diff --git "a/latest_checkpoint/trainer_state.json" "b/latest_checkpoint/trainer_state.json" new file mode 100644--- /dev/null +++ "b/latest_checkpoint/trainer_state.json" @@ -0,0 +1,7041 @@ +{ + "best_global_step": null, + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 34.01362862010222, + "eval_steps": 500, + "global_step": 10000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0034071550255536627, + "grad_norm": 500.9735107421875, + "learning_rate": 0.0001, + "loss": 12.947656631469727, + "step": 1 + }, + { + "epoch": 0.034071550255536626, + "grad_norm": 7.958178520202637, + "learning_rate": 9.999980014064402e-05, + "loss": 5.756121317545573, + "step": 10 + }, + { + "epoch": 0.06814310051107325, + "grad_norm": 7.650475025177002, + "learning_rate": 9.999910927084749e-05, + "loss": 3.2973617553710937, + "step": 20 + }, + { + "epoch": 0.10221465076660988, + "grad_norm": 6.018689155578613, + "learning_rate": 9.999792493002788e-05, + "loss": 2.9039283752441407, + "step": 30 + }, + { + "epoch": 0.1362862010221465, + "grad_norm": 4.7625932693481445, + "learning_rate": 9.99962471298742e-05, + "loss": 2.800827980041504, + "step": 40 + }, + { + "epoch": 0.17035775127768313, + "grad_norm": 4.674959659576416, + "learning_rate": 9.999407588694567e-05, + "loss": 2.730032730102539, + "step": 50 + }, + { + "epoch": 0.20442930153321975, + "grad_norm": 4.226341724395752, + "learning_rate": 9.999141122267154e-05, + "loss": 2.7617685317993166, + "step": 60 + }, + { + "epoch": 0.23850085178875638, + "grad_norm": 5.014325141906738, + "learning_rate": 9.998825316335099e-05, + "loss": 2.6507848739624023, + "step": 70 + }, + { + "epoch": 0.272572402044293, + "grad_norm": 4.564136505126953, + "learning_rate": 9.99846017401528e-05, + "loss": 2.5295478820800783, + "step": 80 + }, + { + "epoch": 0.30664395229982966, + "grad_norm": 4.037264347076416, + "learning_rate": 9.998045698911505e-05, + "loss": 2.5867191314697267, + "step": 90 + }, + { + "epoch": 0.34071550255536626, + "grad_norm": 4.819895267486572, + "learning_rate": 9.997581895114472e-05, + "loss": 2.504091262817383, + "step": 100 + }, + { + "epoch": 0.3747870528109029, + "grad_norm": 7.4263458251953125, + "learning_rate": 9.99706876720174e-05, + "loss": 2.3738216400146483, + "step": 110 + }, + { + "epoch": 0.4088586030664395, + "grad_norm": 7.7423624992370605, + "learning_rate": 9.996506320237675e-05, + "loss": 2.1092508316040037, + "step": 120 + }, + { + "epoch": 0.44293015332197616, + "grad_norm": 9.078983306884766, + "learning_rate": 9.9958945597734e-05, + "loss": 1.4427892684936523, + "step": 130 + }, + { + "epoch": 0.47700170357751276, + "grad_norm": 6.083858489990234, + "learning_rate": 9.995233491846744e-05, + "loss": 0.9778816223144531, + "step": 140 + }, + { + "epoch": 0.5110732538330494, + "grad_norm": 5.164566516876221, + "learning_rate": 9.99452312298218e-05, + "loss": 0.7422083854675293, + "step": 150 + }, + { + "epoch": 0.545144804088586, + "grad_norm": 5.00791597366333, + "learning_rate": 9.993763460190763e-05, + "loss": 0.6625918388366699, + "step": 160 + }, + { + "epoch": 0.5792163543441227, + "grad_norm": 5.410003662109375, + "learning_rate": 9.99295451097006e-05, + "loss": 0.6380346298217774, + "step": 170 + }, + { + "epoch": 0.6132879045996593, + "grad_norm": 4.947885513305664, + "learning_rate": 9.992096283304069e-05, + "loss": 0.5670146465301513, + "step": 180 + }, + { + "epoch": 0.6473594548551959, + "grad_norm": 3.298889636993408, + "learning_rate": 9.991188785663153e-05, + "loss": 0.5779364109039307, + "step": 190 + }, + { + "epoch": 0.6814310051107325, + "grad_norm": 3.480926513671875, + "learning_rate": 9.990232027003947e-05, + "loss": 0.5716145038604736, + "step": 200 + }, + { + "epoch": 0.7155025553662692, + "grad_norm": 3.2379493713378906, + "learning_rate": 9.989226016769273e-05, + "loss": 0.493179988861084, + "step": 210 + }, + { + "epoch": 0.7495741056218058, + "grad_norm": 5.55947208404541, + "learning_rate": 9.988170764888045e-05, + "loss": 0.5329806804656982, + "step": 220 + }, + { + "epoch": 0.7836456558773425, + "grad_norm": 4.637864112854004, + "learning_rate": 9.987066281775174e-05, + "loss": 0.4877805233001709, + "step": 230 + }, + { + "epoch": 0.817717206132879, + "grad_norm": 3.430638551712036, + "learning_rate": 9.985912578331463e-05, + "loss": 0.4545762062072754, + "step": 240 + }, + { + "epoch": 0.8517887563884157, + "grad_norm": 3.589728593826294, + "learning_rate": 9.984709665943494e-05, + "loss": 0.39948291778564454, + "step": 250 + }, + { + "epoch": 0.8858603066439523, + "grad_norm": 4.287285804748535, + "learning_rate": 9.983457556483536e-05, + "loss": 0.4370544910430908, + "step": 260 + }, + { + "epoch": 0.919931856899489, + "grad_norm": 3.479282855987549, + "learning_rate": 9.982156262309395e-05, + "loss": 0.4726566314697266, + "step": 270 + }, + { + "epoch": 0.9540034071550255, + "grad_norm": 4.917163372039795, + "learning_rate": 9.980805796264325e-05, + "loss": 0.419099235534668, + "step": 280 + }, + { + "epoch": 0.9880749574105622, + "grad_norm": 2.7116401195526123, + "learning_rate": 9.979406171676877e-05, + "loss": 0.4242133617401123, + "step": 290 + }, + { + "epoch": 1.020442930153322, + "grad_norm": 2.691772937774658, + "learning_rate": 9.977957402360783e-05, + "loss": 0.3366670608520508, + "step": 300 + }, + { + "epoch": 1.0545144804088586, + "grad_norm": 2.624289035797119, + "learning_rate": 9.976459502614811e-05, + "loss": 0.2996376991271973, + "step": 310 + }, + { + "epoch": 1.0885860306643953, + "grad_norm": 2.505502939224243, + "learning_rate": 9.974912487222626e-05, + "loss": 0.2905572414398193, + "step": 320 + }, + { + "epoch": 1.122657580919932, + "grad_norm": 3.138418674468994, + "learning_rate": 9.973316371452646e-05, + "loss": 0.318436336517334, + "step": 330 + }, + { + "epoch": 1.1567291311754686, + "grad_norm": 2.702422618865967, + "learning_rate": 9.971671171057886e-05, + "loss": 0.27060980796813966, + "step": 340 + }, + { + "epoch": 1.190800681431005, + "grad_norm": 2.5750415325164795, + "learning_rate": 9.969976902275815e-05, + "loss": 0.27452983856201174, + "step": 350 + }, + { + "epoch": 1.2248722316865417, + "grad_norm": 3.4507346153259277, + "learning_rate": 9.968233581828179e-05, + "loss": 0.33898699283599854, + "step": 360 + }, + { + "epoch": 1.2589437819420783, + "grad_norm": 2.734762668609619, + "learning_rate": 9.966441226920848e-05, + "loss": 0.2835655450820923, + "step": 370 + }, + { + "epoch": 1.293015332197615, + "grad_norm": 4.383528709411621, + "learning_rate": 9.964599855243639e-05, + "loss": 0.24261927604675293, + "step": 380 + }, + { + "epoch": 1.3270868824531517, + "grad_norm": 2.645526647567749, + "learning_rate": 9.96270948497015e-05, + "loss": 0.31325743198394773, + "step": 390 + }, + { + "epoch": 1.3611584327086883, + "grad_norm": 2.6146252155303955, + "learning_rate": 9.960770134757572e-05, + "loss": 0.31850156784057615, + "step": 400 + }, + { + "epoch": 1.3952299829642247, + "grad_norm": 2.889542579650879, + "learning_rate": 9.958781823746505e-05, + "loss": 0.2688532590866089, + "step": 410 + }, + { + "epoch": 1.4293015332197614, + "grad_norm": 3.4360809326171875, + "learning_rate": 9.95674457156078e-05, + "loss": 0.296610689163208, + "step": 420 + }, + { + "epoch": 1.463373083475298, + "grad_norm": 2.8792216777801514, + "learning_rate": 9.954658398307255e-05, + "loss": 0.27008695602416993, + "step": 430 + }, + { + "epoch": 1.4974446337308347, + "grad_norm": 3.3919405937194824, + "learning_rate": 9.952523324575613e-05, + "loss": 0.2359508752822876, + "step": 440 + }, + { + "epoch": 1.5315161839863713, + "grad_norm": 2.511129856109619, + "learning_rate": 9.950339371438172e-05, + "loss": 0.3027271032333374, + "step": 450 + }, + { + "epoch": 1.565587734241908, + "grad_norm": 2.4446396827697754, + "learning_rate": 9.948106560449668e-05, + "loss": 0.26052088737487794, + "step": 460 + }, + { + "epoch": 1.5996592844974447, + "grad_norm": 1.984153151512146, + "learning_rate": 9.945824913647044e-05, + "loss": 0.30088469982147215, + "step": 470 + }, + { + "epoch": 1.6337308347529813, + "grad_norm": 2.9256038665771484, + "learning_rate": 9.943494453549231e-05, + "loss": 0.2888188362121582, + "step": 480 + }, + { + "epoch": 1.667802385008518, + "grad_norm": 2.2333827018737793, + "learning_rate": 9.941115203156933e-05, + "loss": 0.288084602355957, + "step": 490 + }, + { + "epoch": 1.7018739352640546, + "grad_norm": 2.226630210876465, + "learning_rate": 9.93868718595239e-05, + "loss": 0.2429983139038086, + "step": 500 + }, + { + "epoch": 1.7359454855195913, + "grad_norm": 2.5038540363311768, + "learning_rate": 9.936210425899146e-05, + "loss": 0.29560184478759766, + "step": 510 + }, + { + "epoch": 1.770017035775128, + "grad_norm": 1.9872697591781616, + "learning_rate": 9.93368494744183e-05, + "loss": 0.24810783863067626, + "step": 520 + }, + { + "epoch": 1.8040885860306644, + "grad_norm": 2.762561798095703, + "learning_rate": 9.931110775505892e-05, + "loss": 0.2591356039047241, + "step": 530 + }, + { + "epoch": 1.838160136286201, + "grad_norm": 1.9225414991378784, + "learning_rate": 9.928487935497369e-05, + "loss": 0.2622621774673462, + "step": 540 + }, + { + "epoch": 1.8722316865417377, + "grad_norm": 3.4131791591644287, + "learning_rate": 9.925816453302634e-05, + "loss": 0.3099789619445801, + "step": 550 + }, + { + "epoch": 1.9063032367972743, + "grad_norm": 2.7211992740631104, + "learning_rate": 9.923096355288137e-05, + "loss": 0.3053356409072876, + "step": 560 + }, + { + "epoch": 1.9403747870528107, + "grad_norm": 1.9942728281021118, + "learning_rate": 9.92032766830015e-05, + "loss": 0.271621561050415, + "step": 570 + }, + { + "epoch": 1.9744463373083474, + "grad_norm": 1.9207346439361572, + "learning_rate": 9.917510419664491e-05, + "loss": 0.27333619594573977, + "step": 580 + }, + { + "epoch": 2.0068143100511073, + "grad_norm": 2.055387258529663, + "learning_rate": 9.914644637186269e-05, + "loss": 0.20745866298675536, + "step": 590 + }, + { + "epoch": 2.040885860306644, + "grad_norm": 1.4524292945861816, + "learning_rate": 9.911730349149602e-05, + "loss": 0.13981777429580688, + "step": 600 + }, + { + "epoch": 2.0749574105621806, + "grad_norm": 2.8865458965301514, + "learning_rate": 9.908767584317333e-05, + "loss": 0.11298246383666992, + "step": 610 + }, + { + "epoch": 2.1090289608177173, + "grad_norm": 1.1985541582107544, + "learning_rate": 9.905756371930757e-05, + "loss": 0.11284482479095459, + "step": 620 + }, + { + "epoch": 2.143100511073254, + "grad_norm": 1.981881856918335, + "learning_rate": 9.902696741709324e-05, + "loss": 0.13170770406723023, + "step": 630 + }, + { + "epoch": 2.1771720613287906, + "grad_norm": 1.8628884553909302, + "learning_rate": 9.899588723850349e-05, + "loss": 0.14049352407455445, + "step": 640 + }, + { + "epoch": 2.2112436115843273, + "grad_norm": 1.4338645935058594, + "learning_rate": 9.896432349028715e-05, + "loss": 0.1263017773628235, + "step": 650 + }, + { + "epoch": 2.245315161839864, + "grad_norm": 1.722861886024475, + "learning_rate": 9.893227648396564e-05, + "loss": 0.12233226299285889, + "step": 660 + }, + { + "epoch": 2.2793867120954, + "grad_norm": 1.607896089553833, + "learning_rate": 9.889974653583e-05, + "loss": 0.13087828159332277, + "step": 670 + }, + { + "epoch": 2.313458262350937, + "grad_norm": 2.036442756652832, + "learning_rate": 9.886673396693768e-05, + "loss": 0.12004598379135131, + "step": 680 + }, + { + "epoch": 2.3475298126064734, + "grad_norm": 1.2492954730987549, + "learning_rate": 9.88332391031094e-05, + "loss": 0.09768519401550294, + "step": 690 + }, + { + "epoch": 2.38160136286201, + "grad_norm": 1.5268019437789917, + "learning_rate": 9.879926227492595e-05, + "loss": 0.11229777336120605, + "step": 700 + }, + { + "epoch": 2.4156729131175467, + "grad_norm": 2.206562042236328, + "learning_rate": 9.876480381772491e-05, + "loss": 0.12867317199707032, + "step": 710 + }, + { + "epoch": 2.4497444633730834, + "grad_norm": 1.8609817028045654, + "learning_rate": 9.872986407159732e-05, + "loss": 0.12029577493667602, + "step": 720 + }, + { + "epoch": 2.48381601362862, + "grad_norm": 1.9563343524932861, + "learning_rate": 9.869444338138441e-05, + "loss": 0.1118655800819397, + "step": 730 + }, + { + "epoch": 2.5178875638841567, + "grad_norm": 2.2154388427734375, + "learning_rate": 9.865854209667404e-05, + "loss": 0.12033280134201049, + "step": 740 + }, + { + "epoch": 2.5519591141396933, + "grad_norm": 1.3538248538970947, + "learning_rate": 9.862216057179745e-05, + "loss": 0.1174347996711731, + "step": 750 + }, + { + "epoch": 2.58603066439523, + "grad_norm": 4.028043746948242, + "learning_rate": 9.858529916582557e-05, + "loss": 0.12441229820251465, + "step": 760 + }, + { + "epoch": 2.6201022146507666, + "grad_norm": 2.651259422302246, + "learning_rate": 9.85479582425656e-05, + "loss": 0.122542405128479, + "step": 770 + }, + { + "epoch": 2.6541737649063033, + "grad_norm": 1.9554541110992432, + "learning_rate": 9.851013817055737e-05, + "loss": 0.1307147264480591, + "step": 780 + }, + { + "epoch": 2.68824531516184, + "grad_norm": 1.90165376663208, + "learning_rate": 9.847183932306973e-05, + "loss": 0.1521606206893921, + "step": 790 + }, + { + "epoch": 2.7223168654173766, + "grad_norm": 2.0774049758911133, + "learning_rate": 9.843306207809687e-05, + "loss": 0.12136094570159912, + "step": 800 + }, + { + "epoch": 2.7563884156729133, + "grad_norm": 1.1312235593795776, + "learning_rate": 9.83938068183545e-05, + "loss": 0.11534208059310913, + "step": 810 + }, + { + "epoch": 2.7904599659284495, + "grad_norm": 2.1498188972473145, + "learning_rate": 9.835407393127621e-05, + "loss": 0.14543124437332153, + "step": 820 + }, + { + "epoch": 2.8245315161839866, + "grad_norm": 1.941612720489502, + "learning_rate": 9.831386380900953e-05, + "loss": 0.13850215673446656, + "step": 830 + }, + { + "epoch": 2.8586030664395228, + "grad_norm": 1.6838805675506592, + "learning_rate": 9.827317684841216e-05, + "loss": 0.11665056943893433, + "step": 840 + }, + { + "epoch": 2.89267461669506, + "grad_norm": 1.8915469646453857, + "learning_rate": 9.823201345104798e-05, + "loss": 0.12856583595275878, + "step": 850 + }, + { + "epoch": 2.926746166950596, + "grad_norm": 2.020820140838623, + "learning_rate": 9.819037402318308e-05, + "loss": 0.14009920358657837, + "step": 860 + }, + { + "epoch": 2.9608177172061327, + "grad_norm": 1.9529690742492676, + "learning_rate": 9.814825897578179e-05, + "loss": 0.13464512825012206, + "step": 870 + }, + { + "epoch": 2.9948892674616694, + "grad_norm": 3.5946202278137207, + "learning_rate": 9.810566872450269e-05, + "loss": 0.1252979040145874, + "step": 880 + }, + { + "epoch": 3.0272572402044293, + "grad_norm": 1.897910237312317, + "learning_rate": 9.806260368969429e-05, + "loss": 0.08018935918807983, + "step": 890 + }, + { + "epoch": 3.061328790459966, + "grad_norm": 1.5855873823165894, + "learning_rate": 9.801906429639114e-05, + "loss": 0.07474420666694641, + "step": 900 + }, + { + "epoch": 3.0954003407155026, + "grad_norm": 2.107872724533081, + "learning_rate": 9.797505097430947e-05, + "loss": 0.06437333822250366, + "step": 910 + }, + { + "epoch": 3.1294718909710393, + "grad_norm": 2.684037446975708, + "learning_rate": 9.793056415784298e-05, + "loss": 0.07652322649955749, + "step": 920 + }, + { + "epoch": 3.163543441226576, + "grad_norm": 0.9326214790344238, + "learning_rate": 9.788560428605863e-05, + "loss": 0.06489463448524475, + "step": 930 + }, + { + "epoch": 3.1976149914821126, + "grad_norm": 1.6175618171691895, + "learning_rate": 9.784017180269216e-05, + "loss": 0.0631401777267456, + "step": 940 + }, + { + "epoch": 3.2316865417376492, + "grad_norm": 1.7096298933029175, + "learning_rate": 9.779426715614387e-05, + "loss": 0.05567095875740051, + "step": 950 + }, + { + "epoch": 3.2657580919931855, + "grad_norm": 1.9190527200698853, + "learning_rate": 9.774789079947406e-05, + "loss": 0.054007500410079956, + "step": 960 + }, + { + "epoch": 3.2998296422487225, + "grad_norm": 0.7853822708129883, + "learning_rate": 9.770104319039866e-05, + "loss": 0.06382976174354553, + "step": 970 + }, + { + "epoch": 3.3339011925042588, + "grad_norm": 1.1801362037658691, + "learning_rate": 9.765372479128467e-05, + "loss": 0.0761179804801941, + "step": 980 + }, + { + "epoch": 3.3679727427597954, + "grad_norm": 1.2298702001571655, + "learning_rate": 9.760593606914559e-05, + "loss": 0.07370460033416748, + "step": 990 + }, + { + "epoch": 3.402044293015332, + "grad_norm": 3.23696231842041, + "learning_rate": 9.755767749563678e-05, + "loss": 0.059821367263793945, + "step": 1000 + }, + { + "epoch": 3.4361158432708687, + "grad_norm": 4.531341552734375, + "learning_rate": 9.75089495470509e-05, + "loss": 0.0858765959739685, + "step": 1010 + }, + { + "epoch": 3.4701873935264054, + "grad_norm": 2.426567792892456, + "learning_rate": 9.745975270431313e-05, + "loss": 0.0622145414352417, + "step": 1020 + }, + { + "epoch": 3.504258943781942, + "grad_norm": 1.7279127836227417, + "learning_rate": 9.741008745297646e-05, + "loss": 0.07164235711097718, + "step": 1030 + }, + { + "epoch": 3.5383304940374787, + "grad_norm": 2.53861927986145, + "learning_rate": 9.735995428321685e-05, + "loss": 0.06337780952453613, + "step": 1040 + }, + { + "epoch": 3.5724020442930153, + "grad_norm": 2.6302757263183594, + "learning_rate": 9.730935368982846e-05, + "loss": 0.0805334985256195, + "step": 1050 + }, + { + "epoch": 3.606473594548552, + "grad_norm": 3.420285940170288, + "learning_rate": 9.725828617221868e-05, + "loss": 0.0849171221256256, + "step": 1060 + }, + { + "epoch": 3.6405451448040886, + "grad_norm": 1.9331766366958618, + "learning_rate": 9.720675223440334e-05, + "loss": 0.09144250750541687, + "step": 1070 + }, + { + "epoch": 3.6746166950596253, + "grad_norm": 4.827858924865723, + "learning_rate": 9.715475238500159e-05, + "loss": 0.06713611483573914, + "step": 1080 + }, + { + "epoch": 3.708688245315162, + "grad_norm": 0.8885518312454224, + "learning_rate": 9.710228713723092e-05, + "loss": 0.07812190055847168, + "step": 1090 + }, + { + "epoch": 3.7427597955706986, + "grad_norm": 4.049871921539307, + "learning_rate": 9.70493570089022e-05, + "loss": 0.06180936098098755, + "step": 1100 + }, + { + "epoch": 3.7768313458262353, + "grad_norm": 2.131444215774536, + "learning_rate": 9.699596252241438e-05, + "loss": 0.076810222864151, + "step": 1110 + }, + { + "epoch": 3.810902896081772, + "grad_norm": 1.4954314231872559, + "learning_rate": 9.69421042047495e-05, + "loss": 0.059336304664611816, + "step": 1120 + }, + { + "epoch": 3.844974446337308, + "grad_norm": 1.3157131671905518, + "learning_rate": 9.688778258746743e-05, + "loss": 0.07308164238929749, + "step": 1130 + }, + { + "epoch": 3.879045996592845, + "grad_norm": 1.3763610124588013, + "learning_rate": 9.683299820670056e-05, + "loss": 0.06363306045532227, + "step": 1140 + }, + { + "epoch": 3.9131175468483814, + "grad_norm": 1.484768271446228, + "learning_rate": 9.677775160314865e-05, + "loss": 0.06998052597045898, + "step": 1150 + }, + { + "epoch": 3.947189097103918, + "grad_norm": 1.7339237928390503, + "learning_rate": 9.672204332207336e-05, + "loss": 0.0732116937637329, + "step": 1160 + }, + { + "epoch": 3.9812606473594547, + "grad_norm": 0.8836018443107605, + "learning_rate": 9.666587391329294e-05, + "loss": 0.07932557463645935, + "step": 1170 + }, + { + "epoch": 4.013628620102215, + "grad_norm": 0.7781262993812561, + "learning_rate": 9.660924393117675e-05, + "loss": 0.07557746767997742, + "step": 1180 + }, + { + "epoch": 4.047700170357751, + "grad_norm": 2.6470348834991455, + "learning_rate": 9.655215393463989e-05, + "loss": 0.049822351336479186, + "step": 1190 + }, + { + "epoch": 4.081771720613288, + "grad_norm": 1.581343650817871, + "learning_rate": 9.649460448713755e-05, + "loss": 0.042271080613136294, + "step": 1200 + }, + { + "epoch": 4.115843270868824, + "grad_norm": 0.7867306470870972, + "learning_rate": 9.643659615665957e-05, + "loss": 0.03615880012512207, + "step": 1210 + }, + { + "epoch": 4.149914821124361, + "grad_norm": 1.4775359630584717, + "learning_rate": 9.637812951572471e-05, + "loss": 0.04361388087272644, + "step": 1220 + }, + { + "epoch": 4.1839863713798975, + "grad_norm": 1.9977412223815918, + "learning_rate": 9.631920514137516e-05, + "loss": 0.04011210501194, + "step": 1230 + }, + { + "epoch": 4.218057921635435, + "grad_norm": 1.7829431295394897, + "learning_rate": 9.625982361517068e-05, + "loss": 0.03548445403575897, + "step": 1240 + }, + { + "epoch": 4.252129471890971, + "grad_norm": 1.090541958808899, + "learning_rate": 9.619998552318295e-05, + "loss": 0.04351125359535217, + "step": 1250 + }, + { + "epoch": 4.286201022146508, + "grad_norm": 1.0267326831817627, + "learning_rate": 9.61396914559898e-05, + "loss": 