 + }, + { + "epoch": 0.2321764642515414, + "grad_norm": 3.79011607170105, + "learning_rate": 1.7934199632021795e-05, + "loss": 0.1640559434890747, + "step": 2290 + }, + { + "epoch": 0.2331903352744739, + "grad_norm": 2.500624179840088, + "learning_rate": 1.7914171451143235e-05, + "loss": 0.17032144069671631, + "step": 2300 + }, + { + "epoch": 0.2342042062974064, + "grad_norm": 1.8544747829437256, + "learning_rate": 1.7894057947979923e-05, + "loss": 0.15956176519393922, + "step": 2310 + }, + { + "epoch": 0.2352180773203389, + "grad_norm": 3.5280699729919434, + "learning_rate": 1.7873859339374528e-05, + "loss": 0.1741471767425537, + "step": 2320 + }, + { + "epoch": 0.23623194834327138, + "grad_norm": 6.208831310272217, + "learning_rate": 1.7853575843087227e-05, + "loss": 0.15680291652679443, + "step": 2330 + }, + { + "epoch": 0.23724581936620387, + "grad_norm": 1.8986748456954956, + "learning_rate": 1.7833207677793382e-05, + "loss": 0.1563281536102295, + "step": 2340 + }, + { + "epoch": 0.23825969038913636, + "grad_norm": 1.5344898700714111, + "learning_rate": 1.7812755063081156e-05, + "loss": 0.161000657081604, + "step": 2350 + }, + { + "epoch": 0.23927356141206887, + "grad_norm": 2.973267078399658, + "learning_rate": 1.7792218219449165e-05, + "loss": 0.15237464904785156, + "step": 2360 + }, + { + "epoch": 0.24028743243500136, + "grad_norm": 3.7273366451263428, + "learning_rate": 1.777159736830408e-05, + "loss": 0.15424546003341674, + "step": 2370 + }, + { + "epoch": 0.24130130345793385, + "grad_norm": 2.469637155532837, + "learning_rate": 1.7750892731958268e-05, + "loss": 0.15047731399536132, + "step": 2380 + }, + { + "epoch": 0.24231517448086634, + "grad_norm": 2.0232887268066406, + "learning_rate": 1.773010453362737e-05, + "loss": 0.15930893421173095, + "step": 2390 + }, + { + "epoch": 0.24332904550379886, + "grad_norm": 1.7955238819122314, + "learning_rate": 1.770923299742791e-05, + "loss": 0.15245296955108642, + "step": 2400 + }, + { + "epoch": 0.24434291652673135, + "grad_norm": 1.8035132884979248, + "learning_rate": 1.768827834837487e-05, + "loss": 0.15747556686401368, + "step": 2410 + }, + { + "epoch": 0.24535678754966384, + "grad_norm": 2.879544258117676, + "learning_rate": 1.7667240812379266e-05, + "loss": 0.1422521948814392, + "step": 2420 + }, + { + "epoch": 0.24637065857259632, + "grad_norm": 4.262474536895752, + "learning_rate": 1.7646120616245724e-05, + "loss": 0.16299595832824706, + "step": 2430 + }, + { + "epoch": 0.24738452959552884, + "grad_norm": 2.440554141998291, + "learning_rate": 1.7624917987670006e-05, + "loss": 0.17202863693237305, + "step": 2440 + }, + { + "epoch": 0.24839840061846133, + "grad_norm": 2.7783010005950928, + "learning_rate": 1.7603633155236595e-05, + "loss": 0.15401304960250856, + "step": 2450 + }, + { + "epoch": 0.24941227164139382, + "grad_norm": 3.267669439315796, + "learning_rate": 1.7582266348416194e-05, + "loss": 0.17188758850097657, + "step": 2460 + }, + { + "epoch": 0.25042614266432633, + "grad_norm": 2.020472764968872, + "learning_rate": 1.7560817797563275e-05, + "loss": 0.15761997699737548, + "step": 2470 + }, + { + "epoch": 0.2514400136872588, + "grad_norm": 2.1722605228424072, + "learning_rate": 1.7539287733913592e-05, + "loss": 0.16235024929046632, + "step": 2480 + }, + { + "epoch": 0.2524538847101913, + "grad_norm": 4.620025634765625, + "learning_rate": 1.7517676389581675e-05, + "loss": 0.1612251043319702, + "step": 2490 + }, + { + "epoch": 0.2534677557331238, + "grad_norm": 4.868301868438721, + "learning_rate": 1.7495983997558342e-05, + "loss": 0.1509876012802124, + "step": 2500 + }, + { + "epoch": 0.2544816267560563, + "grad_norm": 3.7973430156707764, + "learning_rate": 1.7474210791708177e-05, + "loss": 0.16187913417816163, + "step": 2510 + }, + { + "epoch": 0.2554954977789888, + "grad_norm": 3.036173105239868, + "learning_rate": 1.7452357006767026e-05, + "loss": 0.15743194818496703, + "step": 2520 + }, + { + "epoch": 0.25650936880192127, + "grad_norm": 2.3557236194610596, + "learning_rate": 1.7430422878339443e-05, + "loss": 0.159504234790802, + "step": 2530 + }, + { + "epoch": 0.25752323982485376, + "grad_norm": 2.251978635787964, + "learning_rate": 1.740840864289616e-05, + "loss": 0.15690701007843016, + "step": 2540 + }, + { + "epoch": 0.2585371108477863, + "grad_norm": 2.0876102447509766, + "learning_rate": 1.7386314537771553e-05, + "loss": 0.153395414352417, + "step": 2550 + }, + { + "epoch": 0.2595509818707188, + "grad_norm": 4.1484246253967285, + "learning_rate": 1.736414080116105e-05, + "loss": 0.16843408346176147, + "step": 2560 + }, + { + "epoch": 0.2605648528936513, + "grad_norm": 2.486083507537842, + "learning_rate": 1.7341887672118596e-05, + "loss": 0.15694656372070312, + "step": 2570 + }, + { + "epoch": 0.26157872391658377, + "grad_norm": 2.366814374923706, + "learning_rate": 1.7319555390554054e-05, + "loss": 0.17814427614212036, + "step": 2580 + }, + { + "epoch": 0.26259259493951626, + "grad_norm": 2.5121498107910156, + "learning_rate": 1.7297144197230637e-05, + "loss": 0.16293206214904785, + "step": 2590 + }, + { + "epoch": 0.26360646596244874, + "grad_norm": 2.656708240509033, + "learning_rate": 1.7274654333762284e-05, + "loss": 0.14932069778442383, + "step": 2600 + }, + { + "epoch": 0.26462033698538123, + "grad_norm": 4.215566158294678, + "learning_rate": 1.7252086042611086e-05, + "loss": 0.1669047236442566, + "step": 2610 + }, + { + "epoch": 0.2656342080083137, + "grad_norm": 2.3468356132507324, + "learning_rate": 1.722943956708466e-05, + "loss": 0.14947493076324464, + "step": 2620 + }, + { + "epoch": 0.2666480790312462, + "grad_norm": 2.33480167388916, + "learning_rate": 1.720671515133352e-05, + "loss": 0.15656208992004395, + "step": 2630 + }, + { + "epoch": 0.26766195005417875, + "grad_norm": 4.482690334320068, + "learning_rate": 1.7183913040348447e-05, + "loss": 0.16331549882888793, + "step": 2640 + }, + { + "epoch": 0.26867582107711124, + "grad_norm": 3.6132500171661377, + "learning_rate": 1.716103347995785e-05, + "loss": 0.1586040139198303, + "step": 2650 + }, + { + "epoch": 0.26968969210004373, + "grad_norm": 2.0165648460388184, + "learning_rate": 1.713807671682513e-05, + "loss": 0.14562156200408935, + "step": 2660 + }, + { + "epoch": 0.2707035631229762, + "grad_norm": 2.3356478214263916, + "learning_rate": 1.711504299844599e-05, + "loss": 0.15407261848449708, + "step": 2670 + }, + { + "epoch": 0.2717174341459087, + "grad_norm": 1.663870096206665, + "learning_rate": 1.7091932573145798e-05, + "loss": 0.16880929470062256, + "step": 2680 + }, + { + "epoch": 0.2727313051688412, + "grad_norm": 5.44504976272583, + "learning_rate": 1.706874569007688e-05, + "loss": 0.1500933885574341, + "step": 2690 + }, + { + "epoch": 0.2737451761917737, + "grad_norm": 1.3699334859848022, + "learning_rate": 1.7045482599215856e-05, + "loss": 0.16571455001831054, + "step": 2700 + }, + { + "epoch": 0.2747590472147062, + "grad_norm": 3.8864905834198, + "learning_rate": 1.7022143551360947e-05, + "loss": 0.1511744260787964, + "step": 2710 + }, + { + "epoch": 0.2757729182376387, + "grad_norm": 3.1794416904449463, + "learning_rate": 1.699872879812925e-05, + "loss": 0.16612961292266845, + "step": 2720 + }, + { + "epoch": 0.2767867892605712, + "grad_norm": 4.5570597648620605, + "learning_rate": 1.6975238591954048e-05, + "loss": 0.178527569770813, + "step": 2730 + }, + { + "epoch": 0.2778006602835037, + "grad_norm": 2.7645223140716553, + "learning_rate": 1.695167318608206e-05, + "loss": 0.16305990219116212, + "step": 2740 + }, + { + "epoch": 0.2788145313064362, + "grad_norm": 2.147143602371216, + "learning_rate": 1.6928032834570767e-05, + "loss": 0.16551218032836915, + "step": 2750 + }, + { + "epoch": 0.2798284023293687, + "grad_norm": 3.0803322792053223, + "learning_rate": 1.690431779228559e-05, + "loss": 0.16492114067077637, + "step": 2760 + }, + { + "epoch": 0.28084227335230116, + "grad_norm": 4.966836452484131, + "learning_rate": 1.6880528314897212e-05, + "loss": 0.1623425841331482, + "step": 2770 + }, + { + "epoch": 0.28185614437523365, + "grad_norm": 4.413325786590576, + "learning_rate": 1.6856664658878797e-05, + "loss": 0.15777130126953126, + "step": 2780 + }, + { + "epoch": 0.28287001539816614, + "grad_norm": 2.946309804916382, + "learning_rate": 1.6832727081503227e-05, + "loss": 0.16015386581420898, + "step": 2790 + }, + { + "epoch": 0.28388388642109863, + "grad_norm": 2.983022451400757, + "learning_rate": 1.6808715840840308e-05, + "loss": 0.1487659215927124, + "step": 2800 + }, + { + "epoch": 0.2848977574440312, + "grad_norm": 4.7747697830200195, + "learning_rate": 1.6784631195754027e-05, + "loss": 0.1590745210647583, + "step": 2810 + }, + { + "epoch": 0.28591162846696366, + "grad_norm": 2.849536180496216, + "learning_rate": 1.6760473405899733e-05, + "loss": 0.1399478554725647, + "step": 2820 + }, + { + "epoch": 0.28692549948989615, + "grad_norm": 3.156346321105957, + "learning_rate": 1.673624273172134e-05, + "loss": 0.15951600074768066, + "step": 2830 + }, + { + "epoch": 0.28793937051282864, + "grad_norm": 1.9825046062469482, + "learning_rate": 1.6711939434448537e-05, + "loss": 0.14981322288513182, + "step": 2840 + }, + { + "epoch": 0.28895324153576113, + "grad_norm": 4.588251113891602, + "learning_rate": 1.6687563776093943e-05, + "loss": 0.1673480749130249, + "step": 2850 + }, + { + "epoch": 0.2899671125586936, + "grad_norm": 2.469059944152832, + "learning_rate": 1.6663116019450306e-05, + "loss": 0.14798378944396973, + "step": 2860 + }, + { + "epoch": 0.2909809835816261, + "grad_norm": 1.6991090774536133, + "learning_rate": 1.6638596428087662e-05, + "loss": 0.1673067569732666, + "step": 2870 + }, + { + "epoch": 0.2919948546045586, + "grad_norm": 4.821193695068359, + "learning_rate": 1.66140052663505e-05, + "loss": 0.15859930515289306, + "step": 2880 + }, + { + "epoch": 0.29300872562749114, + "grad_norm": 2.6565518379211426, + "learning_rate": 1.6589342799354895e-05, + "loss": 0.1571721076965332, + "step": 2890 + }, + { + "epoch": 0.29402259665042363, + "grad_norm": 1.807891607284546, + "learning_rate": 1.6564609292985663e-05, + "loss": 0.1556539535522461, + "step": 2900 + }, + { + "epoch": 0.2950364676733561, + "grad_norm": 4.116388320922852, + "learning_rate": 1.6539805013893493e-05, + "loss": 0.16832677125930787, + "step": 2910 + }, + { + "epoch": 0.2960503386962886, + "grad_norm": 6.09294319152832, + "learning_rate": 1.6514930229492073e-05, + "loss": 0.1712158203125, + "step": 2920 + }, + { + "epoch": 0.2970642097192211, + "grad_norm": 4.736142158508301, + "learning_rate": 1.6489985207955207e-05, + "loss": 0.16103771924972535, + "step": 2930 + }, + { + "epoch": 0.2980780807421536, + "grad_norm": 3.4793078899383545, + "learning_rate": 1.6464970218213912e-05, + "loss": 0.1659522533416748, + "step": 2940 + }, + { + "epoch": 0.2990919517650861, + "grad_norm": 3.375481605529785, + "learning_rate": 1.643988552995354e-05, + "loss": 0.16271287202835083, + "step": 2950 + }, + { + "epoch": 0.30010582278801856, + "grad_norm": 3.86802077293396, + "learning_rate": 1.6414731413610856e-05, + "loss": 0.14652674198150634, + "step": 2960 + }, + { + "epoch": 0.30111969381095105, + "grad_norm": 2.423678159713745, + "learning_rate": 1.638950814037112e-05, + "loss": 0.16136612892150878, + "step": 2970 + }, + { + "epoch": 0.3021335648338836, + "grad_norm": 3.0169012546539307, + "learning_rate": 1.6364215982165178e-05, + "loss": 0.15249059200286866, + "step": 2980 + }, + { + "epoch": 0.3031474358568161, + "grad_norm": 1.9516414403915405, + "learning_rate": 1.633885521166651e-05, + "loss": 0.14589663743972778, + "step": 2990 + }, + { + "epoch": 0.30416130687974857, + "grad_norm": 2.120316505432129, + "learning_rate": 1.6313426102288315e-05, + "loss": 0.15704749822616576, + "step": 3000 + }, + { + "epoch": 0.30517517790268106, + "grad_norm": 5.109221935272217, + "learning_rate": 1.6287928928180544e-05, + "loss": 0.16726391315460204, + "step": 3010 + }, + { + "epoch": 0.30618904892561355, + "grad_norm": 1.8231446743011475, + "learning_rate": 1.6262363964226943e-05, + "loss": 0.14267534017562866, + "step": 3020 + }, + { + "epoch": 0.30720291994854604, + "grad_norm": 3.893432140350342, + "learning_rate": 1.623673148604211e-05, + "loss": 0.14939746856689454, + "step": 3030 + }, + { + "epoch": 0.3082167909714785, + "grad_norm": 2.441767930984497, + "learning_rate": 1.6211031769968503e-05, + "loss": 0.1522477388381958, + "step": 3040 + }, + { + "epoch": 0.309230661994411, + "grad_norm": 3.3933162689208984, + "learning_rate": 1.6185265093073468e-05, + "loss": 0.1500061869621277, + "step": 3050 + }, + { + "epoch": 0.3102445330173435, + "grad_norm": 3.2442116737365723, + "learning_rate": 1.6159431733146255e-05, + "loss": 0.15661410093307496, + "step": 3060 + }, + { + "epoch": 0.31125840404027605, + "grad_norm": 2.3887383937835693, + "learning_rate": 1.613353196869502e-05, + "loss": 0.15814971923828125, + "step": 3070 + }, + { + "epoch": 0.31227227506320854, + "grad_norm": 2.406839609146118, + "learning_rate": 1.6107566078943818e-05, + "loss": 0.16517699956893922, + "step": 3080 + }, + { + "epoch": 0.313286146086141, + "grad_norm": 2.143404960632324, + "learning_rate": 1.608153434382961e-05, + "loss": 0.16133458614349366, + "step": 3090 + }, + { + "epoch": 0.3143000171090735, + "grad_norm": 1.3094817399978638, + "learning_rate": 1.6055437043999215e-05, + "loss": 0.14347445964813232, + "step": 3100 + }, + { + "epoch": 0.315313888132006, + "grad_norm": 1.9134424924850464, + "learning_rate": 1.6029274460806318e-05, + "loss": 0.1571304202079773, + "step": 3110 + }, + { + "epoch": 0.3163277591549385, + "grad_norm": 2.6114089488983154, + "learning_rate": 1.6003046876308413e-05, + "loss": 0.16014952659606935, + "step": 3120 + }, + { + "epoch": 0.317341630177871, + "grad_norm": 3.087104558944702, + "learning_rate": 1.5976754573263767e-05, + "loss": 0.13797740936279296, + "step": 3130 + }, + { + "epoch": 0.31835550120080347, + "grad_norm": 3.3546719551086426, + "learning_rate": 1.5950397835128378e-05, + "loss": 0.15225110054016114, + "step": 3140 + }, + { + "epoch": 0.319369372223736, + "grad_norm": 2.468059539794922, + "learning_rate": 1.5923976946052918e-05, + "loss": 0.13575925827026367, + "step": 3150 + }, + { + "epoch": 0.3203832432466685, + "grad_norm": 4.817075252532959, + "learning_rate": 1.5897492190879667e-05, + "loss": 0.1568129062652588, + "step": 3160 + }, + { + "epoch": 0.321397114269601, + "grad_norm": 2.0758056640625, + "learning_rate": 1.5870943855139437e-05, + "loss": 0.16206610202789307, + "step": 3170 + }, + { + "epoch": 0.3224109852925335, + "grad_norm": 2.5459113121032715, + "learning_rate": 1.5844332225048496e-05, + "loss": 0.1420430898666382, + "step": 3180 + }, + { + "epoch": 0.32342485631546597, + "grad_norm": 3.854602575302124, + "learning_rate": 1.5817657587505503e-05, + "loss": 0.15001358985900878, + "step": 3190 + }, + { + "epoch": 0.32443872733839846, + "grad_norm": 2.9747657775878906, + "learning_rate": 1.5790920230088375e-05, + "loss": 0.1558264374732971, + "step": 3200 + }, + { + "epoch": 0.32545259836133095, + "grad_norm": 5.2626848220825195, + "learning_rate": 1.576412044105122e-05, + "loss": 0.14873929023742677, + "step": 3210 + }, + { + "epoch": 0.32646646938426344, + "grad_norm": 3.998882532119751, + "learning_rate": 1.5737258509321224e-05, + "loss": 0.146403706073761, + "step": 3220 + }, + { + "epoch": 0.3274803404071959, + "grad_norm": 4.411140441894531, + "learning_rate": 1.5710334724495514e-05, + "loss": 