0.04062550365924835, + "step": 1260 + }, + { + "epoch": 4.320272572402044, + "grad_norm": 0.9282559752464294, + "learning_rate": 9.607894200866934e-05, + "loss": 0.03696491420269012, + "step": 1270 + }, + { + "epoch": 4.354344122657581, + "grad_norm": 1.3170775175094604, + "learning_rate": 9.601773778079406e-05, + "loss": 0.042220795154571535, + "step": 1280 + }, + { + "epoch": 4.388415672913117, + "grad_norm": 1.0883209705352783, + "learning_rate": 9.595607937642502e-05, + "loss": 0.04217888712882996, + "step": 1290 + }, + { + "epoch": 4.4224872231686545, + "grad_norm": 1.9638941287994385, + "learning_rate": 9.589396740410573e-05, + "loss": 0.03247036635875702, + "step": 1300 + }, + { + "epoch": 4.456558773424191, + "grad_norm": 1.5573837757110596, + "learning_rate": 9.583140247685631e-05, + "loss": 0.04170369505882263, + "step": 1310 + }, + { + "epoch": 4.490630323679728, + "grad_norm": 1.1391565799713135, + "learning_rate": 9.576838521216734e-05, + "loss": 0.046879589557647705, + "step": 1320 + }, + { + "epoch": 4.524701873935264, + "grad_norm": 2.0404841899871826, + "learning_rate": 9.570491623199376e-05, + "loss": 0.05514304637908936, + "step": 1330 + }, + { + "epoch": 4.5587734241908, + "grad_norm": 1.531995415687561, + "learning_rate": 9.564099616274879e-05, + "loss": 0.04663235545158386, + "step": 1340 + }, + { + "epoch": 4.592844974446337, + "grad_norm": 2.1684346199035645, + "learning_rate": 9.557662563529771e-05, + "loss": 0.036445483565330505, + "step": 1350 + }, + { + "epoch": 4.626916524701874, + "grad_norm": 1.308480143547058, + "learning_rate": 9.551180528495165e-05, + "loss": 0.0443327784538269, + "step": 1360 + }, + { + "epoch": 4.660988074957411, + "grad_norm": 1.3180073499679565, + "learning_rate": 9.544653575146127e-05, + "loss": 0.048572239279747007, + "step": 1370 + }, + { + "epoch": 4.695059625212947, + "grad_norm": 2.428232192993164, + "learning_rate": 9.538081767901054e-05, + "loss": 0.04644596576690674, + "step": 1380 + }, + { + "epoch": 4.729131175468484, + "grad_norm": 1.2626821994781494, + "learning_rate": 9.53146517162103e-05, + "loss": 0.05016099214553833, + "step": 1390 + }, + { + "epoch": 4.76320272572402, + "grad_norm": 1.386539340019226, + "learning_rate": 9.524803851609188e-05, + "loss": 0.03919519484043121, + "step": 1400 + }, + { + "epoch": 4.797274275979557, + "grad_norm": 1.5076797008514404, + "learning_rate": 9.518097873610068e-05, + "loss": 0.050628775358200075, + "step": 1410 + }, + { + "epoch": 4.8313458262350935, + "grad_norm": 1.275985598564148, + "learning_rate": 9.511347303808965e-05, + "loss": 0.04660462737083435, + "step": 1420 + }, + { + "epoch": 4.8654173764906306, + "grad_norm": 1.569631814956665, + "learning_rate": 9.504552208831279e-05, + "loss": 0.03862286806106567, + "step": 1430 + }, + { + "epoch": 4.899488926746167, + "grad_norm": 2.065896511077881, + "learning_rate": 9.497712655741851e-05, + "loss": 0.044455283880233766, + "step": 1440 + }, + { + "epoch": 4.933560477001704, + "grad_norm": 1.1074596643447876, + "learning_rate": 9.490828712044312e-05, + "loss": 0.04773740172386169, + "step": 1450 + }, + { + "epoch": 4.96763202725724, + "grad_norm": 1.4310007095336914, + "learning_rate": 9.483900445680407e-05, + "loss": 0.04274417459964752, + "step": 1460 + }, + { + "epoch": 5.0, + "grad_norm": 4.064968585968018, + "learning_rate": 9.476927925029326e-05, + "loss": 0.05169486999511719, + "step": 1470 + }, + { + "epoch": 5.034071550255536, + "grad_norm": 1.5751832723617554, + "learning_rate": 9.469911218907035e-05, + "loss": 0.029029345512390135, + "step": 1480 + }, + { + "epoch": 5.068143100511073, + "grad_norm": 0.6263830065727234, + "learning_rate": 9.462850396565588e-05, + "loss": 0.028271237015724184, + "step": 1490 + }, + { + "epoch": 5.1022146507666095, + "grad_norm": 0.6697548627853394, + "learning_rate": 9.455745527692454e-05, + "loss": 0.026191535592079162, + "step": 1500 + }, + { + "epoch": 5.136286201022147, + "grad_norm": 3.957118034362793, + "learning_rate": 9.448596682409819e-05, + "loss": 0.025492912530899046, + "step": 1510 + }, + { + "epoch": 5.170357751277683, + "grad_norm": 1.7963947057724, + "learning_rate": 9.441403931273897e-05, + "loss": 0.025594806671142577, + "step": 1520 + }, + { + "epoch": 5.20442930153322, + "grad_norm": 1.0430747270584106, + "learning_rate": 9.434167345274244e-05, + "loss": 0.025667831301689148, + "step": 1530 + }, + { + "epoch": 5.238500851788756, + "grad_norm": 2.3541901111602783, + "learning_rate": 9.426886995833038e-05, + "loss": 0.02917010188102722, + "step": 1540 + }, + { + "epoch": 5.272572402044293, + "grad_norm": 0.6873838305473328, + "learning_rate": 9.419562954804388e-05, + "loss": 0.02541271448135376, + "step": 1550 + }, + { + "epoch": 5.306643952299829, + "grad_norm": 0.6828526258468628, + "learning_rate": 9.412195294473625e-05, + "loss": 0.029961639642715455, + "step": 1560 + }, + { + "epoch": 5.3407155025553665, + "grad_norm": 0.8544797897338867, + "learning_rate": 9.404784087556582e-05, + "loss": 0.020680305361747742, + "step": 1570 + }, + { + "epoch": 5.374787052810903, + "grad_norm": 4.307332515716553, + "learning_rate": 9.397329407198875e-05, + "loss": 0.0257159948348999, + "step": 1580 + }, + { + "epoch": 5.40885860306644, + "grad_norm": 0.46097907423973083, + "learning_rate": 9.389831326975196e-05, + "loss": 0.027997365593910216, + "step": 1590 + }, + { + "epoch": 5.442930153321976, + "grad_norm": 1.3980218172073364, + "learning_rate": 9.382289920888565e-05, + "loss": 0.03344095945358276, + "step": 1600 + }, + { + "epoch": 5.477001703577513, + "grad_norm": 0.8466151356697083, + "learning_rate": 9.374705263369617e-05, + "loss": 0.02301342785358429, + "step": 1610 + }, + { + "epoch": 5.511073253833049, + "grad_norm": 1.0830349922180176, + "learning_rate": 9.367077429275859e-05, + "loss": 0.03395429253578186, + "step": 1620 + }, + { + "epoch": 5.5451448040885865, + "grad_norm": 0.4795171618461609, + "learning_rate": 9.359406493890934e-05, + "loss": 0.0325812041759491, + "step": 1630 + }, + { + "epoch": 5.579216354344123, + "grad_norm": 0.9528490900993347, + "learning_rate": 9.351692532923879e-05, + "loss": 0.02982507348060608, + "step": 1640 + }, + { + "epoch": 5.613287904599659, + "grad_norm": 2.5168068408966064, + "learning_rate": 9.343935622508373e-05, + "loss": 0.03338657021522522, + "step": 1650 + }, + { + "epoch": 5.647359454855196, + "grad_norm": 0.9510769844055176, + "learning_rate": 9.336135839201992e-05, + "loss": 0.029187119007110594, + "step": 1660 + }, + { + "epoch": 5.681431005110732, + "grad_norm": 1.0623546838760376, + "learning_rate": 9.328293259985446e-05, + "loss": 0.03202317357063293, + "step": 1670 + }, + { + "epoch": 5.715502555366269, + "grad_norm": 0.6826435327529907, + "learning_rate": 9.320407962261826e-05, + "loss": 0.030450505018234254, + "step": 1680 + }, + { + "epoch": 5.7495741056218055, + "grad_norm": 0.5522363185882568, + "learning_rate": 9.312480023855836e-05, + "loss": 0.021564315259456634, + "step": 1690 + }, + { + "epoch": 5.783645655877343, + "grad_norm": 1.7273434400558472, + "learning_rate": 9.304509523013031e-05, + "loss": 0.028131285309791566, + "step": 1700 + }, + { + "epoch": 5.817717206132879, + "grad_norm": 1.4397395849227905, + "learning_rate": 9.296496538399036e-05, + "loss": 0.029852986335754395, + "step": 1710 + }, + { + "epoch": 5.851788756388416, + "grad_norm": 1.4878379106521606, + "learning_rate": 9.288441149098773e-05, + "loss": 0.027045699954032897, + "step": 1720 + }, + { + "epoch": 5.885860306643952, + "grad_norm": 0.9599165320396423, + "learning_rate": 9.280343434615679e-05, + "loss": 0.029828810691833497, + "step": 1730 + }, + { + "epoch": 5.919931856899489, + "grad_norm": 0.36500489711761475, + "learning_rate": 9.272203474870932e-05, + "loss": 0.03575279414653778, + "step": 1740 + }, + { + "epoch": 5.954003407155025, + "grad_norm": 0.5604597926139832, + "learning_rate": 9.264021350202646e-05, + "loss": 0.03006734549999237, + "step": 1750 + }, + { + "epoch": 5.9880749574105625, + "grad_norm": 1.697237253189087, + "learning_rate": 9.25579714136509e-05, + "loss": 0.0326397567987442, + "step": 1760 + }, + { + "epoch": 6.0204429301533215, + "grad_norm": 0.3036922216415405, + "learning_rate": 9.247530929527881e-05, + "loss": 0.03297402858734131, + "step": 1770 + }, + { + "epoch": 6.054514480408859, + "grad_norm": 0.7871415019035339, + "learning_rate": 9.239222796275195e-05, + "loss": 0.02531479299068451, + "step": 1780 + }, + { + "epoch": 6.088586030664395, + "grad_norm": 0.6982287764549255, + "learning_rate": 9.230872823604956e-05, + "loss": 0.027960145473480226, + "step": 1790 + }, + { + "epoch": 6.122657580919932, + "grad_norm": 2.9335663318634033, + "learning_rate": 9.22248109392802e-05, + "loss": 0.019796872138977052, + "step": 1800 + }, + { + "epoch": 6.156729131175468, + "grad_norm": 0.914587676525116, + "learning_rate": 9.214047690067369e-05, + "loss": 0.01856117695569992, + "step": 1810 + }, + { + "epoch": 6.190800681431005, + "grad_norm": 1.0381855964660645, + "learning_rate": 9.205572695257298e-05, + "loss": 0.023674681782722473, + "step": 1820 + }, + { + "epoch": 6.2248722316865415, + "grad_norm": 0.7270963788032532, + "learning_rate": 9.197056193142583e-05, + "loss": 0.02533746659755707, + "step": 1830 + }, + { + "epoch": 6.258943781942079, + "grad_norm": 0.6352244019508362, + "learning_rate": 9.188498267777664e-05, + "loss": 0.02358451783657074, + "step": 1840 + }, + { + "epoch": 6.293015332197615, + "grad_norm": 1.1831856966018677, + "learning_rate": 9.179899003625804e-05, + "loss": 0.018904535472393035, + "step": 1850 + }, + { + "epoch": 6.327086882453152, + "grad_norm": 0.8320885896682739, + "learning_rate": 9.171258485558274e-05, + "loss": 0.020784467458724976, + "step": 1860 + }, + { + "epoch": 6.361158432708688, + "grad_norm": 0.3053925335407257, + "learning_rate": 9.162576798853497e-05, + "loss": 0.021431468427181244, + "step": 1870 + }, + { + "epoch": 6.395229982964225, + "grad_norm": 1.034461498260498, + "learning_rate": 9.153854029196213e-05, + "loss": 0.02396380603313446, + "step": 1880 + }, + { + "epoch": 6.429301533219761, + "grad_norm": 2.365105390548706, + "learning_rate": 9.145090262676642e-05, + "loss": 0.01987527161836624, + "step": 1890 + }, + { + "epoch": 6.4633730834752985, + "grad_norm": 1.0095535516738892, + "learning_rate": 9.136285585789619e-05, + "loss": 0.01691812127828598, + "step": 1900 + }, + { + "epoch": 6.497444633730835, + "grad_norm": 0.9776434302330017, + "learning_rate": 9.12744008543375e-05, + "loss": 0.022355957329273222, + "step": 1910 + }, + { + "epoch": 6.531516183986371, + "grad_norm": 1.4393912553787231, + "learning_rate": 9.118553848910552e-05, + "loss": 0.016300460696220397, + "step": 1920 + }, + { + "epoch": 6.565587734241908, + "grad_norm": 1.2278251647949219, + "learning_rate": 9.109626963923592e-05, + "loss": 0.02234429270029068, + "step": 1930 + }, + { + "epoch": 6.599659284497445, + "grad_norm": 1.0951917171478271, + "learning_rate": 9.100659518577624e-05, + "loss": 0.028519749641418457, + "step": 1940 + }, + { + "epoch": 6.633730834752981, + "grad_norm": 0.7138704061508179, + "learning_rate": 9.091651601377709e-05, + "loss": 0.018571604788303376, + "step": 1950 + }, + { + "epoch": 6.6678023850085175, + "grad_norm": 1.1325918436050415, + "learning_rate": 9.082603301228354e-05, + "loss": 0.024752214550971985, + "step": 1960 + }, + { + "epoch": 6.701873935264055, + "grad_norm": 0.9077216982841492, + "learning_rate": 9.073514707432631e-05, + "loss": 0.03127498924732208, + "step": 1970 + }, + { + "epoch": 6.735945485519591, + "grad_norm": 1.4316587448120117, + "learning_rate": 9.06438590969129e-05, + "loss": 0.025211939215660097, + "step": 1980 + }, + { + "epoch": 6.770017035775128, + "grad_norm": 1.0027835369110107, + "learning_rate": 9.055216998101879e-05, + "loss": 0.01745348423719406, + "step": 1990 + }, + { + "epoch": 6.804088586030664, + "grad_norm": 1.3488839864730835, + "learning_rate": 9.046008063157855e-05, + "loss": 0.022722427546977998, + "step": 2000 + }, + { + "epoch": 6.838160136286201, + "grad_norm": 0.9204393029212952, + "learning_rate": 9.036759195747686e-05, + "loss": 0.030331790447235107, + "step": 2010 + }, + { + "epoch": 6.872231686541737, + "grad_norm": 0.519404411315918, + "learning_rate": 9.027470487153962e-05, + "loss": 0.025021228194236755, + "step": 2020 + }, + { + "epoch": 6.9063032367972745, + "grad_norm": 1.0315072536468506, + "learning_rate": 9.018142029052483e-05, + "loss": 0.0225812166929245, + "step": 2030 + }, + { + "epoch": 6.940374787052811, + "grad_norm": 2.8452701568603516, + "learning_rate": 9.008773913511369e-05, + "loss": 0.028797915577888487, + "step": 2040 + }, + { + "epoch": 6.974446337308348, + "grad_norm": 0.726943850517273, + "learning_rate": 8.999366232990136e-05, + "loss": 0.021427948772907258, + "step": 2050 + }, + { + "epoch": 7.006814310051107, + "grad_norm": 1.4920276403427124, + "learning_rate": 8.989919080338792e-05, + "loss": 0.02137400358915329, + "step": 2060 + }, + { + "epoch": 7.040885860306644, + "grad_norm": 2.463482141494751, + "learning_rate": 8.98043254879692e-05, + "loss": 0.01812605857849121, + "step": 2070 + }, + { + "epoch": 7.07495741056218, + "grad_norm": 0.8011264801025391, + "learning_rate": 8.970906731992759e-05, + "loss": 0.02192564457654953, + "step": 2080 + }, + { + "epoch": 7.109028960817717, + "grad_norm": 0.16289736330509186, + "learning_rate": 8.961341723942271e-05, + "loss": 0.014302457869052886, + "step": 2090 + }, + { + "epoch": 7.1431005110732535, + "grad_norm": 0.5535656809806824, + "learning_rate": 8.951737619048228e-05, + "loss": 0.015091146528720855, + "step": 2100 + }, + { + "epoch": 7.177172061328791, + "grad_norm": 0.16370069980621338, + "learning_rate": 8.942094512099264e-05, + "loss": 0.016434633731842042, + "step": 2110 + }, + { + "epoch": 7.211243611584327, + "grad_norm": 1.4876139163970947, + "learning_rate": 8.932412498268955e-05, + "loss": 0.011326295137405396, + "step": 2120 + }, + { + "epoch": 7.245315161839864, + "grad_norm": 0.654828667640686, + "learning_rate": 8.922691673114866e-05, + "loss": 0.015352894365787507, + "step": 2130 + }, + { + "epoch": 7.2793867120954, + "grad_norm": 1.1809929609298706, + "learning_rate": 8.912932132577619e-05, + "loss": 0.017652250826358795, + "step": 2140 + }, + { + "epoch": 7.313458262350937, + "grad_norm": 0.6106181740760803, + "learning_rate": 8.903133972979938e-05, + "loss": 0.013139715790748597, + "step": 2150 + }, + { + "epoch": 7.347529812606473, + "grad_norm": 0.6049255728721619, + "learning_rate": 8.893297291025703e-05, + "loss": 0.012686268985271454, + "step": 2160 + }, + { + "epoch": 7.3816013628620105, + "grad_norm": 1.0129069089889526, + "learning_rate": 8.883422183798992e-05, + "loss": 0.016695281863212584, + "step": 2170 + }, + { + "epoch": 7.415672913117547, + "grad_norm": 1.217993140220642, + "learning_rate": 8.873508748763129e-05, + "loss": 0.013319727778434754, + "step": 2180 + }, + { + "epoch": 7.449744463373084, + "grad_norm": 0.5124009847640991, + "learning_rate": 8.863557083759714e-05, + "loss": 0.011351729929447173, + "step": 2190 + }, + { + "epoch": 7.48381601362862, + "grad_norm": 0.9370232224464417, + "learning_rate": 8.853567287007665e-05, + "loss": 0.016303011775016786, + "step": 2200 + }, + { + "epoch": 7.517887563884157, + "grad_norm": 0.5177021622657776, + "learning_rate": 8.843539457102238e-05, + "loss": 0.013715451955795288, + "step": 2210 + }, + { + "epoch": 7.551959114139693, + "grad_norm": 1.2320998907089233, + "learning_rate": 8.833473693014071e-05, + "loss": 0.01952739357948303, + "step": 2220 + }, + { + "epoch": 7.5860306643952296, + "grad_norm": 1.3520736694335938, + "learning_rate": 8.823370094088191e-05, + "loss": 0.018738554418087007, + "step": 2230 + }, + { + "epoch": 7.620102214650767, + "grad_norm": 1.1438735723495483, + "learning_rate": 8.813228760043037e-05, + "loss": 0.012850122153759002, + "step": 2240 + }, + { + "epoch": 7.654173764906303, + "grad_norm": 1.3085774183273315, + "learning_rate": 8.803049790969486e-05, + "loss": 0.022417041659355163, + "step": 2250 + }, + { + "epoch": 7.68824531516184, + "grad_norm": 0.6747460961341858, + "learning_rate": 8.792833287329851e-05, + "loss": 0.015353672206401825, + "step": 2260 + }, + { + "epoch": 7.722316865417376, + "grad_norm": 0.24661444127559662, + "learning_rate": 8.782579349956897e-05, + "loss": 0.015990708768367768, + "step": 2270 + }, + { + "epoch": 7.756388415672913, + "grad_norm": 0.6662095189094543, + "learning_rate": 8.77228808005285e-05, + "loss": 0.017545297741889954, + "step": 2280 + }, + { + "epoch": 7.7904599659284495, + "grad_norm": 0.3242599070072174, + "learning_rate": 8.761959579188386e-05, + "loss": 0.01379595249891281, + "step": 2290 + }, + { + "epoch": 7.824531516183987, + "grad_norm": 0.45924779772758484, + "learning_rate": 8.75159394930164e-05, + "loss": 0.013600316643714905, + "step": 2300 + }, + { + "epoch": 7.858603066439523, + "grad_norm": 0.5505822896957397, + "learning_rate": 8.741191292697195e-05, + "loss": 0.01868756115436554, + "step": 2310 + }, + { + "epoch": 7.89267461669506, + "grad_norm": 1.0311226844787598, + "learning_rate": 8.73075171204507e-05, + "loss": 0.013703931868076325, + "step": 2320 + }, + { + "epoch": 7.926746166950596, + "grad_norm": 1.0836596488952637, + "learning_rate": 8.720275310379713e-05, + "loss": 0.015289519727230073, + "step": 2330 + }, + { + "epoch": 7.960817717206133, + "grad_norm": 0.6845241189002991, + "learning_rate": 8.70976219109898e-05, + "loss": 0.017059548199176787, + "step": 2340 + }, + { + "epoch": 7.994889267461669, + "grad_norm": 1.0206208229064941, + "learning_rate": 8.699212457963111e-05, + "loss": 0.020337219536304473, + "step": 2350 + }, + { + "epoch": 8.02725724020443, + "grad_norm": 1.043688178062439, + "learning_rate": 8.688626215093714e-05, + "loss": 0.011061477661132812, + "step": 2360 + }, + { + "epoch": 8.061328790459966, + "grad_norm": 1.5376160144805908, + "learning_rate": 8.678003566972735e-05, + "loss": 0.0165196493268013, + "step": 2370 + }, + { + "epoch": 8.095400340715502, + "grad_norm": 0.8935771584510803, + "learning_rate": 8.667344618441418e-05, + "loss": 0.01602596491575241, + "step": 2380 + }, + { + "epoch": 8.12947189097104, + "grad_norm": 0.6318204998970032, + "learning_rate": 8.656649474699283e-05, + "loss": 0.011065221577882766, + "step": 2390 + }, + { + "epoch": 8.163543441226576, + "grad_norm": 0.9502279162406921, + "learning_rate": 8.645918241303084e-05, + "loss": 0.014841337502002717, + "step": 2400 + }, + { + "epoch": 8.197614991482112, + "grad_norm": 0.3303418755531311, + "learning_rate": 8.635151024165759e-05, + "loss": 0.011852187663316726, + "step": 2410 + }, + { + "epoch": 8.231686541737648, + "grad_norm": 0.361967533826828, + "learning_rate": 8.624347929555394e-05, + "loss": 0.015762053430080414, + "step": 2420 + }, + { + "epoch": 8.265758091993186, + "grad_norm": 0.6196819543838501, + "learning_rate": 8.613509064094175e-05, + "loss": 0.009194750338792801, + "step": 2430 + }, + { + "epoch": 8.299829642248723, + "grad_norm": 1.6100952625274658, + "learning_rate": 8.602634534757326e-05, + "loss": 0.014374491572380067, + "step": 2440 + }, + { + "epoch": 8.333901192504259, + "grad_norm": 0.3869682848453522, + "learning_rate": 8.591724448872061e-05, + "loss": 0.012869016826152801, + "step": 2450 + }, + { + "epoch": 8.367972742759795, + "grad_norm": 0.7329438924789429, + "learning_rate": 8.580778914116522e-05, + "loss": 0.009707724303007126, + "step": 2460 + }, + { + "epoch": 8.402044293015333, + "grad_norm": 0.527068555355072, + "learning_rate": 8.569798038518719e-05, + "loss": 0.013562756776809692, + "step": 2470 + }, + { + "epoch": 8.43611584327087, + "grad_norm": 0.21004585921764374, + "learning_rate": 8.558781930455464e-05, + "loss": 0.010475656390190125, + "step": 2480 + }, + { + "epoch": 8.470187393526405, + "grad_norm": 0.5486114025115967, + "learning_rate": 8.547730698651293e-05, + "loss": 0.011294472217559814, + "step": 2490 + }, + { + "epoch": 8.504258943781942, + "grad_norm": 0.8511176705360413, + "learning_rate": 8.536644452177404e-05, + "loss": 0.012653800845146179, + "step": 2500 + }, + { + "epoch": 8.538330494037478, + "grad_norm": 0.47672024369239807, + "learning_rate": 8.525523300450571e-05, + "loss": 0.011530681699514388, + "step": 2510 + }, + { + "epoch": 8.572402044293016, + "grad_norm": 0.327532023191452, + "learning_rate": 8.514367353232076e-05, + "loss": 0.01135161817073822, + "step": 2520 + }, + { + "epoch": 8.606473594548552, + "grad_norm": 0.11722666770219803, + "learning_rate": 8.50317672062661e-05, + "loss": 0.007618208974599838, + "step": 2530 + }, + { + "epoch": 8.640545144804088, + "grad_norm": 1.409010887145996, + "learning_rate": 8.491951513081203e-05, + "loss": 0.01017497032880783, + "step": 2540 + }, + { + "epoch": 8.674616695059624, + "grad_norm": 0.6355599164962769, + "learning_rate": 8.480691841384118e-05, + "loss": 0.01415388435125351, + "step": 2550 + }, + { + "epoch": 8.708688245315162, + "grad_norm": 0.22371913492679596, + "learning_rate": 8.469397816663773e-05, + "loss": 0.011060314625501633, + "step": 2560 + }, + { + "epoch": 8.742759795570699, + "grad_norm": 1.234665036201477, + "learning_rate": 8.458069550387629e-05, + "loss": 0.01573746055364609, + "step": 2570 + }, + { + "epoch": 8.776831345826235, + "grad_norm": 0.4359232187271118, + "learning_rate": 8.446707154361102e-05, + "loss": 0.014107954502105714, + "step": 2580 + }, + { + "epoch": 8.810902896081771, + "grad_norm": 0.341754287481308, + "learning_rate": 8.435310740726452e-05, + "loss": 0.010107460618019103, + "step": 2590 + }, + { + "epoch": 8.844974446337309, + "grad_norm": 1.1128056049346924, + "learning_rate": 8.423880421961683e-05, + "loss": 0.0148219034075737, + "step": 2600 + }, + { + "epoch": 8.879045996592845, + "grad_norm": 0.6765305995941162, + "learning_rate": 8.412416310879426e-05, + "loss": 0.01568043977022171, + "step": 2610 + }, + { + "epoch": 8.913117546848381, + "grad_norm": 1.98653244972229, + "learning_rate": 8.400918520625829e-05, + "loss": 0.01641627252101898, + "step": 2620 + }, + { + "epoch": 8.947189097103918, + "grad_norm": 1.2335822582244873, + "learning_rate": 8.389387164679439e-05, + "loss": 0.012778672575950622, + "step": 2630 + }, + { + "epoch": 8.981260647359456, + "grad_norm": 0.49869757890701294, + "learning_rate": 8.377822356850084e-05, + "loss": 0.013412460684776306, + "step": 2640 + }, + { + "epoch": 9.013628620102214, + "grad_norm": 0.274105042219162, + "learning_rate": 8.366224211277751e-05, + "loss": 0.015021094679832458, + "step": 2650 + }, + { + "epoch": 9.047700170357752, + "grad_norm": 0.09190282225608826, + "learning_rate": 8.354592842431452e-05, + "loss": 0.009592237323522568, + "step": 2660 + }, + { + "epoch": 9.081771720613288, + "grad_norm": 0.7742070555686951, + "learning_rate": 8.342928365108102e-05, + "loss": 0.010364116728305816, + "step": 2670 + }, + { + "epoch": 9.115843270868824, + "grad_norm": 0.26427945494651794, + "learning_rate": 8.331230894431381e-05, + "loss": 0.009706391394138337, + "step": 2680 + }, + { + "epoch": 9.14991482112436, + "grad_norm": 0.30318769812583923, + "learning_rate": 8.319500545850608e-05, + "loss": 0.007016823440790176, + "step": 2690 + }, + { + "epoch": 9.183986371379898, + "grad_norm": 2.089289665222168, + "learning_rate": 8.30773743513958e-05, + "loss": 0.009532840549945831, + "step": 2700 + }, + { + "epoch": 9.218057921635435, + "grad_norm": 0.4548279941082001, + "learning_rate": 8.295941678395457e-05, + "loss": 0.007914836704730987, + "step": 2710 + }, + { + "epoch": 9.25212947189097, + "grad_norm": 0.7105698585510254, + "learning_rate": 8.284113392037593e-05, + "loss": 0.007088977098464966, + "step": 2720 + }, + { + "epoch": 9.286201022146507, + "grad_norm": 0.48185616731643677, + "learning_rate": 8.272252692806402e-05, + "loss": 0.00796378254890442, + "step": 2730 + }, + { + "epoch": 9.320272572402045, + "grad_norm": 0.24422025680541992, + "learning_rate": 8.260359697762193e-05, + "loss": 0.008733968436717986, + "step": 2740 + }, + { + "epoch": 9.354344122657581, + "grad_norm": 0.4369703233242035, + "learning_rate": 8.248434524284028e-05, + "loss": 0.01090761125087738, + "step": 2750 + }, + { + "epoch": 9.388415672913117, + "grad_norm": 0.10231366008520126, + "learning_rate": 8.236477290068554e-05, + "loss": 0.0058899007737636564, + "step": 2760 + }, + { + "epoch": 9.422487223168654, + "grad_norm": 0.6973859071731567, + "learning_rate": 8.224488113128845e-05, + "loss": 0.010979560017585755, + "step": 2770 + }, + { + "epoch": 9.456558773424192, + "grad_norm": 0.22373375296592712, + "learning_rate": 8.212467111793239e-05, + "loss": 0.01606069803237915, + "step": 2780 + }, + { + "epoch": 9.490630323679728, + "grad_norm": 0.5596587061882019, + "learning_rate": 8.200414404704164e-05, + "loss": 0.005761120095849037, + "step": 2790 + }, + { + "epoch": 9.524701873935264, + "grad_norm": 0.4550517499446869, + "learning_rate": 8.188330110816976e-05, + "loss": 0.007079315185546875, + "step": 2800 + }, + { + "epoch": 9.5587734241908, + "grad_norm": 0.7034183144569397, + "learning_rate": 8.176214349398774e-05, + "loss": 0.009715914726257324, + "step": 2810 + }, + { + "epoch": 9.592844974446336, + "grad_norm": 0.9386186003684998, + "learning_rate": 8.164067240027233e-05, + "loss": 0.007927747070789337, + "step": 2820 + }, + { + "epoch": 9.626916524701874, + "grad_norm": 0.23752354085445404, + "learning_rate": 8.151888902589419e-05, + "loss": 0.009600657969713211, + "step": 2830 + }, + { + "epoch": 9.66098807495741, + "grad_norm": 0.24582235515117645, + "learning_rate": 8.139679457280605e-05, + "loss": 0.015585379302501678, + "step": 2840 + }, + { + "epoch": 9.695059625212947, + "grad_norm": 0.37661418318748474, + "learning_rate": 8.127439024603088e-05, + "loss": 0.0069290541112422945, + "step": 2850 + }, + { + "epoch": 9.729131175468483, + "grad_norm": 0.179131418466568, + "learning_rate": 8.115167725364995e-05, + "loss": 0.010592963546514511, + "step": 2860 + }, + { + "epoch": 9.763202725724021, + "grad_norm": 1.1144113540649414, + "learning_rate": 8.102865680679099e-05, + "loss": 0.013466599583625793, + "step": 2870 + }, + { + "epoch": 9.797274275979557, + "grad_norm": 0.13078942894935608, + "learning_rate": 8.090533011961609e-05, + "loss": 0.009784573316574096, + "step": 2880 + }, + { + "epoch": 9.831345826235093, + "grad_norm": 1.0190367698669434, + "learning_rate": 8.078169840930993e-05, + "loss": 0.007894467562437057, + "step": 2890 + }, + { + "epoch": 9.86541737649063, + "grad_norm": 0.5391379594802856, + "learning_rate": 8.065776289606752e-05, + "loss": 0.00961867868900299, + "step": 2900 + }, + { + "epoch": 9.899488926746168, + "grad_norm": 0.18672017753124237, + "learning_rate": 8.053352480308239e-05, + "loss": 0.014848986268043518, + "step": 2910 + }, + { + "epoch": 9.933560477001704, + "grad_norm": 15.940567016601562, + "learning_rate": 8.040898535653432e-05, + "loss": 0.020153032243251802, + "step": 2920 + }, + { + "epoch": 9.96763202725724, + "grad_norm": 0.9096807241439819, + "learning_rate": 8.028414578557738e-05, + "loss": 0.016014789044857026, + "step": 2930 + }, + { + "epoch": 10.0, + "grad_norm": 0.15251773595809937, + "learning_rate": 8.015900732232775e-05, + "loss": 0.008510295301675797, + "step": 2940 + }, + { + "epoch": 10.034071550255536, + "grad_norm": 4.133975505828857, + "learning_rate": 8.003357120185154e-05, + "loss": 0.011156510561704636, + "step": 2950 + }, + { + "epoch": 10.068143100511072, + "grad_norm": 0.11909753829240799, + "learning_rate": 7.990783866215259e-05, + "loss": 0.0054241478443145755, + "step": 2960 + }, + { + "epoch": 10.10221465076661, + "grad_norm": 1.1266493797302246, + "learning_rate": 7.978181094416033e-05, + "loss": 0.010347043722867965, + "step": 2970 + }, + { + "epoch": 10.136286201022147, + "grad_norm": 0.14523354172706604, + "learning_rate": 7.965548929171746e-05, + "loss": 0.00852731689810753, + "step": 2980 + }, + { + "epoch": 10.170357751277683, + "grad_norm": 0.3608001470565796, + "learning_rate": 7.95288749515677e-05, + "loss": 0.010224699974060059, + "step": 2990 + }, + { + "epoch": 10.204429301533219, + "grad_norm": 0.41857728362083435, + "learning_rate": 7.940196917334343e-05, + "loss": 0.00889056697487831, + "step": 3000 + }, + { + "epoch": 10.238500851788757, + "grad_norm": 0.7375580668449402, + "learning_rate": 7.927477320955351e-05, + "loss": 0.007346871495246887, + "step": 3010 + }, + { + "epoch": 10.272572402044293, + "grad_norm": 0.6988178491592407, + "learning_rate": 7.91472883155707e-05, + "loss": 0.005931505188345909, + "step": 3020 + }, + { + "epoch": 10.30664395229983, + "grad_norm": 0.5991840362548828, + "learning_rate": 7.901951574961943e-05, + "loss": 0.00799979567527771, + "step": 3030 + }, + { + "epoch": 10.340715502555366, + "grad_norm": 0.15230441093444824, + "learning_rate": 7.88914567727634e-05, + "loss": 0.0060288384556770325, + "step": 3040 + }, + { + "epoch": 10.374787052810904, + "grad_norm": 0.20510618388652802, + "learning_rate": 7.876311264889296e-05, + "loss": 0.0052495885640382765, + "step": 3050 + }, + { + "epoch": 10.40885860306644, + "grad_norm": 0.4645010232925415, + "learning_rate": 7.863448464471281e-05, + "loss": 0.007955972850322724, + "step": 3060 + }, + { + "epoch": 10.442930153321976, + "grad_norm": 0.2566268742084503, + "learning_rate": 7.850557402972942e-05, + "loss": 0.015420112013816833, + "step": 3070 + }, + { + "epoch": 10.477001703577512, + "grad_norm": 0.8585833311080933, + "learning_rate": 7.837638207623854e-05, + "loss": 0.013999661803245545, + "step": 3080 + }, + { + "epoch": 10.51107325383305, + "grad_norm": 0.38296154141426086, + "learning_rate": 7.824691005931255e-05, + "loss": 0.006875631958246231, + "step": 3090 + }, + { + "epoch": 10.545144804088586, + "grad_norm": 0.20931589603424072, + "learning_rate": 7.811715925678804e-05, + "loss": 0.008229302614927292, + "step": 3100 + }, + { + "epoch": 10.579216354344123, + "grad_norm": 0.10362900048494339, + "learning_rate": 7.798713094925301e-05, + "loss": 0.007631321996450424, + "step": 3110 + }, + { + "epoch": 10.613287904599659, + "grad_norm": 1.1679847240447998, + "learning_rate": 7.785682642003437e-05, + "loss": 0.01506734937429428, + "step": 3120 + }, + { + "epoch": 10.647359454855195, + "grad_norm": 0.9852617979049683, + "learning_rate": 7.772624695518522e-05, + "loss": 0.006735594570636749, + "step": 3130 + }, + { + "epoch": 10.681431005110733, + "grad_norm": 3.766563892364502, + "learning_rate": 7.759539384347216e-05, + "loss": 0.008183671534061432, + "step": 3140 + }, + { + "epoch": 10.71550255536627, + "grad_norm": 0.19659624993801117, + "learning_rate": 7.746426837636259e-05, + "loss": 0.0077319532632827755, + "step": 3150 + }, + { + "epoch": 10.749574105621805, + "grad_norm": 0.146270751953125, + "learning_rate": 7.733287184801191e-05, + "loss": 0.004887140542268753, + "step": 3160 + }, + { + "epoch": 10.783645655877342, + "grad_norm": 0.9391510486602783, + "learning_rate": 7.720120555525081e-05, + "loss": 0.007096558064222336, + "step": 3170 + }, + { + "epoch": 10.81771720613288, + "grad_norm": 1.5230553150177002, + "learning_rate": 7.706927079757247e-05, + "loss": 0.009223876893520356, + "step": 3180 + }, + { + "epoch": 10.851788756388416, + "grad_norm": 0.21125009655952454, + "learning_rate": 7.693706887711965e-05, + "loss": 0.01302051693201065, + "step": 3190 + }, + { + "epoch": 10.885860306643952, + "grad_norm": 1.1184583902359009, + "learning_rate": 7.680460109867195e-05, + "loss": 0.0068347200751304625, + "step": 3200 + }, + { + "epoch": 10.919931856899488, + "grad_norm": 0.27241578698158264, + "learning_rate": 7.667186876963286e-05, + "loss": 0.008140674978494643, + "step": 3210 + }, + { + "epoch": 10.954003407155026, + "grad_norm": 1.1789417266845703, + "learning_rate": 7.653887320001688e-05, + "loss": 0.009393244981765747, + "step": 3220 + }, + { + "epoch": 10.988074957410563, + "grad_norm": 0.39907607436180115, + "learning_rate": 7.640561570243661e-05, + "loss": 0.008160093426704406, + "step": 3230 + }, + { + "epoch": 11.020442930153322, + "grad_norm": 1.4396462440490723, + "learning_rate": 7.627209759208971e-05, + "loss": 0.008179085701704026, + "step": 3240 + }, + { + "epoch": 11.054514480408859, + "grad_norm": 0.1519806832075119, + "learning_rate": 7.613832018674606e-05, + "loss": 0.006314434856176376, + "step": 3250 + }, + { + "epoch": 11.088586030664395, + "grad_norm": 0.3013201355934143, + "learning_rate": 7.600428480673463e-05, + "loss": 0.0059268161654472355, + "step": 3260 + }, + { + "epoch": 11.122657580919931, + "grad_norm": 1.1412639617919922, + "learning_rate": 7.586999277493053e-05, + "loss": 0.005475656688213348, + "step": 3270 + }, + { + "epoch": 11.156729131175469, + "grad_norm": 0.3565169870853424, + "learning_rate": 7.573544541674187e-05, + "loss": 0.008369363844394684, + "step": 3280 + }, + { + "epoch": 11.190800681431005, + "grad_norm": 0.14332878589630127, + "learning_rate": 7.560064406009677e-05, + "loss": 0.006354317814111709, + "step": 3290 + }, + { + "epoch": 11.224872231686541, + "grad_norm": 0.19285030663013458, + "learning_rate": 7.546559003543019e-05, + "loss": 0.008741730451583862, + "step": 3300 + }, + { + "epoch": 11.258943781942078, + "grad_norm": 0.07420668751001358, + "learning_rate": 7.533028467567083e-05, + "loss": 0.008161399513483047, + "step": 3310 + }, + { + "epoch": 11.293015332197616, + "grad_norm": 0.7485439777374268, + "learning_rate": 7.519472931622797e-05, + "loss": 0.00587712861597538, + "step": 3320 + }, + { + "epoch": 11.327086882453152, + "grad_norm": 0.2415221929550171, + "learning_rate": 7.50589252949783e-05, + "loss": 0.004889140278100968, + "step": 3330 + }, + { + "epoch": 11.361158432708688, + "grad_norm": 1.4422087669372559, + "learning_rate": 7.492287395225265e-05, + "loss": 0.008794192224740982, + "step": 3340 + }, + { + "epoch": 11.395229982964224, + "grad_norm": 1.1440876722335815, + "learning_rate": 7.478657663082287e-05, + "loss": 0.0076603077352046965, + "step": 3350 + }, + { + "epoch": 11.429301533219762, + "grad_norm": 0.6747322678565979, + "learning_rate": 7.465003467588847e-05, + "loss": 0.00787583738565445, + "step": 3360 + }, + { + "epoch": 11.463373083475298, + "grad_norm": 0.11730386316776276, + "learning_rate": 7.451324943506347e-05, + "loss": 0.012481527030467987, + "step": 3370 + }, + { + "epoch": 11.497444633730835, + "grad_norm": 0.17758522927761078, + "learning_rate": 7.437622225836294e-05, + "loss": 0.008745304495096206, + "step": 3380 + }, + { + "epoch": 11.531516183986371, + "grad_norm": 0.6659291982650757, + "learning_rate": 7.423895449818979e-05, + "loss": 0.009991095960140228, + "step": 3390 + }, + { + "epoch": 11.565587734241909, + "grad_norm": 0.12057994306087494, + "learning_rate": 7.41014475093214e-05, + "loss": 0.006519577652215958, + "step": 3400 + }, + { + "epoch": 11.599659284497445, + "grad_norm": 0.27850911021232605, + "learning_rate": 7.396370264889624e-05, + "loss": 0.005896145477890968, + "step": 3410 + }, + { + "epoch": 11.633730834752981, + "grad_norm": 0.13032549619674683, + "learning_rate": 7.382572127640048e-05, + "loss": 0.004059913754463196, + "step": 3420 + }, + { + "epoch": 11.667802385008518, + "grad_norm": 1.1132233142852783, + "learning_rate": 7.368750475365455e-05, + "loss": 0.009586858004331589, + "step": 3430 + }, + { + "epoch": 11.701873935264054, + "grad_norm": 1.294803261756897, + "learning_rate": 7.354905444479975e-05, + "loss": 0.007687809318304062, + "step": 3440 + }, + { + "epoch": 11.735945485519592, + "grad_norm": 1.8427352905273438, + "learning_rate": 7.34103717162847e-05, + "loss": 0.009158594161272049, + "step": 3450 + }, + { + "epoch": 11.770017035775128, + "grad_norm": 0.9688911437988281, + "learning_rate": 7.327145793685197e-05, + "loss": 0.009636565297842025, + "step": 3460 + }, + { + "epoch": 11.804088586030664, + "grad_norm": 0.3230285346508026, + "learning_rate": 7.313231447752447e-05, + "loss": 0.011190295219421387, + "step": 3470 + }, + { + "epoch": 11.8381601362862, + "grad_norm": 0.581055223941803, + "learning_rate": 7.299294271159196e-05, + "loss": 0.011519234627485275, + "step": 3480 + }, + { + "epoch": 11.872231686541738, + "grad_norm": 1.193308711051941, + "learning_rate": 7.285334401459752e-05, + "loss": 0.008448301255702973, + "step": 3490 + }, + { + "epoch": 11.906303236797275, + "grad_norm": 1.002673625946045, + "learning_rate": 7.271351976432393e-05, + "loss": 0.012346546351909637, + "step": 3500 + }, + { + "epoch": 11.94037478705281, + "grad_norm": 0.2775571048259735, + "learning_rate": 7.257347134078007e-05, + "loss": 0.007344400137662887, + "step": 3510 + }, + { + "epoch": 11.974446337308347, + "grad_norm": 0.19726519286632538, + "learning_rate": 7.243320012618737e-05, + "loss": 0.0080881729722023, + "step": 3520 + }, + { + "epoch": 12.006814310051107, + "grad_norm": 0.09112777560949326, + "learning_rate": 7.229270750496606e-05, + "loss": 0.005721393600106239, + "step": 3530 + }, + { + "epoch": 12.040885860306643, + "grad_norm": 0.2795824706554413, + "learning_rate": 7.215199486372162e-05, + "loss": 0.008645810931921006, + "step": 3540 + }, + { + "epoch": 12.074957410562181, + "grad_norm": 0.3062441051006317, + "learning_rate": 7.2011063591231e-05, + "loss": 0.010251733660697936, + "step": 3550 + }, + { + "epoch": 12.109028960817717, + "grad_norm": 0.16469214856624603, + "learning_rate": 7.186991507842894e-05, + "loss": 0.011274916678667068, + "step": 3560 + }, + { + "epoch": 12.143100511073254, + "grad_norm": 0.07611066102981567, + "learning_rate": 7.172855071839432e-05, + "loss": 0.005245493724942207, + "step": 3570 + }, + { + "epoch": 12.17717206132879, + "grad_norm": 0.08895934373140335, + "learning_rate": 7.158697190633627e-05, + "loss": 0.017691905796527862, + "step": 3580 + }, + { + "epoch": 12.211243611584328, + "grad_norm": 0.12958474457263947, + "learning_rate": 7.144518003958051e-05, + "loss": 0.006613949686288834, + "step": 3590 + }, + { + "epoch": 12.245315161839864, + "grad_norm": 0.5355010032653809, + "learning_rate": 7.130317651755554e-05, + "loss": 0.006677114963531494, + "step": 3600 + }, + { + "epoch": 12.2793867120954, + "grad_norm": 0.3897210955619812, + "learning_rate": 7.116096274177878e-05, + "loss": 0.008327482640743256, + "step": 3610 + }, + { + "epoch": 12.313458262350936, + "grad_norm": 0.1993381232023239, + "learning_rate": 7.101854011584277e-05, + "loss": 0.008164171129465103, + "step": 3620 + }, + { + "epoch": 12.347529812606474, + "grad_norm": 1.007405161857605, + "learning_rate": 7.087591004540137e-05, + "loss": 0.008335817605257034, + "step": 3630 + }, + { + "epoch": 12.38160136286201, + "grad_norm": 0.4568595588207245, + "learning_rate": 7.073307393815576e-05, + "loss": 0.007994977384805679, + "step": 3640 + }, + { + "epoch": 12.415672913117547, + "grad_norm": 0.07542114704847336, + "learning_rate": 7.059003320384065e-05, + "loss": 0.006418896466493606, + "step": 3650 + }, + { + "epoch": 12.449744463373083, + "grad_norm": 0.14576096832752228, + "learning_rate": 7.044678925421036e-05, + "loss": 0.006898117810487747, + "step": 3660 + }, + { + "epoch": 12.483816013628621, + "grad_norm": 0.19279853999614716, + "learning_rate": 7.030334350302483e-05, + "loss": 0.007405627518892288, + "step": 3670 + }, + { + "epoch": 12.517887563884157, + "grad_norm": 0.3989012837409973, + "learning_rate": 7.015969736603573e-05, + "loss": 0.004340417310595512, + "step": 3680 + }, + { + "epoch": 12.551959114139693, + "grad_norm": 0.9645401835441589, + "learning_rate": 7.00158522609724e-05, + "loss": 0.005963471531867981, + "step": 3690 + }, + { + "epoch": 12.58603066439523, + "grad_norm": 0.1078684851527214, + "learning_rate": 6.9871809607528e-05, + "loss": 0.007095368206501007, + "step": 3700 + }, + { + "epoch": 12.620102214650768, + "grad_norm": 0.17458277940750122, + "learning_rate": 6.972757082734534e-05, + "loss": 0.005171228945255279, + "step": 3710 + }, + { + "epoch": 12.654173764906304, + "grad_norm": 2.5305137634277344, + "learning_rate": 6.958313734400295e-05, + "loss": 0.006702803075313568, + "step": 3720 + }, + { + "epoch": 12.68824531516184, + "grad_norm": 1.3492250442504883, + "learning_rate": 6.943851058300103e-05, + "loss": 0.005169990658760071, + "step": 3730 + }, + { + "epoch": 12.722316865417376, + "grad_norm": 0.2801433503627777, + "learning_rate": 6.929369197174728e-05, + "loss": 0.0059206090867519375, + "step": 3740 + }, + { + "epoch": 12.756388415672912, + "grad_norm": 0.588880717754364, + "learning_rate": 6.914868293954297e-05, + "loss": 0.00973268821835518, + "step": 3750 + }, + { + "epoch": 12.79045996592845, + "grad_norm": 0.17676682770252228, + "learning_rate": 6.900348491756867e-05, + "loss": 0.005406035482883454, + "step": 3760 + }, + { + "epoch": 12.824531516183987, + "grad_norm": 0.08467921614646912, + "learning_rate": 6.885809933887023e-05, + "loss": 0.0045086205005645755, + "step": 3770 + }, + { + "epoch": 12.858603066439523, + "grad_norm": 0.09700228273868561, + "learning_rate": 6.871252763834465e-05, + "loss": 0.004319820553064346, + "step": 3780 + }, + { + "epoch": 12.892674616695059, + "grad_norm": 0.05134895071387291, + "learning_rate": 6.856677125272585e-05, + "loss": 0.004656513035297394, + "step": 3790 + }, + { + "epoch": 12.926746166950597, + "grad_norm": 0.17641349136829376, + "learning_rate": 6.842083162057047e-05, + "loss": 0.004298872128129005, + "step": 