0.14902397394180297, + "step": 3230 + }, + { + "epoch": 0.32849421143012847, + "grad_norm": 1.9681049585342407, + "learning_rate": 1.5683349376838067e-05, + "loss": 0.15362627506256105, + "step": 3240 + }, + { + "epoch": 0.32950808245306096, + "grad_norm": 2.869114637374878, + "learning_rate": 1.565630275727656e-05, + "loss": 0.14706436395645142, + "step": 3250 + }, + { + "epoch": 0.33052195347599345, + "grad_norm": 3.1659183502197266, + "learning_rate": 1.5629195157399236e-05, + "loss": 0.1660413146018982, + "step": 3260 + }, + { + "epoch": 0.33153582449892594, + "grad_norm": 3.8156678676605225, + "learning_rate": 1.5602026869451773e-05, + "loss": 0.15087927579879762, + "step": 3270 + }, + { + "epoch": 0.3325496955218584, + "grad_norm": 2.207087516784668, + "learning_rate": 1.5574798186334112e-05, + "loss": 0.14515562057495118, + "step": 3280 + }, + { + "epoch": 0.3335635665447909, + "grad_norm": 2.301206588745117, + "learning_rate": 1.5547509401597322e-05, + "loss": 0.1525092601776123, + "step": 3290 + }, + { + "epoch": 0.3345774375677234, + "grad_norm": 1.7333687543869019, + "learning_rate": 1.552016080944042e-05, + "loss": 0.147652530670166, + "step": 3300 + }, + { + "epoch": 0.3355913085906559, + "grad_norm": 3.2019591331481934, + "learning_rate": 1.5492752704707198e-05, + "loss": 0.15862106084823607, + "step": 3310 + }, + { + "epoch": 0.3366051796135884, + "grad_norm": 4.577446937561035, + "learning_rate": 1.5465285382883065e-05, + "loss": 0.16009078025817872, + "step": 3320 + }, + { + "epoch": 0.3376190506365209, + "grad_norm": 5.188497543334961, + "learning_rate": 1.5437759140091827e-05, + "loss": 0.16193052530288696, + "step": 3330 + }, + { + "epoch": 0.3386329216594534, + "grad_norm": 2.946054458618164, + "learning_rate": 1.541017427309254e-05, + "loss": 0.15738890171051026, + "step": 3340 + }, + { + "epoch": 0.3396467926823859, + "grad_norm": 2.484792947769165, + "learning_rate": 1.538253107927625e-05, + "loss": 0.14138500690460204, + "step": 3350 + }, + { + "epoch": 0.3406606637053184, + "grad_norm": 3.500901937484741, + "learning_rate": 1.535482985666285e-05, + "loss": 0.1425542950630188, + "step": 3360 + }, + { + "epoch": 0.3416745347282509, + "grad_norm": 4.271117687225342, + "learning_rate": 1.5327070903897832e-05, + "loss": 0.16243972778320312, + "step": 3370 + }, + { + "epoch": 0.34268840575118337, + "grad_norm": 4.930404186248779, + "learning_rate": 1.529925452024907e-05, + "loss": 0.15196540355682372, + "step": 3380 + }, + { + "epoch": 0.34370227677411586, + "grad_norm": 4.2457098960876465, + "learning_rate": 1.5271381005603593e-05, + "loss": 0.15773863792419435, + "step": 3390 + }, + { + "epoch": 0.34471614779704834, + "grad_norm": 4.004637241363525, + "learning_rate": 1.5243450660464368e-05, + "loss": 0.14284372329711914, + "step": 3400 + }, + { + "epoch": 0.3457300188199809, + "grad_norm": 3.1746811866760254, + "learning_rate": 1.5215463785947036e-05, + "loss": 0.15854647159576415, + "step": 3410 + }, + { + "epoch": 0.3467438898429134, + "grad_norm": 2.0277047157287598, + "learning_rate": 1.518742068377669e-05, + "loss": 0.14736793041229249, + "step": 3420 + }, + { + "epoch": 0.34775776086584587, + "grad_norm": 2.482897996902466, + "learning_rate": 1.5159321656284602e-05, + "loss": 0.149668550491333, + "step": 3430 + }, + { + "epoch": 0.34877163188877836, + "grad_norm": 2.8284385204315186, + "learning_rate": 1.5131167006404978e-05, + "loss": 0.1610569715499878, + "step": 3440 + }, + { + "epoch": 0.34978550291171084, + "grad_norm": 6.1309494972229, + "learning_rate": 1.5102957037671689e-05, + "loss": 0.1485116720199585, + "step": 3450 + }, + { + "epoch": 0.35079937393464333, + "grad_norm": 2.461714267730713, + "learning_rate": 1.5074692054214977e-05, + "loss": 0.15174973011016846, + "step": 3460 + }, + { + "epoch": 0.3518132449575758, + "grad_norm": 3.3416595458984375, + "learning_rate": 1.504637236075822e-05, + "loss": 0.13843655586242676, + "step": 3470 + }, + { + "epoch": 0.3528271159805083, + "grad_norm": 3.9959776401519775, + "learning_rate": 1.5017998262614609e-05, + "loss": 0.1539894938468933, + "step": 3480 + }, + { + "epoch": 0.3538409870034408, + "grad_norm": 1.432332158088684, + "learning_rate": 1.498957006568386e-05, + "loss": 0.1476918339729309, + "step": 3490 + }, + { + "epoch": 0.35485485802637334, + "grad_norm": 4.456820011138916, + "learning_rate": 1.4961088076448947e-05, + "loss": 0.14358968734741212, + "step": 3500 + }, + { + "epoch": 0.35586872904930583, + "grad_norm": 2.1591198444366455, + "learning_rate": 1.4932552601972763e-05, + "loss": 0.15828763246536254, + "step": 3510 + }, + { + "epoch": 0.3568826000722383, + "grad_norm": 4.589548110961914, + "learning_rate": 1.490396394989482e-05, + "loss": 0.1551862597465515, + "step": 3520 + }, + { + "epoch": 0.3578964710951708, + "grad_norm": 2.3587028980255127, + "learning_rate": 1.4875322428427943e-05, + "loss": 0.14813508987426757, + "step": 3530 + }, + { + "epoch": 0.3589103421181033, + "grad_norm": 1.9275258779525757, + "learning_rate": 1.4846628346354934e-05, + "loss": 0.1424973726272583, + "step": 3540 + }, + { + "epoch": 0.3599242131410358, + "grad_norm": 3.939291000366211, + "learning_rate": 1.4817882013025253e-05, + "loss": 0.166415011882782, + "step": 3550 + }, + { + "epoch": 0.3609380841639683, + "grad_norm": 2.0860707759857178, + "learning_rate": 1.478908373835167e-05, + "loss": 0.16250795125961304, + "step": 3560 + }, + { + "epoch": 0.36195195518690076, + "grad_norm": 2.3876686096191406, + "learning_rate": 1.476023383280694e-05, + "loss": 0.14899988174438478, + "step": 3570 + }, + { + "epoch": 0.3629658262098333, + "grad_norm": 4.007850646972656, + "learning_rate": 1.473133260742045e-05, + "loss": 0.1472618818283081, + "step": 3580 + }, + { + "epoch": 0.3639796972327658, + "grad_norm": 3.6748263835906982, + "learning_rate": 1.4702380373774852e-05, + "loss": 0.1574415922164917, + "step": 3590 + }, + { + "epoch": 0.3649935682556983, + "grad_norm": 3.9576895236968994, + "learning_rate": 1.4673377444002724e-05, + "loss": 0.1385578393936157, + "step": 3600 + }, + { + "epoch": 0.3660074392786308, + "grad_norm": 5.127712726593018, + "learning_rate": 1.4644324130783197e-05, + "loss": 0.14661288261413574, + "step": 3610 + }, + { + "epoch": 0.36702131030156326, + "grad_norm": 1.5397738218307495, + "learning_rate": 1.4615220747338574e-05, + "loss": 0.16506681442260743, + "step": 3620 + }, + { + "epoch": 0.36803518132449575, + "grad_norm": 3.9013655185699463, + "learning_rate": 1.4586067607430976e-05, + "loss": 0.15868569612503053, + "step": 3630 + }, + { + "epoch": 0.36904905234742824, + "grad_norm": 3.287862539291382, + "learning_rate": 1.4556865025358942e-05, + "loss": 0.15742528438568115, + "step": 3640 + }, + { + "epoch": 0.37006292337036073, + "grad_norm": 2.3225042819976807, + "learning_rate": 1.4527613315954035e-05, + "loss": 0.14953329563140869, + "step": 3650 + }, + { + "epoch": 0.3710767943932932, + "grad_norm": 2.5588650703430176, + "learning_rate": 1.4498312794577465e-05, + "loss": 0.1377772569656372, + "step": 3660 + }, + { + "epoch": 0.37209066541622576, + "grad_norm": 2.131206750869751, + "learning_rate": 1.4468963777116686e-05, + "loss": 0.16717950105667115, + "step": 3670 + }, + { + "epoch": 0.37310453643915825, + "grad_norm": 3.6540908813476562, + "learning_rate": 1.4439566579981979e-05, + "loss": 0.15226175785064697, + "step": 3680 + }, + { + "epoch": 0.37411840746209074, + "grad_norm": 2.3808586597442627, + "learning_rate": 1.4410121520103045e-05, + "loss": 0.14323383569717407, + "step": 3690 + }, + { + "epoch": 0.37513227848502323, + "grad_norm": 2.266160488128662, + "learning_rate": 1.4380628914925597e-05, + "loss": 0.13526065349578859, + "step": 3700 + }, + { + "epoch": 0.3761461495079557, + "grad_norm": 3.2769813537597656, + "learning_rate": 1.4351089082407929e-05, + "loss": 0.1488935351371765, + "step": 3710 + }, + { + "epoch": 0.3771600205308882, + "grad_norm": 3.619873523712158, + "learning_rate": 1.4321502341017496e-05, + "loss": 0.15701539516448976, + "step": 3720 + }, + { + "epoch": 0.3781738915538207, + "grad_norm": 3.7231035232543945, + "learning_rate": 1.4291869009727463e-05, + "loss": 0.15080446004867554, + "step": 3730 + }, + { + "epoch": 0.3791877625767532, + "grad_norm": 3.057117223739624, + "learning_rate": 1.4262189408013294e-05, + "loss": 0.14088962078094483, + "step": 3740 + }, + { + "epoch": 0.3802016335996857, + "grad_norm": 4.1840081214904785, + "learning_rate": 1.4232463855849275e-05, + "loss": 0.1632833242416382, + "step": 3750 + }, + { + "epoch": 0.3812155046226182, + "grad_norm": 1.6274490356445312, + "learning_rate": 1.4202692673705097e-05, + "loss": 0.16335034370422363, + "step": 3760 + }, + { + "epoch": 0.3822293756455507, + "grad_norm": 2.3689301013946533, + "learning_rate": 1.4172876182542372e-05, + "loss": 0.14884527921676635, + "step": 3770 + }, + { + "epoch": 0.3832432466684832, + "grad_norm": 3.346585273742676, + "learning_rate": 1.41430147038112e-05, + "loss": 0.14609973430633544, + "step": 3780 + }, + { + "epoch": 0.3842571176914157, + "grad_norm": 2.684267997741699, + "learning_rate": 1.411310855944668e-05, + "loss": 0.15893809795379638, + "step": 3790 + }, + { + "epoch": 0.3852709887143482, + "grad_norm": 2.9765217304229736, + "learning_rate": 1.4083158071865451e-05, + "loss": 0.15987597703933715, + "step": 3800 + }, + { + "epoch": 0.38628485973728066, + "grad_norm": 2.1168746948242188, + "learning_rate": 1.4053163563962215e-05, + "loss": 0.1542700171470642, + "step": 3810 + }, + { + "epoch": 0.38729873076021315, + "grad_norm": 3.052720069885254, + "learning_rate": 1.4023125359106253e-05, + "loss": 0.1487845301628113, + "step": 3820 + }, + { + "epoch": 0.38831260178314564, + "grad_norm": 2.674450635910034, + "learning_rate": 1.3993043781137945e-05, + "loss": 0.14977353811264038, + "step": 3830 + }, + { + "epoch": 0.3893264728060782, + "grad_norm": 2.2986960411071777, + "learning_rate": 1.3962919154365273e-05, + "loss": 0.1593056082725525, + "step": 3840 + }, + { + "epoch": 0.39034034382901067, + "grad_norm": 1.720723271369934, + "learning_rate": 1.393275180356032e-05, + "loss": 0.15123729705810546, + "step": 3850 + }, + { + "epoch": 0.39135421485194316, + "grad_norm": 4.039378643035889, + "learning_rate": 1.3902542053955781e-05, + "loss": 0.15710922479629516, + "step": 3860 + }, + { + "epoch": 0.39236808587487565, + "grad_norm": 2.1258432865142822, + "learning_rate": 1.387229023124145e-05, + "loss": 0.15607482194900513, + "step": 3870 + }, + { + "epoch": 0.39338195689780814, + "grad_norm": 2.0659918785095215, + "learning_rate": 1.3841996661560706e-05, + "loss": 0.14743976593017577, + "step": 3880 + }, + { + "epoch": 0.3943958279207406, + "grad_norm": 4.122640132904053, + "learning_rate": 1.3811661671507004e-05, + "loss": 0.13856662511825563, + "step": 3890 + }, + { + "epoch": 0.3954096989436731, + "grad_norm": 5.116918563842773, + "learning_rate": 1.3781285588120343e-05, + "loss": 0.14389435052871705, + "step": 3900 + }, + { + "epoch": 0.3964235699666056, + "grad_norm": 2.2652273178100586, + "learning_rate": 1.3750868738883755e-05, + "loss": 0.1469268798828125, + "step": 3910 + }, + { + "epoch": 0.3974374409895381, + "grad_norm": 3.995129346847534, + "learning_rate": 1.3720411451719762e-05, + "loss": 0.15586447715759277, + "step": 3920 + }, + { + "epoch": 0.39845131201247064, + "grad_norm": 1.9743337631225586, + "learning_rate": 1.3689914054986849e-05, + "loss": 0.1541958212852478, + "step": 3930 + }, + { + "epoch": 0.3994651830354031, + "grad_norm": 3.7179393768310547, + "learning_rate": 1.3659376877475916e-05, + "loss": 0.15046162605285646, + "step": 3940 + }, + { + "epoch": 0.4004790540583356, + "grad_norm": 4.0756120681762695, + "learning_rate": 1.3628800248406738e-05, + "loss": 0.1484516143798828, + "step": 3950 + }, + { + "epoch": 0.4014929250812681, + "grad_norm": 3.1328577995300293, + "learning_rate": 1.3598184497424416e-05, + "loss": 0.1405585527420044, + "step": 3960 + }, + { + "epoch": 0.4025067961042006, + "grad_norm": 2.1790614128112793, + "learning_rate": 1.3567529954595826e-05, + "loss": 0.14826864004135132, + "step": 3970 + }, + { + "epoch": 0.4035206671271331, + "grad_norm": 2.844003438949585, + "learning_rate": 1.3536836950406057e-05, + "loss": 0.1474652886390686, + "step": 3980 + }, + { + "epoch": 0.40453453815006557, + "grad_norm": 2.1645991802215576, + "learning_rate": 1.3506105815754836e-05, + "loss": 0.1415319561958313, + "step": 3990 + }, + { + "epoch": 0.40554840917299806, + "grad_norm": 3.7088541984558105, + "learning_rate": 1.3475336881952988e-05, + "loss": 0.17246508598327637, + "step": 4000 + }, + { + "epoch": 0.40656228019593055, + "grad_norm": 1.6019930839538574, + "learning_rate": 1.3444530480718838e-05, + "loss": 0.15497559309005737, + "step": 4010 + }, + { + "epoch": 0.4075761512188631, + "grad_norm": 1.8616230487823486, + "learning_rate": 1.3413686944174655e-05, + "loss": 0.1340821146965027, + "step": 4020 + }, + { + "epoch": 0.4085900222417956, + "grad_norm": 3.318916082382202, + "learning_rate": 1.3382806604843053e-05, + "loss": 0.1546313524246216, + "step": 4030 + }, + { + "epoch": 0.40960389326472807, + "grad_norm": 2.3759145736694336, + "learning_rate": 1.3351889795643423e-05, + "loss": 0.1341363787651062, + "step": 4040 + }, + { + "epoch": 0.41061776428766056, + "grad_norm": 4.953732490539551, + "learning_rate": 1.3320936849888322e-05, + "loss": 0.15554239749908447, + "step": 4050 + }, + { + "epoch": 0.41163163531059305, + "grad_norm": 2.1146607398986816, + "learning_rate": 1.3289948101279912e-05, + "loss": 0.16502437591552735, + "step": 4060 + }, + { + "epoch": 0.41264550633352554, + "grad_norm": 3.7556099891662598, + "learning_rate": 1.3258923883906337e-05, + "loss": 0.14651042222976685, + "step": 4070 + }, + { + "epoch": 0.413659377356458, + "grad_norm": 1.4182454347610474, + "learning_rate": 1.3227864532238113e-05, + "loss": 0.15607728958129882, + "step": 4080 + }, + { + "epoch": 0.4146732483793905, + "grad_norm": 2.2242259979248047, + "learning_rate": 1.3196770381124557e-05, + "loss": 0.1487821578979492, + "step": 4090 + }, + { + "epoch": 0.41568711940232306, + "grad_norm": 2.7310028076171875, + "learning_rate": 1.3165641765790153e-05, + "loss": 0.1324383497238159, + "step": 4100 + }, + { + "epoch": 0.41670099042525555, + "grad_norm": 1.658709168434143, + "learning_rate": 1.3134479021830938e-05, + "loss": 0.15048023462295532, + "step": 4110 + }, + { + "epoch": 0.41771486144818804, + "grad_norm": 3.928611993789673, + "learning_rate": 1.3103282485210882e-05, + "loss": 0.1399364709854126, + "step": 4120 + }, + { + "epoch": 0.4187287324711205, + "grad_norm": 2.2957942485809326, + "learning_rate": 1.3072052492258286e-05, + "loss": 0.13578213453292848, + "step": 4130 + }, + { + "epoch": 0.419742603494053, + "grad_norm": 2.7933764457702637, + "learning_rate": 1.3040789379662131e-05, + "loss": 0.12907679080963136, + "step": 4140 + }, + { + "epoch": 0.4207564745169855, + "grad_norm": 1.9445477724075317, + "learning_rate": 1.3009493484468464e-05, + "loss": 0.14400614500045777, + "step": 4150 + }, + { + "epoch": 0.421770345539918, + "grad_norm": 1.7721681594848633, + "learning_rate": 1.2978165144076757e-05, + "loss": 0.14715178012847902, + "step": 4160 + }, + { + "epoch": 0.4227842165628505, + "grad_norm": 3.174835681915283, + "learning_rate": 1.2946804696236277e-05, + "loss": 0.1374817371368408, + "step": 4170 + }, + { + "epoch": 0.42379808758578297, + "grad_norm": 3.6904873847961426, + "learning_rate": 1.2915412479042436e-05, + "loss": 0.14422943592071533, + "step": 4180 + }, + { + "epoch": 0.4248119586087155, + "grad_norm": 1.5207897424697876, + "learning_rate": 1.2883988830933148e-05, + "loss": 0.1435297727584839, + "step": 4190 + }, + { + "epoch": 0.425825829631648, + "grad_norm": 2.571296453475952, + "learning_rate": 1.2852534090685184e-05, + "loss": 0.1451401710510254, + "step": 4200 + }, + { + "epoch": 0.4268397006545805, + "grad_norm": 2.346543550491333, + "learning_rate": 1.282104859741052e-05, + "loss": 0.15196511745452881, + "step": 4210 + }, + { + "epoch": 0.427853571677513, + "grad_norm": 1.605188250541687, + "learning_rate": 1.2789532690552673e-05, + "loss": 0.15199148654937744, + "step": 4220 + }, + { + "epoch": 0.42886744270044547, + "grad_norm": 5.959034442901611, + "learning_rate": 1.2757986709883059e-05, + "loss": 0.14361968040466308, + "step": 4230 + }, + { + "epoch": 0.42988131372337796, + "grad_norm": 3.0900332927703857, + "learning_rate": 1.2726410995497304e-05, + "loss": 0.14757080078125, + "step": 4240 + }, + { + "epoch": 0.43089518474631044, + "grad_norm": 2.7943034172058105, + "learning_rate": 1.2694805887811601e-05, + "loss": 0.1240424394607544, + "step": 4250 + }, + { + "epoch": 0.43190905576924293, + "grad_norm": 2.490248203277588, + "learning_rate": 1.2663171727559025e-05, + "loss": 0.1430919051170349, + "step": 4260 + }, + { + "epoch": 0.4329229267921755, + "grad_norm": 1.9061025381088257, + "learning_rate": 1.2631508855785877e-05, + "loss": 0.13898229598999023, + "step": 4270 + }, + { + "epoch": 0.43393679781510797, + "grad_norm": 3.2105226516723633, + "learning_rate": 1.2599817613847973e-05, + "loss": 0.1547446846961975, + "step": 4280 + }, + { + "epoch": 0.43495066883804046, + "grad_norm": 2.7589781284332275, + "learning_rate": 1.2568098343407007e-05, + "loss": 0.14320020675659179, + "step": 4290 + }, + { + "epoch": 0.43596453986097294, + "grad_norm": 4.019106864929199, + "learning_rate": 1.2536351386426844e-05, + "loss": 0.14277291297912598, + "step": 4300 + }, + { + "epoch": 0.43697841088390543, + "grad_norm": 2.293457508087158, + "learning_rate": 1.2504577085169826e-05, + "loss": 0.1494807481765747, + "step": 4310 + }, + { + "epoch": 0.4379922819068379, + "grad_norm": 2.7457549571990967, + "learning_rate": 1.2472775782193099e-05, + "loss": 0.15933771133422853, + "step": 4320 + }, + { + "epoch": 0.4390061529297704, + "grad_norm": 3.7955281734466553, + "learning_rate": 1.2440947820344916e-05, + "loss": 0.1455759048461914, + "step": 4330 + }, + { + "epoch": 0.4400200239527029, + "grad_norm": 2.5522046089172363, + "learning_rate": 1.2409093542760925e-05, + "loss": 0.13894047737121581, + "step": 4340 + }, + { + "epoch": 0.4410338949756354, + "grad_norm": 4.951901912689209, + "learning_rate": 1.2377213292860506e-05, + "loss": 0.16205700635910034, + "step": 4350 + }, + { + "epoch": 0.44204776599856793, + "grad_norm": 4.953357219696045, + "learning_rate": 1.2345307414343021e-05, + "loss": 0.1446847438812256, + "step": 4360 + }, + { + "epoch": 0.4430616370215004, + "grad_norm": 3.079133987426758, + "learning_rate": 1.2313376251184148e-05, + "loss": 0.1417895197868347, + "step": 4370 + }, + { + "epoch": 0.4440755080444329, + "grad_norm": 2.386073589324951, + "learning_rate": 1.2281420147632148e-05, + "loss": 0.13770065307617188, + "step": 4380 + }, + { + "epoch": 0.4450893790673654, + "grad_norm": 4.551647186279297, + "learning_rate": 1.2249439448204172e-05, + "loss": 0.130433988571167, + "step": 4390 + }, + { + "epoch": 0.4461032500902979, + "grad_norm": 2.4367423057556152, + "learning_rate": 1.2217434497682537e-05, + "loss": 0.14291188716888428, + "step": 4400 + }, + { + "epoch": 0.4471171211132304, + "grad_norm": 5.101945877075195, + "learning_rate": 1.2185405641110997e-05, + "loss": 0.14585447311401367, + "step": 4410 + }, + { + "epoch": 0.44813099213616286, + "grad_norm": 4.126468658447266, + "learning_rate": 1.2153353223791049e-05, + "loss": 0.1413429856300354, + "step": 4420 + }, + { + "epoch": 0.44914486315909535, + "grad_norm": 3.858680009841919, + "learning_rate": 1.2121277591278197e-05, + "loss": 0.15877158641815187, + "step": 4430 + }, + { + "epoch": 0.45015873418202784, + "grad_norm": 4.082738399505615, + "learning_rate": 1.2089179089378222e-05, + "loss": 0.1410611867904663, + "step": 4440 + }, + { + "epoch": 0.4511726052049604, + "grad_norm": 4.9232025146484375, + "learning_rate": 1.2057058064143456e-05, + "loss": 0.1375490188598633, + "step": 4450 + }, + { + "epoch": 0.4521864762278929, + "grad_norm": 5.1778998374938965, + "learning_rate": 1.2024914861869064e-05, + "loss": 0.15429674386978148, + "step": 4460 + }, + { + "epoch": 0.45320034725082536, + "grad_norm": 2.189485549926758, + "learning_rate": 1.199274982908929e-05, + "loss": 0.14973764419555663, + "step": 4470 + }, + { + "epoch": 0.45421421827375785, + "grad_norm": 2.9805305004119873, + "learning_rate": 1.196056331257374e-05, + "loss": 0.14422680139541627, + "step": 4480 + }, + { + "epoch": 0.45522808929669034, + "grad_norm": 4.711015701293945, + "learning_rate": 1.1928355659323627e-05, + "loss": 0.13877198696136475, + "step": 4490 + }, + { + "epoch": 0.45624196031962283, + "grad_norm": 2.468822479248047, + "learning_rate": 1.1896127216568047e-05, + "loss": 0.16027402877807617, + "step": 4500 + }, + { + "epoch": 0.4572558313425553, + "grad_norm": 1.7075681686401367, + "learning_rate": 1.1863878331760217e-05, + "loss": 0.14952528476715088, + "step": 4510 + }, + { + "epoch": 0.4582697023654878, + "grad_norm": 1.9382368326187134, + "learning_rate": 1.183160935257375e-05, + "loss": 0.1465291738510132, + "step": 4520 + }, + { + "epoch": 0.45928357338842035, + "grad_norm": 2.11478853225708, + "learning_rate": 1.1799320626898883e-05, + "loss": 0.12838668823242189, + "step": 4530 + }, + { + "epoch": 0.46029744441135284, + "grad_norm": 2.4004440307617188, + "learning_rate": 1.1767012502838746e-05, + "loss": 0.14215359687805176, + "step": 4540 + }, + { + "epoch": 0.46131131543428533, + "grad_norm": 5.318121433258057, + "learning_rate": 1.1734685328705604e-05, + "loss": 0.14960155487060547, + "step": 4550 + }, + { + "epoch": 0.4623251864572178, + "grad_norm": 1.8953633308410645, + "learning_rate": 1.17023394530171e-05, + "loss": 0.16846923828125, + "step": 4560 + }, + { + "epoch": 0.4633390574801503, + "grad_norm": 2.9742345809936523, + "learning_rate": 1.1669975224492486e-05, + "loss": 0.14449639320373536, + "step": 4570 + }, + { + "epoch": 0.4643529285030828, + "grad_norm": 2.8249671459198, + "learning_rate": 1.1637592992048888e-05, + "loss": 0.13185099363327027, + "step": 4580 + }, + { + "epoch": 0.4653667995260153, + "grad_norm": 4.623602867126465, + "learning_rate": 1.160519310479753e-05, + "loss": 0.1598207712173462, + "step": 4590 + }, + { + "epoch": 0.4663806705489478, + "grad_norm": 1.5297914743423462, + "learning_rate": 1.157277591203996e-05, + "loss": 0.13980872631073, + "step": 4600 + }, + { + "epoch": 0.46739454157188026, + "grad_norm": 4.045602798461914, + "learning_rate": 1.1540341763264314e-05, + "loss": 0.13908884525299073, + "step": 4610 + }, + { + "epoch": 0.4684084125948128, + "grad_norm": 2.7758519649505615, + "learning_rate": 1.1507891008141512e-05, + "loss": 0.13622967004776002, + "step": 4620 + }, + { + "epoch": 0.4694222836177453, + "grad_norm": 1.505149006843567, + "learning_rate": 1.147542399652152e-05, + "loss": 0.14128323793411254, + "step": 4630 + }, + { + "epoch": 0.4704361546406778, + "grad_norm": 1.9750386476516724, + "learning_rate": 1.1442941078429554e-05, + "loss": 0.12925097942352295, + "step": 4640 + }, + { + "epoch": 0.4714500256636103, + "grad_norm": 3.0825517177581787, + "learning_rate": 1.141044260406233e-05, + "loss": 0.14765868186950684, + "step": 4650 + }, + { + "epoch": 0.47246389668654276, + "grad_norm": 4.581333160400391, + "learning_rate": 1.1377928923784261e-05, + "loss": 0.1401889681816101, + "step": 4660 + }, + { + "epoch": 0.47347776770947525, + "grad_norm": 2.9602789878845215, + "learning_rate": 1.1345400388123698e-05, + "loss": 0.13256590366363524, + "step": 4670 + }, + { + "epoch": 0.47449163873240774, + "grad_norm": 2.249246597290039, + "learning_rate": 1.1312857347769153e-05, + "loss": 0.1375988483428955, + "step": 4680 + }, + { + "epoch": 0.4755055097553402, + "grad_norm": 2.257286310195923, + "learning_rate": 1.128030015356551e-05, + "loss": 0.14341552257537843, + "step": 4690 + }, + { + "epoch": 0.4765193807782727, + "grad_norm": 1.8892070055007935, + "learning_rate": 1.1247729156510245e-05, + "loss": 0.14488463401794432, + "step": 4700 + }, + { + "epoch": 0.47753325180120526, + "grad_norm": 2.5849761962890625, + "learning_rate": 1.1215144707749635e-05, + "loss": 0.13902533054351807, + "step": 4710 + }, + { + "epoch": 0.47854712282413775, + "grad_norm": 2.4534788131713867, + "learning_rate": 1.1182547158574988e-05, + "loss": 0.12838621139526368, + "step": 4720 + }, + { + "epoch": 0.47956099384707024, + "grad_norm": 2.331000328063965, + "learning_rate": 1.1149936860418846e-05, + "loss": 0.14466395378112792, + "step": 4730 + }, + { + "epoch": 0.4805748648700027, + "grad_norm": 1.8673733472824097, + "learning_rate": 1.1117314164851191e-05, + "loss": 0.13672213554382323, + "step": 4740 + }, + { + "epoch": 0.4815887358929352, + "grad_norm": 3.9680521488189697, + "learning_rate": 1.1084679423575666e-05, + "loss": 0.1361147403717041, + "step": 4750 + }, + { + "epoch": 0.4826026069158677, + "grad_norm": 2.8554317951202393, + "learning_rate": 1.1052032988425778e-05, + "loss": 0.12761780023574829, + "step": 4760 + }, + { + "epoch": 0.4836164779388002, + "grad_norm": 5.791364669799805, + "learning_rate": 1.1019375211361104e-05, + "loss": 0.12505840063095092, + "step": 4770 + }, + { + "epoch": 0.4846303489617327, + "grad_norm": 1.7616297006607056, + "learning_rate": 1.0986706444463496e-05, + "loss": 0.15178056955337524, + "step": 4780 + }, + { + "epoch": 0.4856442199846652, + "grad_norm": 3.7878222465515137, + "learning_rate": 1.0954027039933288e-05, + "loss": 0.1333021879196167, + "step": 4790 + }, + { + "epoch": 0.4866580910075977, + "grad_norm": 5.28480863571167, + "learning_rate": 1.09213373500855e-05, + "loss": 0.14376311302185057, + "step": 4800 + }, + { + "epoch": 0.4876719620305302, + "grad_norm": 1.7212532758712769, + "learning_rate": 1.0888637727346032e-05, + "loss": 0.15401549339294435, + "step": 4810 + }, + { + "epoch": 0.4886858330534627, + "grad_norm": 4.17246150970459, + "learning_rate": 1.0855928524247884e-05, + "loss": 0.15330212116241454, + "step": 4820 + }, + { + "epoch": 0.4896997040763952, + "grad_norm": 2.411529541015625, + "learning_rate": 1.0823210093427318e-05, + "loss": 0.12667444944381714, + "step": 4830 + }, + { + "epoch": 0.49071357509932767, + "grad_norm": 2.0121703147888184, + "learning_rate": 1.07904827876201e-05, + "loss": 0.13882408142089844, + "step": 4840 + }, + { + "epoch": 0.49172744612226016, + "grad_norm": 3.7966158390045166, + "learning_rate": 1.0757746959657666e-05, + "loss": 0.14201059341430664, + "step": 4850 + }, + { + "epoch": 0.49274131714519265, + "grad_norm": 6.346258163452148, + "learning_rate": 1.072500296246334e-05, + "loss": 0.13271256685256957, + "step": 4860 + }, + { + "epoch": 0.49375518816812514, + "grad_norm": 4.615936279296875, + "learning_rate": 1.06922511490485e-05, + "loss": 0.131577467918396, + "step": 4870 + }, + { + "epoch": 0.4947690591910577, + "grad_norm": 2.2846786975860596, + "learning_rate": 1.0659491872508807e-05, + "loss": 0.1408820629119873, + "step": 4880 + }, + { + "epoch": 0.49578293021399017, + "grad_norm": 6.527868270874023, + "learning_rate": 1.0626725486020374e-05, + "loss": 0.13936009407043456, + "step": 4890 + }, + { + "epoch": 0.49679680123692266, + "grad_norm": 3.9807016849517822, + "learning_rate": 1.059395234283597e-05, + "loss": 0.13570343255996703, + "step": 4900 + }, + { + "epoch": 0.49781067225985515, + "grad_norm": 2.011223554611206, + "learning_rate": 1.0561172796281196e-05, + "loss": 0.1482239246368408, + "step": 4910 + }, + { + "epoch": 0.49882454328278764, + "grad_norm": 2.0938832759857178, + "learning_rate": 1.0528387199750706e-05, + "loss": 0.14073108434677123, + "step": 4920 + }, + { + "epoch": 0.4998384143057201, + "grad_norm": 2.6859755516052246, + "learning_rate": 1.0495595906704365e-05, + "loss": 0.13691926002502441, + "step": 4930 + }, + { + "epoch": 0.5008522853286527, + "grad_norm": 3.243030309677124, + "learning_rate": 1.0462799270663457e-05, + "loss": 0.12626502513885499, + "step": 4940 + }, + { + "epoch": 0.5018661563515852, + "grad_norm": 2.3998830318450928, + "learning_rate": 1.0429997645206865e-05, + "loss": 0.12610852718353271, + "step": 4950 + }, + { + "epoch": 0.5028800273745176, + "grad_norm": 2.6368649005889893, + "learning_rate": 1.0397191383967267e-05, + "loss": 0.12857861518859864, + "step": 4960 + }, + { + "epoch": 0.5038938983974501, + "grad_norm": 2.581754684448242, + "learning_rate": 1.0364380840627315e-05, + "loss": 0.13839619159698485, + "step": 4970 + }, + { + "epoch": 0.5049077694203826, + "grad_norm": 3.7578508853912354, + "learning_rate": 1.033156636891583e-05, + "loss": 0.14772312641143798, + "step": 4980 + }, + { + "epoch": 0.5059216404433151, + "grad_norm": 1.727844476699829, + "learning_rate": 1.0298748322603982e-05, + "loss": 0.13031535148620604, + "step": 4990 + }, + { + "epoch": 0.5069355114662476, + "grad_norm": 3.5588784217834473, + "learning_rate": 1.0265927055501476e-05, + "loss": 0.14915614128112792, + "step": 5000 + }, + { + "epoch": 0.5079493824891801, + "grad_norm": 1.7651562690734863, + "learning_rate": 1.023310292145275e-05, + "loss": 0.12886136770248413, + "step": 5010 + }, + { + "epoch": 0.5089632535121126, + "grad_norm": 6.098463535308838, + "learning_rate": 1.020027627433314e-05, + "loss": 0.1340983748435974, + "step": 5020 + }, + { + "epoch": 0.5099771245350451, + "grad_norm": 3.1103854179382324, + "learning_rate": 1.0167447468045083e-05, + "loss": 0.12727158069610595, + "step": 5030 + }, + { + "epoch": 0.5109909955579776, + "grad_norm": 3.9124486446380615, + "learning_rate": 1.0134616856514282e-05, + "loss": 0.13732383251190186, + "step": 5040 + }, + { + "epoch": 0.51200486658091, + "grad_norm": 3.0352559089660645, + "learning_rate": 1.0101784793685914e-05, + "loss": 0.1397580623626709, + "step": 5050 + }, + { + "epoch": 0.5130187376038425, + "grad_norm": 4.240923881530762, + "learning_rate": 1.0068951633520802e-05, + "loss": 0.12731703519821166, + "step": 5060 + }, + { + "epoch": 0.514032608626775, + "grad_norm": 2.9493002891540527, + "learning_rate": 1.0036117729991595e-05, + "loss": 0.13342556953430176, + "step": 5070 + }, + { + "epoch": 0.5150464796497075, + "grad_norm": 4.238946914672852, + "learning_rate": 1.0003283437078951e-05, + "loss": 0.1369422197341919, + "step": 5080 + }, + { + "epoch": 0.5160603506726401, + "grad_norm": 2.9656577110290527, + "learning_rate": 9.970449108767742e-06, + "loss": 0.13607466220855713, + "step": 5090 + }, + { + "epoch": 0.5170742216955726, + "grad_norm": 2.6521146297454834, + "learning_rate": 9.937615099043206e-06, + "loss": 0.1272137999534607, + "step": 5100 + }, + { + "epoch": 0.5180880927185051, + "grad_norm": 2.9922194480895996, + "learning_rate": 9.904781761887147e-06, + "loss": 0.14072468280792236, + "step": 5110 + }, + { + "epoch": 0.5191019637414376, + "grad_norm": 1.7018963098526, + "learning_rate": 9.871949451274137e-06, + "loss": 0.13426640033721923, + "step": 5120 + }, + { + "epoch": 0.5201158347643701, + "grad_norm": 