3800 + }, + { + "epoch": 12.960817717206133, + "grad_norm": 0.08891180902719498, + "learning_rate": 6.82747101822438e-05, + "loss": 0.004157590493559837, + "step": 3810 + }, + { + "epoch": 12.99488926746167, + "grad_norm": 1.2321891784667969, + "learning_rate": 6.812840837990542e-05, + "loss": 0.007444197684526444, + "step": 3820 + }, + { + "epoch": 13.02725724020443, + "grad_norm": 0.26937195658683777, + "learning_rate": 6.798192765749508e-05, + "loss": 0.005770929902791977, + "step": 3830 + }, + { + "epoch": 13.061328790459966, + "grad_norm": 0.31624147295951843, + "learning_rate": 6.783526946071835e-05, + "loss": 0.01097165048122406, + "step": 3840 + }, + { + "epoch": 13.095400340715502, + "grad_norm": 2.0840344429016113, + "learning_rate": 6.768843523703244e-05, + "loss": 0.015105555951595306, + "step": 3850 + }, + { + "epoch": 13.12947189097104, + "grad_norm": 0.23075535893440247, + "learning_rate": 6.754142643563185e-05, + "loss": 0.005851412191987038, + "step": 3860 + }, + { + "epoch": 13.163543441226576, + "grad_norm": 0.08374005556106567, + "learning_rate": 6.739424450743409e-05, + "loss": 0.007998282462358475, + "step": 3870 + }, + { + "epoch": 13.197614991482112, + "grad_norm": 0.4589119851589203, + "learning_rate": 6.724689090506539e-05, + "loss": 0.0033220205456018447, + "step": 3880 + }, + { + "epoch": 13.231686541737648, + "grad_norm": 0.38335326313972473, + "learning_rate": 6.70993670828463e-05, + "loss": 0.00552159808576107, + "step": 3890 + }, + { + "epoch": 13.265758091993186, + "grad_norm": 0.36032038927078247, + "learning_rate": 6.69516744967774e-05, + "loss": 0.007507776468992233, + "step": 3900 + }, + { + "epoch": 13.299829642248723, + "grad_norm": 0.05128803849220276, + "learning_rate": 6.680381460452488e-05, + "loss": 0.006090529635548592, + "step": 3910 + }, + { + "epoch": 13.333901192504259, + "grad_norm": 0.10289156436920166, + "learning_rate": 6.66557888654062e-05, + "loss": 0.008611975610256195, + "step": 3920 + }, + { + "epoch": 13.367972742759795, + "grad_norm": 0.10389873385429382, + "learning_rate": 6.650759874037562e-05, + "loss": 0.006964873522520065, + "step": 3930 + }, + { + "epoch": 13.402044293015333, + "grad_norm": 0.09713359922170639, + "learning_rate": 6.635924569200989e-05, + "loss": 0.005909853056073189, + "step": 3940 + }, + { + "epoch": 13.43611584327087, + "grad_norm": 0.1851266473531723, + "learning_rate": 6.621073118449364e-05, + "loss": 0.004036874324083328, + "step": 3950 + }, + { + "epoch": 13.470187393526405, + "grad_norm": 0.05056624859571457, + "learning_rate": 6.606205668360516e-05, + "loss": 0.0055687643587589266, + "step": 3960 + }, + { + "epoch": 13.504258943781942, + "grad_norm": 0.11166422814130783, + "learning_rate": 6.591322365670171e-05, + "loss": 0.00469561256468296, + "step": 3970 + }, + { + "epoch": 13.538330494037478, + "grad_norm": 0.12096738070249557, + "learning_rate": 6.57642335727052e-05, + "loss": 0.004750548303127289, + "step": 3980 + }, + { + "epoch": 13.572402044293016, + "grad_norm": 0.12073399126529694, + "learning_rate": 6.56150879020876e-05, + "loss": 0.0059893205761909485, + "step": 3990 + }, + { + "epoch": 13.606473594548552, + "grad_norm": 0.9339401721954346, + "learning_rate": 6.546578811685648e-05, + "loss": 0.01335892677307129, + "step": 4000 + }, + { + "epoch": 13.640545144804088, + "grad_norm": 0.5179684162139893, + "learning_rate": 6.531633569054044e-05, + "loss": 0.007028652727603913, + "step": 4010 + }, + { + "epoch": 13.674616695059624, + "grad_norm": 0.315481573343277, + "learning_rate": 6.516673209817458e-05, + "loss": 0.0076746389269828795, + "step": 4020 + }, + { + "epoch": 13.708688245315162, + "grad_norm": 0.19218428432941437, + "learning_rate": 6.501697881628597e-05, + "loss": 0.00918719694018364, + "step": 4030 + }, + { + "epoch": 13.742759795570699, + "grad_norm": 0.5113493204116821, + "learning_rate": 6.486707732287906e-05, + "loss": 0.008242735266685485, + "step": 4040 + }, + { + "epoch": 13.776831345826235, + "grad_norm": 0.09307843446731567, + "learning_rate": 6.471702909742104e-05, + "loss": 0.005062764137983322, + "step": 4050 + }, + { + "epoch": 13.810902896081771, + "grad_norm": 0.7766565084457397, + "learning_rate": 6.456683562082734e-05, + "loss": 0.004186952859163285, + "step": 4060 + }, + { + "epoch": 13.844974446337309, + "grad_norm": 0.583267331123352, + "learning_rate": 6.441649837544694e-05, + "loss": 0.005122175067663192, + "step": 4070 + }, + { + "epoch": 13.879045996592845, + "grad_norm": 0.053830988705158234, + "learning_rate": 6.426601884504773e-05, + "loss": 0.007482742518186569, + "step": 4080 + }, + { + "epoch": 13.913117546848381, + "grad_norm": 1.3554465770721436, + "learning_rate": 6.411539851480195e-05, + "loss": 0.004661658406257629, + "step": 4090 + }, + { + "epoch": 13.947189097103918, + "grad_norm": 0.04218238592147827, + "learning_rate": 6.396463887127144e-05, + "loss": 0.003842930123209953, + "step": 4100 + }, + { + "epoch": 13.981260647359456, + "grad_norm": 0.9284960031509399, + "learning_rate": 6.381374140239303e-05, + "loss": 0.006544541567564011, + "step": 4110 + }, + { + "epoch": 14.013628620102214, + "grad_norm": 0.04177882522344589, + "learning_rate": 6.366270759746382e-05, + "loss": 0.0044401433318853375, + "step": 4120 + }, + { + "epoch": 14.047700170357752, + "grad_norm": 0.04898664727807045, + "learning_rate": 6.351153894712647e-05, + "loss": 0.004527781158685684, + "step": 4130 + }, + { + "epoch": 14.081771720613288, + "grad_norm": 0.09982441365718842, + "learning_rate": 6.336023694335456e-05, + "loss": 0.0024997029453516007, + "step": 4140 + }, + { + "epoch": 14.115843270868824, + "grad_norm": 0.0643860250711441, + "learning_rate": 6.320880307943774e-05, + "loss": 0.0032753311097621918, + "step": 4150 + }, + { + "epoch": 14.14991482112436, + "grad_norm": 1.2694058418273926, + "learning_rate": 6.305723884996715e-05, + "loss": 0.00369129404425621, + "step": 4160 + }, + { + "epoch": 14.183986371379898, + "grad_norm": 1.2078691720962524, + "learning_rate": 6.290554575082052e-05, + "loss": 0.002693903259932995, + "step": 4170 + }, + { + "epoch": 14.218057921635435, + "grad_norm": 0.4831646978855133, + "learning_rate": 6.275372527914752e-05, + "loss": 0.0029770523309707642, + "step": 4180 + }, + { + "epoch": 14.25212947189097, + "grad_norm": 0.601255476474762, + "learning_rate": 6.260177893335489e-05, + "loss": 0.011412195861339569, + "step": 4190 + }, + { + "epoch": 14.286201022146507, + "grad_norm": 0.09007900208234787, + "learning_rate": 6.244970821309174e-05, + "loss": 0.004860934242606163, + "step": 4200 + }, + { + "epoch": 14.320272572402045, + "grad_norm": 0.06651585549116135, + "learning_rate": 6.229751461923468e-05, + "loss": 0.009242969751358032, + "step": 4210 + }, + { + "epoch": 14.354344122657581, + "grad_norm": 0.11677233129739761, + "learning_rate": 6.214519965387303e-05, + "loss": 0.0030382387340068817, + "step": 4220 + }, + { + "epoch": 14.388415672913117, + "grad_norm": 0.05548478290438652, + "learning_rate": 6.199276482029401e-05, + "loss": 0.0038005076348781586, + "step": 4230 + }, + { + "epoch": 14.422487223168654, + "grad_norm": 0.10228066146373749, + "learning_rate": 6.18402116229679e-05, + "loss": 0.0026105983182787896, + "step": 4240 + }, + { + "epoch": 14.456558773424192, + "grad_norm": 0.04359025880694389, + "learning_rate": 6.168754156753316e-05, + "loss": 0.0022529620677232743, + "step": 4250 + }, + { + "epoch": 14.490630323679728, + "grad_norm": 0.3089030981063843, + "learning_rate": 6.153475616078158e-05, + "loss": 0.002916494943201542, + "step": 4260 + }, + { + "epoch": 14.524701873935264, + "grad_norm": 0.5625247359275818, + "learning_rate": 6.138185691064348e-05, + "loss": 0.0038366343826055525, + "step": 4270 + }, + { + "epoch": 14.5587734241908, + "grad_norm": 0.09490062296390533, + "learning_rate": 6.12288453261727e-05, + "loss": 0.0048129215836524965, + "step": 4280 + }, + { + "epoch": 14.592844974446336, + "grad_norm": 0.0946214497089386, + "learning_rate": 6.107572291753183e-05, + "loss": 0.003979400172829628, + "step": 4290 + }, + { + "epoch": 14.626916524701874, + "grad_norm": 1.290644884109497, + "learning_rate": 6.092249119597721e-05, + "loss": 0.004271459579467773, + "step": 4300 + }, + { + "epoch": 14.66098807495741, + "grad_norm": 0.04261275380849838, + "learning_rate": 6.076915167384407e-05, + "loss": 0.002717731706798077, + "step": 4310 + }, + { + "epoch": 14.695059625212947, + "grad_norm": 0.057752665132284164, + "learning_rate": 6.0615705864531594e-05, + "loss": 0.003598403185606003, + "step": 4320 + }, + { + "epoch": 14.729131175468483, + "grad_norm": 0.11382733285427094, + "learning_rate": 6.046215528248798e-05, + "loss": 0.0051520105451345445, + "step": 4330 + }, + { + "epoch": 14.763202725724021, + "grad_norm": 0.30120739340782166, + "learning_rate": 6.0308501443195464e-05, + "loss": 0.003737257793545723, + "step": 4340 + }, + { + "epoch": 14.797274275979557, + "grad_norm": 0.15558502078056335, + "learning_rate": 6.0154745863155405e-05, + "loss": 0.0033736743032932283, + "step": 4350 + }, + { + "epoch": 14.831345826235093, + "grad_norm": 1.0245122909545898, + "learning_rate": 6.000089005987331e-05, + "loss": 0.004328646138310433, + "step": 4360 + }, + { + "epoch": 14.86541737649063, + "grad_norm": 0.06399249285459518, + "learning_rate": 5.984693555184386e-05, + "loss": 0.0029862189665436746, + "step": 4370 + }, + { + "epoch": 14.899488926746168, + "grad_norm": 0.050825756043195724, + "learning_rate": 5.9692883858535884e-05, + "loss": 0.006860410422086715, + "step": 4380 + }, + { + "epoch": 14.933560477001704, + "grad_norm": 0.07893214374780655, + "learning_rate": 5.953873650037739e-05, + "loss": 0.005103478953242302, + "step": 4390 + }, + { + "epoch": 14.96763202725724, + "grad_norm": 0.7493880987167358, + "learning_rate": 5.938449499874059e-05, + "loss": 0.008201855421066283, + "step": 4400 + }, + { + "epoch": 15.0, + "grad_norm": 3.547017812728882, + "learning_rate": 5.923016087592681e-05, + "loss": 0.005634984374046326, + "step": 4410 + }, + { + "epoch": 15.034071550255536, + "grad_norm": 0.12131509929895401, + "learning_rate": 5.9075735655151575e-05, + "loss": 0.004040924087166786, + "step": 4420 + }, + { + "epoch": 15.068143100511072, + "grad_norm": 0.03410695865750313, + "learning_rate": 5.892122086052944e-05, + "loss": 0.00687539279460907, + "step": 4430 + }, + { + "epoch": 15.10221465076661, + "grad_norm": 0.04947010800242424, + "learning_rate": 5.876661801705905e-05, + "loss": 0.0032738476991653444, + "step": 4440 + }, + { + "epoch": 15.136286201022147, + "grad_norm": 0.8641470074653625, + "learning_rate": 5.861192865060805e-05, + "loss": 0.008090145885944366, + "step": 4450 + }, + { + "epoch": 15.170357751277683, + "grad_norm": 0.09471169859170914, + "learning_rate": 5.845715428789805e-05, + "loss": 0.002891349606215954, + "step": 4460 + }, + { + "epoch": 15.204429301533219, + "grad_norm": 0.061726536601781845, + "learning_rate": 5.830229645648952e-05, + "loss": 0.003794405609369278, + "step": 4470 + }, + { + "epoch": 15.238500851788757, + "grad_norm": 0.5859624743461609, + "learning_rate": 5.814735668476674e-05, + "loss": 0.0029357653111219406, + "step": 4480 + }, + { + "epoch": 15.272572402044293, + "grad_norm": 0.03562292829155922, + "learning_rate": 5.79923365019227e-05, + "loss": 0.010583274066448212, + "step": 4490 + }, + { + "epoch": 15.30664395229983, + "grad_norm": 0.06037396192550659, + "learning_rate": 5.783723743794402e-05, + "loss": 0.0022368822246789932, + "step": 4500 + }, + { + "epoch": 15.340715502555366, + "grad_norm": 0.06155624985694885, + "learning_rate": 5.7682061023595845e-05, + "loss": 0.004347968474030494, + "step": 4510 + }, + { + "epoch": 15.374787052810904, + "grad_norm": 0.06593833863735199, + "learning_rate": 5.752680879040676e-05, + "loss": 0.006146045774221421, + "step": 4520 + }, + { + "epoch": 15.40885860306644, + "grad_norm": 0.04347379505634308, + "learning_rate": 5.73714822706536e-05, + "loss": 0.00284601878374815, + "step": 4530 + }, + { + "epoch": 15.442930153321976, + "grad_norm": 0.04662046581506729, + "learning_rate": 5.721608299734641e-05, + "loss": 0.007786727696657181, + "step": 4540 + }, + { + "epoch": 15.477001703577512, + "grad_norm": 0.34194523096084595, + "learning_rate": 5.706061250421328e-05, + "loss": 0.009722773730754853, + "step": 4550 + }, + { + "epoch": 15.51107325383305, + "grad_norm": 0.1249171644449234, + "learning_rate": 5.690507232568524e-05, + "loss": 0.0033404458314180373, + "step": 4560 + }, + { + "epoch": 15.545144804088586, + "grad_norm": 0.0931662991642952, + "learning_rate": 5.6749463996881026e-05, + "loss": 0.002273181825876236, + "step": 4570 + }, + { + "epoch": 15.579216354344123, + "grad_norm": 0.045032959431409836, + "learning_rate": 5.659378905359203e-05, + "loss": 0.002775738388299942, + "step": 4580 + }, + { + "epoch": 15.613287904599659, + "grad_norm": 0.2108381986618042, + "learning_rate": 5.6438049032267095e-05, + "loss": 0.0032906975597143174, + "step": 4590 + }, + { + "epoch": 15.647359454855195, + "grad_norm": 0.07233528047800064, + "learning_rate": 5.628224546999735e-05, + "loss": 0.004193896055221557, + "step": 4600 + }, + { + "epoch": 15.681431005110733, + "grad_norm": 0.034625567495822906, + "learning_rate": 5.612637990450107e-05, + "loss": 0.0040230326354503635, + "step": 4610 + }, + { + "epoch": 15.71550255536627, + "grad_norm": 0.05149779096245766, + "learning_rate": 5.597045387410845e-05, + "loss": 0.0028049644082784654, + "step": 4620 + }, + { + "epoch": 15.749574105621805, + "grad_norm": 0.9837811589241028, + "learning_rate": 5.5814468917746475e-05, + "loss": 0.004248417913913727, + "step": 4630 + }, + { + "epoch": 15.783645655877342, + "grad_norm": 0.04748018831014633, + "learning_rate": 5.565842657492366e-05, + "loss": 0.0023705245926976203, + "step": 4640 + }, + { + "epoch": 15.81771720613288, + "grad_norm": 0.02733342908322811, + "learning_rate": 5.5502328385714954e-05, + "loss": 0.004021480306982994, + "step": 4650 + }, + { + "epoch": 15.851788756388416, + "grad_norm": 0.028498021885752678, + "learning_rate": 5.534617589074647e-05, + "loss": 0.0026785416528582575, + "step": 4660 + }, + { + "epoch": 15.885860306643952, + "grad_norm": 0.10506656020879745, + "learning_rate": 5.5189970631180265e-05, + "loss": 0.002691369690001011, + "step": 4670 + }, + { + "epoch": 15.919931856899488, + "grad_norm": 0.21124304831027985, + "learning_rate": 5.503371414869922e-05, + "loss": 0.0054808568209409715, + "step": 4680 + }, + { + "epoch": 15.954003407155026, + "grad_norm": 0.04369935393333435, + "learning_rate": 5.4877407985491714e-05, + "loss": 0.002211957797408104, + "step": 4690 + }, + { + "epoch": 15.988074957410563, + "grad_norm": 0.040269821882247925, + "learning_rate": 5.472105368423648e-05, + "loss": 0.002709386870265007, + "step": 4700 + }, + { + "epoch": 16.020442930153322, + "grad_norm": 0.028259102255105972, + "learning_rate": 5.4564652788087353e-05, + "loss": 0.0016449641436338425, + "step": 4710 + }, + { + "epoch": 16.05451448040886, + "grad_norm": 0.03208509832620621, + "learning_rate": 5.440820684065804e-05, + "loss": 0.002032238245010376, + "step": 4720 + }, + { + "epoch": 16.088586030664395, + "grad_norm": 0.08327673375606537, + "learning_rate": 5.425171738600686e-05, + "loss": 0.0017761236056685447, + "step": 4730 + }, + { + "epoch": 16.12265758091993, + "grad_norm": 0.030065344646573067, + "learning_rate": 5.4095185968621575e-05, + "loss": 0.0038551613688468933, + "step": 4740 + }, + { + "epoch": 16.156729131175467, + "grad_norm": 0.02324601449072361, + "learning_rate": 5.3938614133404075e-05, + "loss": 0.002114511840045452, + "step": 4750 + }, + { + "epoch": 16.190800681431003, + "grad_norm": 0.036281730979681015, + "learning_rate": 5.378200342565516e-05, + "loss": 0.00202888622879982, + "step": 4760 + }, + { + "epoch": 16.224872231686543, + "grad_norm": 1.3998425006866455, + "learning_rate": 5.362535539105928e-05, + "loss": 0.002383740805089474, + "step": 4770 + }, + { + "epoch": 16.25894378194208, + "grad_norm": 0.5290566086769104, + "learning_rate": 5.346867157566931e-05, + "loss": 0.0023939421400427817, + "step": 4780 + }, + { + "epoch": 16.293015332197616, + "grad_norm": 0.07418588548898697, + "learning_rate": 5.331195352589126e-05, + "loss": 0.001642211526632309, + "step": 4790 + }, + { + "epoch": 16.327086882453152, + "grad_norm": 0.027548249810934067, + "learning_rate": 5.315520278846898e-05, + "loss": 0.0016159243881702424, + "step": 4800 + }, + { + "epoch": 16.361158432708688, + "grad_norm": 0.028864076361060143, + "learning_rate": 5.299842091046899e-05, + "loss": 0.0017805464565753936, + "step": 4810 + }, + { + "epoch": 16.395229982964224, + "grad_norm": 0.02843449078500271, + "learning_rate": 5.284160943926515e-05, + "loss": 0.001770513504743576, + "step": 4820 + }, + { + "epoch": 16.42930153321976, + "grad_norm": 0.018429072573781013, + "learning_rate": 5.2684769922523325e-05, + "loss": 0.0042362667620182036, + "step": 4830 + }, + { + "epoch": 16.463373083475297, + "grad_norm": 0.021774500608444214, + "learning_rate": 5.252790390818625e-05, + "loss": 0.002549749054014683, + "step": 4840 + }, + { + "epoch": 16.497444633730836, + "grad_norm": 0.1091708391904831, + "learning_rate": 5.2371012944458165e-05, + "loss": 0.001800367422401905, + "step": 4850 + }, + { + "epoch": 16.531516183986373, + "grad_norm": 0.021982407197356224, + "learning_rate": 5.2214098579789526e-05, + "loss": 0.0017428381368517875, + "step": 4860 + }, + { + "epoch": 16.56558773424191, + "grad_norm": 0.029410257935523987, + "learning_rate": 5.2057162362861775e-05, + "loss": 0.0015747616067528724, + "step": 4870 + }, + { + "epoch": 16.599659284497445, + "grad_norm": 0.027157938107848167, + "learning_rate": 5.190020584257201e-05, + "loss": 0.0014591478742659093, + "step": 4880 + }, + { + "epoch": 16.63373083475298, + "grad_norm": 0.025023752823472023, + "learning_rate": 5.174323056801773e-05, + "loss": 0.0014232179149985314, + "step": 4890 + }, + { + "epoch": 16.667802385008518, + "grad_norm": 0.01952843740582466, + "learning_rate": 5.158623808848152e-05, + "loss": 0.0013258514925837516, + "step": 4900 + }, + { + "epoch": 16.701873935264054, + "grad_norm": 0.03049285151064396, + "learning_rate": 5.142922995341576e-05, + "loss": 0.003258787468075752, + "step": 4910 + }, + { + "epoch": 16.73594548551959, + "grad_norm": 0.056466344743967056, + "learning_rate": 5.127220771242737e-05, + "loss": 0.0014892812818288804, + "step": 4920 + }, + { + "epoch": 16.770017035775126, + "grad_norm": 0.06346353143453598, + "learning_rate": 5.111517291526245e-05, + "loss": 0.005571158602833748, + "step": 4930 + }, + { + "epoch": 16.804088586030666, + "grad_norm": 0.040035296231508255, + "learning_rate": 5.0958127111791086e-05, + "loss": 0.0017191551625728608, + "step": 4940 + }, + { + "epoch": 16.838160136286202, + "grad_norm": 0.029128847643733025, + "learning_rate": 5.080107185199193e-05, + "loss": 0.0015416602604091168, + "step": 4950 + }, + { + "epoch": 16.87223168654174, + "grad_norm": 0.03200097382068634, + "learning_rate": 5.0644008685937005e-05, + "loss": 0.001368429884314537, + "step": 4960 + }, + { + "epoch": 16.906303236797275, + "grad_norm": 0.027507126331329346, + "learning_rate": 5.048693916377635e-05, + "loss": 0.0013849607668817044, + "step": 4970 + }, + { + "epoch": 16.94037478705281, + "grad_norm": 0.11205657571554184, + "learning_rate": 5.032986483572273e-05, + "loss": 0.0047020137310028074, + "step": 4980 + }, + { + "epoch": 16.974446337308347, + "grad_norm": 0.04292678087949753, + "learning_rate": 5.017278725203636e-05, + "loss": 0.001518221851438284, + "step": 4990 + }, + { + "epoch": 17.00681431005111, + "grad_norm": 0.020551787689328194, + "learning_rate": 5.0015707963009575e-05, + "loss": 0.004414630308747292, + "step": 5000 + }, + { + "epoch": 17.040885860306645, + "grad_norm": 0.02067749574780464, + "learning_rate": 4.985862851895152e-05, + "loss": 0.0026002321392297745, + "step": 5010 + }, + { + "epoch": 17.07495741056218, + "grad_norm": 0.038795456290245056, + "learning_rate": 4.9701550470172914e-05, + "loss": 0.004252338781952858, + "step": 5020 + }, + { + "epoch": 17.109028960817717, + "grad_norm": 0.024375731125473976, + "learning_rate": 4.954447536697068e-05, + "loss": 0.0017102709040045737, + "step": 5030 + }, + { + "epoch": 17.143100511073254, + "grad_norm": 0.22505810856819153, + "learning_rate": 4.9387404759612674e-05, + "loss": 0.0014576041139662266, + "step": 5040 + }, + { + "epoch": 17.17717206132879, + "grad_norm": 0.028041576966643333, + "learning_rate": 4.923034019832237e-05, + "loss": 0.002051759511232376, + "step": 5050 + }, + { + "epoch": 17.211243611584326, + "grad_norm": 