2.133653163909912, + "learning_rate": 9.839118521167656e-06, + "loss": 0.13162192106246948, + "step": 5130 + }, + { + "epoch": 0.5211297057873026, + "grad_norm": 3.561145544052124, + "learning_rate": 9.80628932551631e-06, + "loss": 0.14092371463775635, + "step": 5140 + }, + { + "epoch": 0.522143576810235, + "grad_norm": 3.841282367706299, + "learning_rate": 9.773462218250014e-06, + "loss": 0.12865025997161866, + "step": 5150 + }, + { + "epoch": 0.5231574478331675, + "grad_norm": 3.77545166015625, + "learning_rate": 9.740637553276162e-06, + "loss": 0.14333443641662597, + "step": 5160 + }, + { + "epoch": 0.5241713188561, + "grad_norm": 2.878243923187256, + "learning_rate": 9.707815684475817e-06, + "loss": 0.12886457443237304, + "step": 5170 + }, + { + "epoch": 0.5251851898790325, + "grad_norm": 3.1325747966766357, + "learning_rate": 9.674996965699895e-06, + "loss": 0.13027002811431884, + "step": 5180 + }, + { + "epoch": 0.526199060901965, + "grad_norm": 3.253485918045044, + "learning_rate": 9.642181750765354e-06, + "loss": 0.13881994485855104, + "step": 5190 + }, + { + "epoch": 0.5272129319248975, + "grad_norm": 2.436286449432373, + "learning_rate": 9.60937039345138e-06, + "loss": 0.1452277421951294, + "step": 5200 + }, + { + "epoch": 0.52822680294783, + "grad_norm": 1.9443204402923584, + "learning_rate": 9.576563247495566e-06, + "loss": 0.12879693508148193, + "step": 5210 + }, + { + "epoch": 0.5292406739707625, + "grad_norm": 2.795990467071533, + "learning_rate": 9.543760666590105e-06, + "loss": 0.13727729320526122, + "step": 5220 + }, + { + "epoch": 0.530254544993695, + "grad_norm": 2.7152020931243896, + "learning_rate": 9.510963004377974e-06, + "loss": 0.12838644981384278, + "step": 5230 + }, + { + "epoch": 0.5312684160166274, + "grad_norm": 2.4083058834075928, + "learning_rate": 9.478170614449114e-06, + "loss": 0.14684834480285644, + "step": 5240 + }, + { + "epoch": 0.5322822870395599, + "grad_norm": 2.4841010570526123, + "learning_rate": 9.445383850336648e-06, + "loss": 0.13429325819015503, + "step": 5250 + }, + { + "epoch": 0.5332961580624924, + "grad_norm": 3.7223289012908936, + "learning_rate": 9.412603065513021e-06, + "loss": 0.13675734996795655, + "step": 5260 + }, + { + "epoch": 0.534310029085425, + "grad_norm": 2.5427308082580566, + "learning_rate": 9.379828613386225e-06, + "loss": 0.14670679569244385, + "step": 5270 + }, + { + "epoch": 0.5353239001083575, + "grad_norm": 2.207491874694824, + "learning_rate": 9.347060847295991e-06, + "loss": 0.13578200340270996, + "step": 5280 + }, + { + "epoch": 0.53633777113129, + "grad_norm": 2.456340789794922, + "learning_rate": 9.314300120509952e-06, + "loss": 0.13536118268966674, + "step": 5290 + }, + { + "epoch": 0.5373516421542225, + "grad_norm": 2.156038522720337, + "learning_rate": 9.281546786219861e-06, + "loss": 0.13045032024383546, + "step": 5300 + }, + { + "epoch": 0.538365513177155, + "grad_norm": 2.632943630218506, + "learning_rate": 9.24880119753777e-06, + "loss": 0.1314428210258484, + "step": 5310 + }, + { + "epoch": 0.5393793842000875, + "grad_norm": 5.040396213531494, + "learning_rate": 9.216063707492218e-06, + "loss": 0.13456228971481324, + "step": 5320 + }, + { + "epoch": 0.54039325522302, + "grad_norm": 3.1989004611968994, + "learning_rate": 9.183334669024449e-06, + "loss": 0.1422601103782654, + "step": 5330 + }, + { + "epoch": 0.5414071262459524, + "grad_norm": 2.6820249557495117, + "learning_rate": 9.150614434984578e-06, + "loss": 0.13722591400146483, + "step": 5340 + }, + { + "epoch": 0.5424209972688849, + "grad_norm": 3.8938260078430176, + "learning_rate": 9.117903358127801e-06, + "loss": 0.12815515995025634, + "step": 5350 + }, + { + "epoch": 0.5434348682918174, + "grad_norm": 2.4761505126953125, + "learning_rate": 9.085201791110602e-06, + "loss": 0.12028461694717407, + "step": 5360 + }, + { + "epoch": 0.5444487393147499, + "grad_norm": 3.561060667037964, + "learning_rate": 9.052510086486922e-06, + "loss": 0.12784595489501954, + "step": 5370 + }, + { + "epoch": 0.5454626103376824, + "grad_norm": 1.6509217023849487, + "learning_rate": 9.019828596704394e-06, + "loss": 0.1283278226852417, + "step": 5380 + }, + { + "epoch": 0.5464764813606149, + "grad_norm": 3.509165048599243, + "learning_rate": 8.987157674100519e-06, + "loss": 0.13272740840911865, + "step": 5390 + }, + { + "epoch": 0.5474903523835474, + "grad_norm": 2.9297759532928467, + "learning_rate": 8.954497670898858e-06, + "loss": 0.12141624689102173, + "step": 5400 + }, + { + "epoch": 0.5485042234064799, + "grad_norm": 3.651435375213623, + "learning_rate": 8.92184893920528e-06, + "loss": 0.12331033945083618, + "step": 5410 + }, + { + "epoch": 0.5495180944294124, + "grad_norm": 2.782097578048706, + "learning_rate": 8.889211831004112e-06, + "loss": 0.13164777755737306, + "step": 5420 + }, + { + "epoch": 0.5505319654523448, + "grad_norm": 4.683573246002197, + "learning_rate": 8.856586698154382e-06, + "loss": 0.13341509103775023, + "step": 5430 + }, + { + "epoch": 0.5515458364752774, + "grad_norm": 3.0757555961608887, + "learning_rate": 8.823973892386007e-06, + "loss": 0.1441299796104431, + "step": 5440 + }, + { + "epoch": 0.5525597074982099, + "grad_norm": 3.8742945194244385, + "learning_rate": 8.791373765296006e-06, + "loss": 0.12618000507354737, + "step": 5450 + }, + { + "epoch": 0.5535735785211424, + "grad_norm": 3.0782253742218018, + "learning_rate": 8.758786668344711e-06, + "loss": 0.13583043813705445, + "step": 5460 + }, + { + "epoch": 0.5545874495440749, + "grad_norm": 2.461061954498291, + "learning_rate": 8.726212952851978e-06, + "loss": 0.1408136487007141, + "step": 5470 + }, + { + "epoch": 0.5556013205670074, + "grad_norm": 2.9635531902313232, + "learning_rate": 8.693652969993396e-06, + "loss": 0.1266424059867859, + "step": 5480 + }, + { + "epoch": 0.5566151915899399, + "grad_norm": 2.266136646270752, + "learning_rate": 8.661107070796504e-06, + "loss": 0.1361314058303833, + "step": 5490 + }, + { + "epoch": 0.5576290626128724, + "grad_norm": 2.9469313621520996, + "learning_rate": 8.628575606137005e-06, + "loss": 0.1346881866455078, + "step": 5500 + }, + { + "epoch": 0.5586429336358049, + "grad_norm": 4.113528251647949, + "learning_rate": 8.59605892673499e-06, + "loss": 0.13883583545684813, + "step": 5510 + }, + { + "epoch": 0.5596568046587374, + "grad_norm": 4.621987819671631, + "learning_rate": 8.563557383151143e-06, + "loss": 0.13232572078704835, + "step": 5520 + }, + { + "epoch": 0.5606706756816698, + "grad_norm": 2.6912143230438232, + "learning_rate": 8.531071325782965e-06, + "loss": 0.13418498039245605, + "step": 5530 + }, + { + "epoch": 0.5616845467046023, + "grad_norm": 2.9830398559570312, + "learning_rate": 8.49860110486102e-06, + "loss": 0.11645398139953614, + "step": 5540 + }, + { + "epoch": 0.5626984177275348, + "grad_norm": 2.9361493587493896, + "learning_rate": 8.466147070445123e-06, + "loss": 0.12407329082489013, + "step": 5550 + }, + { + "epoch": 0.5637122887504673, + "grad_norm": 4.406589508056641, + "learning_rate": 8.433709572420586e-06, + "loss": 0.12237588167190552, + "step": 5560 + }, + { + "epoch": 0.5647261597733998, + "grad_norm": 2.8003907203674316, + "learning_rate": 8.40128896049445e-06, + "loss": 0.12786319255828857, + "step": 5570 + }, + { + "epoch": 0.5657400307963323, + "grad_norm": 2.4227280616760254, + "learning_rate": 8.368885584191698e-06, + "loss": 0.11714946031570435, + "step": 5580 + }, + { + "epoch": 0.5667539018192648, + "grad_norm": 3.120598793029785, + "learning_rate": 8.336499792851505e-06, + "loss": 0.11894145011901855, + "step": 5590 + }, + { + "epoch": 0.5677677728421973, + "grad_norm": 2.3636646270751953, + "learning_rate": 8.304131935623461e-06, + "loss": 0.13403282165527344, + "step": 5600 + }, + { + "epoch": 0.5687816438651299, + "grad_norm": 2.7188661098480225, + "learning_rate": 8.271782361463806e-06, + "loss": 0.1258084774017334, + "step": 5610 + }, + { + "epoch": 0.5697955148880623, + "grad_norm": 2.785005807876587, + "learning_rate": 8.239451419131676e-06, + "loss": 0.12704088687896728, + "step": 5620 + }, + { + "epoch": 0.5708093859109948, + "grad_norm": 3.798413038253784, + "learning_rate": 8.207139457185328e-06, + "loss": 0.12981817722320557, + "step": 5630 + }, + { + "epoch": 0.5718232569339273, + "grad_norm": 5.369892120361328, + "learning_rate": 8.174846823978412e-06, + "loss": 0.12903120517730712, + "step": 5640 + }, + { + "epoch": 0.5728371279568598, + "grad_norm": 3.1835269927978516, + "learning_rate": 8.142573867656182e-06, + "loss": 0.11969665288925171, + "step": 5650 + }, + { + "epoch": 0.5738509989797923, + "grad_norm": 3.732945442199707, + "learning_rate": 8.110320936151746e-06, + "loss": 0.11435606479644775, + "step": 5660 + }, + { + "epoch": 0.5748648700027248, + "grad_norm": 3.486506223678589, + "learning_rate": 8.078088377182352e-06, + "loss": 0.10843586921691895, + "step": 5670 + }, + { + "epoch": 0.5758787410256573, + "grad_norm": 2.5324549674987793, + "learning_rate": 8.045876538245595e-06, + "loss": 0.1160558819770813, + "step": 5680 + }, + { + "epoch": 0.5768926120485898, + "grad_norm": 2.6564769744873047, + "learning_rate": 8.013685766615687e-06, + "loss": 0.1261429786682129, + "step": 5690 + }, + { + "epoch": 0.5779064830715223, + "grad_norm": 3.403454065322876, + "learning_rate": 7.981516409339726e-06, + "loss": 0.1376142382621765, + "step": 5700 + }, + { + "epoch": 0.5789203540944547, + "grad_norm": 2.8978850841522217, + "learning_rate": 7.94936881323393e-06, + "loss": 0.11550960540771485, + "step": 5710 + }, + { + "epoch": 0.5799342251173872, + "grad_norm": 3.2867023944854736, + "learning_rate": 7.917243324879924e-06, + "loss": 0.1350696086883545, + "step": 5720 + }, + { + "epoch": 0.5809480961403197, + "grad_norm": 2.915381669998169, + "learning_rate": 7.88514029062098e-06, + "loss": 0.12485302686691284, + "step": 5730 + }, + { + "epoch": 0.5819619671632522, + "grad_norm": 2.577099561691284, + "learning_rate": 7.853060056558294e-06, + "loss": 0.12092335224151611, + "step": 5740 + }, + { + "epoch": 0.5829758381861847, + "grad_norm": 3.481149435043335, + "learning_rate": 7.821002968547266e-06, + "loss": 0.13634812831878662, + "step": 5750 + }, + { + "epoch": 0.5839897092091172, + "grad_norm": 2.2351715564727783, + "learning_rate": 7.788969372193744e-06, + "loss": 0.11540305614471436, + "step": 5760 + }, + { + "epoch": 0.5850035802320497, + "grad_norm": 3.000387668609619, + "learning_rate": 7.75695961285032e-06, + "loss": 0.11687803268432617, + "step": 5770 + }, + { + "epoch": 0.5860174512549823, + "grad_norm": 2.4926598072052, + "learning_rate": 7.724974035612611e-06, + "loss": 0.12135491371154786, + "step": 5780 + }, + { + "epoch": 0.5870313222779148, + "grad_norm": 3.1591737270355225, + "learning_rate": 7.693012985315503e-06, + "loss": 0.1372124433517456, + "step": 5790 + }, + { + "epoch": 0.5880451933008473, + "grad_norm": 4.414812088012695, + "learning_rate": 7.661076806529486e-06, + "loss": 0.11833515167236328, + "step": 5800 + }, + { + "epoch": 0.5890590643237797, + "grad_norm": 4.384387969970703, + "learning_rate": 7.629165843556896e-06, + "loss": 0.12182070016860962, + "step": 5810 + }, + { + "epoch": 0.5900729353467122, + "grad_norm": 3.093618154525757, + "learning_rate": 7.597280440428216e-06, + "loss": 0.12524832487106324, + "step": 5820 + }, + { + "epoch": 0.5910868063696447, + "grad_norm": 2.730748176574707, + "learning_rate": 7.565420940898381e-06, + "loss": 0.11574351787567139, + "step": 5830 + }, + { + "epoch": 0.5921006773925772, + "grad_norm": 4.333120822906494, + "learning_rate": 7.5335876884430495e-06, + "loss": 0.1335510492324829, + "step": 5840 + }, + { + "epoch": 0.5931145484155097, + "grad_norm": 3.205434799194336, + "learning_rate": 7.501781026254922e-06, + "loss": 0.11877676248550414, + "step": 5850 + }, + { + "epoch": 0.5941284194384422, + "grad_norm": 3.765887975692749, + "learning_rate": 7.4700012972400224e-06, + "loss": 0.1355007529258728, + "step": 5860 + }, + { + "epoch": 0.5951422904613747, + "grad_norm": 2.515082359313965, + "learning_rate": 7.438248844014011e-06, + "loss": 0.12562137842178345, + "step": 5870 + }, + { + "epoch": 0.5961561614843072, + "grad_norm": 3.134382963180542, + "learning_rate": 7.4065240088984944e-06, + "loss": 0.11263692378997803, + "step": 5880 + }, + { + "epoch": 0.5971700325072397, + "grad_norm": 3.355600118637085, + "learning_rate": 7.3748271339173224e-06, + "loss": 0.12848684787750245, + "step": 5890 + }, + { + "epoch": 0.5981839035301721, + "grad_norm": 3.3887929916381836, + "learning_rate": 7.34315856079291e-06, + "loss": 0.14063920974731445, + "step": 5900 + }, + { + "epoch": 0.5991977745531046, + "grad_norm": 3.806663990020752, + "learning_rate": 7.311518630942562e-06, + "loss": 0.13027294874191284, + "step": 5910 + }, + { + "epoch": 0.6002116455760371, + "grad_norm": 3.0723605155944824, + "learning_rate": 7.279907685474759e-06, + "loss": 0.1212736964225769, + "step": 5920 + }, + { + "epoch": 0.6012255165989696, + "grad_norm": 3.5551059246063232, + "learning_rate": 7.248326065185527e-06, + "loss": 0.1317020058631897, + "step": 5930 + }, + { + "epoch": 0.6022393876219021, + "grad_norm": 2.420109748840332, + "learning_rate": 7.216774110554724e-06, + "loss": 0.1191293478012085, + "step": 5940 + }, + { + "epoch": 0.6032532586448346, + "grad_norm": 3.485996961593628, + "learning_rate": 7.185252161742385e-06, + "loss": 0.12419047355651855, + "step": 5950 + }, + { + "epoch": 0.6042671296677672, + "grad_norm": 2.946549415588379, + "learning_rate": 7.153760558585059e-06, + "loss": 0.09880372285842895, + "step": 5960 + }, + { + "epoch": 0.6052810006906997, + "grad_norm": 3.77720308303833, + "learning_rate": 7.122299640592135e-06, + "loss": 0.11933207511901855, + "step": 5970 + }, + { + "epoch": 0.6062948717136322, + "grad_norm": 2.7840023040771484, + "learning_rate": 7.090869746942191e-06, + "loss": 0.11739814281463623, + "step": 5980 + }, + { + "epoch": 0.6073087427365647, + "grad_norm": 3.340381622314453, + "learning_rate": 7.05947121647933e-06, + "loss": 0.1126866340637207, + "step": 5990 + }, + { + "epoch": 0.6083226137594971, + "grad_norm": 3.765659809112549, + "learning_rate": 7.028104387709531e-06, + "loss": 0.1145860195159912, + "step": 6000 + }, + { + "epoch": 0.6093364847824296, + "grad_norm": 3.4316091537475586, + "learning_rate": 6.996769598797001e-06, + "loss": 0.1165287971496582, + "step": 6010 + }, + { + "epoch": 0.6103503558053621, + "grad_norm": 3.6676077842712402, + "learning_rate": 6.965467187560524e-06, + "loss": 0.10932881832122802, + "step": 6020 + }, + { + "epoch": 0.6113642268282946, + "grad_norm": 3.8429625034332275, + "learning_rate": 6.934197491469818e-06, + "loss": 0.10596539974212646, + "step": 6030 + }, + { + "epoch": 0.6123780978512271, + "grad_norm": 4.4424214363098145, + "learning_rate": 6.902960847641915e-06, + "loss": 0.11706302165985108, + "step": 6040 + }, + { + "epoch": 0.6133919688741596, + "grad_norm": 3.4496212005615234, + "learning_rate": 6.8717575928374915e-06, + "loss": 0.12355380058288574, + "step": 6050 + }, + { + "epoch": 0.6144058398970921, + "grad_norm": 2.4938673973083496, + "learning_rate": 6.840588063457278e-06, + "loss": 0.11547729969024659, + "step": 6060 + }, + { + "epoch": 0.6154197109200246, + "grad_norm": 2.814981698989868, + "learning_rate": 6.809452595538403e-06, + "loss": 0.12263381481170654, + "step": 6070 + }, + { + "epoch": 0.616433581942957, + "grad_norm": 2.491060972213745, + "learning_rate": 6.778351524750777e-06, + "loss": 0.10702656507492066, + "step": 6080 + }, + { + "epoch": 0.6174474529658895, + "grad_norm": 3.4924604892730713, + "learning_rate": 6.747285186393488e-06, + "loss": 0.11343364715576172, + "step": 6090 + }, + { + "epoch": 0.618461323988822, + "grad_norm": 6.075539588928223, + "learning_rate": 6.716253915391167e-06, + "loss": 0.1371016263961792, + "step": 6100 + }, + { + "epoch": 0.6194751950117545, + "grad_norm": 2.8840835094451904, + "learning_rate": 6.685258046290381e-06, + "loss": 0.12178136110305786, + "step": 6110 + }, + { + "epoch": 0.620489066034687, + "grad_norm": 3.9059081077575684, + "learning_rate": 6.654297913256047e-06, + "loss": 0.13028035163879395, + "step": 6120 + }, + { + "epoch": 0.6215029370576196, + "grad_norm": 3.4152772426605225, + "learning_rate": 6.623373850067794e-06, + "loss": 0.13074930906295776, + "step": 6130 + }, + { + "epoch": 0.6225168080805521, + "grad_norm": 4.023529529571533, + "learning_rate": 6.592486190116405e-06, + "loss": 0.12868292331695558, + "step": 6140 + }, + { + "epoch": 0.6235306791034846, + "grad_norm": 5.504587650299072, + "learning_rate": 6.561635266400181e-06, + "loss": 0.10948985815048218, + "step": 6150 + }, + { + "epoch": 0.6245445501264171, + "grad_norm": 4.5505900382995605, + "learning_rate": 6.5308214115213785e-06, + "loss": 0.12432782649993897, + "step": 6160 + }, + { + "epoch": 0.6255584211493496, + "grad_norm": 2.610111951828003, + "learning_rate": 6.500044957682627e-06, + "loss": 0.12032835483551026, + "step": 6170 + }, + { + "epoch": 0.626572292172282, + "grad_norm": 3.2917075157165527, + "learning_rate": 6.469306236683324e-06, + "loss": 0.10741511583328248, + "step": 6180 + }, + { + "epoch": 0.6275861631952145, + "grad_norm": 3.0540552139282227, + "learning_rate": 6.438605579916078e-06, + "loss": 0.13316771984100342, + "step": 6190 + }, + { + "epoch": 0.628600034218147, + "grad_norm": 4.435017108917236, + "learning_rate": 6.4079433183631255e-06, + "loss": 0.11245529651641846, + "step": 6200 + }, + { + "epoch": 0.6296139052410795, + "grad_norm": 2.968557119369507, + "learning_rate": 6.377319782592767e-06, + "loss": 0.11640489101409912, + "step": 6210 + }, + { + "epoch": 0.630627776264012, + "grad_norm": 3.9529054164886475, + "learning_rate": 6.346735302755805e-06, + "loss": 0.1235194444656372, + "step": 6220 + }, + { + "epoch": 0.6316416472869445, + "grad_norm": 3.5193979740142822, + "learning_rate": 6.316190208581979e-06, + "loss": 0.11049515008926392, + "step": 6230 + }, + { + "epoch": 0.632655518309877, + "grad_norm": 2.510277509689331, + "learning_rate": 6.285684829376413e-06, + "loss": 0.11749703884124756, + "step": 6240 + }, + { + "epoch": 0.6336693893328095, + "grad_norm": 3.0587687492370605, + "learning_rate": 6.25521949401607e-06, + "loss": 0.10767536163330078, + "step": 6250 + }, + { + "epoch": 0.634683260355742, + "grad_norm": 4.516772747039795, + "learning_rate": 6.2247945309461955e-06, + "loss": 0.1101502776145935, + "step": 6260 + }, + { + "epoch": 0.6356971313786745, + "grad_norm": 3.4169180393218994, + "learning_rate": 6.194410268176796e-06, + "loss": 0.11560378074645997, + "step": 6270 + }, + { + "epoch": 0.6367110024016069, + "grad_norm": 3.7133498191833496, + "learning_rate": 6.164067033279073e-06, + "loss": 0.11294629573822021, + "step": 6280 + }, + { + "epoch": 0.6377248734245394, + "grad_norm": 2.992820978164673, + "learning_rate": 6.133765153381918e-06, + "loss": 0.12664527893066407, + "step": 6290 + }, + { + "epoch": 0.638738744447472, + "grad_norm": 3.8269197940826416, + "learning_rate": 6.103504955168382e-06, + "loss": 0.11149437427520752, + "step": 6300 + }, + { + "epoch": 0.6397526154704045, + "grad_norm": 3.5067291259765625, + "learning_rate": 6.07328676487214e-06, + "loss": 0.12186002731323242, + "step": 6310 + }, + { + "epoch": 0.640766486493337, + "grad_norm": 3.001103401184082, + "learning_rate": 6.043110908273973e-06, + "loss": 0.11294422149658204, + "step": 6320 + }, + { + "epoch": 0.6417803575162695, + "grad_norm": 4.304075241088867, + "learning_rate": 6.012977710698288e-06, + "loss": 0.10929944515228271, + "step": 6330 + }, + { + "epoch": 0.642794228539202, + "grad_norm": 4.965757846832275, + "learning_rate": 5.982887497009566e-06, + "loss": 0.12052617073059083, + "step": 6340 + }, + { + "epoch": 0.6438080995621345, + "grad_norm": 3.560875415802002, + "learning_rate": 5.9528405916088926e-06, + "loss": 0.10822978019714355, + "step": 6350 + }, + { + "epoch": 0.644821970585067, + "grad_norm": 4.201086521148682, + "learning_rate": 5.9228373184304455e-06, + "loss": 0.11204761266708374, + "step": 6360 + }, + { + "epoch": 0.6458358416079995, + "grad_norm": 5.024351119995117, + "learning_rate": 5.892878000938001e-06, + "loss": 0.11234544515609741, + "step": 6370 + }, + { + "epoch": 0.6468497126309319, + "grad_norm": 3.0084140300750732, + "learning_rate": 5.86296296212146e-06, + "loss": 0.11078963279724122, + "step": 6380 + }, + { + "epoch": 0.6478635836538644, + "grad_norm": 3.0973682403564453, + "learning_rate": 5.833092524493349e-06, + "loss": 0.10360321998596192, + "step": 6390 + }, + { + "epoch": 0.6488774546767969, + "grad_norm": 4.345056533813477, + "learning_rate": 5.80326701008536e-06, + "loss": 0.11852946281433105, + "step": 6400 + }, + { + "epoch": 0.6498913256997294, + "grad_norm": 2.602504253387451, + "learning_rate": 5.773486740444862e-06, + "loss": 0.09967931509017944, + "step": 6410 + }, + { + "epoch": 0.6509051967226619, + "grad_norm": 3.1056089401245117, + "learning_rate": 5.743752036631443e-06, + "loss": 0.11507558822631836, + "step": 6420 + }, + { + "epoch": 0.6519190677455944, + "grad_norm": 3.6924421787261963, + "learning_rate": 5.714063219213456e-06, + "loss": 0.09169327020645142, + "step": 6430 + }, + { + "epoch": 0.6529329387685269, + "grad_norm": 4.179924964904785, + "learning_rate": 5.684420608264549e-06, + "loss": 0.11617479324340821, + "step": 6440 + }, + { + "epoch": 0.6539468097914594, + "grad_norm": 3.3177223205566406, + "learning_rate": 5.654824523360219e-06, + "loss": 0.10017070770263672, + "step": 6450 + }, + { + "epoch": 0.6549606808143918, + "grad_norm": 4.340502738952637, + "learning_rate": 5.625275283574372e-06, + "loss": 0.10695762634277343, + "step": 6460 + }, + { + "epoch": 0.6559745518373244, + "grad_norm": 2.8455255031585693, + "learning_rate": 5.595773207475872e-06, + "loss": 0.11225184202194213, + "step": 6470 + }, + { + "epoch": 0.6569884228602569, + "grad_norm": 4.011252403259277, + "learning_rate": 5.566318613125128e-06, + "loss": 0.10962972640991211, + "step": 6480 + }, + { + "epoch": 0.6580022938831894, + "grad_norm": 4.013031959533691, + "learning_rate": 5.536911818070636e-06, + "loss": 0.11159608364105225, + "step": 6490 + }, + { + "epoch": 0.6590161649061219, + "grad_norm": 2.9336776733398438, + "learning_rate": 5.507553139345574e-06, + "loss": 0.1061591386795044, + "step": 6500 + }, + { + "epoch": 0.6600300359290544, + "grad_norm": 4.472209453582764, + "learning_rate": 5.478242893464389e-06, + "loss": 0.10536766052246094, + "step": 6510 + }, + { + "epoch": 0.6610439069519869, + "grad_norm": 3.8235745429992676, + "learning_rate": 5.448981396419366e-06, + "loss": 0.11015594005584717, + "step": 6520 + }, + { + "epoch": 0.6620577779749194, + "grad_norm": 3.116565465927124, + "learning_rate": 5.419768963677233e-06, + "loss": 0.10620653629302979, + "step": 6530 + }, + { + "epoch": 0.6630716489978519, + "grad_norm": 3.2516772747039795, + "learning_rate": 5.39060591017576e-06, + "loss": 0.11769680976867676, + "step": 6540 + }, + { + "epoch": 0.6640855200207844, + "grad_norm": 2.6058149337768555, + "learning_rate": 5.3614925503203586e-06, + "loss": 0.11139934062957764, + "step": 6550 + }, + { + "epoch": 0.6650993910437168, + "grad_norm": 3.3498332500457764, + "learning_rate": 5.332429197980706e-06, + "loss": 0.1041823148727417, + "step": 6560 + }, + { + "epoch": 0.6661132620666493, + "grad_norm": 4.169055461883545, + "learning_rate": 5.303416166487337e-06, + "loss": 0.10531578063964844, + "step": 6570 + }, + { + "epoch": 0.6671271330895818, + "grad_norm": 5.180627346038818, + "learning_rate": 5.274453768628289e-06, + "loss": 0.11977331638336182, + "step": 6580 + }, + { + "epoch": 0.6681410041125143, + "grad_norm": 2.940885066986084, + "learning_rate": 5.245542316645718e-06, + "loss": 0.10504686832427979, + "step": 6590 + }, + { + "epoch": 0.6691548751354468, + "grad_norm": 3.87593150138855, + "learning_rate": 5.2166821222325306e-06, + "loss": 0.09505177736282348, + "step": 6600 + }, + { + "epoch": 0.6701687461583793, + "grad_norm": 3.2273948192596436, + "learning_rate": 5.187873496529041e-06, + "loss": 0.09401922225952149, + "step": 6610 + }, + { + "epoch": 0.6711826171813118, + "grad_norm": 4.136530876159668, + "learning_rate": 5.15911675011959e-06, + "loss": 0.11609668731689453, + "step": 6620 + }, + { + "epoch": 0.6721964882042443, + "grad_norm": 3.8510000705718994, + "learning_rate": 5.130412193029213e-06, + "loss": 0.10241303443908692, + "step": 6630 + }, + { + "epoch": 0.6732103592271768, + "grad_norm": 4.273486137390137, + "learning_rate": 5.101760134720306e-06, + "loss": 0.10996030569076538, + "step": 6640 + }, + { + "epoch": 0.6742242302501094, + "grad_norm": 3.257629871368408, + "learning_rate": 5.073160884089263e-06, + "loss": 0.10912792682647705, + "step": 6650 + }, + { + "epoch": 0.6752381012730418, + "grad_norm": 3.964616060256958, + "learning_rate": 5.044614749463173e-06, + "loss": 0.10700547695159912, + "step": 6660 + }, + { + "epoch": 0.6762519722959743, + "grad_norm": 2.444891929626465, + "learning_rate": 5.0161220385964745e-06, + "loss": 0.09193029403686523, + "step": 6670 + }, + { + "epoch": 0.6772658433189068, + "grad_norm": 3.4299352169036865, + "learning_rate": 4.987683058667651e-06, + "loss": 0.10039536952972412, + "step": 6680 + }, + { + "epoch": 0.6782797143418393, + "grad_norm": 4.580758571624756, + "learning_rate": 4.9592981162759215e-06, + "loss": 0.10943892002105712, + "step": 6690 + }, + { + "epoch": 0.6792935853647718, + "grad_norm": 3.0384387969970703, + "learning_rate": 4.930967517437922e-06, + "loss": 0.09660143852233886, + "step": 6700 + }, + { + "epoch": 0.6803074563877043, + "grad_norm": 5.231195449829102, + "learning_rate": 4.902691567584412e-06, + "loss": 0.10751786231994628, + "step": 6710 + }, + { + "epoch": 0.6813213274106368, + "grad_norm": 3.767773151397705, + "learning_rate": 4.87447057155699e-06, + "loss": 0.10292325019836426, + "step": 6720 + }, + { + "epoch": 0.6823351984335693, + "grad_norm": 2.7454493045806885, + "learning_rate": 4.846304833604792e-06, + "loss": 0.09774127006530761, + "step": 6730 + }, + { + "epoch": 0.6833490694565018, + "grad_norm": 4.279491424560547, + "learning_rate": 4.8181946573812236e-06, + "loss": 0.11527223587036133, + "step": 6740 + }, + { + "epoch": 0.6843629404794342, + "grad_norm": 3.5353314876556396, + "learning_rate": 4.790140345940688e-06, + "loss": 0.0970741331577301, + "step": 6750 + }, + { + "epoch": 0.6853768115023667, + "grad_norm": 4.993627548217773, + "learning_rate": 4.762142201735299e-06, + "loss": 0.10593552589416504, + "step": 6760 + }, + { + "epoch": 0.6863906825252992, + "grad_norm": 3.715656280517578, + "learning_rate": 4.734200526611652e-06, + "loss": 0.09863643646240235, + "step": 6770 + }, + { + "epoch": 0.6874045535482317, + "grad_norm": 3.5685930252075195, + "learning_rate": 4.706315621807537e-06, + "loss": 0.10195640325546265, + "step": 6780 + }, + { + "epoch": 0.6884184245711642, + "grad_norm": 3.719881296157837, + "learning_rate": 4.678487787948711e-06, + "loss": 0.09518191814422608, + "step": 6790 + }, + { + "epoch": 0.6894322955940967, + "grad_norm": 3.870194673538208, + "learning_rate": 4.650717325045655e-06, + "loss": 0.09561514854431152, + "step": 6800 + }, + { + "epoch": 0.6904461666170292, + "grad_norm": 3.9872376918792725, + "learning_rate": 4.623004532490328e-06, + "loss": 0.1054659366607666, + "step": 6810 + }, + { + "epoch": 0.6914600376399618, + "grad_norm": 3.9999423027038574, + "learning_rate": 4.595349709052964e-06, + "loss": 0.10201894044876099, + "step": 6820 + }, + { + "epoch": 0.6924739086628943, + "grad_norm": 3.8844408988952637, + "learning_rate": 4.5677531528788196e-06, + "loss": 0.09797661304473877, + "step": 6830 + }, + { + "epoch": 0.6934877796858268, + "grad_norm": 4.107059001922607, + "learning_rate": 4.54021516148498e-06, + "loss": 0.10272367000579834, + "step": 6840 + }, + { + "epoch": 0.6945016507087592, + "grad_norm": 4.020380020141602, + "learning_rate": 4.5127360317571475e-06, + "loss": 0.10402097702026367, + "step": 6850 + }, + { + "epoch": 0.6955155217316917, + "grad_norm": 3.8428847789764404, + "learning_rate": 4.485316059946438e-06, + "loss": 0.0922930359840393, + "step": 6860 + }, + { + "epoch": 0.6965293927546242, + "grad_norm": 5.432012557983398, + "learning_rate": 4.457955541666188e-06, + "loss": 0.10665203332901001, + "step": 6870 + }, + { + "epoch": 0.6975432637775567, + "grad_norm": 3.582287073135376, + "learning_rate": 4.430654771888776e-06, + "loss": 0.09460080862045288, + "step": 6880 + }, + { + "epoch": 0.6985571348004892, + "grad_norm": 5.047722816467285, + "learning_rate": 4.403414044942422e-06, + "loss": 0.08819352388381958, + "step": 6890 + }, + { + "epoch": 0.6995710058234217, + "grad_norm": 3.799659490585327, + "learning_rate": 4.376233654508042e-06, + "loss": 0.09641538858413697, + "step": 6900 + }, + { + "epoch": 0.7005848768463542, + "grad_norm": 5.5390753746032715, + "learning_rate": 4.349113893616056e-06, + "loss": 0.0932919442653656, + "step": 6910 + }, + { + "epoch": 0.7015987478692867, + "grad_norm": 4.343361854553223, + "learning_rate": 4.322055054643244e-06, + "loss": 0.09172606468200684, + "step": 6920 + }, + { + "epoch": 0.7026126188922192, + "grad_norm": 5.08474588394165, + "learning_rate": 4.295057429309589e-06, + "loss": 0.10368661880493164, + "step": 6930 + }, + { + "epoch": 0.7036264899151516, + "grad_norm": 4.198439121246338, + "learning_rate": 4.268121308675132e-06, + "loss": 0.0870087742805481, + "step": 6940 + }, + { + "epoch": 0.7046403609380841, + "grad_norm": 4.15541934967041, + "learning_rate": 4.241246983136834e-06, + "loss": 0.1024333119392395, + "step": 6950 + }, + { + "epoch": 0.7056542319610166, + "grad_norm": 3.7183258533477783, + "learning_rate": 4.214434742425455e-06, + "loss": 0.09056981801986694, + "step": 6960 + }, + { + "epoch": 0.7066681029839491, + "grad_norm": 4.42243766784668, + "learning_rate": 4.1876848756024116e-06, + "loss": 0.09452618956565857, + "step": 6970 + }, + { + "epoch": 0.7076819740068816, + "grad_norm": 3.0845441818237305, + "learning_rate": 4.1609976710566735e-06, + "loss": 0.10820035934448242, + "step": 6980 + }, + { + "epoch": 0.7086958450298142, + "grad_norm": 2.5538697242736816, + "learning_rate": 4.134373416501652e-06, + "loss": 0.09790681600570679, + "step": 6990 + }, + { + "epoch": 0.7097097160527467, + "grad_norm": 3.598191738128662, + "learning_rate": 4.107812398972093e-06, + "loss": 0.10495233535766602, + "step": 7000 + }, + { + "epoch": 0.7107235870756792, + "grad_norm": 3.3377981185913086, + "learning_rate": 4.081314904821001e-06, + "loss": 0.0863314688205719, + "step": 7010 + }, + { + "epoch": 0.7117374580986117, + "grad_norm": 3.975931406021118, + "learning_rate": 4.054881219716521e-06, + "loss": 0.10136170387268066, + "step": 7020 + }, + { + "epoch": 0.7127513291215442, + "grad_norm": 4.7231903076171875, + "learning_rate": 4.0285116286388914e-06, + "loss": 0.09371399879455566, + "step": 7030 + }, + { + "epoch": 0.7137652001444766, + "grad_norm": 