0.21869738399982452, + "learning_rate": 4.907328323326359e-05, + "loss": 0.001416478306055069, + "step": 5060 + }, + { + "epoch": 17.245315161839862, + "grad_norm": 0.02534378692507744, + "learning_rate": 4.891623541452516e-05, + "loss": 0.001756133884191513, + "step": 5070 + }, + { + "epoch": 17.279386712095402, + "grad_norm": 0.022471051663160324, + "learning_rate": 4.8759198292105655e-05, + "loss": 0.0013532243669033051, + "step": 5080 + }, + { + "epoch": 17.313458262350938, + "grad_norm": 0.02355014532804489, + "learning_rate": 4.8602173415898076e-05, + "loss": 0.001428174041211605, + "step": 5090 + }, + { + "epoch": 17.347529812606474, + "grad_norm": 0.023271949961781502, + "learning_rate": 4.8445162335674574e-05, + "loss": 0.001497176382690668, + "step": 5100 + }, + { + "epoch": 17.38160136286201, + "grad_norm": 0.09229321032762527, + "learning_rate": 4.828816660107109e-05, + "loss": 0.0013473049737513065, + "step": 5110 + }, + { + "epoch": 17.415672913117547, + "grad_norm": 0.020730620250105858, + "learning_rate": 4.813118776157216e-05, + "loss": 0.0012679336592555045, + "step": 5120 + }, + { + "epoch": 17.449744463373083, + "grad_norm": 0.020511914044618607, + "learning_rate": 4.797422736649555e-05, + "loss": 0.0012394772842526437, + "step": 5130 + }, + { + "epoch": 17.48381601362862, + "grad_norm": 0.020659010857343674, + "learning_rate": 4.7817286964976995e-05, + "loss": 0.0012148208916187286, + "step": 5140 + }, + { + "epoch": 17.517887563884155, + "grad_norm": 0.021254630759358406, + "learning_rate": 4.7660368105954905e-05, + "loss": 0.0011381401680409908, + "step": 5150 + }, + { + "epoch": 17.55195911413969, + "grad_norm": 0.020152298733592033, + "learning_rate": 4.750347233815506e-05, + "loss": 0.0013540409505367278, + "step": 5160 + }, + { + "epoch": 17.58603066439523, + "grad_norm": 0.015300494618713856, + "learning_rate": 4.734660121007534e-05, + "loss": 0.0011846719309687615, + "step": 5170 + }, + { + "epoch": 17.620102214650768, + "grad_norm": 0.040884144604206085, + "learning_rate": 4.718975626997047e-05, + "loss": 0.0011502589099109174, + "step": 5180 + }, + { + "epoch": 17.654173764906304, + "grad_norm": 0.0147259421646595, + "learning_rate": 4.7032939065836674e-05, + "loss": 0.0010198661126196384, + "step": 5190 + }, + { + "epoch": 17.68824531516184, + "grad_norm": 0.017110861837863922, + "learning_rate": 4.687615114539644e-05, + "loss": 0.006115290522575379, + "step": 5200 + }, + { + "epoch": 17.722316865417376, + "grad_norm": 0.06430409103631973, + "learning_rate": 4.6719394056083246e-05, + "loss": 0.0011276694945991038, + "step": 5210 + }, + { + "epoch": 17.756388415672912, + "grad_norm": 0.02295953780412674, + "learning_rate": 4.65626693450263e-05, + "loss": 0.001248331554234028, + "step": 5220 + }, + { + "epoch": 17.79045996592845, + "grad_norm": 0.017782188951969147, + "learning_rate": 4.6405978559035195e-05, + "loss": 0.0011788049712777137, + "step": 5230 + }, + { + "epoch": 17.824531516183985, + "grad_norm": 4.459109783172607, + "learning_rate": 4.6249323244584756e-05, + "loss": 0.0015946637839078903, + "step": 5240 + }, + { + "epoch": 17.858603066439525, + "grad_norm": 0.017394445836544037, + "learning_rate": 4.6092704947799684e-05, + "loss": 0.0010795393958687783, + "step": 5250 + }, + { + "epoch": 17.89267461669506, + "grad_norm": 0.054457612335681915, + "learning_rate": 4.593612521443933e-05, + "loss": 0.0012965104542672634, + "step": 5260 + }, + { + "epoch": 17.926746166950597, + "grad_norm": 0.01878967694938183, + "learning_rate": 4.577958558988247e-05, + "loss": 0.0021525075659155845, + "step": 5270 + }, + { + "epoch": 17.960817717206133, + "grad_norm": 0.019479788839817047, + "learning_rate": 4.5623087619111985e-05, + "loss": 0.0011511732824146747, + "step": 5280 + }, + { + "epoch": 17.99488926746167, + "grad_norm": 0.02074812538921833, + "learning_rate": 4.546663284669967e-05, + "loss": 0.003984493762254715, + "step": 5290 + }, + { + "epoch": 18.027257240204428, + "grad_norm": 0.08662942051887512, + "learning_rate": 4.5310222816790956e-05, + "loss": 0.0010945863090455533, + "step": 5300 + }, + { + "epoch": 18.061328790459967, + "grad_norm": 0.014388803392648697, + "learning_rate": 4.5153859073089696e-05, + "loss": 0.0010205313563346862, + "step": 5310 + }, + { + "epoch": 18.095400340715504, + "grad_norm": 1.6488099098205566, + "learning_rate": 4.499754315884293e-05, + "loss": 0.0011971708387136458, + "step": 5320 + }, + { + "epoch": 18.12947189097104, + "grad_norm": 0.014312258921563625, + "learning_rate": 4.484127661682561e-05, + "loss": 0.0011194176971912385, + "step": 5330 + }, + { + "epoch": 18.163543441226576, + "grad_norm": 0.017495693638920784, + "learning_rate": 4.468506098932541e-05, + "loss": 0.0010449805296957493, + "step": 5340 + }, + { + "epoch": 18.197614991482112, + "grad_norm": 0.03344535082578659, + "learning_rate": 4.452889781812753e-05, + "loss": 0.001077896449714899, + "step": 5350 + }, + { + "epoch": 18.23168654173765, + "grad_norm": 0.09552326053380966, + "learning_rate": 4.437278864449941e-05, + "loss": 0.0011071257293224334, + "step": 5360 + }, + { + "epoch": 18.265758091993185, + "grad_norm": 0.015476961620151997, + "learning_rate": 4.421673500917558e-05, + "loss": 0.0010746978223323822, + "step": 5370 + }, + { + "epoch": 18.29982964224872, + "grad_norm": 0.01175448764115572, + "learning_rate": 4.40607384523424e-05, + "loss": 0.0009136024862527847, + "step": 5380 + }, + { + "epoch": 18.33390119250426, + "grad_norm": 0.015426024794578552, + "learning_rate": 4.390480051362292e-05, + "loss": 0.002662300691008568, + "step": 5390 + }, + { + "epoch": 18.367972742759797, + "grad_norm": 0.016918612644076347, + "learning_rate": 4.374892273206164e-05, + "loss": 0.0009650848805904388, + "step": 5400 + }, + { + "epoch": 18.402044293015333, + "grad_norm": 0.0167019572108984, + "learning_rate": 4.359310664610933e-05, + "loss": 0.0009537719190120697, + "step": 5410 + }, + { + "epoch": 18.43611584327087, + "grad_norm": 0.014781847596168518, + "learning_rate": 4.343735379360785e-05, + "loss": 0.0009710612706840038, + "step": 5420 + }, + { + "epoch": 18.470187393526405, + "grad_norm": 0.018976446241140366, + "learning_rate": 4.3281665711774986e-05, + "loss": 0.0010830318555235864, + "step": 5430 + }, + { + "epoch": 18.50425894378194, + "grad_norm": 0.014287130907177925, + "learning_rate": 4.312604393718925e-05, + "loss": 0.0009636364877223969, + "step": 5440 + }, + { + "epoch": 18.538330494037478, + "grad_norm": 0.026724105700850487, + "learning_rate": 4.2970490005774716e-05, + "loss": 0.0009737368673086166, + "step": 5450 + }, + { + "epoch": 18.572402044293014, + "grad_norm": 0.012648231349885464, + "learning_rate": 4.28150054527859e-05, + "loss": 0.003246837854385376, + "step": 5460 + }, + { + "epoch": 18.606473594548554, + "grad_norm": 0.014268019236624241, + "learning_rate": 4.2659591812792585e-05, + "loss": 0.00107917794957757, + "step": 5470 + }, + { + "epoch": 18.64054514480409, + "grad_norm": 0.01439786609262228, + "learning_rate": 4.250425061966461e-05, + "loss": 0.0010169627144932746, + "step": 5480 + }, + { + "epoch": 18.674616695059626, + "grad_norm": 0.011672698892652988, + "learning_rate": 4.234898340655687e-05, + "loss": 0.0010213958099484443, + "step": 5490 + }, + { + "epoch": 18.708688245315162, + "grad_norm": 0.016902077943086624, + "learning_rate": 4.219379170589407e-05, + "loss": 0.0009170673787593841, + "step": 5500 + }, + { + "epoch": 18.7427597955707, + "grad_norm": 0.014487321488559246, + "learning_rate": 4.203867704935566e-05, + "loss": 0.000946090929210186, + "step": 5510 + }, + { + "epoch": 18.776831345826235, + "grad_norm": 0.014147425070405006, + "learning_rate": 4.188364096786063e-05, + "loss": 0.000928798783570528, + "step": 5520 + }, + { + "epoch": 18.81090289608177, + "grad_norm": 0.013360686600208282, + "learning_rate": 4.172868499155254e-05, + "loss": 0.007861136645078658, + "step": 5530 + }, + { + "epoch": 18.844974446337307, + "grad_norm": 0.013410428538918495, + "learning_rate": 4.157381064978432e-05, + "loss": 0.0009810113348066806, + "step": 5540 + }, + { + "epoch": 18.879045996592843, + "grad_norm": 0.017660394310951233, + "learning_rate": 4.14190194711032e-05, + "loss": 0.0014070034958422184, + "step": 5550 + }, + { + "epoch": 18.913117546848383, + "grad_norm": 0.03081461787223816, + "learning_rate": 4.126431298323561e-05, + "loss": 0.0011111219413578511, + "step": 5560 + }, + { + "epoch": 18.94718909710392, + "grad_norm": 0.013971783220767975, + "learning_rate": 4.110969271307215e-05, + "loss": 0.001013590395450592, + "step": 5570 + }, + { + "epoch": 18.981260647359456, + "grad_norm": 0.26775357127189636, + "learning_rate": 4.095516018665244e-05, + "loss": 0.0011561580002307891, + "step": 5580 + }, + { + "epoch": 19.013628620102214, + "grad_norm": 0.023728186264634132, + "learning_rate": 4.080071692915014e-05, + "loss": 0.0014442360028624535, + "step": 5590 + }, + { + "epoch": 19.04770017035775, + "grad_norm": 0.02013670653104782, + "learning_rate": 4.0646364464857854e-05, + "loss": 0.0009878143668174745, + "step": 5600 + }, + { + "epoch": 19.081771720613286, + "grad_norm": 0.013261818327009678, + "learning_rate": 4.049210431717208e-05, + "loss": 0.0009241646155714988, + "step": 5610 + }, + { + "epoch": 19.115843270868826, + "grad_norm": 0.018740715458989143, + "learning_rate": 4.0337938008578204e-05, + "loss": 0.0009923766367137433, + "step": 5620 + }, + { + "epoch": 19.149914821124362, + "grad_norm": 0.011729925870895386, + "learning_rate": 4.0183867060635446e-05, + "loss": 0.0009203552268445492, + "step": 5630 + }, + { + "epoch": 19.1839863713799, + "grad_norm": 0.013013752177357674, + "learning_rate": 4.002989299396187e-05, + "loss": 0.0008703425526618957, + "step": 5640 + }, + { + "epoch": 19.218057921635435, + "grad_norm": 0.012250490486621857, + "learning_rate": 3.987601732821935e-05, + "loss": 0.000981604866683483, + "step": 5650 + }, + { + "epoch": 19.25212947189097, + "grad_norm": 0.012271963991224766, + "learning_rate": 3.9722241582098574e-05, + "loss": 0.0008470822125673294, + "step": 5660 + }, + { + "epoch": 19.286201022146507, + "grad_norm": 0.010430771857500076, + "learning_rate": 3.9568567273304094e-05, + "loss": 0.0022851787507534027, + "step": 5670 + }, + { + "epoch": 19.320272572402043, + "grad_norm": 0.011908025480806828, + "learning_rate": 3.941499591853928e-05, + "loss": 0.0008988802321255207, + "step": 5680 + }, + { + "epoch": 19.35434412265758, + "grad_norm": 0.010292290709912777, + "learning_rate": 3.926152903349143e-05, + "loss": 0.0007751309778541327, + "step": 5690 + }, + { + "epoch": 19.38841567291312, + "grad_norm": 0.012682536616921425, + "learning_rate": 3.91081681328167e-05, + "loss": 0.0009175571613013744, + "step": 5700 + }, + { + "epoch": 19.422487223168655, + "grad_norm": 0.014314210042357445, + "learning_rate": 3.8954914730125304e-05, + "loss": 0.000846368633210659, + "step": 5710 + }, + { + "epoch": 19.45655877342419, + "grad_norm": 0.019071565940976143, + "learning_rate": 3.880177033796643e-05, + "loss": 0.0008073801174759865, + "step": 5720 + }, + { + "epoch": 19.490630323679728, + "grad_norm": 0.012006244622170925, + "learning_rate": 3.864873646781341e-05, + "loss": 0.0008969792164862156, + "step": 5730 + }, + { + "epoch": 19.524701873935264, + "grad_norm": 0.013704975135624409, + "learning_rate": 3.849581463004877e-05, + "loss": 0.0007448746357113123, + "step": 5740 + }, + { + "epoch": 19.5587734241908, + "grad_norm": 0.010101907886564732, + "learning_rate": 3.83430063339493e-05, + "loss": 0.0008202052675187587, + "step": 5750 + }, + { + "epoch": 19.592844974446336, + "grad_norm": 0.01134677417576313, + "learning_rate": 3.819031308767119e-05, + "loss": 0.0008172311820089817, + "step": 5760 + }, + { + "epoch": 19.626916524701873, + "grad_norm": 0.010277379304170609, + "learning_rate": 3.803773639823515e-05, + "loss": 0.0007495251018553972, + "step": 5770 + }, + { + "epoch": 19.66098807495741, + "grad_norm": 0.01130878645926714, + "learning_rate": 3.7885277771511495e-05, + "loss": 0.0008373687975108624, + "step": 5780 + }, + { + "epoch": 19.69505962521295, + "grad_norm": 2.182074785232544, + "learning_rate": 3.7732938712205336e-05, + "loss": 0.001305585727095604, + "step": 5790 + }, + { + "epoch": 19.729131175468485, + "grad_norm": 0.010290215723216534, + "learning_rate": 3.758072072384167e-05, + "loss": 0.0008105946704745293, + "step": 5800 + }, + { + "epoch": 19.76320272572402, + "grad_norm": 0.01141059398651123, + "learning_rate": 3.742862530875058e-05, + "loss": 0.0008374320343136788, + "step": 5810 + }, + { + "epoch": 19.797274275979557, + "grad_norm": 0.014809155836701393, + "learning_rate": 3.7276653968052434e-05, + "loss": 0.0009145875461399555, + "step": 5820 + }, + { + "epoch": 19.831345826235093, + "grad_norm": 0.014833973720669746, + "learning_rate": 3.712480820164299e-05, + "loss": 0.0008336975239217282, + "step": 5830 + }, + { + "epoch": 19.86541737649063, + "grad_norm": 0.0111404312774539, + "learning_rate": 3.697308950817868e-05, + "loss": 0.0030387239530682564, + "step": 5840 + }, + { + "epoch": 19.899488926746166, + "grad_norm": 0.013064036145806313, + "learning_rate": 3.6821499385061744e-05, + "loss": 0.0008362457156181335, + "step": 5850 + }, + { + "epoch": 19.933560477001702, + "grad_norm": 0.022499095648527145, + "learning_rate": 3.6670039328425505e-05, + "loss": 0.0013565801084041596, + "step": 5860 + }, + { + "epoch": 19.967632027257242, + "grad_norm": 0.019801879301667213, + "learning_rate": 3.651871083311957e-05, + "loss": 0.000828612968325615, + "step": 5870 + }, + { + "epoch": 20.0, + "grad_norm": 0.02100897952914238, + "learning_rate": 3.636751539269511e-05, + "loss": 0.0009760197252035141, + "step": 5880 + }, + { + "epoch": 20.034071550255536, + "grad_norm": 0.014585711993277073, + "learning_rate": 3.621645449939005e-05, + "loss": 0.0016422362998127938, + "step": 5890 + }, + { + "epoch": 20.068143100511072, + "grad_norm": 0.013150216080248356, + "learning_rate": 3.6065529644114436e-05, + "loss": 0.0008103516884148121, + "step": 5900 + }, + { + "epoch": 20.10221465076661, + "grad_norm": 0.04175172373652458, + "learning_rate": 3.5914742316435665e-05, + "loss": 0.0007623786572366953, + "step": 5910 + }, + { + "epoch": 20.136286201022145, + "grad_norm": 0.015947718173265457, + "learning_rate": 3.576409400456379e-05, + "loss": 0.0007993712089955807, + "step": 5920 + }, + { + "epoch": 20.170357751277685, + "grad_norm": 0.012471335940063, + "learning_rate": 3.561358619533681e-05, + "loss": 0.0008018011227250099, + "step": 5930 + }, + { + "epoch": 20.20442930153322, + "grad_norm": 0.01128925196826458, + "learning_rate": 3.5463220374206055e-05, + "loss": 0.0016793888062238693, + "step": 5940 + }, + { + "epoch": 20.238500851788757, + "grad_norm": 0.016027703881263733, + "learning_rate": 3.531299802522148e-05, + "loss": 0.0009128442965447902, + "step": 5950 + }, + { + "epoch": 20.272572402044293, + "grad_norm": 0.011450287885963917, + "learning_rate": 3.5162920631017004e-05, + "loss": 0.0008219889365136624, + "step": 5960 + }, + { + "epoch": 20.30664395229983, + "grad_norm": 0.011038366705179214, + "learning_rate": 3.501298967279594e-05, + "loss": 0.0008107548579573632, + "step": 5970 + }, + { + "epoch": 20.340715502555366, + "grad_norm": 0.009187030605971813, + "learning_rate": 3.486320663031629e-05, + "loss": 0.0007637172937393188, + "step": 5980 + }, + { + "epoch": 20.374787052810902, + "grad_norm": 0.012092473916709423, + "learning_rate": 3.471357298187624e-05, + "loss": 0.000820909533649683, + "step": 5990 + }, + { + "epoch": 20.408858603066438, + "grad_norm": 0.012225104495882988, + "learning_rate": 3.456409020429947e-05, + "loss": 0.00079767731949687, + "step": 6000 + }, + { + "epoch": 20.442930153321978, + "grad_norm": 0.009418906643986702, + "learning_rate": 3.441475977292066e-05, + "loss": 0.0007463518530130386, + "step": 6010 + }, + { + "epoch": 20.477001703577514, + "grad_norm": 0.013578281737864017, + "learning_rate": 3.426558316157087e-05, + "loss": 0.0007727259304374457, + "step": 6020 + }, + { + "epoch": 20.51107325383305, + "grad_norm": 0.012921578250825405, + "learning_rate": 3.411656184256304e-05, + "loss": 0.0008030150085687638, + "step": 6030 + }, + { + "epoch": 20.545144804088586, + "grad_norm": 0.01194257102906704, + "learning_rate": 3.3967697286677415e-05, + "loss": 0.0007611874025315046, + "step": 6040 + }, + { + "epoch": 20.579216354344123, + "grad_norm": 0.010260739363729954, + "learning_rate": 3.3818990963147065e-05, + "loss": 0.0007778072264045477, + "step": 6050 + }, + { + "epoch": 20.61328790459966, + "grad_norm": 0.009722772054374218, + "learning_rate": 3.367044433964338e-05, + "loss": 0.0007769105955958367, + "step": 6060 + }, + { + "epoch": 20.647359454855195, + "grad_norm": 0.012824216857552528, + "learning_rate": 3.352205888226156e-05, + "loss": 0.0007471825927495956, + "step": 6070 + }, + { + "epoch": 20.68143100511073, + "grad_norm": 0.01087024062871933, + "learning_rate": 3.337383605550615e-05, + "loss": 0.000748574547469616, + "step": 6080 + }, + { + "epoch": 20.715502555366268, + "grad_norm": 0.011229326948523521, + "learning_rate": 3.3225777322276616e-05, + "loss": 0.0007508057169616222, + "step": 6090 + }, + { + "epoch": 20.749574105621807, + "grad_norm": 0.011968586593866348, + "learning_rate": 3.3077884143852886e-05, + "loss": 0.0007394180633127689, + "step": 6100 + }, + { + "epoch": 20.783645655877343, + "grad_norm": 0.012026949785649776, + "learning_rate": 3.2930157979880925e-05, + "loss": 0.000751538434997201, + "step": 6110 + }, + { + "epoch": 20.81771720613288, + "grad_norm": 0.009909541346132755, + "learning_rate": 3.278260028835832e-05, + "loss": 0.0007217846345156431, + "step": 6120 + }, + { + "epoch": 20.851788756388416, + "grad_norm": 0.010897542349994183, + "learning_rate": 3.263521252561993e-05, + "loss": 0.0007196740712970495, + "step": 6130 + }, + { + "epoch": 20.885860306643952, + "grad_norm": 0.008069319650530815, + "learning_rate": 3.2487996146323465e-05, + "loss": 0.0007380717433989048, + "step": 6140 + }, + { + "epoch": 20.91993185689949, + "grad_norm": 0.00971982441842556, + "learning_rate": 3.234095260343515e-05, + "loss": 0.0007068548817187548, + "step": 6150 + }, + { + "epoch": 20.954003407155025, + "grad_norm": 0.008872823789715767, + "learning_rate": 3.21940833482154e-05, + "loss": 0.0007374713197350502, + "step": 6160 + }, + { + "epoch": 20.98807495741056, + "grad_norm": 0.011516732163727283, + "learning_rate": 3.204738983020445e-05, + "loss": 0.0006894583813846112, + "step": 6170 + }, + { + "epoch": 21.020442930153322, + "grad_norm": 0.008599553257226944, + "learning_rate": 3.190087349720813e-05, + "loss": 0.0006507616024464368, + "step": 6180 + }, + { + "epoch": 21.05451448040886, + "grad_norm": 0.006778911221772432, + "learning_rate": 3.175453579528347e-05, + "loss": 0.000645681144669652, + "step": 6190 + }, + { + "epoch": 21.088586030664395, + "grad_norm": 0.010211668908596039, + "learning_rate": 3.160837816872452e-05, + "loss": 0.0007401840761303901, + "step": 6200 + }, + { + "epoch": 21.12265758091993, + "grad_norm": 0.011292105540633202, + "learning_rate": 3.146240206004805e-05, + "loss": 0.0006437677424401045, + "step": 6210 + }, + { + "epoch": 21.156729131175467, + "grad_norm": 0.008116209879517555, + "learning_rate": 3.1316608909979306e-05, + "loss": 0.0006522554904222488, + "step": 6220 + }, + { + "epoch": 21.190800681431003, + "grad_norm": 0.009086155332624912, + "learning_rate": 3.1171000157437835e-05, + "loss": 0.0006358759943395853, + "step": 6230 + }, + { + "epoch": 21.224872231686543, + "grad_norm": 0.008314697071909904, + "learning_rate": 3.102557723952324e-05, + "loss": 0.0006720889359712601, + "step": 6240 + }, + { + "epoch": 21.25894378194208, + "grad_norm": 0.007924322038888931, + "learning_rate": 3.088034159150102e-05, + "loss": 0.0006351920310407877, + "step": 6250 + }, + { + "epoch": 21.293015332197616, + "grad_norm": 0.006983079947531223, + "learning_rate": 3.073529464678835e-05, + "loss": 0.0006468505132943391, + "step": 6260 + }, + { + "epoch": 21.327086882453152, + "grad_norm": 0.009893915615975857, + "learning_rate": 3.059043783694005e-05, + "loss": 0.0006170779000967741, + "step": 6270 + }, + { + "epoch": 21.361158432708688, + "grad_norm": 0.009398180991411209, + "learning_rate": 3.0445772591634336e-05, + "loss": 0.000634445482864976, + "step": 6280 + }, + { + "epoch": 21.395229982964224, + "grad_norm": 0.007430765312165022, + "learning_rate": 3.030130033865879e-05, + "loss": 0.0006386633031070232, + "step": 6290 + }, + { + "epoch": 21.42930153321976, + "grad_norm": 0.009238178841769695, + "learning_rate": 3.0157022503896203e-05, + "loss": 0.0006485390476882458, + "step": 