6.137765407562256, + "learning_rate": 4.0022064158773445e-06, + "loss": 0.1050760269165039, + "step": 7040 + }, + { + "epoch": 0.7147790711674091, + "grad_norm": 4.281856060028076, + "learning_rate": 3.9759658650270575e-06, + "loss": 0.1076965093612671, + "step": 7050 + }, + { + "epoch": 0.7157929421903416, + "grad_norm": 4.439468860626221, + "learning_rate": 3.94979025898609e-06, + "loss": 0.08502861857414246, + "step": 7060 + }, + { + "epoch": 0.7168068132132741, + "grad_norm": 4.579451560974121, + "learning_rate": 3.9236798799523375e-06, + "loss": 0.09553117156028748, + "step": 7070 + }, + { + "epoch": 0.7178206842362066, + "grad_norm": 4.263151168823242, + "learning_rate": 3.89763500942048e-06, + "loss": 0.11094107627868652, + "step": 7080 + }, + { + "epoch": 0.7188345552591391, + "grad_norm": 4.712898254394531, + "learning_rate": 3.871655928178962e-06, + "loss": 0.09270687699317932, + "step": 7090 + }, + { + "epoch": 0.7198484262820716, + "grad_norm": 3.5212020874023438, + "learning_rate": 3.8457429163069534e-06, + "loss": 0.10604369640350342, + "step": 7100 + }, + { + "epoch": 0.7208622973050041, + "grad_norm": 3.5095746517181396, + "learning_rate": 3.819896253171332e-06, + "loss": 0.10351967811584473, + "step": 7110 + }, + { + "epoch": 0.7218761683279366, + "grad_norm": 3.5819551944732666, + "learning_rate": 3.7941162174236745e-06, + "loss": 0.08718013763427734, + "step": 7120 + }, + { + "epoch": 0.722890039350869, + "grad_norm": 3.278212308883667, + "learning_rate": 3.7684030869972464e-06, + "loss": 0.08557592034339905, + "step": 7130 + }, + { + "epoch": 0.7239039103738015, + "grad_norm": 3.8637046813964844, + "learning_rate": 3.742757139104023e-06, + "loss": 0.1110339879989624, + "step": 7140 + }, + { + "epoch": 0.724917781396734, + "grad_norm": 3.3959288597106934, + "learning_rate": 3.7171786502316767e-06, + "loss": 0.09853792786598206, + "step": 7150 + }, + { + "epoch": 0.7259316524196666, + "grad_norm": 2.1100659370422363, + "learning_rate": 3.691667896140605e-06, + "loss": 0.0855784296989441, + "step": 7160 + }, + { + "epoch": 0.7269455234425991, + "grad_norm": 4.360743522644043, + "learning_rate": 3.6662251518609747e-06, + "loss": 0.08750160932540893, + "step": 7170 + }, + { + "epoch": 0.7279593944655316, + "grad_norm": 4.560605049133301, + "learning_rate": 3.64085069168973e-06, + "loss": 0.09529680609703065, + "step": 7180 + }, + { + "epoch": 0.7289732654884641, + "grad_norm": 4.598474025726318, + "learning_rate": 3.6155447891876518e-06, + "loss": 0.08418370485305786, + "step": 7190 + }, + { + "epoch": 0.7299871365113966, + "grad_norm": 2.8007967472076416, + "learning_rate": 3.590307717176401e-06, + "loss": 0.07898926734924316, + "step": 7200 + }, + { + "epoch": 0.7310010075343291, + "grad_norm": 4.978453636169434, + "learning_rate": 3.565139747735581e-06, + "loss": 0.08115924596786499, + "step": 7210 + }, + { + "epoch": 0.7320148785572616, + "grad_norm": 3.8893840312957764, + "learning_rate": 3.5400411521998123e-06, + "loss": 0.07174283862113953, + "step": 7220 + }, + { + "epoch": 0.733028749580194, + "grad_norm": 7.114648342132568, + "learning_rate": 3.5150122011557888e-06, + "loss": 0.09366913437843323, + "step": 7230 + }, + { + "epoch": 0.7340426206031265, + "grad_norm": 4.572839260101318, + "learning_rate": 3.4900531644393744e-06, + "loss": 0.08273743391036988, + "step": 7240 + }, + { + "epoch": 0.735056491626059, + "grad_norm": 4.599020004272461, + "learning_rate": 3.4651643111326895e-06, + "loss": 0.09210007786750793, + "step": 7250 + }, + { + "epoch": 0.7360703626489915, + "grad_norm": 3.138455629348755, + "learning_rate": 3.440345909561208e-06, + "loss": 0.07570738792419433, + "step": 7260 + }, + { + "epoch": 0.737084233671924, + "grad_norm": 2.776540994644165, + "learning_rate": 3.415598227290877e-06, + "loss": 0.08272864818572997, + "step": 7270 + }, + { + "epoch": 0.7380981046948565, + "grad_norm": 3.589670181274414, + "learning_rate": 3.390921531125211e-06, + "loss": 0.09235111474990845, + "step": 7280 + }, + { + "epoch": 0.739111975717789, + "grad_norm": 3.834231376647949, + "learning_rate": 3.366316087102429e-06, + "loss": 0.07656435370445251, + "step": 7290 + }, + { + "epoch": 0.7401258467407215, + "grad_norm": 3.978571891784668, + "learning_rate": 3.3417821604925937e-06, + "loss": 0.09106248617172241, + "step": 7300 + }, + { + "epoch": 0.741139717763654, + "grad_norm": 4.302156925201416, + "learning_rate": 3.3173200157947315e-06, + "loss": 0.10016658306121826, + "step": 7310 + }, + { + "epoch": 0.7421535887865864, + "grad_norm": 5.071165561676025, + "learning_rate": 3.2929299167339955e-06, + "loss": 0.09221208095550537, + "step": 7320 + }, + { + "epoch": 0.7431674598095189, + "grad_norm": 3.581084966659546, + "learning_rate": 3.2686121262588165e-06, + "loss": 0.0917863130569458, + "step": 7330 + }, + { + "epoch": 0.7441813308324515, + "grad_norm": 3.7746267318725586, + "learning_rate": 3.244366906538071e-06, + "loss": 0.0912899911403656, + "step": 7340 + }, + { + "epoch": 0.745195201855384, + "grad_norm": 2.7246875762939453, + "learning_rate": 3.2201945189582597e-06, + "loss": 0.08881901502609253, + "step": 7350 + }, + { + "epoch": 0.7462090728783165, + "grad_norm": 3.757465124130249, + "learning_rate": 3.1960952241206745e-06, + "loss": 0.08509420156478882, + "step": 7360 + }, + { + "epoch": 0.747222943901249, + "grad_norm": 5.802705764770508, + "learning_rate": 3.172069281838603e-06, + "loss": 0.09016928672790528, + "step": 7370 + }, + { + "epoch": 0.7482368149241815, + "grad_norm": 5.926902770996094, + "learning_rate": 3.148116951134522e-06, + "loss": 0.08701608777046203, + "step": 7380 + }, + { + "epoch": 0.749250685947114, + "grad_norm": 3.0803403854370117, + "learning_rate": 3.1242384902373024e-06, + "loss": 0.09025301933288574, + "step": 7390 + }, + { + "epoch": 0.7502645569700465, + "grad_norm": 3.2145934104919434, + "learning_rate": 3.1004341565794373e-06, + "loss": 0.09012865424156188, + "step": 7400 + }, + { + "epoch": 0.751278427992979, + "grad_norm": 4.195135593414307, + "learning_rate": 3.0767042067942475e-06, + "loss": 0.08236073255538941, + "step": 7410 + }, + { + "epoch": 0.7522922990159114, + "grad_norm": 4.608127117156982, + "learning_rate": 3.0530488967131265e-06, + "loss": 0.08907614946365357, + "step": 7420 + }, + { + "epoch": 0.7533061700388439, + "grad_norm": 4.196930408477783, + "learning_rate": 3.0294684813627917e-06, + "loss": 0.07983638048171997, + "step": 7430 + }, + { + "epoch": 0.7543200410617764, + "grad_norm": 5.582228183746338, + "learning_rate": 3.0059632149625118e-06, + "loss": 0.0856090784072876, + "step": 7440 + }, + { + "epoch": 0.7553339120847089, + "grad_norm": 4.249148368835449, + "learning_rate": 2.982533350921383e-06, + "loss": 0.09263153076171875, + "step": 7450 + }, + { + "epoch": 0.7563477831076414, + "grad_norm": 2.0879597663879395, + "learning_rate": 2.959179141835591e-06, + "loss": 0.07193584442138672, + "step": 7460 + }, + { + "epoch": 0.7573616541305739, + "grad_norm": 3.5875535011291504, + "learning_rate": 2.935900839485688e-06, + "loss": 0.08644048571586609, + "step": 7470 + }, + { + "epoch": 0.7583755251535064, + "grad_norm": 5.607058048248291, + "learning_rate": 2.9126986948338898e-06, + "loss": 0.07533917427062989, + "step": 7480 + }, + { + "epoch": 0.7593893961764389, + "grad_norm": 3.010120153427124, + "learning_rate": 2.889572958021347e-06, + "loss": 0.09604118466377258, + "step": 7490 + }, + { + "epoch": 0.7604032671993713, + "grad_norm": 3.651982545852661, + "learning_rate": 2.866523878365468e-06, + "loss": 0.07544807195663453, + "step": 7500 + }, + { + "epoch": 0.761417138222304, + "grad_norm": 3.233186960220337, + "learning_rate": 2.8435517043572204e-06, + "loss": 0.07849785089492797, + "step": 7510 + }, + { + "epoch": 0.7624310092452364, + "grad_norm": 4.150699138641357, + "learning_rate": 2.820656683658456e-06, + "loss": 0.0910403311252594, + "step": 7520 + }, + { + "epoch": 0.7634448802681689, + "grad_norm": 4.493210792541504, + "learning_rate": 2.797839063099249e-06, + "loss": 0.08287804126739502, + "step": 7530 + }, + { + "epoch": 0.7644587512911014, + "grad_norm": 2.9788002967834473, + "learning_rate": 2.775099088675216e-06, + "loss": 0.09334096312522888, + "step": 7540 + }, + { + "epoch": 0.7654726223140339, + "grad_norm": 4.3548665046691895, + "learning_rate": 2.7524370055448755e-06, + "loss": 0.08318241834640502, + "step": 7550 + }, + { + "epoch": 0.7664864933369664, + "grad_norm": 3.9257402420043945, + "learning_rate": 2.7298530580270154e-06, + "loss": 0.08051464557647706, + "step": 7560 + }, + { + "epoch": 0.7675003643598989, + "grad_norm": 4.017498970031738, + "learning_rate": 2.7073474895980333e-06, + "loss": 0.08062975406646729, + "step": 7570 + }, + { + "epoch": 0.7685142353828314, + "grad_norm": 4.329975605010986, + "learning_rate": 2.6849205428893334e-06, + "loss": 0.08122682571411133, + "step": 7580 + }, + { + "epoch": 0.7695281064057639, + "grad_norm": 2.74079966545105, + "learning_rate": 2.662572459684699e-06, + "loss": 0.08176946640014648, + "step": 7590 + }, + { + "epoch": 0.7705419774286963, + "grad_norm": 3.3427393436431885, + "learning_rate": 2.640303480917693e-06, + "loss": 0.08642507195472718, + "step": 7600 + }, + { + "epoch": 0.7715558484516288, + "grad_norm": 3.319228172302246, + "learning_rate": 2.618113846669059e-06, + "loss": 0.07937125563621521, + "step": 7610 + }, + { + "epoch": 0.7725697194745613, + "grad_norm": 4.3847880363464355, + "learning_rate": 2.596003796164127e-06, + "loss": 0.08316410779953003, + "step": 7620 + }, + { + "epoch": 0.7735835904974938, + "grad_norm": 3.90621280670166, + "learning_rate": 2.5739735677702416e-06, + "loss": 0.08201948404312134, + "step": 7630 + }, + { + "epoch": 0.7745974615204263, + "grad_norm": 3.744443893432617, + "learning_rate": 2.552023398994188e-06, + "loss": 0.08914436101913452, + "step": 7640 + }, + { + "epoch": 0.7756113325433588, + "grad_norm": 4.821990966796875, + "learning_rate": 2.5301535264796274e-06, + "loss": 0.09046002626419067, + "step": 7650 + }, + { + "epoch": 0.7766252035662913, + "grad_norm": 3.8368961811065674, + "learning_rate": 2.508364186004566e-06, + "loss": 0.08538991212844849, + "step": 7660 + }, + { + "epoch": 0.7776390745892238, + "grad_norm": 4.17926025390625, + "learning_rate": 2.486655612478782e-06, + "loss": 0.07871721386909485, + "step": 7670 + }, + { + "epoch": 0.7786529456121564, + "grad_norm": 3.97518253326416, + "learning_rate": 2.465028039941316e-06, + "loss": 0.06599408984184266, + "step": 7680 + }, + { + "epoch": 0.7796668166350889, + "grad_norm": 4.071691036224365, + "learning_rate": 2.443481701557946e-06, + "loss": 0.07802656888961793, + "step": 7690 + }, + { + "epoch": 0.7806806876580213, + "grad_norm": 3.2859222888946533, + "learning_rate": 2.4220168296186587e-06, + "loss": 0.08231407999992371, + "step": 7700 + }, + { + "epoch": 0.7816945586809538, + "grad_norm": 4.081238746643066, + "learning_rate": 2.4006336555351615e-06, + "loss": 0.07257992029190063, + "step": 7710 + }, + { + "epoch": 0.7827084297038863, + "grad_norm": 4.816384792327881, + "learning_rate": 2.3793324098383796e-06, + "loss": 0.0758434772491455, + "step": 7720 + }, + { + "epoch": 0.7837223007268188, + "grad_norm": 4.90199089050293, + "learning_rate": 2.35811332217597e-06, + "loss": 0.06388907432556153, + "step": 7730 + }, + { + "epoch": 0.7847361717497513, + "grad_norm": 4.301286220550537, + "learning_rate": 2.3369766213098553e-06, + "loss": 0.07262455224990845, + "step": 7740 + }, + { + "epoch": 0.7857500427726838, + "grad_norm": 4.864070415496826, + "learning_rate": 2.315922535113743e-06, + "loss": 0.07637689113616944, + "step": 7750 + }, + { + "epoch": 0.7867639137956163, + "grad_norm": 4.626368045806885, + "learning_rate": 2.2949512905706793e-06, + "loss": 0.07640719413757324, + "step": 7760 + }, + { + "epoch": 0.7877777848185488, + "grad_norm": 5.83775520324707, + "learning_rate": 2.274063113770597e-06, + "loss": 0.06869802474975586, + "step": 7770 + }, + { + "epoch": 0.7887916558414813, + "grad_norm": 4.25607442855835, + "learning_rate": 2.2532582299078775e-06, + "loss": 0.082994943857193, + "step": 7780 + }, + { + "epoch": 0.7898055268644137, + "grad_norm": 3.2623932361602783, + "learning_rate": 2.2325368632789345e-06, + "loss": 0.07475184202194214, + "step": 7790 + }, + { + "epoch": 0.7908193978873462, + "grad_norm": 5.235363483428955, + "learning_rate": 2.211899237279779e-06, + "loss": 0.07325462102890015, + "step": 7800 + }, + { + "epoch": 0.7918332689102787, + "grad_norm": 4.662465572357178, + "learning_rate": 2.1913455744036138e-06, + "loss": 0.10266615152359009, + "step": 7810 + }, + { + "epoch": 0.7928471399332112, + "grad_norm": 3.6242082118988037, + "learning_rate": 2.1708760962384545e-06, + "loss": 0.08586451411247253, + "step": 7820 + }, + { + "epoch": 0.7938610109561437, + "grad_norm": 4.0032806396484375, + "learning_rate": 2.1504910234647112e-06, + "loss": 0.07612470984458923, + "step": 7830 + }, + { + "epoch": 0.7948748819790762, + "grad_norm": 3.2457780838012695, + "learning_rate": 2.130190575852832e-06, + "loss": 0.0685884952545166, + "step": 7840 + }, + { + "epoch": 0.7958887530020088, + "grad_norm": 3.5957343578338623, + "learning_rate": 2.109974972260921e-06, + "loss": 0.061274802684783934, + "step": 7850 + }, + { + "epoch": 0.7969026240249413, + "grad_norm": 4.264318943023682, + "learning_rate": 2.0898444306323807e-06, + "loss": 0.06579298377037049, + "step": 7860 + }, + { + "epoch": 0.7979164950478738, + "grad_norm": 3.5217931270599365, + "learning_rate": 2.069799167993577e-06, + "loss": 0.07335643768310547, + "step": 7870 + }, + { + "epoch": 0.7989303660708063, + "grad_norm": 6.662380218505859, + "learning_rate": 2.0498394004514745e-06, + "loss": 0.07476885318756103, + "step": 7880 + }, + { + "epoch": 0.7999442370937387, + "grad_norm": 4.564207077026367, + "learning_rate": 2.029965343191326e-06, + "loss": 0.0841230034828186, + "step": 7890 + }, + { + "epoch": 0.8009581081166712, + "grad_norm": 4.797231674194336, + "learning_rate": 2.0101772104743422e-06, + "loss": 0.08010620474815369, + "step": 7900 + }, + { + "epoch": 0.8019719791396037, + "grad_norm": 5.493629455566406, + "learning_rate": 1.990475215635388e-06, + "loss": 0.07487180233001708, + "step": 7910 + }, + { + "epoch": 0.8029858501625362, + "grad_norm": 3.7394256591796875, + "learning_rate": 1.9708595710806787e-06, + "loss": 0.07076634168624878, + "step": 7920 + }, + { + "epoch": 0.8039997211854687, + "grad_norm": 3.870835304260254, + "learning_rate": 1.9513304882854966e-06, + "loss": 0.0680532455444336, + "step": 7930 + }, + { + "epoch": 0.8050135922084012, + "grad_norm": 4.345033645629883, + "learning_rate": 1.9318881777918973e-06, + "loss": 0.06445815563201904, + "step": 7940 + }, + { + "epoch": 0.8060274632313337, + "grad_norm": 5.748864650726318, + "learning_rate": 1.9125328492064575e-06, + "loss": 0.08083143234252929, + "step": 7950 + }, + { + "epoch": 0.8070413342542662, + "grad_norm": 3.6761703491210938, + "learning_rate": 1.8932647111980007e-06, + "loss": 0.06737288236618041, + "step": 7960 + }, + { + "epoch": 0.8080552052771987, + "grad_norm": 3.810264825820923, + "learning_rate": 1.8740839714953552e-06, + "loss": 0.07562073469161987, + "step": 7970 + }, + { + "epoch": 0.8090690763001311, + "grad_norm": 5.799306869506836, + "learning_rate": 1.8549908368851099e-06, + "loss": 0.08068746328353882, + "step": 7980 + }, + { + "epoch": 0.8100829473230636, + "grad_norm": 3.4200022220611572, + "learning_rate": 1.8359855132093884e-06, + "loss": 0.06974664330482483, + "step": 7990 + }, + { + "epoch": 0.8110968183459961, + "grad_norm": 3.3165814876556396, + "learning_rate": 1.8170682053636358e-06, + "loss": 0.08143278956413269, + "step": 8000 + }, + { + "epoch": 0.8121106893689286, + "grad_norm": 4.209066867828369, + "learning_rate": 1.7982391172943936e-06, + "loss": 0.07681785821914673, + "step": 8010 + }, + { + "epoch": 0.8131245603918611, + "grad_norm": 4.192827224731445, + "learning_rate": 1.7794984519971126e-06, + "loss": 0.08347246646881104, + "step": 8020 + }, + { + "epoch": 0.8141384314147937, + "grad_norm": 3.1553447246551514, + "learning_rate": 1.7608464115139712e-06, + "loss": 0.0671984851360321, + "step": 8030 + }, + { + "epoch": 0.8151523024377262, + "grad_norm": 4.333932876586914, + "learning_rate": 1.7422831969316724e-06, + "loss": 0.08003370761871338, + "step": 8040 + }, + { + "epoch": 0.8161661734606587, + "grad_norm": 4.82602596282959, + "learning_rate": 1.7238090083793003e-06, + "loss": 0.07150027751922608, + "step": 8050 + }, + { + "epoch": 0.8171800444835912, + "grad_norm": 4.34873104095459, + "learning_rate": 1.7054240450261594e-06, + "loss": 0.06122281551361084, + "step": 8060 + }, + { + "epoch": 0.8181939155065237, + "grad_norm": 4.144717216491699, + "learning_rate": 1.6871285050796093e-06, + "loss": 0.07338210344314575, + "step": 8070 + }, + { + "epoch": 0.8192077865294561, + "grad_norm": 3.824894666671753, + "learning_rate": 1.6689225857829527e-06, + "loss": 0.06478950381278992, + "step": 8080 + }, + { + "epoch": 0.8202216575523886, + "grad_norm": 4.097350120544434, + "learning_rate": 1.6508064834132875e-06, + "loss": 0.07208226323127746, + "step": 8090 + }, + { + "epoch": 0.8212355285753211, + "grad_norm": 4.492333889007568, + "learning_rate": 1.632780393279404e-06, + "loss": 0.07045000791549683, + "step": 8100 + }, + { + "epoch": 0.8222493995982536, + "grad_norm": 3.946300983428955, + "learning_rate": 1.614844509719674e-06, + "loss": 0.06987136602401733, + "step": 8110 + }, + { + "epoch": 0.8232632706211861, + "grad_norm": 4.634923934936523, + "learning_rate": 1.5969990260999569e-06, + "loss": 0.07804993391036988, + "step": 8120 + }, + { + "epoch": 0.8242771416441186, + "grad_norm": 2.8438479900360107, + "learning_rate": 1.5792441348115116e-06, + "loss": 0.05947579145431518, + "step": 8130 + }, + { + "epoch": 0.8252910126670511, + "grad_norm": 4.83300256729126, + "learning_rate": 1.5615800272689352e-06, + "loss": 0.06559336185455322, + "step": 8140 + }, + { + "epoch": 0.8263048836899836, + "grad_norm": 4.646676540374756, + "learning_rate": 1.5440068939080776e-06, + "loss": 0.06457459926605225, + "step": 8150 + }, + { + "epoch": 0.827318754712916, + "grad_norm": 4.632523059844971, + "learning_rate": 1.5265249241840153e-06, + "loss": 0.07536451816558838, + "step": 8160 + }, + { + "epoch": 0.8283326257358485, + "grad_norm": 4.729177474975586, + "learning_rate": 1.5091343065689756e-06, + "loss": 0.07242452502250671, + "step": 8170 + }, + { + "epoch": 0.829346496758781, + "grad_norm": 4.673914432525635, + "learning_rate": 1.4918352285503357e-06, + "loss": 0.06610317230224609, + "step": 8180 + }, + { + "epoch": 0.8303603677817135, + "grad_norm": 4.032249450683594, + "learning_rate": 1.4746278766285905e-06, + "loss": 0.07159507870674134, + "step": 8190 + }, + { + "epoch": 0.8313742388046461, + "grad_norm": 3.6339027881622314, + "learning_rate": 1.4575124363153315e-06, + "loss": 0.07360475659370422, + "step": 8200 + }, + { + "epoch": 0.8323881098275786, + "grad_norm": 3.8315789699554443, + "learning_rate": 1.4404890921312654e-06, + "loss": 0.06600156426429749, + "step": 8210 + }, + { + "epoch": 0.8334019808505111, + "grad_norm": 5.03754997253418, + "learning_rate": 1.4235580276042071e-06, + "loss": 0.05969521403312683, + "step": 8220 + }, + { + "epoch": 0.8344158518734436, + "grad_norm": 3.446748733520508, + "learning_rate": 1.4067194252671157e-06, + "loss": 0.0568367063999176, + "step": 8230 + }, + { + "epoch": 0.8354297228963761, + "grad_norm": 4.020167827606201, + "learning_rate": 1.3899734666561138e-06, + "loss": 0.0709646224975586, + "step": 8240 + }, + { + "epoch": 0.8364435939193086, + "grad_norm": 5.661200046539307, + "learning_rate": 1.373320332308541e-06, + "loss": 0.0712781548500061, + "step": 8250 + }, + { + "epoch": 0.837457464942241, + "grad_norm": 3.544123888015747, + "learning_rate": 1.3567602017609981e-06, + "loss": 0.06652522087097168, + "step": 8260 + }, + { + "epoch": 0.8384713359651735, + "grad_norm": 4.671868801116943, + "learning_rate": 1.3402932535474277e-06, + "loss": 0.0645371913909912, + "step": 8270 + }, + { + "epoch": 0.839485206988106, + "grad_norm": 4.484004974365234, + "learning_rate": 1.3239196651971654e-06, + "loss": 0.07063988447189332, + "step": 8280 + }, + { + "epoch": 0.8404990780110385, + "grad_norm": 2.618915557861328, + "learning_rate": 1.3076396132330516e-06, + "loss": 0.05529460310935974, + "step": 8290 + }, + { + "epoch": 0.841512949033971, + "grad_norm": 4.026424884796143, + "learning_rate": 1.2914532731694996e-06, + "loss": 0.059697973728179934, + "step": 8300 + }, + { + "epoch": 0.8425268200569035, + "grad_norm": 4.404993057250977, + "learning_rate": 1.2753608195106314e-06, + "loss": 0.05888262987136841, + "step": 8310 + }, + { + "epoch": 0.843540691079836, + "grad_norm": 5.22847843170166, + "learning_rate": 1.2593624257483838e-06, + "loss": 0.07349226474761963, + "step": 8320 + }, + { + "epoch": 0.8445545621027685, + "grad_norm": 4.128693580627441, + "learning_rate": 1.243458264360634e-06, + "loss": 0.06507245302200318, + "step": 8330 + }, + { + "epoch": 0.845568433125701, + "grad_norm": 3.6669297218322754, + "learning_rate": 1.2276485068093447e-06, + "loss": 0.06138986349105835, + "step": 8340 + }, + { + "epoch": 0.8465823041486334, + "grad_norm": 5.250206470489502, + "learning_rate": 1.2119333235387232e-06, + "loss": 0.058102619647979734, + "step": 8350 + }, + { + "epoch": 0.8475961751715659, + "grad_norm": 4.161450386047363, + "learning_rate": 1.1963128839733674e-06, + "loss": 0.0497779905796051, + "step": 8360 + }, + { + "epoch": 0.8486100461944985, + "grad_norm": 3.999025821685791, + "learning_rate": 1.1807873565164507e-06, + "loss": 0.06247832775115967, + "step": 8370 + }, + { + "epoch": 0.849623917217431, + "grad_norm": 4.787498474121094, + "learning_rate": 1.1653569085479067e-06, + "loss": 0.06528663635253906, + "step": 8380 + }, + { + "epoch": 0.8506377882403635, + "grad_norm": 3.644591808319092, + "learning_rate": 1.150021706422615e-06, + "loss": 0.0669103741645813, + "step": 8390 + }, + { + "epoch": 0.851651659263296, + "grad_norm": 4.6746110916137695, + "learning_rate": 1.1347819154686246e-06, + "loss": 0.07205952405929565, + "step": 8400 + }, + { + "epoch": 0.8526655302862285, + "grad_norm": 3.396259307861328, + "learning_rate": 1.1196376999853531e-06, + "loss": 0.05701977014541626, + "step": 8410 + }, + { + "epoch": 0.853679401309161, + "grad_norm": 4.710214614868164, + "learning_rate": 1.104589223241831e-06, + "loss": 0.06529380679130554, + "step": 8420 + }, + { + "epoch": 0.8546932723320935, + "grad_norm": 3.3613412380218506, + "learning_rate": 1.089636647474923e-06, + "loss": 0.06597838401794434, + "step": 8430 + }, + { + "epoch": 0.855707143355026, + "grad_norm": 2.977033853530884, + "learning_rate": 1.0747801338876006e-06, + "loss": 0.055444192886352536, + "step": 8440 + }, + { + "epoch": 0.8567210143779584, + "grad_norm": 4.760598182678223, + "learning_rate": 1.0600198426471908e-06, + "loss": 0.059856307506561277, + "step": 8450 + }, + { + "epoch": 0.8577348854008909, + "grad_norm": 4.506906032562256, + "learning_rate": 1.0453559328836517e-06, + "loss": 0.06994937658309937, + "step": 8460 + }, + { + "epoch": 0.8587487564238234, + "grad_norm": 4.489019870758057, + "learning_rate": 1.0307885626878555e-06, + "loss": 0.05568776726722717, + "step": 8470 + }, + { + "epoch": 0.8597626274467559, + "grad_norm": 4.3067755699157715, + "learning_rate": 1.0163178891098912e-06, + "loss": 0.05839992761611938, + "step": 8480 + }, + { + "epoch": 0.8607764984696884, + "grad_norm": 4.984467506408691, + "learning_rate": 1.0019440681573633e-06, + "loss": 0.05739668607711792, + "step": 8490 + }, + { + "epoch": 0.8617903694926209, + "grad_norm": 2.168604612350464, + "learning_rate": 9.876672547937117e-07, + "loss": 0.05533984899520874, + "step": 8500 + }, + { + "epoch": 0.8628042405155534, + "grad_norm": 4.860452175140381, + "learning_rate": 9.734876029365448e-07, + "loss": 0.050375497341156004, + "step": 8510 + }, + { + "epoch": 0.8638181115384859, + "grad_norm": 5.294694900512695, + "learning_rate": 9.594052654559737e-07, + "loss": 0.05289696455001831, + "step": 8520 + }, + { + "epoch": 0.8648319825614184, + "grad_norm": 5.216662406921387, + "learning_rate": 9.454203941729767e-07, + "loss": 0.054437273740768434, + "step": 8530 + }, + { + "epoch": 0.865845853584351, + "grad_norm": 4.552338600158691, + "learning_rate": 9.315331398577454e-07, + "loss": 0.05201611518859863, + "step": 8540 + }, + { + "epoch": 0.8668597246072834, + "grad_norm": 4.903604507446289, + "learning_rate": 9.177436522280714e-07, + "loss": 0.055954158306121826, + "step": 8550 + }, + { + "epoch": 0.8678735956302159, + "grad_norm": 6.8809380531311035, + "learning_rate": 9.04052079947727e-07, + "loss": 0.06602383852005005, + "step": 8560 + }, + { + "epoch": 0.8688874666531484, + "grad_norm": 5.51625919342041, + "learning_rate": 8.904585706248637e-07, + "loss": 0.053892111778259276, + "step": 8570 + }, + { + "epoch": 0.8699013376760809, + "grad_norm": 4.694087028503418, + "learning_rate": 8.76963270810427e-07, + "loss": 0.05995824933052063, + "step": 8580 + }, + { + "epoch": 0.8709152086990134, + "grad_norm": 4.8453593254089355, + "learning_rate": 8.635663259965632e-07, + "loss": 0.06062227487564087, + "step": 8590 + }, + { + "epoch": 0.8719290797219459, + "grad_norm": 2.581042766571045, + "learning_rate": 8.502678806150588e-07, + "loss": 0.04872031211853027, + "step": 8600 + }, + { + "epoch": 0.8729429507448784, + "grad_norm": 4.1101274490356445, + "learning_rate": 8.370680780357909e-07, + "loss": 0.05745810866355896, + "step": 8610 + }, + { + "epoch": 0.8739568217678109, + "grad_norm": 4.185912609100342, + "learning_rate": 8.23967060565164e-07, + "loss": 0.05633423328399658, + "step": 8620 + }, + { + "epoch": 0.8749706927907434, + "grad_norm": 3.71832537651062, + "learning_rate": 8.109649694445898e-07, + "loss": 0.055595004558563234, + "step": 8630 + }, + { + "epoch": 0.8759845638136758, + "grad_norm": 5.364109992980957, + "learning_rate": 7.980619448489601e-07, + "loss": 0.06206071376800537, + "step": 8640 + }, + { + "epoch": 0.8769984348366083, + "grad_norm": 3.568004608154297, + "learning_rate": 7.852581258851333e-07, + "loss": 0.04831468462944031, + "step": 8650 + }, + { + "epoch": 0.8780123058595408, + "grad_norm": 4.095893859863281, + "learning_rate": 7.725536505904396e-07, + "loss": 0.04539310336112976, + "step": 8660 + }, + { + "epoch": 0.8790261768824733, + "grad_norm": 3.463761568069458, + "learning_rate": 7.599486559311887e-07, + "loss": 0.056469178199768065, + "step": 8670 + }, + { + "epoch": 0.8800400479054058, + "grad_norm": 4.052176475524902, + "learning_rate": 7.474432778011953e-07, + "loss": 0.057718807458877565, + "step": 8680 + }, + { + "epoch": 0.8810539189283383, + "grad_norm": 3.9141082763671875, + "learning_rate": 7.350376510203105e-07, + "loss": 0.0517420768737793, + "step": 8690 + }, + { + "epoch": 0.8820677899512708, + "grad_norm": 3.884742259979248, + "learning_rate": 7.227319093329732e-07, + "loss": 0.06175707578659058, + "step": 8700 + }, + { + "epoch": 0.8830816609742033, + "grad_norm": 3.5338048934936523, + "learning_rate": 7.105261854067691e-07, + "loss": 0.0454587310552597, + "step": 8710 + }, + { + "epoch": 0.8840955319971359, + "grad_norm": 3.973628282546997, + "learning_rate": 6.984206108309932e-07, + "loss": 0.056200063228607176, + "step": 8720 + }, + { + "epoch": 0.8851094030200684, + "grad_norm": 2.807649612426758, + "learning_rate": 6.864153161152354e-07, + "loss": 0.03583030700683594, + "step": 8730 + }, + { + "epoch": 0.8861232740430008, + "grad_norm": 3.9924516677856445, + "learning_rate": 6.745104306879769e-07, + "loss": 0.06085458993911743, + "step": 8740 + }, + { + "epoch": 0.8871371450659333, + "grad_norm": 5.401541709899902, + "learning_rate": 6.627060828951881e-07, + "loss": 0.06207215189933777, + "step": 8750 + }, + { + "epoch": 0.8881510160888658, + "grad_norm": 3.9619009494781494, + "learning_rate": 6.510023999989501e-07, + "loss": 0.061995464563369754, + "step": 8760 + }, + { + "epoch": 0.8891648871117983, + "grad_norm": 8.907490730285645, + "learning_rate": 6.393995081760795e-07, + "loss": 0.047555601596832274, + "step": 8770 + }, + { + "epoch": 0.8901787581347308, + "grad_norm": 3.590949058532715, + "learning_rate": 6.278975325167691e-07, + "loss": 0.05746371746063232, + "step": 8780 + }, + { + "epoch": 0.8911926291576633, + "grad_norm": 6.2922468185424805, + "learning_rate": 6.164965970232417e-07, + "loss": 0.07090592384338379, + "step": 8790 + }, + { + "epoch": 0.8922065001805958, + "grad_norm": 3.9544894695281982, + "learning_rate": 6.051968246084117e-07, + "loss": 0.05637362599372864, + "step": 8800 + }, + { + "epoch": 0.8932203712035283, + "grad_norm": 5.967639923095703, + "learning_rate": 5.939983370945535e-07, + "loss": 0.062176215648651126, + "step": 8810 + }, + { + "epoch": 0.8942342422264608, + "grad_norm": 2.185020685195923, + "learning_rate": 5.829012552120039e-07, + "loss": 0.05582418441772461, + "step": 8820 + }, + { + "epoch": 0.8952481132493932, + "grad_norm": 4.049367904663086, + "learning_rate": 5.719056985978388e-07, + "loss": 0.047624561190605166, + "step": 8830 + }, + { + "epoch": 0.8962619842723257, + "grad_norm": 4.620866775512695, + "learning_rate": 5.610117857946062e-07, + "loss": 0.05343813896179199, + "step": 8840 + }, + { + "epoch": 0.8972758552952582, + "grad_norm": 4.959354877471924, + "learning_rate": 5.502196342490318e-07, + "loss": 0.050389498472213745, + "step": 8850 + }, + { + "epoch": 0.8982897263181907, + "grad_norm": 4.164300918579102, + "learning_rate": 5.395293603107566e-07, + "loss": 0.045359176397323606, + "step": 8860 + }, + { + "epoch": 0.8993035973411232, + "grad_norm": 3.666952610015869, + "learning_rate": 5.2894107923109e-07, + "loss": 0.06125727295875549, + "step": 8870 + }, + { + "epoch": 0.9003174683640557, + "grad_norm": 4.082065582275391, + "learning_rate": 5.184549051617571e-07, + "loss": 0.060871231555938723, + "step": 8880 + }, + { + "epoch": 0.9013313393869883, + "grad_norm": 4.89211893081665, + "learning_rate": 5.08070951153673e-07, + "loss": 0.048236137628555296, + "step": 8890 + }, + { + "epoch": 0.9023452104099208, + "grad_norm": 3.656381130218506, + "learning_rate": 4.977893291557256e-07, + "loss": 0.05188581347465515, + "step": 8900 + }, + { + "epoch": 0.9033590814328533, + "grad_norm": 3.654435396194458, + "learning_rate": 4.876101500135621e-07, + "loss": 0.05569586753845215, + "step": 8910 + }, + { + "epoch": 0.9043729524557858, + "grad_norm": 4.415435791015625, + "learning_rate": 4.775335234684042e-07, + "loss": 0.04953027069568634, + "step": 8920 + }, + { + "epoch": 0.9053868234787182, + "grad_norm": 3.363234758377075, + "learning_rate": 4.675595581558556e-07, + "loss": 0.051300299167633054, + "step": 8930 + }, + { + "epoch": 0.9064006945016507, + "grad_norm": 4.058452129364014, + "learning_rate": 4.576883616047345e-07, + "loss": 