6300 + }, + { + "epoch": 21.463373083475297, + "grad_norm": 0.010283173061907291, + "learning_rate": 3.001294051131058e-05, + "loss": 0.0006661145482212306, + "step": 6310 + }, + { + "epoch": 21.497444633730836, + "grad_norm": 0.008291657082736492, + "learning_rate": 2.9869055782933e-05, + "loss": 0.000661746459081769, + "step": 6320 + }, + { + "epoch": 21.531516183986373, + "grad_norm": 0.009388615377247334, + "learning_rate": 2.9725369738847653e-05, + "loss": 0.0006240284070372582, + "step": 6330 + }, + { + "epoch": 21.56558773424191, + "grad_norm": 0.008348040282726288, + "learning_rate": 2.958188379717779e-05, + "loss": 0.0006332620978355408, + "step": 6340 + }, + { + "epoch": 21.599659284497445, + "grad_norm": 0.00879604835063219, + "learning_rate": 2.9438599374071725e-05, + "loss": 0.0007023872807621956, + "step": 6350 + }, + { + "epoch": 21.63373083475298, + "grad_norm": 0.009466956369578838, + "learning_rate": 2.9295517883688855e-05, + "loss": 0.0006332709453999996, + "step": 6360 + }, + { + "epoch": 21.667802385008518, + "grad_norm": 0.008182782679796219, + "learning_rate": 2.9152640738185733e-05, + "loss": 0.0006574525032192469, + "step": 6370 + }, + { + "epoch": 21.701873935264054, + "grad_norm": 0.00854550488293171, + "learning_rate": 2.9009969347702116e-05, + "loss": 0.0006626255810260773, + "step": 6380 + }, + { + "epoch": 21.73594548551959, + "grad_norm": 0.009672138839960098, + "learning_rate": 2.886750512034703e-05, + "loss": 0.0006541370879858732, + "step": 6390 + }, + { + "epoch": 21.770017035775126, + "grad_norm": 0.007534137461334467, + "learning_rate": 2.872524946218484e-05, + "loss": 0.0005920753348618746, + "step": 6400 + }, + { + "epoch": 21.804088586030666, + "grad_norm": 0.00771123543381691, + "learning_rate": 2.8583203777221524e-05, + "loss": 0.0005872510373592377, + "step": 6410 + }, + { + "epoch": 21.838160136286202, + "grad_norm": 0.009504510089755058, + "learning_rate": 2.844136946739059e-05, + "loss": 0.0006363318767398596, + "step": 6420 + }, + { + "epoch": 21.87223168654174, + "grad_norm": 0.009700518101453781, + "learning_rate": 2.8299747932539468e-05, + "loss": 0.0006262643262743949, + "step": 6430 + }, + { + "epoch": 21.906303236797275, + "grad_norm": 0.010623015463352203, + "learning_rate": 2.8158340570415477e-05, + "loss": 0.0006218648049980402, + "step": 6440 + }, + { + "epoch": 21.94037478705281, + "grad_norm": 0.008022764697670937, + "learning_rate": 2.8017148776652246e-05, + "loss": 0.0006008680909872055, + "step": 6450 + }, + { + "epoch": 21.974446337308347, + "grad_norm": 0.0077269612811505795, + "learning_rate": 2.7876173944755733e-05, + "loss": 0.0006492588203400373, + "step": 6460 + }, + { + "epoch": 22.00681431005111, + "grad_norm": 0.007539920508861542, + "learning_rate": 2.7735417466090652e-05, + "loss": 0.0006455070339143276, + "step": 6470 + }, + { + "epoch": 22.040885860306645, + "grad_norm": 0.00820513628423214, + "learning_rate": 2.759488072986661e-05, + "loss": 0.0005961808376014233, + "step": 6480 + }, + { + "epoch": 22.07495741056218, + "grad_norm": 0.007629638072103262, + "learning_rate": 2.7454565123124432e-05, + "loss": 0.0005448696203529835, + "step": 6490 + }, + { + "epoch": 22.109028960817717, + "grad_norm": 0.007591860368847847, + "learning_rate": 2.731447203072255e-05, + "loss": 0.0005985318217426539, + "step": 6500 + }, + { + "epoch": 22.143100511073254, + "grad_norm": 0.008149676024913788, + "learning_rate": 2.7174602835323182e-05, + "loss": 0.0006121289450675249, + "step": 6510 + }, + { + "epoch": 22.17717206132879, + "grad_norm": 0.006638308987021446, + "learning_rate": 2.7034958917378867e-05, + "loss": 0.0005916266702115536, + "step": 6520 + }, + { + "epoch": 22.211243611584326, + "grad_norm": 0.007917654700577259, + "learning_rate": 2.6895541655118652e-05, + "loss": 0.0006176072172820568, + "step": 6530 + }, + { + "epoch": 22.245315161839862, + "grad_norm": 0.007654066663235426, + "learning_rate": 2.6756352424534674e-05, + "loss": 0.0006128148175776005, + "step": 6540 + }, + { + "epoch": 22.279386712095402, + "grad_norm": 0.008756019175052643, + "learning_rate": 2.6617392599368406e-05, + "loss": 0.000577111728489399, + "step": 6550 + }, + { + "epoch": 22.313458262350938, + "grad_norm": 0.00724271172657609, + "learning_rate": 2.6478663551097267e-05, + "loss": 0.0005699345842003822, + "step": 6560 + }, + { + "epoch": 22.347529812606474, + "grad_norm": 0.007359336595982313, + "learning_rate": 2.6340166648920905e-05, + "loss": 0.0005645026452839374, + "step": 6570 + }, + { + "epoch": 22.38160136286201, + "grad_norm": 0.00949008483439684, + "learning_rate": 2.620190325974788e-05, + "loss": 0.0005713486578315496, + "step": 6580 + }, + { + "epoch": 22.415672913117547, + "grad_norm": 0.009292861446738243, + "learning_rate": 2.6063874748182e-05, + "loss": 0.00060190767981112, + "step": 6590 + }, + { + "epoch": 22.449744463373083, + "grad_norm": 0.006737905088812113, + "learning_rate": 2.5926082476508928e-05, + "loss": 0.0005614507477730513, + "step": 6600 + }, + { + "epoch": 22.48381601362862, + "grad_norm": 0.0063851517625153065, + "learning_rate": 2.5788527804682804e-05, + "loss": 0.0005616650916635991, + "step": 6610 + }, + { + "epoch": 22.517887563884155, + "grad_norm": 0.007453413680195808, + "learning_rate": 2.565121209031266e-05, + "loss": 0.0005592905450612307, + "step": 6620 + }, + { + "epoch": 22.55195911413969, + "grad_norm": 0.006294765509665012, + "learning_rate": 2.55141366886492e-05, + "loss": 0.0005961489398032427, + "step": 6630 + }, + { + "epoch": 22.58603066439523, + "grad_norm": 0.007724790833890438, + "learning_rate": 2.5377302952571257e-05, + "loss": 0.0005565430037677288, + "step": 6640 + }, + { + "epoch": 22.620102214650768, + "grad_norm": 0.010858775116503239, + "learning_rate": 2.5240712232572618e-05, + "loss": 0.0006014471407979726, + "step": 6650 + }, + { + "epoch": 22.654173764906304, + "grad_norm": 0.007493149023503065, + "learning_rate": 2.510436587674848e-05, + "loss": 0.000552827725186944, + "step": 6660 + }, + { + "epoch": 22.68824531516184, + "grad_norm": 0.008717391639947891, + "learning_rate": 2.4968265230782394e-05, + "loss": 0.0005557754542678594, + "step": 6670 + }, + { + "epoch": 22.722316865417376, + "grad_norm": 0.008045868948101997, + "learning_rate": 2.4832411637932757e-05, + "loss": 0.0005558242090046405, + "step": 6680 + }, + { + "epoch": 22.756388415672912, + "grad_norm": 0.008209879510104656, + "learning_rate": 2.4696806439019664e-05, + "loss": 0.0005993363447487354, + "step": 6690 + }, + { + "epoch": 22.79045996592845, + "grad_norm": 0.007669121492654085, + "learning_rate": 2.456145097241173e-05, + "loss": 0.000534927286207676, + "step": 6700 + }, + { + "epoch": 22.824531516183985, + "grad_norm": 0.007663552183657885, + "learning_rate": 2.442634657401272e-05, + "loss": 0.0005680276546627284, + "step": 6710 + }, + { + "epoch": 22.858603066439525, + "grad_norm": 0.00621365662664175, + "learning_rate": 2.4291494577248545e-05, + "loss": 0.0005532774608582258, + "step": 6720 + }, + { + "epoch": 22.89267461669506, + "grad_norm": 0.00827084481716156, + "learning_rate": 2.415689631305393e-05, + "loss": 0.0005640100222080946, + "step": 6730 + }, + { + "epoch": 22.926746166950597, + "grad_norm": 0.008009429089725018, + "learning_rate": 2.4022553109859435e-05, + "loss": 0.0005609836429357528, + "step": 6740 + }, + { + "epoch": 22.960817717206133, + "grad_norm": 0.006816904991865158, + "learning_rate": 2.388846629357821e-05, + "loss": 0.0005942148622125387, + "step": 6750 + }, + { + "epoch": 22.99488926746167, + "grad_norm": 0.012693436816334724, + "learning_rate": 2.3754637187593053e-05, + "loss": 0.0006112535484135151, + "step": 6760 + }, + { + "epoch": 23.027257240204428, + "grad_norm": 0.006464693695306778, + "learning_rate": 2.3621067112743134e-05, + "loss": 0.0005183239933103323, + "step": 6770 + }, + { + "epoch": 23.061328790459967, + "grad_norm": 0.007488445844501257, + "learning_rate": 2.3487757387311227e-05, + "loss": 0.0005513842683285475, + "step": 6780 + }, + { + "epoch": 23.095400340715504, + "grad_norm": 0.005201038904488087, + "learning_rate": 2.335470932701046e-05, + "loss": 0.0005259967874735594, + "step": 6790 + }, + { + "epoch": 23.12947189097104, + "grad_norm": 0.006821426097303629, + "learning_rate": 2.322192424497152e-05, + "loss": 0.0005407142918556929, + "step": 6800 + }, + { + "epoch": 23.163543441226576, + "grad_norm": 0.007693462539464235, + "learning_rate": 2.3089403451729536e-05, + "loss": 0.0005513488315045834, + "step": 6810 + }, + { + "epoch": 23.197614991482112, + "grad_norm": 0.008380657993257046, + "learning_rate": 2.2957148255211215e-05, + "loss": 0.0005209746304899454, + "step": 6820 + }, + { + "epoch": 23.23168654173765, + "grad_norm": 0.0061517683789134026, + "learning_rate": 2.2825159960721988e-05, + "loss": 0.0005351031199097633, + "step": 6830 + }, + { + "epoch": 23.265758091993185, + "grad_norm": 0.00597794633358717, + "learning_rate": 2.2693439870933007e-05, + "loss": 0.0005202882923185826, + "step": 6840 + }, + { + "epoch": 23.29982964224872, + "grad_norm": 0.008325228467583656, + "learning_rate": 2.2561989285868417e-05, + "loss": 0.0005216157529503107, + "step": 6850 + }, + { + "epoch": 23.33390119250426, + "grad_norm": 0.006761698052287102, + "learning_rate": 2.2430809502892382e-05, + "loss": 0.0005191454198211432, + "step": 6860 + }, + { + "epoch": 23.367972742759797, + "grad_norm": 0.007872886024415493, + "learning_rate": 2.2299901816696438e-05, + "loss": 0.0004972329363226891, + "step": 6870 + }, + { + "epoch": 23.402044293015333, + "grad_norm": 0.007851509377360344, + "learning_rate": 2.2169267519286563e-05, + "loss": 0.0005328899715095758, + "step": 6880 + }, + { + "epoch": 23.43611584327087, + "grad_norm": 0.0064047458581626415, + "learning_rate": 2.2038907899970568e-05, + "loss": 0.0005287681706249714, + "step": 6890 + }, + { + "epoch": 23.470187393526405, + "grad_norm": 0.005854478571563959, + "learning_rate": 2.1908824245345257e-05, + "loss": 0.0004978588782250881, + "step": 6900 + }, + { + "epoch": 23.50425894378194, + "grad_norm": 0.007258667144924402, + "learning_rate": 2.1779017839283754e-05, + "loss": 0.0005199769511818886, + "step": 6910 + }, + { + "epoch": 23.538330494037478, + "grad_norm": 0.008400578051805496, + "learning_rate": 2.1649489962922924e-05, + "loss": 0.0005408741533756256, + "step": 6920 + }, + { + "epoch": 23.572402044293014, + "grad_norm": 0.006624211557209492, + "learning_rate": 2.1520241894650585e-05, + "loss": 0.0005609713029116392, + "step": 6930 + }, + { + "epoch": 23.606473594548554, + "grad_norm": 0.007774651516228914, + "learning_rate": 2.139127491009301e-05, + "loss": 0.000538663286715746, + "step": 6940 + }, + { + "epoch": 23.64054514480409, + "grad_norm": 0.0059646302834153175, + "learning_rate": 2.126259028210225e-05, + "loss": 0.0005345881450921297, + "step": 6950 + }, + { + "epoch": 23.674616695059626, + "grad_norm": 0.007310234010219574, + "learning_rate": 2.1134189280743665e-05, + "loss": 0.0005549527239054441, + "step": 6960 + }, + { + "epoch": 23.708688245315162, + "grad_norm": 0.007316111586987972, + "learning_rate": 2.1006073173283257e-05, + "loss": 0.0005422893911600113, + "step": 6970 + }, + { + "epoch": 23.7427597955707, + "grad_norm": 0.0071487282402813435, + "learning_rate": 2.0878243224175327e-05, + "loss": 0.0005402588285505772, + "step": 6980 + }, + { + "epoch": 23.776831345826235, + "grad_norm": 0.007640121970325708, + "learning_rate": 2.0750700695049847e-05, + "loss": 0.0005570085719227791, + "step": 6990 + }, + { + "epoch": 23.81090289608177, + "grad_norm": 0.007600583601742983, + "learning_rate": 2.062344684470007e-05, + "loss": 0.0005309465806931258, + "step": 7000 + }, + { + "epoch": 23.844974446337307, + "grad_norm": 0.007132083177566528, + "learning_rate": 2.0496482929070155e-05, + "loss": 0.0005037210881710053, + "step": 7010 + }, + { + "epoch": 23.879045996592843, + "grad_norm": 0.007993524894118309, + "learning_rate": 2.036981020124266e-05, + "loss": 0.0005222490988671779, + "step": 7020 + }, + { + "epoch": 23.913117546848383, + "grad_norm": 0.007271486334502697, + "learning_rate": 2.024342991142631e-05, + "loss": 0.0005064224824309349, + "step": 7030 + }, + { + "epoch": 23.94718909710392, + "grad_norm": 0.0062245652079582214, + "learning_rate": 2.01173433069435e-05, + "loss": 0.0005136637017130851, + "step": 7040 + }, + { + "epoch": 23.981260647359456, + "grad_norm": 0.007592536974698305, + "learning_rate": 1.9991551632218154e-05, + "loss": 0.0005042028613388538, + "step": 7050 + }, + { + "epoch": 24.013628620102214, + "grad_norm": 0.007376953959465027, + "learning_rate": 1.9866056128763278e-05, + "loss": 0.0005108125973492861, + "step": 7060 + }, + { + "epoch": 24.04770017035775, + "grad_norm": 0.0060317316092550755, + "learning_rate": 1.9740858035168857e-05, + "loss": 0.0004809682257473469, + "step": 7070 + }, + { + "epoch": 24.081771720613286, + "grad_norm": 0.005852894391864538, + "learning_rate": 1.961595858708953e-05, + "loss": 0.0004623830318450928, + "step": 7080 + }, + { + "epoch": 24.115843270868826, + "grad_norm": 0.005669396836310625, + "learning_rate": 1.949135901723239e-05, + "loss": 0.00048606968484818933, + "step": 7090 + }, + { + "epoch": 24.149914821124362, + "grad_norm": 0.006334891077131033, + "learning_rate": 1.9367060555344934e-05, + "loss": 0.00046719731763005256, + "step": 7100 + }, + { + "epoch": 24.1839863713799, + "grad_norm": 0.00437652925029397, + "learning_rate": 1.924306442820276e-05, + "loss": 0.0004909945651888848, + "step": 7110 + }, + { + "epoch": 24.218057921635435, + "grad_norm": 0.006320801097899675, + "learning_rate": 1.9119371859597623e-05, + "loss": 0.0004730307962745428, + "step": 7120 + }, + { + "epoch": 24.25212947189097, + "grad_norm": 0.00540345162153244, + "learning_rate": 1.8995984070325197e-05, + "loss": 0.0005037723109126091, + "step": 7130 + }, + { + "epoch": 24.286201022146507, + "grad_norm": 0.005003531463444233, + "learning_rate": 1.8872902278173188e-05, + "loss": 0.00047008455730974675, + "step": 7140 + }, + { + "epoch": 24.320272572402043, + "grad_norm": 0.006603986956179142, + "learning_rate": 1.8750127697909154e-05, + "loss": 0.0005003036931157112, + "step": 7150 + }, + { + "epoch": 24.35434412265758, + "grad_norm": 0.005954455118626356, + "learning_rate": 1.862766154126868e-05, + "loss": 0.0005137852858752012, + "step": 7160 + }, + { + "epoch": 24.38841567291312, + "grad_norm": 0.007608614396303892, + "learning_rate": 1.8505505016943252e-05, + "loss": 0.0005309767089784146, + "step": 7170 + }, + { + "epoch": 24.422487223168655, + "grad_norm": 0.005628710612654686, + "learning_rate": 1.838365933056848e-05, + "loss": 0.0004935880191624164, + "step": 7180 + }, + { + "epoch": 24.45655877342419, + "grad_norm": 0.0062730335630476475, + "learning_rate": 1.8262125684712084e-05, + "loss": 0.0004809846635907888, + "step": 7190 + }, + { + "epoch": 24.490630323679728, + "grad_norm": 0.00636539701372385, + "learning_rate": 1.814090527886206e-05, + "loss": 0.0004986200947314501, + "step": 7200 + }, + { + "epoch": 24.524701873935264, + "grad_norm": 0.007138014771044254, + "learning_rate": 1.8019999309414916e-05, + "loss": 0.0004885056521743536, + "step": 7210 + }, + { + "epoch": 24.5587734241908, + "grad_norm": 0.00615114439278841, + "learning_rate": 1.7899408969663723e-05, + "loss": 0.0004942286293953657, + "step": 7220 + }, + { + "epoch": 24.592844974446336, + "grad_norm": 0.006483121309429407, + "learning_rate": 1.7779135449786482e-05, + "loss": 0.00047507346607744694, + "step": 7230 + }, + { + "epoch": 24.626916524701873, + "grad_norm": 0.005928986705839634, + "learning_rate": 1.7659179936834246e-05, + "loss": 0.0004952204413712025, + "step": 7240 + }, + { + "epoch": 24.66098807495741, + "grad_norm": 0.007677637506276369, + "learning_rate": 1.753954361471954e-05, + "loss": 0.0005071885883808136, + "step": 7250 + }, + { + "epoch": 24.69505962521295, + "grad_norm": 0.00708012655377388, + "learning_rate": 1.7420227664204514e-05, + "loss": 0.0004928736947476864, + "step": 7260 + }, + { + "epoch": 24.729131175468485, + "grad_norm": 0.0056620873510837555, + "learning_rate": 1.7301233262889475e-05, + "loss": 0.0004892120603471994, + "step": 7270 + }, + { + "epoch": 24.76320272572402, + "grad_norm": 0.006079996936023235, + "learning_rate": 1.7182561585201108e-05, + "loss": 0.0004955714568495751, + "step": 7280 + }, + { + "epoch": 24.797274275979557, + "grad_norm": 0.006458075251430273, + "learning_rate": 1.7064213802380942e-05, + "loss": 0.00049555622972548, + "step": 7290 + }, + { + "epoch": 24.831345826235093, + "grad_norm": 0.0053098793141543865, + "learning_rate": 1.6946191082473846e-05, + "loss": 0.0005073010921478272, + "step": 7300 + }, + { + "epoch": 24.86541737649063, + "grad_norm": 0.005432505626231432, + "learning_rate": 1.682849459031639e-05, + "loss": 0.0004960733465850353, + "step": 7310 + }, + { + "epoch": 24.899488926746166, + "grad_norm": 0.0077927736565470695, + "learning_rate": 1.6711125487525464e-05, + "loss": 0.0005189849995076657, + "step": 7320 + }, + { + "epoch": 24.933560477001702, + "grad_norm": 0.00717386556789279, + "learning_rate": 1.6594084932486697e-05, + "loss": 0.0004941137973219156, + "step": 7330 + }, + { + "epoch": 24.967632027257242, + "grad_norm": 0.006342039443552494, + "learning_rate": 1.647737408034315e-05, + "loss": 0.00052413921803236, + "step": 7340 + }, + { + "epoch": 25.0, + "grad_norm": 0.0038004769012331963, + "learning_rate": 1.6360994082983775e-05, + "loss": 0.0004258518572896719, + "step": 7350 + }, + { + "epoch": 25.034071550255536, + "grad_norm": 0.005416833329945803, + "learning_rate": 1.6244946089032216e-05, + "loss": 0.0004529508762061596, + "step": 7360 + }, + { + "epoch": 25.068143100511072, + "grad_norm": 0.005682576447725296, + "learning_rate": 1.612923124383528e-05, + "loss": 0.00044956523925065994, + "step": 7370 + }, + { + "epoch": 25.10221465076661, + "grad_norm": 0.006330459378659725, + "learning_rate": 1.601385068945177e-05, + "loss": 0.00044929394498467446, + "step": 7380 + }, + { + "epoch": 25.136286201022145, + "grad_norm": 0.00562339136376977, + "learning_rate": 1.589880556464121e-05, + "loss": 0.0004864254035055637, + "step": 7390 + }, + { + "epoch": 25.170357751277685, + "grad_norm": 0.005379036534577608, + "learning_rate": 1.5784097004852505e-05, + "loss": 0.0004678622353821993, + "step": 7400 + }, + { + "epoch": 25.20442930153322, + "grad_norm": 0.006202719174325466, + "learning_rate": 1.5669726142212855e-05, + "loss": 0.00045688869431614876, + "step": 7410 + }, + { + "epoch": 25.238500851788757, + "grad_norm": 0.007115985732525587, + "learning_rate": 1.555569410551647e-05, + "loss": 0.0004639118444174528, + "step": 7420 + }, + { + "epoch": 25.272572402044293, + "grad_norm": 0.005236926954239607, + "learning_rate": 1.5442002020213558e-05, + "loss": 0.000463517289608717, + "step": 7430 + }, + { + "epoch": 25.30664395229983, + "grad_norm": 0.006341162603348494, + "learning_rate": 1.5328651008399057e-05, + "loss": 0.0004880161490291357, + "step": 7440 + }, + { + "epoch": 25.340715502555366, + "grad_norm": 0.006201290525496006, + "learning_rate": 1.5215642188801732e-05, + "loss": 0.0004617507569491863, + "step": 7450 + }, + { + "epoch": 25.374787052810902, + "grad_norm": 0.004709569271653891, + "learning_rate": 1.510297667677299e-05, + "loss": 0.0004495989065617323, + "step": 7460 + }, + { + "epoch": 25.408858603066438, + "grad_norm": 0.005513674113899469, + "learning_rate": 1.4990655584275931e-05, + "loss": 0.000475541315972805, + "step": 7470 + }, + { + "epoch": 25.442930153321978, + "grad_norm": 0.007107003126293421, + "learning_rate": 1.4878680019874431e-05, + "loss": 0.0004677797202020884, + "step": 7480 + }, + { + "epoch": 25.477001703577514, + "grad_norm": 0.005166848190128803, + "learning_rate": 1.4767051088722067e-05, + "loss": 0.00047347424551844595, + "step": 7490 + }, + { + "epoch": 25.51107325383305, + "grad_norm": 0.005388825200498104, + "learning_rate": 1.4655769892551357e-05, + "loss": 0.0004688589368015528, + "step": 7500 + }, + { + "epoch": 25.545144804088586, + "grad_norm": 0.005632973276078701, + "learning_rate": 1.4544837529662748e-05, + "loss": 0.0004677624441683292, + "step": 7510 + }, + { + "epoch": 25.579216354344123, + "grad_norm": 0.005388225428760052, + "learning_rate": 1.4434255094913906e-05, + "loss": 0.0004646487534046173, + "step": 7520 + }, + { + "epoch": 25.61328790459966, + "grad_norm": 0.006038753315806389, + "learning_rate": 1.4324023679708792e-05, + "loss": 0.00045312829315662383, + "step": 7530 + }, + { + "epoch": 25.647359454855195, + "grad_norm": 0.005922420881688595, + "learning_rate": 1.4214144371987004e-05, + "loss": 0.00048261820338666437, + "step": 7540 + }, + { + "epoch": 25.68143100511073, + "grad_norm": 0.0061474149115383625, + "learning_rate": 1.4104618256212926e-05, + "loss": 0.00046353056095540526, + "step": 7550 + }, + { + "epoch": 25.715502555366268, + "grad_norm": 0.005116403568536043, + "learning_rate": 1.3995446413365093e-05, + "loss": 0.0004678342491388321, + "step": 7560 + }, + { + "epoch": 25.749574105621807, + "grad_norm": 0.00608143862336874, + "learning_rate": 1.3886629920925548e-05, + "loss": 0.00045505915768444537, + "step": 7570 + }, + { + "epoch": 25.783645655877343, + "grad_norm": 0.005517200101166964, + "learning_rate": 1.377816985286911e-05, + "loss": 0.0004353174939751625, + "step": 7580 + }, + { + "epoch": 25.81771720613288, + "grad_norm": 0.007087164092808962, + "learning_rate": 1.3670067279652899e-05, + "loss": 0.0004676383920013905, + "step": 7590 + }, + { + "epoch": 25.851788756388416, + "grad_norm": 0.005983814597129822, + "learning_rate": 1.3562323268205629e-05, + "loss": 0.0004769494757056236, + "step": 7600 + }, + { + "epoch": 25.885860306643952, + "grad_norm": 0.0058109392412006855, + "learning_rate": 1.345493888191724e-05, + "loss": 0.00048498655669391153, + "step": 7610 + }, + { + "epoch": 25.91993185689949, + "grad_norm": 0.006097118835896254, + "learning_rate": 1.3347915180628228e-05, + "loss": 0.00043746624141931533, + "step": 7620 + }, + { + "epoch": 25.954003407155025, + "grad_norm": 0.005795330740511417, + "learning_rate": 1.3241253220619355e-05, + "loss": 0.00046371580101549624, + "step": 7630 + }, + { + "epoch": 25.98807495741056, + "grad_norm": 0.007155901286751032, + "learning_rate": 1.3134954054601083e-05, + "loss": 0.0004725466016680002, + "step": 7640 + }, + { + "epoch": 26.020442930153322, + "grad_norm": 0.0056617953814566135, + "learning_rate": 1.3029018731703263e-05, + "loss": 0.0004797719419002533, + "step": 7650 + }, + { + "epoch": 26.05451448040886, + "grad_norm": 0.005709235090762377, + "learning_rate": 1.2923448297464785e-05, + "loss": 0.0004465170670300722, + "step": 7660 + }, + { + "epoch": 26.088586030664395, + "grad_norm": 0.006274583749473095, + "learning_rate": 1.2818243793823188e-05, + "loss": 0.0004596684128046036, + "step": 7670 + }, + { + "epoch": 26.12265758091993, + "grad_norm": 0.004958844743669033, + "learning_rate": 1.271340625910447e-05, + "loss": 0.00043697194196283816, + "step": 7680 + }, + { + "epoch": 26.156729131175467, + "grad_norm": 0.005463967099785805, + "learning_rate": 1.260893672801276e-05, + "loss": 0.0004823775961995125, + "step": 7690 + }, + { + "epoch": 26.190800681431003, + "grad_norm": 0.005295682232826948, + "learning_rate": 1.2504836231620176e-05, + "loss": 0.00044223233126103876, + "step": 7700 + }, + { + "epoch": 26.224872231686543, + "grad_norm": 0.005236088298261166, + "learning_rate": 1.2401105797356566e-05, + "loss": 0.0004612309858202934, + "step": 7710 + }, + { + "epoch": 26.25894378194208, + "grad_norm": 0.004934109281748533, + "learning_rate": 1.2297746448999453e-05, + "loss": 0.0004583748523145914, + "step": 7720 + }, + { + "epoch": 26.293015332197616, + "grad_norm": 0.005991233978420496, + "learning_rate": 1.2194759206663881e-05, + "loss": 0.00044856127351522446, + "step": 7730 + }, + { + "epoch": 26.327086882453152, + "grad_norm": 0.004782292991876602, + "learning_rate": 1.2092145086792327e-05, + "loss": 0.00042555080726742746, + "step": 7740 + }, + { + "epoch": 26.361158432708688, + "grad_norm": 0.005826896987855434, + "learning_rate": 1.198990510214476e-05, + "loss": 0.00044993571937084197, + "step": 7750 + }, + { + "epoch": 26.395229982964224, + "grad_norm": 0.005082950461655855, + "learning_rate": 1.188804026178853e-05, + "loss": 0.0004447337705641985, + "step": 7760 + }, + { + "epoch": 26.42930153321976, + "grad_norm": 0.005609357729554176, + "learning_rate": 1.1786551571088506e-05, + "loss": 0.0004195962566882372, + "step": 7770 + }, + { + "epoch": 26.463373083475297, + "grad_norm": 0.005520106293261051, + "learning_rate": 1.168544003169707e-05, + "loss": 0.00043586334213614465, + "step": 7780 + }, + { + "epoch": 26.497444633730836, + "grad_norm": 0.005495567340403795, + "learning_rate": 1.1584706641544319e-05, + "loss": 0.00042241401970386503, + "step": 7790 + }, + { + "epoch": 26.531516183986373, + "grad_norm": 0.005580009426921606, + "learning_rate": 1.1484352394828118e-05, + "loss": 0.00042024566791951654, + "step": 7800 + }, + { + "epoch": 26.56558773424191, + "grad_norm": 0.005603221245110035, + "learning_rate": 1.1384378282004404e-05, + "loss": 0.0004328824579715729, + "step": 7810 + }, + { + "epoch": 26.599659284497445, + "grad_norm": 0.0046456316486001015, + "learning_rate": 1.1284785289777272e-05, + "loss": 0.0004326967056840658, + "step": 7820 + }, + { + "epoch": 26.63373083475298, + "grad_norm": 0.005678926128894091, + "learning_rate": 1.1185574401089383e-05, + "loss": 0.0004535626154392958, + "step": 7830 + }, + { + "epoch": 26.667802385008518, + "grad_norm": 0.00526011735200882, + "learning_rate": 1.1086746595112136e-05, + "loss": 0.00044193663634359834, + "step": 7840 + }, + { + "epoch": 26.701873935264054, + "grad_norm": 0.005378768313676119, + "learning_rate": 1.0988302847236066e-05, + "loss": 0.0004401818383485079, + "step": 7850 + }, + { + "epoch": 26.73594548551959, + "grad_norm": 0.0056685502640903, + "learning_rate": 1.0890244129061245e-05, + "loss": 0.0004342911299318075, + "step": 7860 + }, + { + "epoch": 26.770017035775126, + "grad_norm": 0.005207818001508713, + "learning_rate": 1.0792571408387608e-05, + "loss": 0.000432520778849721, + "step": 7870 + }, + { + "epoch": 26.804088586030666, + "grad_norm": 0.005392038729041815, + "learning_rate": 1.069528564920551e-05, + "loss": 0.0004403261002153158, + "step": 7880 + }, + { + "epoch": 26.838160136286202, + "grad_norm": 0.005832510069012642, + "learning_rate": 1.0598387811686084e-05, + "loss": 0.00042518815025687217, + "step": 7890 + }, + { + "epoch": 26.87223168654174, + "grad_norm": 0.004946849774569273, + "learning_rate": 1.0501878852171893e-05, + "loss": 0.0004303318914026022, + "step": 7900 + }, + { + "epoch": 26.906303236797275, + "grad_norm": 0.0063912533223629, + "learning_rate": 1.0405759723167385e-05, + "loss": 0.00044775954447686674, + "step": 7910 + }, + { + "epoch": 26.94037478705281, + "grad_norm": 0.006100860424339771, + "learning_rate": 1.0310031373329576e-05, + "loss": 0.0004396803677082062, + "step": 7920 + }, + { + "epoch": 26.974446337308347, + "grad_norm": 0.0057065715081989765, + "learning_rate": 1.0214694747458631e-05, + "loss": 0.00042748311534523964, + "step": 7930 + }, + { + "epoch": 27.00681431005111, + "grad_norm": 0.0068487138487398624, + "learning_rate": 1.0119750786488536e-05, + "loss": 0.00044254655949771404, + "step": 7940 + }, + { + "epoch": 27.040885860306645, + "grad_norm": 0.00538451736792922, + "learning_rate": 1.0025200427477876e-05, + "loss": 0.00041661132127046585, + "step": 7950 + }, + { + "epoch": 27.07495741056218, + "grad_norm": 0.004859358537942171, + "learning_rate": 9.931044603600514e-06, + "loss": 0.00041967052966356276, + "step": 7960 + }, + { + "epoch": 27.109028960817717, + "grad_norm": 0.006202720105648041, + "learning_rate": 9.83728424413643e-06, + "loss": 0.00044184853322803974, + "step": 7970 + }, + { + "epoch": 27.143100511073254, + "grad_norm": 0.0047525400295853615, + "learning_rate": 9.743920274462498e-06, + "loss": 0.0004158275201916695, + "step": 7980 + }, + { + "epoch": 27.17717206132879, + "grad_norm": 0.006198120769113302, + "learning_rate": 9.65095361604344e-06, + "loss": 0.00043417327105998993, + "step": 7990 + }, + { + "epoch": 27.211243611584326, + "grad_norm": 0.0066219414584338665, + "learning_rate": 9.558385186422603e-06, + "loss": 0.00042222961783409116, + "step": 8000 + }, + { + "epoch": 27.245315161839862, + "grad_norm": 0.0055702864192426205, + "learning_rate": 9.466215899213065e-06, + "loss": 0.0004271657671779394, + "step": 8010 + }, + { + "epoch": 27.279386712095402, + "grad_norm": 0.004891074262559414, + "learning_rate": 9.374446664088427e-06, + "loss": 0.0004259935114532709, + "step": 8020 + }, + { + "epoch": 27.313458262350938, + "grad_norm": 0.005272470880299807, + "learning_rate": 9.283078386774025e-06, + "loss": 0.00040517104789614675, + "step": 8030 + }, + { + "epoch": 27.347529812606474, + "grad_norm": 0.00502979988232255, + "learning_rate": 9.192111969037841e-06, + "loss": 0.00042363600805401803, + "step": 8040 + }, + { + "epoch": 27.38160136286201, + "grad_norm": 0.005891913548111916, + "learning_rate": 9.101548308681734e-06, + "loss": 0.0004411347210407257, + "step": 8050 + }, + { + "epoch": 27.415672913117547, + "grad_norm": 0.0058002774603664875, + "learning_rate": 9.011388299532441e-06, + "loss": 0.0004146874416619539, + "step": 8060 + }, + { + "epoch": 27.449744463373083, + "grad_norm": 0.005654113832861185, + "learning_rate": 8.921632831432852e-06, + "loss": 0.00042466390877962115, + "step": 8070 + }, + { + "epoch": 27.48381601362862, + "grad_norm": 0.006625378038734198, + "learning_rate": 8.832282790233221e-06, + "loss": 0.0004333822056651115, + "step": 8080 + }, + { + "epoch": 27.517887563884155, + "grad_norm": 0.005088499281555414, + "learning_rate": 8.743339057782351e-06, + "loss": 0.0004219602793455124, + "step": 8090 + }, + { + "epoch": 27.55195911413969, + "grad_norm": 0.005173454061150551, + "learning_rate": 8.654802511919013e-06, + "loss": 0.00042137070558965204, + "step": 8100 + }, + { + "epoch": 27.58603066439523, + "grad_norm": 0.005211536306887865, + "learning_rate": 8.566674026463135e-06, + "loss": 0.00042257718741893767, + "step": 8110 + }, + { + "epoch": 27.620102214650768, + "grad_norm": 0.00504072941839695, + "learning_rate": 8.478954471207317e-06, + "loss": 0.00040095774456858633, + "step": 8120 + }, + { + "epoch": 27.654173764906304, + "grad_norm": 0.004734995309263468, + "learning_rate": 8.391644711908136e-06, + "loss": 0.0004277116153389215, + "step": 8130 + }, + { + "epoch": 27.68824531516184, + "grad_norm": 0.005591129418462515, + "learning_rate": 8.304745610277687e-06, + "loss": 0.00042661316692829134, + "step": 8140 + }, + { + "epoch": 27.722316865417376, + "grad_norm": 0.004635486286133528, + "learning_rate": 8.218258023975006e-06, + "loss": 0.00041025131940841675, + "step": 8150 + }, + { + "epoch": 27.756388415672912, + "grad_norm": 0.005779591389000416, + "learning_rate": 8.13218280659765e-06, + "loss": 0.0004400585778057575, + "step": 8160 + }, + { + "epoch": 27.79045996592845, + "grad_norm": 0.005932126194238663, + "learning_rate": 8.046520807673274e-06, + "loss": 0.0004161299206316471, + "step": 8170 + }, + { + "epoch": 27.824531516183985, + "grad_norm": 0.005546386819332838, + "learning_rate": 7.961272872651221e-06, + "loss": 0.0004326997324824333, + "step": 8180 + }, + { + "epoch": 27.858603066439525, + "grad_norm": 0.004769790451973677, + "learning_rate": 7.876439842894195e-06, + "loss": 0.0003943036776036024, + "step": 8190 + }, + { + "epoch": 27.89267461669506, + "grad_norm": 0.0041409204714000225, + "learning_rate": 7.792022555669947e-06, + "loss": 0.0003955497406423092, + "step": 8200 + }, + { + "epoch": 27.926746166950597, + "grad_norm": 0.005350794177502394, + "learning_rate": 7.708021844143032e-06, + "loss": 0.000398462126031518, + "step": 8210 + }, + { + "epoch": 27.960817717206133, + "grad_norm": 0.005668057128787041, + "learning_rate": 7.6244385373665465e-06, + "loss": 0.0004361903294920921, + "step": 8220 + }, + { + "epoch": 27.99488926746167, + "grad_norm": 0.006333161145448685, + "learning_rate": 7.541273460273996e-06, + "loss": 0.00042402190156280993, + "step": 8230 + }, + { + "epoch": 28.027257240204428, + "grad_norm": 0.005466774106025696, + "learning_rate": 7.458527433671114e-06, + "loss": 0.000400385120883584, + "step": 8240 + }, + { + "epoch": 28.061328790459967, + "grad_norm": 0.0045157442800700665, + "learning_rate": 7.376201274227762e-06, + "loss": 0.00041829477995634077, + "step": 8250 + }, + { + "epoch": 28.095400340715504, + "grad_norm": 0.004654949996620417, + "learning_rate": 7.294295794469924e-06, + "loss": 0.00041680284775793555, + "step": 8260 + }, + { + "epoch": 28.12947189097104, + "grad_norm": 0.004837220534682274, + "learning_rate": 7.212811802771591e-06, + "loss": 0.00041743190959095953, + "step": 8270 + }, + { + "epoch": 28.163543441226576, + "grad_norm": 0.005391380283981562, + "learning_rate": 7.1317501033468905e-06, + "loss": 0.00041745444759726523, + "step": 8280 + }, + { + "epoch": 28.197614991482112, + "grad_norm": 0.006460465025156736, + "learning_rate": 7.051111496242047e-06, + "loss": 0.00040775276720523835, + "step": 8290 + }, + { + "epoch": 28.23168654173765, + "grad_norm": 0.004577051382511854, + "learning_rate": 6.970896777327573e-06, + "loss": 0.00042567341588437555, + "step": 8300 + }, + { + "epoch": 28.265758091993185, + "grad_norm": 0.005753329489380121, + "learning_rate": 6.891106738290348e-06, + "loss": 0.00040739565156400204, + "step": 8310 + }, + { + "epoch": 28.29982964224872, + "grad_norm": 0.006131471134722233, + "learning_rate": 6.811742166625862e-06, + "loss": 0.00041704727336764333, + "step": 8320 + }, + { + "epoch": 28.33390119250426, + "grad_norm": 0.005437177140265703, + "learning_rate": 6.732803845630392e-06, + "loss": 0.0004021385218948126, + "step": 8330 + }, + { + "epoch": 28.367972742759797, + "grad_norm": 0.005158474203199148, + "learning_rate": 6.654292554393276e-06, + "loss": 0.00040545552037656306, + "step": 8340 + }, + { + "epoch": 28.402044293015333, + "grad_norm": 0.004876469261944294, + "learning_rate": 6.576209067789285e-06, + "loss": 0.0003895287169143558, + "step": 8350 + }, + { + "epoch": 28.43611584327087, + "grad_norm": 0.004807023797184229, + "learning_rate": 6.498554156470899e-06, + "loss": 0.0004088780377060175, + "step": 8360 + }, + { + "epoch": 28.470187393526405, + "grad_norm": 0.0056871771812438965, + "learning_rate": 6.421328586860747e-06, + "loss": 0.00041042077355086805, + "step": 8370 + }, + { + "epoch": 28.50425894378194, + "grad_norm": 0.005160088650882244, + "learning_rate": 6.344533121144014e-06, + "loss": 0.0004123701248317957, + "step": 8380 + }, + { + "epoch": 28.538330494037478, + "grad_norm": 0.005306396633386612, + "learning_rate": 6.268168517260964e-06, + "loss": 0.0004113129805773497, + "step": 8390 + }, + { + "epoch": 28.572402044293014, + "grad_norm": 0.004495360888540745, + "learning_rate": 6.192235528899382e-06, + "loss": 0.0003872907254844904, + "step": 8400 + }, + { + "epoch": 28.606473594548554, + "grad_norm": 0.005054249428212643, + "learning_rate": 6.116734905487226e-06, + "loss": 0.0003946823533624411, + "step": 8410 + }, + { + "epoch": 28.64054514480409, + "grad_norm": 0.005848642438650131, + "learning_rate": 6.0416673921851576e-06, + "loss": 0.00039847791194915774, + "step": 8420 + }, + { + "epoch": 28.674616695059626, + "grad_norm": 0.004590380005538464, + "learning_rate": 5.96703372987924e-06, + "loss": 0.00040190936997532847, + "step": 8430 + }, + { + "epoch": 28.708688245315162, + "grad_norm": 0.005066314712166786, + "learning_rate": 5.892834655173579e-06, + "loss": 0.0004121654201298952, + "step": 8440 + }, + { + "epoch": 28.7427597955707, + "grad_norm": 0.005346378311514854, + "learning_rate": 5.819070900383078e-06, + "loss": 0.0004073288291692734, + "step": 8450 + }, + { + "epoch": 28.776831345826235, + "grad_norm": 0.005778873339295387, + "learning_rate": 5.745743193526237e-06, + "loss": 0.00039315088652074336, + "step": 8460 + }, + { + "epoch": 28.81090289608177, + "grad_norm": 0.004868451505899429, + "learning_rate": 5.672852258317901e-06, + "loss": 0.0003976979292929173, + "step": 8470 + }, + { + "epoch": 28.844974446337307, + "grad_norm": 0.005879078526049852, + "learning_rate": 5.600398814162194e-06, + "loss": 0.000414629140868783, + "step": 8480 + }, + { + "epoch": 28.879045996592843, + "grad_norm": 0.004795163869857788, + "learning_rate": 5.528383576145335e-06, + "loss": 0.0004018770530819893, + "step": 8490 + }, + { + "epoch": 28.913117546848383, + "grad_norm": 0.005251680500805378, + "learning_rate": 5.456807255028673e-06, + "loss": 0.0004144621081650257, + "step": 8500 + }, + { + "epoch": 28.94718909710392, + "grad_norm": 0.00586005998775363, + "learning_rate": 5.385670557241579e-06, + "loss": 0.0004176293034106493, + "step": 8510 + }, + { + "epoch": 28.981260647359456, + "grad_norm": 0.004291141871362925, + "learning_rate": 5.31497418487456e-06, + "loss": 0.00039627389051020143, + "step": 8520 + }, + { + "epoch": 29.013628620102214, + "grad_norm": 0.005097679328173399, + "learning_rate": 5.2447188356722636e-06, + "loss": 0.0003812724957242608, + "step": 8530 + }, + { + "epoch": 29.04770017035775, + "grad_norm": 0.0051975855603814125, + "learning_rate": 5.174905203026608e-06, + "loss": 0.0004123745951801538, + "step": 8540 + }, + { + "epoch": 29.081771720613286, + "grad_norm": 0.004907531198114157, + "learning_rate": 5.10553397596999e-06, + "loss": 0.00043122475035488604, + "step": 8550 + }, + { + "epoch": 29.115843270868826, + "grad_norm": 0.004903047811239958, + "learning_rate": 5.03660583916839e-06, + "loss": 0.0003974732011556625, + "step": 8560 + }, + { + "epoch": 29.149914821124362, + "grad_norm": 0.0064574237912893295, + "learning_rate": 4.968121472914705e-06, + "loss": 0.0003998386673629284, + "step": 8570 + }, + { + "epoch": 29.1839863713799, + "grad_norm": 0.004835375584661961, + "learning_rate": 4.900081553121971e-06, + "loss": 0.00038853392470628023, + "step": 8580 + }, + { + "epoch": 29.218057921635435, + "grad_norm": 0.004598340019583702, + "learning_rate": 4.832486751316745e-06, + "loss": 0.0003884763224050403, + "step": 8590 + }, + { + "epoch": 29.25212947189097, + "grad_norm": 0.004990984220057726, + "learning_rate": 4.765337734632408e-06, + "loss": 0.00037395297549664975, + "step": 8600 + }, + { + "epoch": 29.286201022146507, + "grad_norm": 0.005623216275125742, + "learning_rate": 4.698635165802673e-06, + "loss": 0.0003956631291657686, + "step": 8610 + }, + { + "epoch": 29.320272572402043, + "grad_norm": 0.005677004810422659, + "learning_rate": 4.632379703154949e-06, + "loss": 0.00038630112539976834, + "step": 8620 + }, + { + "epoch": 29.35434412265758, + "grad_norm": 0.004108254332095385, + "learning_rate": 4.566572000603892e-06, + "loss": 0.0004083862528204918, + "step": 8630 + }, + { + "epoch": 29.38841567291312, + "grad_norm": 0.005012644454836845, + "learning_rate": 4.501212707644986e-06, + "loss": 0.000413311505690217, + "step": 8640 + }, + { + "epoch": 29.422487223168655, + "grad_norm": 0.004901448264718056, + "learning_rate": 4.436302469348041e-06, + "loss": 0.0003847259562462568, + "step": 8650 + }, + { + "epoch": 29.45655877342419, + "grad_norm": 0.0046895453706383705, + "learning_rate": 4.371841926350922e-06, + "loss": 0.00038478029891848565, + "step": 8660 + }, + { + "epoch": 29.490630323679728, + "grad_norm": 0.004507383331656456, + "learning_rate": 4.30783171485315e-06, + "loss": 0.00041964929550886154, + "step": 8670 + }, + { + "epoch": 29.524701873935264, + "grad_norm": 0.005036781542003155, + "learning_rate": 4.244272466609689e-06, + "loss": 0.00039485073648393154, + "step": 8680 + }, + { + "epoch": 29.5587734241908, + "grad_norm": 0.004939622711390257, + "learning_rate": 4.1811648089246365e-06, + "loss": 0.00037607681006193163, + "step": 8690 + }, + { + "epoch": 29.592844974446336, + "grad_norm": 0.00515342690050602, + "learning_rate": 4.118509364645118e-06, + "loss": 0.00040146284736692907, + "step": 8700 + }, + { + "epoch": 29.626916524701873, + "grad_norm": 0.005482867825776339, + "learning_rate": 4.056306752155059e-06, + "loss": 0.0004012126009911299, + "step": 8710 + }, + { + "epoch": 29.66098807495741, + "grad_norm": 0.0044959550723433495, + "learning_rate": 3.994557585369135e-06, + "loss": 0.00039921659044921397, + "step": 8720 + }, + { + "epoch": 29.69505962521295, + "grad_norm": 0.004613898694515228, + "learning_rate": 3.933262473726701e-06, + "loss": 0.00038427074905484916, + "step": 8730 + }, + { + "epoch": 29.729131175468485, + "grad_norm": 0.005130247678607702, + "learning_rate": 3.872422022185746e-06, + "loss": 0.0003987327218055725, + "step": 8740 + }, + { + "epoch": 29.76320272572402, + "grad_norm": 0.006411183159798384, + "learning_rate": 3.812036831216986e-06, + "loss": 0.0004151281900703907, + "step": 8750 + }, + { + "epoch": 29.797274275979557, + "grad_norm": 0.005595535039901733, + "learning_rate": 3.7521074967978577e-06, + "loss": 0.0004142165184020996, + "step": 8760 + }, + { + "epoch": 29.831345826235093, + "grad_norm": 0.004568733274936676, + "learning_rate": 3.6926346104067143e-06, + "loss": 0.0003730351570993662, + "step": 8770 + }, + { + "epoch": 29.86541737649063, + "grad_norm": 0.005357666872441769, + "learning_rate": 