0.06935640573501586, + "step": 8940 + }, + { + "epoch": 0.9074145655245832, + "grad_norm": 3.230126142501831, + "learning_rate": 4.479200402359174e-07, + "loss": 0.04556883275508881, + "step": 8950 + }, + { + "epoch": 0.9084284365475157, + "grad_norm": 5.397610187530518, + "learning_rate": 4.382546993611836e-07, + "loss": 0.05044255256652832, + "step": 8960 + }, + { + "epoch": 0.9094423075704482, + "grad_norm": 3.867213726043701, + "learning_rate": 4.28692443182086e-07, + "loss": 0.05272663831710815, + "step": 8970 + }, + { + "epoch": 0.9104561785933807, + "grad_norm": 4.495839595794678, + "learning_rate": 4.192333747888311e-07, + "loss": 0.046833467483520505, + "step": 8980 + }, + { + "epoch": 0.9114700496163132, + "grad_norm": 4.011538505554199, + "learning_rate": 4.0987759615915837e-07, + "loss": 0.053583329916000365, + "step": 8990 + }, + { + "epoch": 0.9124839206392457, + "grad_norm": 5.345243453979492, + "learning_rate": 4.0062520815724816e-07, + "loss": 0.051647520065307616, + "step": 9000 + }, + { + "epoch": 0.9134977916621781, + "grad_norm": 4.43966007232666, + "learning_rate": 3.9147631053263027e-07, + "loss": 0.040019291639328006, + "step": 9010 + }, + { + "epoch": 0.9145116626851106, + "grad_norm": 5.278563499450684, + "learning_rate": 3.824310019191102e-07, + "loss": 0.055504173040390015, + "step": 9020 + }, + { + "epoch": 0.9155255337080431, + "grad_norm": 3.1646811962127686, + "learning_rate": 3.734893798337058e-07, + "loss": 0.0428948849439621, + "step": 9030 + }, + { + "epoch": 0.9165394047309756, + "grad_norm": 3.1028330326080322, + "learning_rate": 3.646515406755946e-07, + "loss": 0.05799094438552856, + "step": 9040 + }, + { + "epoch": 0.9175532757539081, + "grad_norm": 4.803033828735352, + "learning_rate": 3.559175797250769e-07, + "loss": 0.052741020917892456, + "step": 9050 + }, + { + "epoch": 0.9185671467768407, + "grad_norm": 5.094013214111328, + "learning_rate": 3.4728759114254774e-07, + "loss": 0.050057971477508546, + "step": 9060 + }, + { + "epoch": 0.9195810177997732, + "grad_norm": 3.9706530570983887, + "learning_rate": 3.3876166796747877e-07, + "loss": 0.06135883331298828, + "step": 9070 + }, + { + "epoch": 0.9205948888227057, + "grad_norm": 2.8856067657470703, + "learning_rate": 3.303399021174214e-07, + "loss": 0.05447368621826172, + "step": 9080 + }, + { + "epoch": 0.9216087598456382, + "grad_norm": 7.216680526733398, + "learning_rate": 3.220223843870096e-07, + "loss": 0.051039642095565795, + "step": 9090 + }, + { + "epoch": 0.9226226308685707, + "grad_norm": 7.048734664916992, + "learning_rate": 3.1380920444698093e-07, + "loss": 0.05923209190368652, + "step": 9100 + }, + { + "epoch": 0.9236365018915031, + "grad_norm": 3.520742893218994, + "learning_rate": 3.057004508432193e-07, + "loss": 0.0446564257144928, + "step": 9110 + }, + { + "epoch": 0.9246503729144356, + "grad_norm": 2.305326223373413, + "learning_rate": 2.976962109957882e-07, + "loss": 0.04739856719970703, + "step": 9120 + }, + { + "epoch": 0.9256642439373681, + "grad_norm": 1.9894465208053589, + "learning_rate": 2.8979657119799667e-07, + "loss": 0.04466649293899536, + "step": 9130 + }, + { + "epoch": 0.9266781149603006, + "grad_norm": 5.532656192779541, + "learning_rate": 2.8200161661546377e-07, + "loss": 0.04986405074596405, + "step": 9140 + }, + { + "epoch": 0.9276919859832331, + "grad_norm": 4.431173324584961, + "learning_rate": 2.7431143128520243e-07, + "loss": 0.04426388144493103, + "step": 9150 + }, + { + "epoch": 0.9287058570061656, + "grad_norm": 3.7239391803741455, + "learning_rate": 2.667260981147157e-07, + "loss": 0.04048580527305603, + "step": 9160 + }, + { + "epoch": 0.9297197280290981, + "grad_norm": 5.4593729972839355, + "learning_rate": 2.5924569888109654e-07, + "loss": 0.05242964029312134, + "step": 9170 + }, + { + "epoch": 0.9307335990520306, + "grad_norm": 5.179825305938721, + "learning_rate": 2.5187031423015597e-07, + "loss": 0.048725205659866336, + "step": 9180 + }, + { + "epoch": 0.9317474700749631, + "grad_norm": 3.8134920597076416, + "learning_rate": 2.4460002367554523e-07, + "loss": 0.040651601552963254, + "step": 9190 + }, + { + "epoch": 0.9327613410978955, + "grad_norm": 5.037268161773682, + "learning_rate": 2.3743490559789973e-07, + "loss": 0.05762947797775268, + "step": 9200 + }, + { + "epoch": 0.933775212120828, + "grad_norm": 3.235048532485962, + "learning_rate": 2.3037503724399946e-07, + "loss": 0.03936000168323517, + "step": 9210 + }, + { + "epoch": 0.9347890831437605, + "grad_norm": 4.979572772979736, + "learning_rate": 2.2342049472592996e-07, + "loss": 0.04393621683120728, + "step": 9220 + }, + { + "epoch": 0.9358029541666931, + "grad_norm": 6.494809150695801, + "learning_rate": 2.1657135302026177e-07, + "loss": 0.050293976068496705, + "step": 9230 + }, + { + "epoch": 0.9368168251896256, + "grad_norm": 3.7971179485321045, + "learning_rate": 2.09827685967251e-07, + "loss": 0.038842815160751346, + "step": 9240 + }, + { + "epoch": 0.9378306962125581, + "grad_norm": 3.0034334659576416, + "learning_rate": 2.0318956627003227e-07, + "loss": 0.037390795350074765, + "step": 9250 + }, + { + "epoch": 0.9388445672354906, + "grad_norm": 3.99235200881958, + "learning_rate": 1.9665706549384044e-07, + "loss": 0.048086100816726686, + "step": 9260 + }, + { + "epoch": 0.9398584382584231, + "grad_norm": 3.6204681396484375, + "learning_rate": 1.9023025406523787e-07, + "loss": 0.054035228490829465, + "step": 9270 + }, + { + "epoch": 0.9408723092813556, + "grad_norm": 4.166232109069824, + "learning_rate": 1.8390920127135613e-07, + "loss": 0.043131071329116824, + "step": 9280 + }, + { + "epoch": 0.9418861803042881, + "grad_norm": 5.064733028411865, + "learning_rate": 1.7769397525914668e-07, + "loss": 0.044551041722297666, + "step": 9290 + }, + { + "epoch": 0.9429000513272205, + "grad_norm": 3.8290116786956787, + "learning_rate": 1.7158464303465017e-07, + "loss": 0.043517133593559264, + "step": 9300 + }, + { + "epoch": 0.943913922350153, + "grad_norm": 5.571969032287598, + "learning_rate": 1.6558127046226945e-07, + "loss": 0.0515766978263855, + "step": 9310 + }, + { + "epoch": 0.9449277933730855, + "grad_norm": 4.338844299316406, + "learning_rate": 1.5968392226406337e-07, + "loss": 0.05126514434814453, + "step": 9320 + }, + { + "epoch": 0.945941664396018, + "grad_norm": 5.510708808898926, + "learning_rate": 1.5389266201904506e-07, + "loss": 0.03968401551246643, + "step": 9330 + }, + { + "epoch": 0.9469555354189505, + "grad_norm": 2.5495848655700684, + "learning_rate": 1.482075521625026e-07, + "loss": 0.05013599991798401, + "step": 9340 + }, + { + "epoch": 0.947969406441883, + "grad_norm": 3.4331796169281006, + "learning_rate": 1.4262865398531834e-07, + "loss": 0.04042300283908844, + "step": 9350 + }, + { + "epoch": 0.9489832774648155, + "grad_norm": 5.587519645690918, + "learning_rate": 1.3715602763331172e-07, + "loss": 0.04711132049560547, + "step": 9360 + }, + { + "epoch": 0.949997148487748, + "grad_norm": 5.029048919677734, + "learning_rate": 1.3178973210659528e-07, + "loss": 0.049567264318466184, + "step": 9370 + }, + { + "epoch": 0.9510110195106805, + "grad_norm": 4.195864200592041, + "learning_rate": 1.2652982525892975e-07, + "loss": 0.05070853233337402, + "step": 9380 + }, + { + "epoch": 0.9520248905336129, + "grad_norm": 6.554836273193359, + "learning_rate": 1.213763637971066e-07, + "loss": 0.05167155265808106, + "step": 9390 + }, + { + "epoch": 0.9530387615565454, + "grad_norm": 4.599090576171875, + "learning_rate": 1.1632940328033526e-07, + "loss": 0.03980710208415985, + "step": 9400 + }, + { + "epoch": 0.954052632579478, + "grad_norm": 5.103481292724609, + "learning_rate": 1.1138899811964477e-07, + "loss": 0.045555830001831055, + "step": 9410 + }, + { + "epoch": 0.9550665036024105, + "grad_norm": 6.243200302124023, + "learning_rate": 1.0655520157729194e-07, + "loss": 0.046390372514724734, + "step": 9420 + }, + { + "epoch": 0.956080374625343, + "grad_norm": 3.6347124576568604, + "learning_rate": 1.0182806576619742e-07, + "loss": 0.04723300933837891, + "step": 9430 + }, + { + "epoch": 0.9570942456482755, + "grad_norm": 4.091207981109619, + "learning_rate": 9.72076416493728e-08, + "loss": 0.03826842904090881, + "step": 9440 + }, + { + "epoch": 0.958108116671208, + "grad_norm": 4.707202911376953, + "learning_rate": 9.269397903937883e-08, + "loss": 0.04955234527587891, + "step": 9450 + }, + { + "epoch": 0.9591219876941405, + "grad_norm": 4.93208646774292, + "learning_rate": 8.828712659778471e-08, + "loss": 0.04361345171928406, + "step": 9460 + }, + { + "epoch": 0.960135858717073, + "grad_norm": 4.273813247680664, + "learning_rate": 8.398713183464857e-08, + "loss": 0.038337743282318114, + "step": 9470 + }, + { + "epoch": 0.9611497297400055, + "grad_norm": 6.449532985687256, + "learning_rate": 7.979404110799338e-08, + "loss": 0.04353020489215851, + "step": 9480 + }, + { + "epoch": 0.9621636007629379, + "grad_norm": 4.734971046447754, + "learning_rate": 7.570789962332181e-08, + "loss": 0.04119510054588318, + "step": 9490 + }, + { + "epoch": 0.9631774717858704, + "grad_norm": 4.298241138458252, + "learning_rate": 7.172875143311996e-08, + "loss": 0.04541396498680115, + "step": 9500 + }, + { + "epoch": 0.9641913428088029, + "grad_norm": 6.5605645179748535, + "learning_rate": 6.785663943638333e-08, + "loss": 0.05138916969299316, + "step": 9510 + }, + { + "epoch": 0.9652052138317354, + "grad_norm": 3.282313108444214, + "learning_rate": 6.409160537815818e-08, + "loss": 0.046571022272109984, + "step": 9520 + }, + { + "epoch": 0.9662190848546679, + "grad_norm": 4.403426647186279, + "learning_rate": 6.043368984908759e-08, + "loss": 0.05289278626441955, + "step": 9530 + }, + { + "epoch": 0.9672329558776004, + "grad_norm": 3.794332981109619, + "learning_rate": 5.688293228497399e-08, + "loss": 0.04766389727592468, + "step": 9540 + }, + { + "epoch": 0.9682468269005329, + "grad_norm": 3.5862913131713867, + "learning_rate": 5.34393709663561e-08, + "loss": 0.04343914389610291, + "step": 9550 + }, + { + "epoch": 0.9692606979234654, + "grad_norm": 3.474169969558716, + "learning_rate": 5.0103043018097096e-08, + "loss": 0.05159216523170471, + "step": 9560 + }, + { + "epoch": 0.9702745689463979, + "grad_norm": 2.750925302505493, + "learning_rate": 4.687398440897828e-08, + "loss": 0.04581067264080048, + "step": 9570 + }, + { + "epoch": 0.9712884399693305, + "grad_norm": 3.368772268295288, + "learning_rate": 4.3752229951321553e-08, + "loss": 0.04315263032913208, + "step": 9580 + }, + { + "epoch": 0.9723023109922629, + "grad_norm": 3.135019063949585, + "learning_rate": 4.0737813300602e-08, + "loss": 0.04959716796875, + "step": 9590 + }, + { + "epoch": 0.9733161820151954, + "grad_norm": 4.5461835861206055, + "learning_rate": 3.783076695509702e-08, + "loss": 0.04830975532531738, + "step": 9600 + }, + { + "epoch": 0.9743300530381279, + "grad_norm": 3.5164551734924316, + "learning_rate": 3.5031122255524406e-08, + "loss": 0.0467154860496521, + "step": 9610 + }, + { + "epoch": 0.9753439240610604, + "grad_norm": 4.50295877456665, + "learning_rate": 3.233890938471373e-08, + "loss": 0.05592261552810669, + "step": 9620 + }, + { + "epoch": 0.9763577950839929, + "grad_norm": 2.4753334522247314, + "learning_rate": 2.9754157367277715e-08, + "loss": 0.05559673309326172, + "step": 9630 + }, + { + "epoch": 0.9773716661069254, + "grad_norm": 4.623264789581299, + "learning_rate": 2.72768940692969e-08, + "loss": 0.04049312472343445, + "step": 9640 + }, + { + "epoch": 0.9783855371298579, + "grad_norm": 6.182003498077393, + "learning_rate": 2.4907146198023256e-08, + "loss": 0.04693756401538849, + "step": 9650 + }, + { + "epoch": 0.9793994081527904, + "grad_norm": 6.662285327911377, + "learning_rate": 2.2644939301588174e-08, + "loss": 0.0470002681016922, + "step": 9660 + }, + { + "epoch": 0.9804132791757229, + "grad_norm": 5.010436058044434, + "learning_rate": 2.049029776873268e-08, + "loss": 0.03270629048347473, + "step": 9670 + }, + { + "epoch": 0.9814271501986553, + "grad_norm": 4.371983051300049, + "learning_rate": 1.8443244828536544e-08, + "loss": 0.05144522190093994, + "step": 9680 + }, + { + "epoch": 0.9824410212215878, + "grad_norm": 4.276113510131836, + "learning_rate": 1.650380255017625e-08, + "loss": 0.056083542108535764, + "step": 9690 + }, + { + "epoch": 0.9834548922445203, + "grad_norm": 2.8994107246398926, + "learning_rate": 1.4671991842680756e-08, + "loss": 0.03495630323886871, + "step": 9700 + }, + { + "epoch": 0.9844687632674528, + "grad_norm": 5.253048896789551, + "learning_rate": 1.294783245470943e-08, + "loss": 0.044024106860160825, + "step": 9710 + }, + { + "epoch": 0.9854826342903853, + "grad_norm": 4.4842681884765625, + "learning_rate": 1.13313429743378e-08, + "loss": 0.043555465340614316, + "step": 9720 + }, + { + "epoch": 0.9864965053133178, + "grad_norm": 2.4563965797424316, + "learning_rate": 9.822540828858807e-09, + "loss": 0.04690899550914764, + "step": 9730 + }, + { + "epoch": 0.9875103763362503, + "grad_norm": 4.723957538604736, + "learning_rate": 8.421442284591852e-09, + "loss": 0.05804799795150757, + "step": 9740 + }, + { + "epoch": 0.9885242473591829, + "grad_norm": 4.236462593078613, + "learning_rate": 7.128062446709605e-09, + "loss": 0.045592135190963744, + "step": 9750 + }, + { + "epoch": 0.9895381183821154, + "grad_norm": 6.025130271911621, + "learning_rate": 5.942415259074796e-09, + "loss": 0.04974843859672547, + "step": 9760 + }, + { + "epoch": 0.9905519894050479, + "grad_norm": 3.7627041339874268, + "learning_rate": 4.86451350409145e-09, + "loss": 0.039305266737937924, + "step": 9770 + }, + { + "epoch": 0.9915658604279803, + "grad_norm": 3.8150980472564697, + "learning_rate": 3.894368802563886e-09, + "loss": 0.03772591650485992, + "step": 9780 + }, + { + "epoch": 0.9925797314509128, + "grad_norm": 4.096359729766846, + "learning_rate": 3.0319916135723713e-09, + "loss": 0.05125555992126465, + "step": 9790 + }, + { + "epoch": 0.9935936024738453, + "grad_norm": 5.047242164611816, + "learning_rate": 2.277391234363213e-09, + "loss": 0.0378339022397995, + "step": 9800 + }, + { + "epoch": 0.9946074734967778, + "grad_norm": 3.583134651184082, + "learning_rate": 1.6305758002443939e-09, + "loss": 0.03901187181472778, + "step": 9810 + }, + { + "epoch": 0.9956213445197103, + "grad_norm": 3.8519444465637207, + "learning_rate": 1.091552284500086e-09, + "loss": 0.04941209256649017, + "step": 9820 + }, + { + "epoch": 0.9966352155426428, + "grad_norm": 5.321685791015625, + "learning_rate": 6.603264983173762e-10, + "loss": 0.05181398391723633, + "step": 9830 + }, + { + "epoch": 0.9976490865655753, + "grad_norm": 3.19205904006958, + "learning_rate": 3.3690309071965355e-10, + "loss": 0.04244235157966614, + "step": 9840 + }, + { + "epoch": 0.9986629575885078, + "grad_norm": 3.2344555854797363, + "learning_rate": 1.2128554851664754e-10, + "loss": 0.040759125351905824, + "step": 9850 + }, + { + "epoch": 0.9996768286114402, + "grad_norm": 5.037717819213867, + "learning_rate": 1.3476196272232955e-11, + "loss": 0.03721001744270325, + "step": 9860 + }, + { + "epoch": 1.0, + "step": 9864, + "total_flos": 5.019574154025134e+19, + "train_loss": 0.12587243983248628, + "train_runtime": 149611.9607, + "train_samples_per_second": 4.219, + "train_steps_per_second": 0.066 + } + ], + "logging_steps": 10, + "max_steps": 9864, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 1000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 5.019574154025134e+19, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}