3.6336187590169214e-06, + "loss": 0.00040193768218159675, + "step": 8780 + }, + { + "epoch": 29.899488926746166, + "grad_norm": 0.0057200901210308075, + "learning_rate": 3.575060525091117e-06, + "loss": 0.00038862563669681547, + "step": 8790 + }, + { + "epoch": 29.933560477001702, + "grad_norm": 0.005241156555712223, + "learning_rate": 3.5169604865754212e-06, + "loss": 0.0003931347280740738, + "step": 8800 + }, + { + "epoch": 29.967632027257242, + "grad_norm": 0.004278367850929499, + "learning_rate": 3.459319216893758e-06, + "loss": 0.0003920105751603842, + "step": 8810 + }, + { + "epoch": 30.0, + "grad_norm": 0.01224629208445549, + "learning_rate": 3.4021372849421972e-06, + "loss": 0.00035213571973145006, + "step": 8820 + }, + { + "epoch": 30.034071550255536, + "grad_norm": 0.004650902468711138, + "learning_rate": 3.3454152550833063e-06, + "loss": 0.00041070561856031417, + "step": 8830 + }, + { + "epoch": 30.068143100511072, + "grad_norm": 0.004452348221093416, + "learning_rate": 3.2891536871406393e-06, + "loss": 0.0003755549201741815, + "step": 8840 + }, + { + "epoch": 30.10221465076661, + "grad_norm": 0.004402755293995142, + "learning_rate": 3.233353136393141e-06, + "loss": 0.0003724572248756886, + "step": 8850 + }, + { + "epoch": 30.136286201022145, + "grad_norm": 0.004652101546525955, + "learning_rate": 3.1780141535697363e-06, + "loss": 0.0003761503379791975, + "step": 8860 + }, + { + "epoch": 30.170357751277685, + "grad_norm": 0.004677773453295231, + "learning_rate": 3.1231372848438243e-06, + "loss": 0.0003732655430212617, + "step": 8870 + }, + { + "epoch": 30.20442930153322, + "grad_norm": 0.0045632254332304, + "learning_rate": 3.0687230718279615e-06, + "loss": 0.0003793169045820832, + "step": 8880 + }, + { + "epoch": 30.238500851788757, + "grad_norm": 0.0050588869489729404, + "learning_rate": 3.0147720515684653e-06, + "loss": 0.00039807846769690514, + "step": 8890 + }, + { + "epoch": 30.272572402044293, + "grad_norm": 0.005203649867326021, + "learning_rate": 2.9612847565401005e-06, + "loss": 0.00038288210052996876, + "step": 8900 + }, + { + "epoch": 30.30664395229983, + "grad_norm": 0.004914138000458479, + "learning_rate": 2.908261714640903e-06, + "loss": 0.000366090820170939, + "step": 8910 + }, + { + "epoch": 30.340715502555366, + "grad_norm": 0.0043633137829601765, + "learning_rate": 2.8557034491868696e-06, + "loss": 0.0003828644519671798, + "step": 8920 + }, + { + "epoch": 30.374787052810902, + "grad_norm": 0.004355897195637226, + "learning_rate": 2.803610478906876e-06, + "loss": 0.00037504469510167836, + "step": 8930 + }, + { + "epoch": 30.408858603066438, + "grad_norm": 0.004408527631312609, + "learning_rate": 2.7519833179374952e-06, + "loss": 0.0003829608671367168, + "step": 8940 + }, + { + "epoch": 30.442930153321978, + "grad_norm": 0.004476489499211311, + "learning_rate": 2.700822475817982e-06, + "loss": 0.0003814706578850746, + "step": 8950 + }, + { + "epoch": 30.477001703577514, + "grad_norm": 0.005382033996284008, + "learning_rate": 2.6501284574851737e-06, + "loss": 0.0003860419150441885, + "step": 8960 + }, + { + "epoch": 30.51107325383305, + "grad_norm": 0.004880188032984734, + "learning_rate": 2.599901763268575e-06, + "loss": 0.0003984314389526844, + "step": 8970 + }, + { + "epoch": 30.545144804088586, + "grad_norm": 0.004750372841954231, + "learning_rate": 2.5501428888853816e-06, + "loss": 0.0004004231188446283, + "step": 8980 + }, + { + "epoch": 30.579216354344123, + "grad_norm": 0.00549035519361496, + "learning_rate": 2.5008523254355878e-06, + "loss": 0.00038932133466005323, + "step": 8990 + }, + { + "epoch": 30.61328790459966, + "grad_norm": 0.0044222003780305386, + "learning_rate": 2.452030559397167e-06, + "loss": 0.00039637424051761625, + "step": 9000 + }, + { + "epoch": 30.647359454855195, + "grad_norm": 0.005450745113193989, + "learning_rate": 2.4036780726212237e-06, + "loss": 0.0003862245939671993, + "step": 9010 + }, + { + "epoch": 30.68143100511073, + "grad_norm": 0.005370100028812885, + "learning_rate": 2.3557953423273048e-06, + "loss": 0.00040809987112879755, + "step": 9020 + }, + { + "epoch": 30.715502555366268, + "grad_norm": 0.005018346011638641, + "learning_rate": 2.3083828410986e-06, + "loss": 0.0004006767179816961, + "step": 9030 + }, + { + "epoch": 30.749574105621807, + "grad_norm": 0.005486037116497755, + "learning_rate": 2.2614410368773765e-06, + "loss": 0.00038786791265010834, + "step": 9040 + }, + { + "epoch": 30.783645655877343, + "grad_norm": 0.004210353828966618, + "learning_rate": 2.214970392960275e-06, + "loss": 0.0003764190711081028, + "step": 9050 + }, + { + "epoch": 30.81771720613288, + "grad_norm": 0.004774926695972681, + "learning_rate": 2.1689713679938105e-06, + "loss": 0.0003962088841944933, + "step": 9060 + }, + { + "epoch": 30.851788756388416, + "grad_norm": 0.00459167268127203, + "learning_rate": 2.1234444159697665e-06, + "loss": 0.0003867223160341382, + "step": 9070 + }, + { + "epoch": 30.885860306643952, + "grad_norm": 0.005524986423552036, + "learning_rate": 2.0783899862207967e-06, + "loss": 0.0003982493653893471, + "step": 9080 + }, + { + "epoch": 30.91993185689949, + "grad_norm": 0.005982302129268646, + "learning_rate": 2.03380852341592e-06, + "loss": 0.00036522489972412586, + "step": 9090 + }, + { + "epoch": 30.954003407155025, + "grad_norm": 0.004889259114861488, + "learning_rate": 1.9897004675561825e-06, + "loss": 0.0004274255596101284, + "step": 9100 + }, + { + "epoch": 30.98807495741056, + "grad_norm": 0.0048242961056530476, + "learning_rate": 1.9460662539702968e-06, + "loss": 0.0003841693280264735, + "step": 9110 + }, + { + "epoch": 31.020442930153322, + "grad_norm": 0.004305668640881777, + "learning_rate": 1.9029063133103153e-06, + "loss": 0.0003937560599297285, + "step": 9120 + }, + { + "epoch": 31.05451448040886, + "grad_norm": 0.005034316331148148, + "learning_rate": 1.8602210715474466e-06, + "loss": 0.00037171714939177036, + "step": 9130 + }, + { + "epoch": 31.088586030664395, + "grad_norm": 0.005186181981116533, + "learning_rate": 1.81801094996778e-06, + "loss": 0.0003877129405736923, + "step": 9140 + }, + { + "epoch": 31.12265758091993, + "grad_norm": 0.004060785751789808, + "learning_rate": 1.7762763651681892e-06, + "loss": 0.00036375392228364943, + "step": 9150 + }, + { + "epoch": 31.156729131175467, + "grad_norm": 0.005512597970664501, + "learning_rate": 1.735017729052163e-06, + "loss": 0.00037468173541128636, + "step": 9160 + }, + { + "epoch": 31.190800681431003, + "grad_norm": 0.004430436994880438, + "learning_rate": 1.6942354488257928e-06, + "loss": 0.00035622487775981425, + "step": 9170 + }, + { + "epoch": 31.224872231686543, + "grad_norm": 0.005033303517848253, + "learning_rate": 1.6539299269937192e-06, + "loss": 0.000362906651571393, + "step": 9180 + }, + { + "epoch": 31.25894378194208, + "grad_norm": 0.0043688369914889336, + "learning_rate": 1.6141015613551636e-06, + "loss": 0.0003627951489761472, + "step": 9190 + }, + { + "epoch": 31.293015332197616, + "grad_norm": 0.004733602982014418, + "learning_rate": 1.5747507450000198e-06, + "loss": 0.000360643956810236, + "step": 9200 + }, + { + "epoch": 31.327086882453152, + "grad_norm": 0.004427157808095217, + "learning_rate": 1.5358778663049579e-06, + "loss": 0.00037265238352119924, + "step": 9210 + }, + { + "epoch": 31.361158432708688, + "grad_norm": 0.005903623998165131, + "learning_rate": 1.4974833089296037e-06, + "loss": 0.00040756086818873884, + "step": 9220 + }, + { + "epoch": 31.395229982964224, + "grad_norm": 0.006067934446036816, + "learning_rate": 1.4595674518127268e-06, + "loss": 0.00038415833842009306, + "step": 9230 + }, + { + "epoch": 31.42930153321976, + "grad_norm": 0.004608978517353535, + "learning_rate": 1.4221306691685366e-06, + "loss": 0.0004104959312826395, + "step": 9240 + }, + { + "epoch": 31.463373083475297, + "grad_norm": 0.00555928610265255, + "learning_rate": 1.385173330482964e-06, + "loss": 0.0003708329284563661, + "step": 9250 + }, + { + "epoch": 31.497444633730836, + "grad_norm": 0.00429227901622653, + "learning_rate": 1.3486958005100247e-06, + "loss": 0.0003841962432488799, + "step": 9260 + }, + { + "epoch": 31.531516183986373, + "grad_norm": 0.004735529888421297, + "learning_rate": 1.3126984392682006e-06, + "loss": 0.0003921882249414921, + "step": 9270 + }, + { + "epoch": 31.56558773424191, + "grad_norm": 0.004542259965091944, + "learning_rate": 1.2771816020369253e-06, + "loss": 0.0003919101320207119, + "step": 9280 + }, + { + "epoch": 31.599659284497445, + "grad_norm": 0.004277036525309086, + "learning_rate": 1.2421456393530372e-06, + "loss": 0.0003835365641862154, + "step": 9290 + }, + { + "epoch": 31.63373083475298, + "grad_norm": 0.005043212324380875, + "learning_rate": 1.207590897007349e-06, + "loss": 0.0003854848211631179, + "step": 9300 + }, + { + "epoch": 31.667802385008518, + "grad_norm": 0.005156104918569326, + "learning_rate": 1.173517716041217e-06, + "loss": 0.00039365091361105443, + "step": 9310 + }, + { + "epoch": 31.701873935264054, + "grad_norm": 0.005229398608207703, + "learning_rate": 1.1399264327431713e-06, + "loss": 0.0003849109401926398, + "step": 9320 + }, + { + "epoch": 31.73594548551959, + "grad_norm": 0.005152896512299776, + "learning_rate": 1.106817378645636e-06, + "loss": 0.00040498794987797737, + "step": 9330 + }, + { + "epoch": 31.770017035775126, + "grad_norm": 0.004509305115789175, + "learning_rate": 1.0741908805215861e-06, + "loss": 0.00039845663122832774, + "step": 9340 + }, + { + "epoch": 31.804088586030666, + "grad_norm": 0.00556230591610074, + "learning_rate": 1.042047260381407e-06, + "loss": 0.0003905012272298336, + "step": 9350 + }, + { + "epoch": 31.838160136286202, + "grad_norm": 0.004550788551568985, + "learning_rate": 1.0103868354696343e-06, + "loss": 0.0003927945625036955, + "step": 9360 + }, + { + "epoch": 31.87223168654174, + "grad_norm": 0.004953572526574135, + "learning_rate": 9.792099182618919e-07, + "loss": 0.00036837037187069653, + "step": 9370 + }, + { + "epoch": 31.906303236797275, + "grad_norm": 0.0052412766963243484, + "learning_rate": 9.485168164617642e-07, + "loss": 0.00037995502352714536, + "step": 9380 + }, + { + "epoch": 31.94037478705281, + "grad_norm": 0.00424719275906682, + "learning_rate": 9.183078329977723e-07, + "loss": 0.000392998056486249, + "step": 9390 + }, + { + "epoch": 31.974446337308347, + "grad_norm": 0.005236937664449215, + "learning_rate": 8.885832660203875e-07, + "loss": 0.0003804670181125402, + "step": 9400 + }, + { + "epoch": 32.006814310051105, + "grad_norm": 0.004403868690133095, + "learning_rate": 8.593434088990882e-07, + "loss": 0.00036736084148287774, + "step": 9410 + }, + { + "epoch": 32.040885860306645, + "grad_norm": 0.004324344918131828, + "learning_rate": 8.305885502194577e-07, + "loss": 0.00037797554396092894, + "step": 9420 + }, + { + "epoch": 32.07495741056218, + "grad_norm": 0.004861414898186922, + "learning_rate": 8.02318973780336e-07, + "loss": 0.0003818760393187404, + "step": 9430 + }, + { + "epoch": 32.10902896081772, + "grad_norm": 0.004530130419880152, + "learning_rate": 7.745349585910389e-07, + "loss": 0.00037451134994626045, + "step": 9440 + }, + { + "epoch": 32.14310051107326, + "grad_norm": 0.004341290332376957, + "learning_rate": 7.472367788685763e-07, + "loss": 0.00037274835631251335, + "step": 9450 + }, + { + "epoch": 32.17717206132879, + "grad_norm": 0.004395690280944109, + "learning_rate": 7.204247040349665e-07, + "loss": 0.0003736852901056409, + "step": 9460 + }, + { + "epoch": 32.21124361158433, + "grad_norm": 0.004745770711451769, + "learning_rate": 6.940989987145485e-07, + "loss": 0.000399275217205286, + "step": 9470 + }, + { + "epoch": 32.24531516183986, + "grad_norm": 0.005878156516700983, + "learning_rate": 6.682599227314234e-07, + "loss": 0.00039322515949606894, + "step": 9480 + }, + { + "epoch": 32.2793867120954, + "grad_norm": 0.0036912173964083195, + "learning_rate": 6.429077311068344e-07, + "loss": 0.0003532448783516884, + "step": 9490 + }, + { + "epoch": 32.313458262350935, + "grad_norm": 0.0043158698827028275, + "learning_rate": 6.18042674056668e-07, + "loss": 0.0003742758883163333, + "step": 9500 + }, + { + "epoch": 32.347529812606474, + "grad_norm": 0.004498814232647419, + "learning_rate": 5.936649969890073e-07, + "loss": 0.0003707940457388759, + "step": 9510 + }, + { + "epoch": 32.38160136286201, + "grad_norm": 0.005161894019693136, + "learning_rate": 5.697749405016828e-07, + "loss": 0.000389894749969244, + "step": 9520 + }, + { + "epoch": 32.41567291311755, + "grad_norm": 0.004696319345384836, + "learning_rate": 5.463727403799079e-07, + "loss": 0.00040489970706403255, + "step": 9530 + }, + { + "epoch": 32.44974446337309, + "grad_norm": 0.004610268399119377, + "learning_rate": 5.234586275939424e-07, + "loss": 0.00036891817580908537, + "step": 9540 + }, + { + "epoch": 32.48381601362862, + "grad_norm": 0.004635406658053398, + "learning_rate": 5.010328282968435e-07, + "loss": 0.00038140476681292057, + "step": 9550 + }, + { + "epoch": 32.51788756388416, + "grad_norm": 0.004985370207577944, + "learning_rate": 4.790955638221795e-07, + "loss": 0.00037632756866514685, + "step": 9560 + }, + { + "epoch": 32.55195911413969, + "grad_norm": 0.006067863665521145, + "learning_rate": 4.576470506819086e-07, + "loss": 0.0003656267188489437, + "step": 9570 + }, + { + "epoch": 32.58603066439523, + "grad_norm": 0.005050371866673231, + "learning_rate": 4.366875005641924e-07, + "loss": 0.00037925925571471455, + "step": 9580 + }, + { + "epoch": 32.620102214650764, + "grad_norm": 0.004783795680850744, + "learning_rate": 4.162171203313192e-07, + "loss": 0.000395351555198431, + "step": 9590 + }, + { + "epoch": 32.654173764906304, + "grad_norm": 0.0053469352424144745, + "learning_rate": 3.9623611201769493e-07, + "loss": 0.00038469338323920965, + "step": 9600 + }, + { + "epoch": 32.688245315161836, + "grad_norm": 0.004364959429949522, + "learning_rate": 3.76744672827789e-07, + "loss": 0.0003702714340761304, + "step": 9610 + }, + { + "epoch": 32.722316865417376, + "grad_norm": 0.00545252813026309, + "learning_rate": 3.5774299513424147e-07, + "loss": 0.0003899726085364819, + "step": 9620 + }, + { + "epoch": 32.756388415672916, + "grad_norm": 0.004598419182002544, + "learning_rate": 3.3923126647594785e-07, + "loss": 0.00039402819238603114, + "step": 9630 + }, + { + "epoch": 32.79045996592845, + "grad_norm": 0.00497333612293005, + "learning_rate": 3.212096695561884e-07, + "loss": 0.00038548940792679785, + "step": 9640 + }, + { + "epoch": 32.82453151618399, + "grad_norm": 0.00417772401124239, + "learning_rate": 3.0367838224084623e-07, + "loss": 0.00037141768261790277, + "step": 9650 + }, + { + "epoch": 32.85860306643952, + "grad_norm": 0.006549280136823654, + "learning_rate": 2.866375775566643e-07, + "loss": 0.00039078202098608017, + "step": 9660 + }, + { + "epoch": 32.89267461669506, + "grad_norm": 0.004185816273093224, + "learning_rate": 2.700874236894857e-07, + "loss": 0.0003707955591380596, + "step": 9670 + }, + { + "epoch": 32.92674616695059, + "grad_norm": 0.004471260588616133, + "learning_rate": 2.540280839826659e-07, + "loss": 0.00037909401580691335, + "step": 9680 + }, + { + "epoch": 32.96081771720613, + "grad_norm": 0.004595866426825523, + "learning_rate": 2.3845971693539083e-07, + "loss": 0.00037641602102667093, + "step": 9690 + }, + { + "epoch": 32.99488926746167, + "grad_norm": 0.0038562098052352667, + "learning_rate": 2.2338247620116714e-07, + "loss": 0.0003711195196956396, + "step": 9700 + }, + { + "epoch": 33.02725724020443, + "grad_norm": 0.005165089853107929, + "learning_rate": 2.087965105862677e-07, + "loss": 0.00035953233018517496, + "step": 9710 + }, + { + "epoch": 33.06132879045997, + "grad_norm": 0.00405836058780551, + "learning_rate": 1.9470196404828278e-07, + "loss": 0.00038429896812886, + "step": 9720 + }, + { + "epoch": 33.0954003407155, + "grad_norm": 0.004763876553624868, + "learning_rate": 1.8109897569470457e-07, + "loss": 0.0003812138456851244, + "step": 9730 + }, + { + "epoch": 33.12947189097104, + "grad_norm": 0.005231298040598631, + "learning_rate": 1.6798767978153386e-07, + "loss": 0.0003790364135056734, + "step": 9740 + }, + { + "epoch": 33.16354344122657, + "grad_norm": 0.004111438058316708, + "learning_rate": 1.553682057119643e-07, + "loss": 0.00037519976031035184, + "step": 9750 + }, + { + "epoch": 33.19761499148211, + "grad_norm": 0.0037241752725094557, + "learning_rate": 1.4324067803511142e-07, + "loss": 0.0003526164684444666, + "step": 9760 + }, + { + "epoch": 33.23168654173765, + "grad_norm": 0.005046004429459572, + "learning_rate": 1.3160521644478008e-07, + "loss": 0.0003639630973339081, + "step": 9770 + }, + { + "epoch": 33.265758091993185, + "grad_norm": 0.004693778697401285, + "learning_rate": 1.2046193577827657e-07, + "loss": 0.0003764047985896468, + "step": 9780 + }, + { + "epoch": 33.299829642248724, + "grad_norm": 0.004903760738670826, + "learning_rate": 1.0981094601528186e-07, + "loss": 0.0003788340603932738, + "step": 9790 + }, + { + "epoch": 33.33390119250426, + "grad_norm": 0.005005911458283663, + "learning_rate": 9.965235227676895e-08, + "loss": 0.0003828839864581823, + "step": 9800 + }, + { + "epoch": 33.3679727427598, + "grad_norm": 0.005354180466383696, + "learning_rate": 8.998625482394829e-08, + "loss": 0.00038404972292482853, + "step": 9810 + }, + { + "epoch": 33.40204429301533, + "grad_norm": 0.004775671288371086, + "learning_rate": 8.08127490573074e-08, + "loss": 0.00037608013954013584, + "step": 9820 + }, + { + "epoch": 33.43611584327087, + "grad_norm": 0.004217393230646849, + "learning_rate": 7.213192551564496e-08, + "loss": 0.00037033415865153076, + "step": 9830 + }, + { + "epoch": 33.47018739352641, + "grad_norm": 0.004750920459628105, + "learning_rate": 6.394386987517708e-08, + "loss": 0.0003689353121444583, + "step": 9840 + }, + { + "epoch": 33.50425894378194, + "grad_norm": 0.0038863159716129303, + "learning_rate": 5.624866294871578e-08, + "loss": 0.0003623348427936435, + "step": 9850 + }, + { + "epoch": 33.53833049403748, + "grad_norm": 0.004313604906201363, + "learning_rate": 4.90463806848418e-08, + "loss": 0.00038169913459569217, + "step": 9860 + }, + { + "epoch": 33.572402044293014, + "grad_norm": 0.005349631421267986, + "learning_rate": 4.233709416717746e-08, + "loss": 0.00040154606103897094, + "step": 9870 + }, + { + "epoch": 33.606473594548554, + "grad_norm": 0.0038566242437809706, + "learning_rate": 3.612086961366501e-08, + "loss": 0.0003621417796239257, + "step": 9880 + }, + { + "epoch": 33.640545144804086, + "grad_norm": 0.005148730240762234, + "learning_rate": 3.0397768375939325e-08, + "loss": 0.00039507062174379823, + "step": 9890 + }, + { + "epoch": 33.674616695059626, + "grad_norm": 0.0049018869176507, + "learning_rate": 2.5167846938689567e-08, + "loss": 0.0003678308567032218, + "step": 9900 + }, + { + "epoch": 33.70868824531516, + "grad_norm": 0.0053993333131074905, + "learning_rate": 2.0431156919137328e-08, + "loss": 0.0003757542232051492, + "step": 9910 + }, + { + "epoch": 33.7427597955707, + "grad_norm": 0.004623922985047102, + "learning_rate": 1.6187745066498228e-08, + "loss": 0.00038737766444683073, + "step": 9920 + }, + { + "epoch": 33.77683134582624, + "grad_norm": 0.0055684191174805164, + "learning_rate": 1.2437653261537785e-08, + "loss": 0.0003868590574711561, + "step": 9930 + }, + { + "epoch": 33.81090289608177, + "grad_norm": 0.00475655822083354, + "learning_rate": 9.180918516138448e-09, + "loss": 0.00036844646092504263, + "step": 9940 + }, + { + "epoch": 33.84497444633731, + "grad_norm": 0.004000869579613209, + "learning_rate": 6.417572972966524e-09, + "loss": 0.0003942666575312614, + "step": 9950 + }, + { + "epoch": 33.87904599659284, + "grad_norm": 0.004669535905122757, + "learning_rate": 4.147643905122456e-09, + "loss": 0.0003999684005975723, + "step": 9960 + }, + { + "epoch": 33.91311754684838, + "grad_norm": 0.005190175957977772, + "learning_rate": 2.371153715891028e-09, + "loss": 0.0003674669889733195, + "step": 9970 + }, + { + "epoch": 33.947189097103916, + "grad_norm": 0.004048798233270645, + "learning_rate": 1.0881199385137652e-09, + "loss": 0.00037811114452779293, + "step": 9980 + }, + { + "epoch": 33.981260647359456, + "grad_norm": 0.004426008556038141, + "learning_rate": 2.985552360168509e-10, + "loss": 0.0003870480926707387, + "step": 9990 + }, + { + "epoch": 34.01362862010222, + "grad_norm": 0.005143159069120884, + "learning_rate": 2.4674010778991117e-12, + "loss": 0.0003771143266931176, + "step": 10000 + } + ], + "logging_steps": 10, + "max_steps": 10000, + "num_input_tokens_seen": 0, + "num_train_epochs": 35, + "save_steps": 500, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 0.0, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}